Ecological Risk Assessment for General Electric (GE)/Housatonic River Site Rest of River Volume 4 Appendix B: Pre-era Appendix C: Supporting Technical Information Appendix D: Assessment Endpoint


U.S. Army
Corps of Engineers

U.S. Environmental
Protection Agency

New England District
Concord, Massachusetts

New England Region
Boston, Massachusetts

ECOLOGICAL RISK ASSESSMENT FOR
GENERAL ELECTRIC (GE)/HOUSATONIC RIVER SITE,

REST OF RIVER

Volume 4
Appendix B: Pre-ERA
Appendix C: Supporting Technical Information
Appendix D: Assessment Endpoint - Benthic Invertebrates

DCN: GE-070703-ABRC
July 2003

Environmental Remediation Contract
GE/Housatonic River Project
Pittsfield, Massachusetts

Contract No. DACW33-00-D-0006

Task Order 0003

03P-0966-4

-------
APPENDIX B
Pre-ERA

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

-------
TABLE OF CONTENTS

Section	Page

B. PRE-ERA	B-l

B. 1 INTRODUCTION	B-l

B.2 DATA	B-l

B.2.1 Primary Study Area	B-2

B.2.2 Below Woods Pond	B-4

B.2.3 Background Data	B-4

B.2.3.1 Sediment Background Data	B-5

B.2.3.2 Surface Water Background Data	B-5

B.2.3.3 Soil Background Data	B-6

B.2.3.4 Fish Background Data	B-6

B.3 DATA SUMMARY	B-7

B.4 PRIMARY STUDY AREA EVALUATION	B-7

B.4.1 Tier I Approach	B-8

B.4.1.1 Frequency of Detection Screening	B-8

B.4.1.2 Benchmark Comparisons	B-8

B.4.1.3 Benchmark Comparison Results	B-19

B.4.1.4 Background Screening	B-19

B.4.2 Tier I Results	B-21

B.4.3 Tier II Approach	B-21

B.4.4 Tier II Results	B-22

B.4.5 Tier III Approach	B-23

B.4.6 Tier III Results	B-23

B.4.6.1 Semivolatiles	B-24

B.4.6.2 Pesticides	B-25

B.4.6.3 Inorganics	B-26

B.5 APPROACH BELOW WOODS POND	B-27

B.6 RESULTS FROM BELOW WOODS POND	B-28

B.7 REFERENCES	B-28

ATTACHMENTS

Attachment B.l—Tables and Figures (on CD)

Attachment B.2—Fish Tissue Re-Analysis and Interpretation Overview

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC	• • •	7/10/2003

-------
LIST OF TABLES

(Tables are presented on CD)

Title

Table B-l	Sediment Background Samples

Table B-2	Surface Water Background Samples

Table B-3	Soil Background Samples

Table B-4	Fish Background Samples

Table B-5	Sediment Chemistry Summary - 5A Main Channel and Aggrading Bars

Table B-6	Sediment Chemistry Summary - 5A SCOX

Table B-7	Sediment Chemistry Summary - 5A Vernal Pools

Table B-8	Sediment Chemistry Summary - 5B Main Channel and Aggrading Bars

Table B-9	Sediment Chemistry Summary - 5B SCOX

Table B-10	Sediment Chemistry Summary - 5B Vernal Pools

Table B-l 1	Sediment Chemistry Summary - 5C Main Channel

Table B-12	Sediment Chemistry Summary - 5C SCOX

Table B-13	Sediment Chemistry Summary - 5C Vernal Pools

Table B-14	Sediment Chemistry Summary - 6AB Main Channel

Table B-l5	Sediment Chemistry Summary - 6AB SCOX

Table B-l6	Sediment Chemistry Summary - 6AB Vernal Pools

Table B-17	Sediment Chemistry Summary - 6CD Pond

Table B-l8	Sediment Chemistry Summary - Woods Pond Dam to
Massachusetts/Connecticut Border

Table B-l9	Sediment Chemistry Summary - Massachusetts/Connecticut Border to
Housatonic Lake

Table B-20	Sediment Chemistry Summary - Background

Table B-21	Surface Water Chemistry Summary - 5A Main Channel

Table B-22	Surface Water Chemistry Summary - 5A SCOX

Table B-23	Surface Water Chemistry Summary - 5A Vernal Pools

Table B-24	Surface Water Chemistry Summary - 5B Main Channel

Table B-25	Surface Water Chemistry Summary - 5B Vernal Pools

Table B-26	Surface Water Chemistry Summary - 5C Main Channel

Table B-27	Surface Water Chemistry Summary - 5C SCOX

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

IV

-------
LIST OF TABLES
(Continued)

(Tables are presented on CD)

Title

Table B-28	Surface Water Chemistry Summary - 5C Vernal Pools

Table B-29	Surface Water Chemistry Summary - 6CD Pond

Table B-30	Surface Water Chemistry Summary - Background

Table B-31	Soil Chemistry Summary - 5A Floodplain

Table B-32	Soil Chemistry Summary - 5A Riverbank

Table B-33	Soil Chemistry Summary - 5B Floodplain

Table B-34	Soil Chemistry Summary - 5B Riverbank

Table B-35	Soil Chemistry Summary - 5C Floodplain

Table B-36	Soil Chemistry Summary - 5C Riverbank

Table B-37	Soil Chemistry Summary - 6AB Floodplain

Table B-38	Soil Chemistry Summary - 6AB Riverbank

Table B-39	Soil Chemistry Summary - 6CD Floodplain

Table B-40	Soil Chemistry Summary - Background

Table B-41	Fish Tissue Chemistry Summary - 5A Small Fish

Table B-42	Fish Tissue Chemistry Summary - 5A Medium Fish

Table B-43	Fish Tissue Chemistry Summary - 5A Large Fish

Table B-44	Fish Tissue Chemistry Summary - 5BC Small Fish

Table B-45	Fish Tissue Chemistry Summary - 5BC Medium Fish

Table B-46	Fish Tissue Chemistry Summary - 5BC Large Fish

Table B-47	Fish Tissue Chemistry Summary - 6CD Small Fish

Table B-48	Fish Tissue Chemistry Summary - 6CD Medium Fish

Table B-49	Fish Tissue Chemistry Summary - 6CD Large Fish

Table B-50	Fish Tissue Chemistry Summary - Background Small Fish

Table B-51	Fish Tissue Chemistry Summary - Background Medium Fish

Table B-52	Fish Tissue Chemistry Summary - Background Large Fish

Table B-53	Pre-ERA Sediment Benchmarks

Table B-54	Pre-ERA Surface Water Benchmarks

Table B-55	Pre-ERA Soil Benchmarks

Table B-56	Fish Concentration Benchmarks for Piscivores

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

-------
LIST OF TABLES
(Continued)

(Tables are presented on CD)

Title

Table B-57 Secondary Sources Reviewed for the Identification of Primary Articles Toxicity
Reference Values

Table B-58 Mammalian Toxicity Reference Values

Table B-59 Avian Toxicity Reference Values

Table B-60 Sediment Comparison to Benchmark Summary - Low - 5A Main Channel and
Aggrading Bars

Table B-61 Sediment Comparison to Benchmark Summary - High - 5A Main Channel and
Aggrading Bars

Table B-62 Sediment Comparison to Benchmark Summary - Low - 5A SCOX

Table B-63 Sediment Comparison to Benchmark Summary - High - 5A SCOX

Table B-64 Sediment Comparison to Benchmark Summary - Low - 5A Vernal Pools

Table B-65 Sediment Comparison to Benchmark Summary - High - 5A Vernal Pools

Table B-66	Sediment Comparison to Benchmark Summary - Low - 5B Main Channel and
Aggrading Bars

Table B-67	Sediment Comparison to Benchmark Summary - High - 5B Main Channel and
Aggrading Bars

Table B-68	Sediment Comparison to Benchmark Summary - Low - 5B SCOX

Table B-69	Sediment Comparison to Benchmark Summary - High - 5B SCOX

Table B-70	Sediment Comparison to Benchmark Summary - Low - 5B Vernal Pools

Table B-71	Sediment Comparison to Benchmark Summary - High - 5B Vernal Pools

Table B-72	Sediment Comparison to Benchmark Summary - Low - 5C Main Channel

Table B-73	Sediment Comparison to Benchmark Summary - High - 5C Main Channel

Table B-74	Sediment Comparison to Benchmark Summary - Low - 5C SCOX

Table B-75	Sediment Comparison to Benchmark Summary - High - 5C SCOX

Table B-76	Sediment Comparison to Benchmark Summary - Low - 5C Vernal Pools

Table B-77	Sediment Comparison to Benchmark Summary - High - 5C Vernal Pools

Table B-78	Sediment Comparison to Benchmark Summary - Low - 6AB Main Channel

Table B-79	Sediment Comparison to Benchmark Summary - High - 6AB Main Channel

Table B-80	Sediment Comparison to Benchmark Summary - Low - 6AB SCOX

Table B-81	Sediment Comparison to Benchmark Summary - High - 6AB SCOX

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

VI

-------
LIST OF TABLES
(Continued)

(Tables are presented on CD)

Title

Table

B-82

Table

B-83

Table

B-84

Table

B-85

Table

B-86

Table

B-87

Table

B-88

Table

B-89

Table

B-90

Table

B-91

Table

B-92

Table

B-93

Table

B-94

Table

B-95

Table

B-96

Table

B-97

Table

B-98

Table

B-99

Table

B-100

Table

B-101

Table

B-102

Table

B-103

Table

B-104

Table

B-105

Table

B-106

Table

B-107

Table

B-108

Table

B-109

Table

B-110

Sediment Comparison to Benchmark Summary - Low - 6AB Vernal Pools
Sediment Comparison to Benchmark Summary - High - 6AB Vernal Pools
Sediment Comparison to Benchmark Summary - Low - 6CD Pond
Sediment Comparison to Benchmark Summary - High - 6CD Pond
Sediment Comparison to Benchmark Summary - Low - Background
Sediment Comparison to Benchmark Summary - High - Background
Surface Water Comparison to Benchmark Summary - 5A Main Channel
Surface Water Comparison to Benchmark Summary - 5A SCOX
Surface Water Comparison to Benchmark Summary - 5A Vernal Pools
Surface Water Comparison to Benchmark Summary - 5B Main Channel
Surface Water Comparison to Benchmark Summary - 5B Vernal Pools
Surface Water Comparison to Benchmark Summary - 5C Main Channel
Surface Water Comparison to Benchmark Summary - 5C SCOX
Surface Water Comparison to Benchmark Summary - 5C Vernal Pools
Surface Water Comparison to Benchmark Summary - 6CD Pond
Surface Water Comparison to Benchmark Summary - Background
Soil Comparison to Benchmark Summary - 5A Floodplain
Soil Comparison to Benchmark Summary - 5A Riverbank
Soil Comparison to Benchmark Summary - 5B Floodplain
Soil Comparison to Benchmark Summary - 5B Riverbank
Soil Comparison to Benchmark Summary - 5C Floodplain
Soil Comparison to Benchmark Summary - 5C Riverbank
Soil Comparison to Benchmark Summary - 6AB Floodplain
Soil Comparison to Benchmark Summary - 6AB Riverbank
Soil Comparison to Benchmark Summary - 6CD Floodplain
Soil Comparison to Benchmark Summary - Background
Fish Tissue Comparison to Benchmark Summary - 5A Small Fish
Fish Tissue Comparison to Benchmark Summary - 5A Medium Fish
Fish Tissue Comparison to Benchmark Summary - 5A Large Fish

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

Vll

-------
LIST OF TABLES
(Continued)

(Tables are presented on CD)

Title

Table B-l 11	Fish Tissue Comparison to Benchmark Summary - 5BC Small Fish

Table B-l 12	Fish Tissue Comparison to Benchmark Summary - 5BC Medium Fish

Table B-l 13	Fish Tissue Comparison to Benchmark Summary - 5BC Large Fish

Table B-l 14	Fish Tissue Comparison to Benchmark Summary - 6CD Small Fish

Table B-l 15	Fish Tissue Comparison to Benchmark Summary - 6CD Medium Fish

Table B-l 16	Fish Tissue Comparison to Benchmark Summary - 6CD Large Fish

Table B-l 17	Fish Tissue Comparison to Benchmark Summary - Background Small Fish

Table B-l 18	Fish Tissue Comparison to Benchmark Summary - Background Medium Fish

Table B-l 19	Fish Tissue Comparison to Benchmark Summary - Background Large Fish

Table B-120	Analysis Options for Sediment Chemicals for Comparison to Background

Table B-121	Background Comparison Results - Sediment

Table B-122	Analysis Options for Surface Water Chemicals for Comparison to Background

Table B-123	Background Comparison Results - Surface Water

Table B-124	Analysis Options for Soil Chemicals for Comparison to Background

Table B-125	Background Comparison Results - Soil

Table B-126	Analysis Options for Fish Chemicals for Comparison to Background

Table B-127	Background Comparison Results - Fish

Table B-128	Tier I Evaluation Results - Sediment

Table B-129	Tier I Evaluation Results - Surface Water

Table B-130	Tier I Evaluation Results - Soil

Table B-131	Tier I Evaluation Results - Fish

Table B-132	Tier II Evaluation Results - Sediment

Table B-133	Tier II Evaluation Results - Surface Water

Table B-134	Tier II Evaluation Results - Soil

Table B-135	Tier II Evaluation Results - Fish

Table B-136	Chemicals Evaluated in Tier III Sediment Evaluation without Statistical

Background Comparisons or that Are Significantly Higher than Background

Table B-l37	Chemicals Evaluated in Tier III Surface Water Evaluation without Statistical
Background Comparisons or that Are Significantly Higher than Background

MK01|O:\20123001.096\ERA_PB\APPB\APPB_PB_PREERA.DOC	•••

-------
LIST OF TABLES
(Continued)

(Tables are presented on CD)

Title

Table B-138	Chemicals Evaluated in Tier III Soil Evaluation without Statistical Background
Comparisons or that Are Significantly Higher than Background

Table B-139	Tier III Evaluation Results - Sediment

Table B-140	Tier III Evaluation Results - Surface Water

Table B-141	Tier III Evaluation Results - Soil

Table B-142	Tier III Evaluation Results - Fish

Table B-143	Nondetected Chemicals with SQLs Greater than Benchmarks

Table B-144 Detected Chemicals with SQLs Greater than Benchmarks Recommended for
Uncertainty Analysis Discussion Instead of Quantitative Evaluation

Table B-145	Final COPCs - Sediment

Table B-146 Final COPCs - Surface Water

Table B-147 Final COPCs - Soil

Table B-148	Final COPCs - Risk Assessment Fish

MK01|O:\20123001.096\ERA PB\APP B\APPB PB PREERA.DOC

I	_ _	lx

-------
LIST OF FIGURES

(Figures are presented on CD)

Title

Figure B -1	Reach Map

Figure B-2	Index to Background Location Maps

Figure B-3	Sediment and Surface Water Background Locations, Upstream of Newell Street

Figure B-4	Sediment and Surface Water Background Locations, Upstream of Newell Street

Figure B-5	Sediment and Surface Water Background Locations, Washington Mountain
Lake

Figure B-6	Sediment and Surface Water Background Locations, Muddy Pond

Figure B-7	Sediment and Surface Water Background Locations, Three-Mile Pond

Figure B-8	Soil Background Locations, Reach 5 and 6, Tile 1/3

Figure B-9	Soil Background Locations, Reach 5 and 6, Tile 2/3

Figure B-10	Soil Background Locations, Reach 5 and 6, Tile 3/3

Figure B-l 1	Fish Background Location Map

Figure B-12	Sediment Samples with Hazard Quotients Exceeding

Figure B-13	Sediment Samples with Hazard Quotients Exceeding

Figure B-14	Sediment Samples with Hazard Quotients Exceeding

Figure B-l5	Sediment Samples with Hazard Quotients Exceeding

Figure B-16	Sediment Samples with Hazard Quotients Exceeding

Figure B-17	Sediment Samples with Hazard Quotients Exceeding

Figure B-l8	Sediment Samples with Hazard Quotients Exceeding

Figure B-19	Sediment Samples with Hazard Quotients Exceeding

Figure B-20	Sediment Samples with Hazard Quotients Exceeding

Figure B-21	Sediment Samples with Hazard Quotients Exceeding

Figure B-22	Sediment Samples with Hazard Quotients Exceeding

Figure B-23	Sediment Samples with Hazard Quotients Exceeding

Figure B-24	Sediment Samples with Hazard Quotients Exceeding

Figure B-25	Sediment Samples with Hazard Quotients Exceeding

Figure B-26	Sediment Samples with Hazard Quotients Exceeding

Figure B-27	Sediment Samples with Hazard Quotients Exceeding

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

X

-------
LIST OF FIGURES
(Continued)

(Figures are presented on CD)

Title

Figure B-28
Figure B-29
Figure B-30
Figure B-31
Figure B-32
Figure B-33
Figure B-34
Figure B-35
Figure B-36
Figure B-37
Figure B-38
Figure B-39
Figure B-40
Figure B-41
Figure B-42
Figure B-43
Figure B-44
Figure B-45
Figure B-46
Figure B-47
Figure B-48
Figure B-49
Figure B-50
Figure B-51
Figure B-52
Figure B-53
Figure B-54

Figure B-55

Sediment Samples with Hazard Quotients Exceeding 1
Sediment Samples with Hazard Quotients Exceeding 1
Sediment Samples with Hazard Quotients Exceeding 1
Sediment Samples with Hazard Quotients Exceeding 1
Sediment Samples with Hazard Quotients Exceeding 1

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Surface

Water

and

Soil

Samples

with

Hazard

Quotient Exceeding

Index to Hazard Quotient Screening Maps, Woods Pond Dam to MA-CT
Border

Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 1/10

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

XI

-------
LIST OF FIGURES
(Continued)

(Figures are presented on CD)

Title

Figure B-56	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 2/10

Figure B-57	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 3/10

Figure B-58	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 4/10

Figure B-59	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 5/10

Figure B-60	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 6/10

Figure B-61	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 7/10

Figure B-62	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 8/10

Figure B-63	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 9/10

Figure B-64	Hazard Quotient Screening, Woods Pond Dam to MA-CT Border, Tile 10/10

Figure B-65	Index to Hazard Quotient Screening Maps, MA-CT Border to Long Island





Sound



















Figure

B-66

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

1/9

Figure

B-67

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

2/9

Figure

B-68

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

3/9

Figure

B-69

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

4/9

Figure

B-70

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

5/9

Figure

B-71

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

6/9

Figure

B-72

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

7/9

Figure

B-73

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

8/9

Figure

B-74

Hazard

Quotient

Screening,

MA-CT

Border to

Long

Island

Sound,

Tile

9/9

MK01|O:\20123001.096\ERA PB\APP B\APPB PB PREERA.DOC

- -	Xll

-------
LIST OF ACRONYMS

ARCS

Assessment and Remediation of Contaminated Sediments

ccc

criteria continuous concentration

COPCs

contaminants of potential concern

EqP

equilibrium partitioning

ER-L

effects range-low

ER-M

effects range-median

HQ

hazard quotient

LEL

lowest effect level

LOAEL

lowest observed adverse effect level

NAWQC

National Ambient Water Quality Criteria

NEC

no effect concentration

NEL

no effect level

NOAEL

no observed adverse effect level

OMEE

Ontario Ministry of Environment and Energy

PEC

probable effect concentration

PEL

probable effect level

PRGs

Preliminary Remediation Goals

PSA

Primary Study Area

PWQOs/Gs

Provincial Water Quality Objectives and Guidelines

SCOX

side channels and oxbows

SECs

sediment effect concentrations

SEL

severe-effect level

SLC

screening level concentration

SQGs

sediment quality guidelines

SSLC

species screening level concentration

TEC

threshold effect concentration

TEL

threshold effect level

TRVs

toxicity reference values

WQGs

Water Quality Guidelines

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC	• • •

X.111

-------
1	APPENDIX B

2

3	PRE-ERA

4	B.1 INTRODUCTION

5	The purpose of the pre-ERA was to narrow the scope of the ecological risk assessment (ERA) by

6	identifying contaminants that pose potential risks to aquatic biota and wildlife in water,

7	sediment, soil, and tissue. The primary objectives of this pre-ERA are two-fold:

8	¦ Identify contaminants of potential ecological concern (COPCs) other than PCBs for

9	the Primary Study Area (PSA).

10	¦ Determine the downstream boundary beyond which PCBs pose a negligible risk to

11	aquatic biota and wildlife.

12	The following discussion provides a detailed description of the steps taken to identify COPCs for

13	the PSA and the downstream extent of the ERA. The pre-ERA is divided into four main

14	sections:

15	¦ Section B.2, Data (describes data sets used in the COPC selection process).

16	¦ Section B.3, Data Summary (describes how data were treated and presents statistical

17	summaries).

18	¦ Section B.4, PSA Evaluation (discussion of the three-tiered evaluation process and

19	COPC selection results).

20	¦ Section B.5, Below Woods Ponds Evaluation (discussion of the evaluation process

21	for the area downstream of Woods Pond and results).

22	B.2 DATA

23	Data sets were developed for the primary media of concern for the PSA, background areas, and

24	for the area below Woods Pond.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-l

-------
1

B.2.1 Primary Study Area

The COPC selection process was limited to data collected as part of the Supplemental
Investigation available as of March 2001. Because data quality determinations were available
only for data collected by EPA as of March 2001, only "A" qualified (acceptable, unrestricted
use) EPA data were used for the PSA COPC selection process. The pre-ERA for the PSA was
limited to the following media:

¦ Surface water.

¦ Sediment (0 to 6 inches [0 to 15 cm]) (Note: vernal pool samples were evaluated in
the pre-ERA as sediments, not soils).

¦ Riverbank and floodplain soil (0 to 12 inches [0 to 30 cm]).

¦ Fish tissue.

Data in the PSA were grouped by media (i.e., sediment, surface water, soil, and fish tissue),
subreach, and geomorphological terrain code. The reaches used for this evaluation are
hydraulically similar sections of the Housatonic River, identified by the project modeling team in
the Modeling Framework Design (WESTON 2003, in preparation) (see Figure B-l). Geomorph
codes are geomorphological terrain descriptions assigned to sediment, soil, and surface water
samples collected by EPA. Each geomorph code represents a depositional or erosional feature or
a terrain type that was formed by a specific geologic process (e.g., main channel, vernal pools,
and side channels). The following are the sediment and water data categories used for the pre-
ERA:

Medium/Reach

Geomorphological Terrain Type

Main Channel/
Aggrading Bars

Side Channels and
Oxbows (SCOX)

Vernal Pools

Pond

Sediment

6AB

6CD

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC g ^ 7/10/2003

-------
Medium/Reach

Geomorphological Terrain Type

Main Channel/
Aggrading Bars

Side Channels and
Oxbows (SCOX)

Vernal Pools

Pond

Surface Water

5A

V

V

V



5B

V



V



5C

V

V

V



6CD







V

1

2	Floodplain and riverbank soils were evaluated separately for Reaches 5A, 5B, 5C, and 6AB.

3	Soils adjacent to Woods Pond (i.e., 6CD) were also evaluated.

4	Fish were grouped based on reach (5A, 5BC, and 6CD) and size class. Three size classes were

5	used, small (< 3 inches [7.6 cm]), medium (> 3 inches [7.6 cm] but < 12 inches [30 cm]), and

6	large (>12 inches [30 cm]).

7	Fish species collected in the size classes are listed in the table below.

Small Fish
(< 3 inches [7.6 cm])

Medium Fish
3 inches [7.6 cm] but < 12
inches [30.5 cm])

Large Fish
12 inches [30.5 cm])

Bluegill

Bluegill

Common carp

Bluntnose minnow

Brown bullhead

Goldfish

Golden shiner

Common shiner

Largemouth bass

Largemouth bass

Fallfish

Yellow perch

Pumpkinseed

Golden shiner





Goldfish





Largemouth bass





Pumpkinseed





Smallmouth bass





Yellow bullhead





Yellow perch



8

9	In general, fish > 12 inches (30 cm) were analyzed as fillet and offal. Fillet samples consist of

10	the large muscle masses located on the sides of the fish between the dorsal spine and the ventral

11	region (two fillets per fish, with skin removed); while offal samples consist of the fish tissue

12	remaining after the fillets are removed (including any skin removed from the fillets). These fish

13	were mathematically "reconstructed" to determine whole body concentrations using the

14	following equation:

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC	g ^	7/10/2003

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

^ _Cf*Wf + Co*Wo

l^/wb — 	

Wf + Wo

where:

Cwb =	Concentration in whole body fish

Cf	=	Concentration in fillet

Wf	=	Weight of fillet

C0	=	Concentration in offal

W0	=	Weight of offal

Data are provided on the accompanying CD.

B.2.2 Below Woods Pond

The data set for determining the spatial extent of the ERA below Woods Pond was limited to
total PCB concentrations in sediment (0 to 6 inches [0 to 15 cm]) only. Of the available data, the
sediment data set had the most comprehensive spatial coverage. In addition, based on the fate
and transport of PCBs released into the river system, the sediment data provided the best
estimate of the potential downstream extent of PCB contamination originating from the GE
facility in Pittsfield. This data set does not contain any data from the tidally influenced portion
of the river extending upstream from Long Island Sound to the Derby-Shelton dam because there
are other Superfund sites that were known to have contributed significantly to the PCB
concentrations observed in this area.

B.2.3 Background Data

Samples that are considered to represent background are media-specific (i.e., sediment, surface
water, soil, and fish tissue) data collected within the Housatonic River watershed that are not
believed to be influenced by contamination originating from the GE Pittsfield facility. The
objective for establishing background concentrations was to identify the media-specific
contaminant concentrations that could be found in the absence of the GE facility. This
information will be used to assist in screening COPCs for the ERA.

Sample identifiers considered to represent background for the ecological risk assessment are
provided in Tables B-l through B-4, and the sample location maps are presented in Figures B-2
through B-ll. (Tables and figures are presented on CD in Attachment B.l.) Factors considered

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC	-pj a	7/10/2003

-------
1	when identifying the background data included whether the location was used as a reference area

2	for any of the biological studies, the potential for influence of GE facility-related releases, and

3	the total PCB concentrations associated with the location (as an indicator of influence from the

4	GE facility). Descriptions of the selected background data are presented below.

5	B.2.3.1 Sediment Background Data

6	Samples collected from the East Branch of the Housatonic River upstream of Unkamet Brook,

7	the Pittsfield Municipal Landfill, and in the West Branch and other waterbodies within the

8	Housatonic River watershed were selected as locations for sediment background samples (see

9	Table B-l and Figures B-3 through B-7). The locations of the 23 sediment background samples
10	(0 to 6 inches [0 to 15 cm]) are as follows:

Geographic Location	Number of Samples

Housatonic River upstream of facility influence	11

Muddy Pond	2

Threemile Pond	3

Washington Mountain Lake	1

WML-1*	2

WML-2*	2

WML-3 *	2

11	* Unnamed ponds separate from but in the vicinity of Washington Mountain Lake.

12

13	B.2.3.2 Surface Water Background Data

14	As with the sediment background data, samples upstream of the GE facility and in other

15	waterbodies in the Housatonic River watershed were selected as surface water background

16	samples (see Table B-2 and Figures B-3 through B-7).

17	The locations of the 39 surface water background samples are as follows:

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-5

-------
Geographic Location Number of Samples

Housatonic River upstream of facility influence 34

Muddy Pond 1

WML-1* 2

WML-2* 1

WML-3* 1

1 * Unnamed ponds separate from but in the vicinity of Washington Mountain Lake.

3 B.2.3.3 Soil Background Data

4 Twenty samples were selected for use as soil background samples. These samples are located

5 below the confluence of the East and West Branches of the Housatonic River (see Table B-3 and

6 Figures B-8 through B-10), and met the following criteria:

7 ¦ PCBs were not detected at a sample quantitation limit of less than 0.6 mg/kg, or

8 detected at a concentration of less than 0.3 mg/kg.

9 ¦ The sample was analyzed for EPA Appendix IX compounds.

10 ¦ The sample was located near the edge of the floodplain, outside the 10-year

11 floodplain, or within a well-defined area within the floodplain that is clearly outside

12 the influence of Housatonic River flooding (usually a high topographical location).

13 ¦ The sample was located at a distance (generally greater than 25 feet), horizontally or

14 vertically, from samples with elevated PCB concentrations (see first criterion).

15 B.2.3.4 Fish Background Data

16 Fish collected from the East Branch of the Housatonic River upstream of Unkamet Brook and

17 the landfill and in Threemile Pond were considered to represent background (see Table B-4 and

18 Figure B-ll). The fish species for which samples are available in each of these areas are

19 presented below.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-6

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

Upstream (n = 76)

Threemile Pond (n = 79)

Bluntnose minnow	Brown bullhead

Brown bullhead	Golden shiner

Common shiner	Largemouth bass

Fallfish	Pumpkinseed

Golden shiner	Yellow perch

Largemouth bass
Pumpkinseed
Yellow perch

B.3 DATA SUMMARY

For the PSA, data summary tables were compiled for each of the categories noted above. For the
contaminants that were detected in at least one sample, tables include frequency of detection, the
range of detected concentrations, the range of sample quantitation limits (SQLs) of the non-
detected samples, the median value, and the 95th percentile value. To be conservative in not
screening out COPCs prematurely, samples where concentrations were non-detects were
assumed to be present at the SQL. Duplicate samples were averaged unless there was a >30%
difference in surface water concentrations or a >50% difference in soil, sediment, or tissue
concentrations, in which case, the higher of the two concentrations was used. Summary tables
for the PSA, below Woods Pond to the tidally influenced portion of the Housatonic River, and
background data sets are provided in Tables B-5 through B-52.

B.4 PRIMARY STUDY AREA EVALUATION

The procedures used to screen potential COPCs were applied to the data groupings as
summarized above. The following is an overview of the three evaluation tiers used to determine
COPCs for the PSA.

¦	Tier I - A three-step process (i.e., screening frequency of detection, exceedance of
benchmarks, and comparisons to background concentrations) was used to establish
the initial COPC list.

¦	Tier II - Further evaluation was performed for the contaminants that were not
removed as a result of the Tier I evaluation including screening contaminants based
on detection and frequency of exceeding benchmarks.

¦	Tier III - The third tier of the COPC selection process evaluated contaminants not
removed in Tier I or II; and consisted of a subjective contaminant evaluation of

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC	g j	7/10/2003

-------
1 spatial extent of contamination, magnitude of benchmark exceedance, if the

2 contaminant was a COPC in another media, and mechanism of toxicity.

3 A more detailed description of the approach is provided in the following subsections.

4 B.4.1 Tier I Approach

5 The initial Tier I screening of potential COPCs was done using the frequency of detection and

6 comparisons with benchmarks and background. Details of the procedures are presented below.

7 B.4.1.1 Frequency of Detection Screening

8 Contaminants with less than a 5% frequency of detection were eliminated as COPCs. However,

9 all contaminants that were eliminated based on less than 5% frequency of detection were

10 subjectively evaluated to ensure that highly ecotoxic or bioaccumulative contaminants (e.g.,

11 dioxins/furans, and mercury) were retained for further consideration as COPCs. Although the

12 frequency of detection approach is not strictly risk-based, the conservative nature of the COPC

13 selection process ensures that contaminants are not screened out without a reasonable certainty

14 that they do not pose unacceptable risks.

15 B.4.1.2 Benchmark Comparisons

16 Comparisons of media concentrations with toxicity benchmarks provide a screening approach to

17 estimate the potential toxicity associated with contaminants at a site. Conservative benchmarks

18 are used to reduce the possibility that contaminants with potentially problematic concentrations

19 are not eliminated from consideration in the risk assessment based upon the benchmark

20 screening. Exhibit B-l provides an overview of the benchmark comparison process used. A

21 detailed discussion of benchmarks used for each media considered in the pre-ERA (i.e.,

22 sediment, surface water, soil, and fish) is presented below, followed by a discussion of the

23 screening results. Uncertainties associated with the use of these benchmarks will be addressed,

24 where appropriate, in other sections of the ERA.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-8

-------
Medium
Concentrations

Benchmark8

Hazard Quotient

Sediment

^ Medium Concentration

~ 	=HQ >\

Benchmark

HQ > 1 Indicates
Potential Risk:
Contaminant retained
for further screening

In general, benchmarks presented in order of priority.

Housatonic River Project
Pittsfield, Massachusetts

EqP = equilibrium partitioning
ERL = effect range low
LEL = lowest effect level

NAWQC = national ambient water quality criteria
NEL = no-effect level

PEC = probable effect concentration
PRG = preliminary remediation goal
SEL = severe effect level
TEC = threshold effect concentration
TEL = threshold effect level

Exhibit B-l
Benchmark Comparison Approach

MKO 1\0:Y20123001.096\ERA_PB\App B\Pre-ERA_AppB_PB_ExhB-1

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

B.4.1.2.1 Sediment

A range of sediment benchmarks was used to screen contaminant concentrations, with both a low
and high effect level benchmark identified (where possible) for each contaminant being
considered. The MacDonald et al. (2000) consensus-based benchmark was used preferentially
for the low-end benchmark, followed by the lowest (most conservative) of the benchmarks
developed by the Ontario Ministry of Environment and Energy (OMEE; Persaud et al. 1996),
Ingersoll et al. (1996), Sediment Preliminary Remediation Goals (PRGs; Efroymson et al.
1997a), and Equilibrium Partitioning Values (EqP; Jones et al. 1997). The high-end benchmark
selected was the one that was derived from the same source as the selected low-end value. For
example, if the low-end value selected was the OMEE lowest effect level (LEL), the high-end
value selected was the OMEE severe effect level (SEL). If the PRG or EqP value was selected
as the low-end, no high-end value was available. Each of the benchmarks is described below.
Sediment benchmark values are listed in Table B-53.

MacDonald et al. (2000) Values—The MacDonald et al. (2000) values are consensus-based
sediment quality guidelines (SQGs) for freshwater ecosystems. The consensus-based values
were developed by compiling published SQGs derived by various investigators and comparing
them with several criteria. If the studies were deemed useful (i.e., methods of derivation readily
apparent, SQGs based on empirical data relating contaminant concentrations to harmful effects
on sediment-dwelling organisms, and the SQGs had been derived, not adopted from another
source), the SQGs were grouped to facilitate derivation of consensus-based SQGs.

The two categories into which the SQGs were grouped were threshold effect concentrations
(TECs) and probable effect concentrations (PECs). TECs were intended to identify
concentrations below which harmful effects on sediment-dwelling organisms were not expected,
while PECs were intended to identify concentrations above which harmful effects were expected
to occur frequently. TECs and PECs were determined by calculating the geometric mean of the
values within their respective categories. Predictability tests (MacDonald et al. 2000) indicate
that the consensus-based values provide a reliable basis for classifying sediments as not toxic or
toxic. TECs and PECs were considered for use in the pre-ERA.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-10

-------
1

Ontario Ministry of Environment and Energy (OMEE)—The OMEE Sediment Quality
Guidelines (Persaud et al. 1996) define three levels of chronic effects on benthic organisms. The no
effect level (NEL) is defined as the level at which no toxic effects have been observed on aquatic
organisms and food chain biomagnification is not expected. The lowest-effect level (LEL) indicates
a level of sediment contamination that can be tolerated by most benthic organisms. The severe-
effect level (SEL) indicates a level of contamination at which pronounced disturbance of sediment-
dwelling organisms will occur and the contaminant concentration will be detrimental to the majority
of benthic species. LELs (NEL in the case of heptachlor) and SELs were considered for use as the
low and high-end benchmarks, respectively (Persaud et al. 1996).

The NEL for non-polar organics is calculated using Provincial Water Quality Objectives and
Guidelines (PWQOs/Gs), which were designed to protect against biomagnification of contaminants
through the food chain as well as all water quality uses and organisms. The PWQO/G is multiplied
by an organic carbon-normalized sediment to water partition coefficient (Koc). The value is
converted to a bulk sediment basis by assuming a 1% TOC concentration. The OMEE selected a
1% TOC level for developing NELs since calculations using the screening level concentration
(SLC) approach have shown that this is the lowest effect level of organic carbon in sediment.

The LEL and SEL are based on SLCs. The SLC uses field data on the co-occurrence of benthic
infaunal species and measured concentrations of contaminants. The SLC is an estimate of the
highest concentration of a contaminant that can be tolerated by a specific proportion of benthic
species. Calculation of the SLC is a two-step process. First, a species SLC (SSLC) is calculated as
the 90th percentile of the frequency distribution of the contaminant concentrations over all of the
sites (n > 10) where the species is present. In the second step, the contaminant-specific SLC is then
calculated as the 5th percentile of the SSLC distributions for a number of species, which represents
the concentrations that 95% of the species can tolerate. The LEL for metals, nutrients, and polar
organics is the 5th percentile of the SLC. For non-polar organics, the SLC is calculated as above,
but the organic carbon-normalized sediment are used instead of the bulk sediment concentrations.
The SLC is then converted back to bulk sediment concentration, assuming a TOC of 1%. The SEL
for metals, nutrients, and polar organics is calculated as the 95th percentile of all SSLCs (i.e., the
level below which 95% of all SSLCs fall). For non-polar organics, the SLC is calculated using the

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-ll

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

organic carbon normalized sediment concentrations. The SEL is converted back to bulk sediment
concentrations based on site-specific TOC concentrations (1% minimum and 10% maximum).

Ingersoll et al. (1996) Values—Ingersoll et al. developed sediment effect concentrations (SECs)
to classify toxicity data for sediment samples tested with Hyalella azteca and Chironomus riparius.
Three types of sediment effect concentrations (SECs) were calculated by Ingersoll et al. (1996): (1)
Effects Range Low (ER-L) and Effects Range-Median (ER-M); (2) Threshold Effect Level (TEL)
and Probable Effect Level (PEL); and (3) No Effect Concentration (NEC). ER-Ls and TELs will be
considered for use in this pre-ERA.

Ingersoll et al. (1996) calculated ER-Ls using the procedures of Long et al. (1995); however, the
data used in the calculations came from individual toxicity tests (e.g., the H. azteca 28-day test) to
have a consistent endpoint for determining a toxic response. ER-Ls refer to the 15th percentile of
effects concentrations. TELs were also calculated from individual toxicity tests to have a consistent
endpoint for toxic responses. The TEL calculations used the procedures in MacDonald (1994) and
MacDonald et al. (1996). Essentially, the concentrations observed or predicted by different
methods to be associated with effects were sorted and the 15th percentile of effects data (ER-L) and
the 50th percentile of the no-effects data (NERM) were calculated. The TEL equals the geometric
mean of the ER-L and NERM.

SECs developed by Ingersoll et al. (1996) were evaluated for their potential to correctly classify
toxic samples as toxic, correctly classify non-toxic samples as non-toxic, incorrectly classify
non-toxic samples as toxic (Type I error), and incorrectly classify toxic samples as not toxic
(Type II error). The Ingersoll et al. (1996) benchmark that was chosen for consideration in this
evaluation was the lowest of the calculated benchmarks that was considered reliable. A
benchmark was considered unreliable if there were less than five samples designated as toxic for
the contaminant or the number of toxic samples with concentrations below the SEC was greater
than the number of toxic samples with concentrations above the SEC.

Sediment Preliminary Remediation Goals (PRGs)—Sediment PRGs (Efroymson et al. 1997a)
are the lowest values of the following sediment toxicity benchmarks:

¦ Sediment quality criteria proposed by EPA.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-12

-------
1

Sediment criteria based on the chronic National Ambient Water Quality Criteria
(NAWQC) using EqP calculations.

¦ Criteria calculated from the lowest chronic value for fish, daphnids, or other
invertebrates in surface waters using EqP calculations.

¦ The NO AA ER-M.

¦ The Florida Department of Environmental Protection probable effect level (PEL).

¦ The Probable Effects Concentration (PEC) selected from the EPA Assessment and
Remediation of Contaminated Sediments (ARCS) Program Report.

If none of the above criteria were available for a contaminant, the PRG is the lower of the
following:

¦ The sediment benchmarks calculated from the secondary chronic value for aquatic
toxicity.

¦ The OMEE Severe Effect Level.

¦ The high No Effect Concentration (NEC) selected from the ARCS report.

A detailed discussion of the specific methods used to develop PRGs is presented in Efroymson et
al (1997a).

Equilibrium Partitioning Values—The equilibrium partitioning (EqP) values used as
benchmarks for this evaluation were taken from Jones et al. (1997). These EqPs were based on
the secondary chronic water quality value and were normalized to organic carbon. The use of
the EqP approach involves four major assumptions:

¦ The partitioning of the organic contaminant between OC and interstitial water is
stable at equilibrium.

¦ The sensitivities of benthic species and species tested to derive the water quality
values are similar.

¦ The levels of protection afforded by the water quality values are appropriate for
benthic organisms.

¦ Exposures are similar regardless of feeding type or habitat.

MK01|O:\20123001.096\ERA_PB\APPB\APPB_PB_PREERA.DOC g 7/10/2003

-------
1 B.4.1.2.2 Surface Water

2 Site-specific surface water concentrations were screened against a single freshwater benchmark.

3 The benchmark selected was based on the following hierarchy:

9 Each of the benchmarks is described below. Surface water benchmark values are listed in Table

10 B-54.

11 National Ambient Water Quality Criteria (NAWQC)—Freshwater chronic values were used

12 as benchmarks to screen surface water concentrations. EPA's Guidelines for Deriving

13 Numerical National Water Quality Criteria for the Protection of Aquatic Organisms and their

14 Uses (Stephan et al. 1985) describes an objective, internally consistent and appropriate way for

15 deriving contaminant-specific, numeric water quality criteria for the protection of the presence

16 of, as well as the uses of, freshwater aquatic organisms. AWQC are derived to protect most of

17 the aquatic communities and their uses most of the time (EPA 1999). Chronic criteria or criteria

18 continuous concentration (CCC) are selected by choosing the most protective value after

19 reviewing and analyzing acute and chronic toxicity information for aquatic organisms, aquatic

20 plants, and tissue concentration studies that demonstrate a water-tissue concentration relationship

21 unacceptable for consumption by humans or wildlife. If data were insufficient to calculate an

22 AWQC, lowest observed adverse effect level (LOAEL)-based benchmarks as presented in Water

23 Criteria Quality Summary (EPA 1991) were used.

24 Tier II Chronic Values—Tier II chronic values (Suter and Tsao 1996) were developed using

25 methods described in EPA's Proposed Water Quality Guidance for the Great Lakes System.

26 Tier II values are specifically developed for contaminants where fewer data than are required for

27 NAWQC are available (Suter and Tsao 1996). Tier II values may be based on data from just one

28 genus. Tier II values are more conservative than NAWQCs because toxicity values are modified

29 based on the number of studies used to determine the value. As more data become available, the

30 value becomes less conservative (EPA 1995). Tier II values are concentrations that would be

31 expected to be higher than NAWQC in no more than 20% of cases (Suter and Tsao 1996). In

¦ NAWQC - chronic.

¦ Tier II Values (Suter and Tsao 1996).

¦ Canadian Water Quality Criteria.

¦ Literature-based values.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-14

7/10/2003

-------
1

general, aquatic organisms are not affected unacceptably if the 4-day average concentration of a
contaminant does not exceed the chronic Tier II value more than once every three years on
average (EPA 1995).

Canadian Water Quality Guidelines—Canadian Water Quality Guidelines (WQGs) are based
on the LOAEL from a chronic exposure study on the most sensitive native Canadian species or
appropriate surrogate. Study values are multiplied by a safety factor of 0.1 to determine the final
guideline concentration. The goals of the WQGs are to protect all life stages during an indefinite
exposure to water.

Literature-Based Values—Contaminants for which none of the aforementioned benchmarks
were available were assessed using the lowest chronic or acute values observed in the literature.
Lowest chronic or acute values were developed in stages:

¦ First, the AQUTRE database was searched.

¦ Where possible, primary references were obtained for freshwater acute and chronic
studies listed in the AQUTRE database. Preference was given to studies with longer
duration, species found within the Housatonic drainage system, and flow-through or
static renewal studies. Studies that did not use a control or reference standard
protocols were rejected. Biological endpoints that could be related to local
population level effects were also given preference. If no site-specific organism data
were available, organisms taxonomically similar to those found within the study areas
were used.

¦ Last, the lowest identified concentration that met the above criteria was selected as a
benchmark.

B.4.1.2.3 Soil

The lowest (most conservative) of the benchmarks described below was used for screening
COPCs in soil. Soil benchmarks are presented in Table B-55.

Soil Preliminary Remediation Goals (PRGs) For Wildlife—Soil PRGs for wildlife were
developed by estimating exposure using different soil concentrations, recommended soil-to-biota
contaminant uptake models, and soil ingestion rates. The soil concentration used to estimate
wildlife exposure was adjusted to produce an exposure estimate equal to the wildlife endpoint-
specific and contaminant-specific LOAEL (obtained from Sample et al. 1996, as noted by

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-15

-------
1

Efroymson et al. 1997a). The specific soil PRGs for wildlife were selected based on the most
sensitive wildlife species evaluated (i.e., typically the short-tailed shrew or American
woodcock). All the soil PRGs for wildlife selected as a soil benchmark for this evaluation were
based on exposure to species observed within the PSA.

Screening Benchmarks for Plants, Earthworms, and Soil Microorganisms and Microbial
Processes—A screening benchmark concentration for the phytotoxicity of contaminants in soil
(Efroymson et al. 1997b) is the 10th percentile of the distribution of various toxic effects
thresholds for different species of terrestrial plants. Like the phytotoxicity value, the earthworm
and microorganisms/microbial processes value is the 10th percentile of the distribution of various
toxic effects thresholds for those receptors (Efroymson et al. 1997c).

Canadian Soil Quality Guidelines—The Canadian Soil Quality Guideline used for benchmark
comparison was the lowest ecologically based value from the four land use categories
(agricultural, residential/parkland, commercial, and industrial). The guideline derivation process
evaluates the potential for adverse effects to occur from both direct and indirect (e.g., food chain
transfer) exposure to contaminants. If an ecologically based value was not available, the Ontario
soil remediation criteria were used.

B.4.1.2.4 Fish Tissue

Benchmark fish tissue concentrations were developed to compare observed fish tissue
concentrations with the concentrations that may be of concern to avian and mammalian
piscivorous receptors. Fish tissue benchmark concentrations were back-calculated using an
exposure model for kingfisher and mink, using class-specific (i.e., mammalian or avian) toxicity
reference values (TRVs). The general equation and assumptions used were as follows.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-16

-------
1

BW * TRV

Fish Benchmark =

Where:

Mink

Kingfisher

Fish

Benchmark

= mg contaminant/kg fish

Calculated (see Table B-56)

= receptor body weight (kg)

1.0 kg
(EPA 1993)

0.15 kg
(EPA 1993)

TRV

= toxicity reference value (mg
contaminant/kg BW-day)

Contaminant-specific

= ingestion rate (kg food
item/day)

0.137 kg/day
(Sample et al. 1996)

0.075 kg/day
(Sample et al. 1996)

3 These species were selected as receptors for the pre-ERA because they either have been observed

4 or are potential inhabitants of the PSA based on known distributions and habitat preferences. In

5 addition, these species are potentially exposed to higher PCB concentrations due to their position

6 in the food chain and dietary habits. The lower of the two calculated fish concentrations was

7 used for the benchmark (see Table B-56). Conservative assumptions were used, such as

8 assuming 100% of the diet is from the study area, maximum ingestion rates and assimilation

9 efficiencies, and minimum adult body weights.

10 The TRVs were derived from doses identified in Sample et al. (1996). If a value was not

11 available in Sample et al. (1996), a literature-based TRV was developed as outlined below.

12 Toxicity Reference Values—TRVs are dose-based levels of contaminants that are not expected

13 to elicit adverse effects. TRVs were derived for specifically selected ecological receptors in the

14 PSA.

15 Study and Dose Selection for Toxicity Reference Values—Doses used for avian and

16 mammalian TRVs were obtained from peer-reviewed primary research articles. The process used

17 to identify primary research articles for use as TRVs includes a review of literature searches,

18 database searches, and secondary sources as listed in Table B-57. If primary research articles

19 could not be obtained, data from secondary sources were used. To qualify for consideration,

20 studies that could be used for TRVs had to meet the following criteria:

MK01|O:\20123001.096\ERA_PB\APPB\APPB_PB_PREERA.DOC g j j 7/10/2003

-------
1

¦ Test species similar to the receptor.

¦ In vivo study.

¦ Oral administration via food, drinking water, or gavage (feeding study preferred).

¦ No observed adverse effect level (NOAEL) or lowest observed adverse effect level
(LOAEL) identifiable.

¦ Effects of potential "ecological significance" evaluated (e.g., lethality and
reproductive effects).

Articles meeting the criteria were summarized, with information noted on study parameters,
effects evaluated, and results. Primary considerations in the selection process included study
species, study duration, effect level, and toxicological endpoint.

Studies using site-specific receptor wildlife species were preferred. However, toxicological data
for the receptor wildlife species were often unavailable; therefore, studies were chosen that, to
the extent possible, used species related to the receptor species, with similar diets and digestive
systems.

Suitable chronic exposure studies were given preference over acute studies. Chronic exposure
represents the extended exposure of an organism to a contaminant, generally greater than one-
tenth of the typical life span of the species (Suter 1993). Acute exposure represents either an
instantaneous single-dose exposure or a continuous exposure of minutes to a few days duration.

For those studies for which both a NOAEL and LOAEL were available, both the NOAEL and
the LOAEL were presented (see Tables B-58 and B-59). By definition, a NOAEL is that dose of
a contaminant at which there is no statistically or biologically significant increase in the
frequency or severity of adverse effects between the exposed population and the control. By
comparison, a LOAEL is the lowest dose of a contaminant in a study or group of studies that
produces biologically significant increases in the frequency or severity of adverse effects
between the exposed population and its appropriate control (Dourson and Stara 1983). Endpoints
that could directly affect the receptor species at the population level were given preference (e.g.,
reproductive effects and mortality of adults or offspring). The next preference was given to
serious histopathological effects (e.g., necrosis or damage to liver, kidney, or brain) that alter
primary body functions. In the absence of preferred data, consideration was given to effects such

MK01|O:\20123001.096\ERA_PB\APPB\APPB_PB_PREERA.DOC B 18 7/10/2003

-------
1	as alterations in biochemical functions of an organ or alterations in normal behavior that could be

2	correlated with decreased survivability. Other effects such as altered body weight, decreased

3	liver size, and changes in blood chemistry are not readily associated with decreased survivability

4	or longevity and were used only in the absence of the preferred toxicity data.

5	Best professional judgment was used to select the most appropriate studies, doses, and endpoints

6	for use in TRV development for the Housatonic River study area. Selected doses were modified

7	using the following safety factors:

8	¦ Acute to chronic duration = 0.01

9	¦ Sub chronic to chronic duration = 0.1

10	¦ LOAEL to NOAEL = 0.1

11

12	Doses and calculated TRVs are presented in Tables B-58 and B-59 for mammalian and avian

13	receptors, respectively.

14	B.4.1.3 Benchmark Comparison Results

15	Each detected concentration was compared with the appropriate medium-specific benchmark.

16	Results are presented in tables indicating the number of detected concentrations exceeding the

17	benchmark versus the number of samples available.

18	The magnitude of the screening hazard quotients (i.e., site-specific concentration/^- benchmark)

19	was grouped into ranges (e.g., l q	7/10/2003

-------
1 whether concentrations exceeding benchmarks are due to anthropogenic sources. For

2 contaminants that exceeded benchmarks, or for which no benchmark was available, site-related

3 concentrations were statistically compared with background concentrations. The methodology

4 used to compare site-related contaminant concentrations to medium-specific background

5 concentrations developed for the PSA is discussed below.

6 Based on the quantity of data available and the number of detected samples within the data sets,

7 various background-screening methods were employed based on following the criteria:

8 Category 1

9 "If more than 25% of the data in both the background data set and the site data set

10 were detected values, and there were at least 3 detected values in each data set, then

11 the following comparison was performed:

12 "A one-way analysis of variance was run on log-transformed concentrations among the

13 data sets (detection limits were substituted for non-detects), and the residuals were

14 tested for approximate normality using Shapiro-Wilks test with alpha=0.05.

15 "If the log-transformed data departed substantially from normality, the ANOVA was

16 re-run on the rankit-transformed data. The rankit transformation is a non-parametric

17 method using normal deviates for ranks of the original data (Conover 1980).

18 "If multiple site data were in this category, Dunnett's multiple comparisons were used

19 on the appropriately transformed data.

20 ¦ If only one data set was in this category, then a t-test was used on the appropriately

21 transformed data to compare the site data set to the reference data set.

22 Category 2

23 ¦ If more than 25% of the data and at least 3 values were detected in the background

24 data, but not in the site data, then a 95% one-sided prediction interval on the

25 background data set was used. The prediction interval assumed a log-normal

26 distribution, but a nonparametric interval was used if the background data did not

27 conform to approximate log-normality. If the data from the site were outside the

28 upper prediction interval around the background data, the contaminant was retained.

29 Category 3

30 ¦ If neither of the two above criteria was met, a statistical comparison was not

31 performed.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-20

-------
1 The statistics used for each comparison and results are presented in Tables B-120 through B-127.

2 Statistical runs are presented in Attachment B.2.

3 B.4.2 Tier I Results

4 As discussed above, the Tier I approach included three procedures for screening the list of

5 COPCs: frequency of detection, comparison to benchmarks, and comparison to background.

6 Results from the Tier I analysis are presented in Tables B-128 through B-131. Contaminants

7 retained advanced to Tier II. Non-metallic inorganics with benchmarks such as ammonia and

8 sulfide were carried through the Tier I evaluation to provide site-specific information, but were

9 not considered as potential COPCs. Non-metallic inorganics for which no benchmarks were

10 available (e.g., TKN, nitrate, and nitrite) were eliminated from further evaluation.

11 The Tier I analysis for sediments resulted in the elimination of six semivolatiles and three

12 inorganics. All PAHs and pesticides and a majority of the metals detected in sediments were

13 retained as COPCs (see Table B-128).

14 In addition to PCBs and dioxins/furans, one volatile, three semivolatiles and six metals were

15 retained as surface water COPCs (see Table B-129).

16 The Tier I analysis for surface soils resulted in the elimination of 20 semivolatiles (including

17 fourteen PAHs) and five metals. Several semivolatiles were retained as COPCs due to the lack

18 of an appropriate benchmark. In addition, five PAHs, four pesticides and 11 inorganics were

19 retained as COPCs along with PCBs and dioxins/furans (see Table B-130).

20 The fish tissue Tier I analysis resulted in the elimination of all five detected semivolatiles, 14

21 pesticides and all three detected metals, including mercury in large fish which was eliminated as

22 a result of comparison with background concentrations. In addition to PCBs and dioxin/furans,

23 nine pesticides were retained as fish tissue COPCs (see Table B-131).

24 B.4.3 Tier II Approach

25 To further refine the potential COPC list and ensure that only the contaminants possibly

26 contributing to risk are further assessed, contaminants retained in Tier I were subjected to the

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC B 21 7/10/2003

-------
1

Tier II approach. In the Tier II evaluation, contaminants were categorized into three classes,
unlikely, likely, and potential COPCs based on the following criteria:

Category

Criteria

Unlikely

Frequency of detection < 25%
or

Exceeds benchmarks in < 10% of samples

Likely

Frequency of detection > 50%
and

Exceeds benchmarks in > 50% of samples

Potential

Frequency of detection > 25% and exceeds benchmarks in < 50% of samples

Frequency of detection > 25% but < 50% and exceeds benchmarks in > 50% of samples

Contaminants were evaluated by reach and terrain as discussed above. If a contaminant was
classified as an "unlikely" COPC within a reach/terrain category, it was eliminated from further
consideration as a COPC within that reach/terrain. If a contaminant was classified as an
"unlikely" COPC in each of the reaches/terrain, it was eliminated as a COPC for the study,
unless SQLs exceeded appropriate benchmarks in at least 10% of the samples, in which case the
contaminant was retained for Tier III analysis. The results of this analysis are presented in
Tables B-132 through B-135. Contaminants identified as "potential" or "likely" COPCs were
evaluated in Tier III.

B.4.4 Tier II Results

Using the Tier II approach presented in Section B.2.5, a preliminary list of contaminants was
developed for inclusion in the Tier III assessment. The following discussion presents an
overview of the Tier II analysis findings; media- and contaminant-specific results of the Tier II
analysis are provided in Tables B-132 through B-135.

The Tier II analysis for sediments resulted in the elimination of five semivolatiles and two
pesticides. After the Tier II sediment evaluation, 22 semivolatiles, 4 pesticides, 16 inorganics,
PCBs, and dioxins/furans were retained as sediment COPCs (see Table B-132).

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC g 7/10/2003

-------
1 The surface water Tier II analysis resulted in the elimination of three semivolatiles and four

2 inorganics. The only surface water COPCs remaining after the Tier II analysis were vinyl

3 chloride, PCBs, dioxins/furans, copper (dissolved and total), lead, and zinc (see Table B-133).

4 The Tier II analysis for soils resulted in the elimination of six semivolatiles, three pesticides and

5 one inorganic. In addition to PCBs and dioxins/furans, six semivolatiles, one pesticide, and nine

6 inorganics were retained as surface soil COPCs (see Table B-134).

7 The fish tissue Tier II analysis resulted in the elimination of one pesticide, endrin. Eight

8 pesticides, PCBs, and dioxins/furans were retained as fish tissue COPCs (see Table B-135).

9 B.4.5 Tier III Approach

10 In Tier III, several factors were evaluated to determine whether "unlikely" contaminants with

11 SQLs greater than benchmarks, or "potential" and "likely" COPCs (as categorized during the

12 Tier II evaluation) would be retained as COPCs in the ERA. Factors considered in the

13 assessment of Tier III COPCs included: the spatial extent of contamination; the magnitude by

14 which the benchmark was exceeded; the likelihood of bioaccumulation/bioconcentration;

15 whether the contaminant was a COPC in sediment; the mechanism of toxicity; and if the

16 contaminant could be associated with an upstream source other than the GE facility (e.g., the

17 Pittsfield Municipal Landfill).

18 The frequency of detection and the number of detected concentrations and SQLs (for the samples

19 that were non-detect) exceeding benchmarks for the affected contaminants in sediment, surface

20 water, and soil are presented in Tables B-136 through B-138.

21 The justification for elimination or inclusion of COPCs in the ERA is presented in Tables B-139

22 through B-142.

23 B.4.6 Tier III Results

24 Using the Tier III approach discussed in Section B.2.7, a final list of COPCs was developed for

25 each reach/terrain within each medium (Tables B-139 through B-142).

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC g ^ 7/10/2003

-------
1

Contaminants that were not detected but had SQLs greater than benchmarks, are presented in
Table B-143. Contaminants that were detected, eliminated as COPCs during the Tier I or Tier II
evaluation, but that had a substantial number of SQLs greater than benchmarks are presented in
Table B-144. While no longer COPCs, these contaminants will be qualitatively evaluated in the
uncertainty analysis. The following discussion presents a summary (by contaminant group) of
the results of the Tier III evaluation. The final COPC lists are presented in Tables B-145 through
B-148.

B.4.6.1 Semivolatiles

Several semivolatiles were evaluated as part of the Tier III sediment assessment. The
semivolatiles other than PAHs that were evaluated included 2-methylphenol, phenol, bis(2-
ethylhexyl) phthalate, dibenzofuran, acetophenone, 1,4-dichlorobenzene, 2-methylnaphthalene
and 4-methylphenol. (Dibenzofuran was analyzed in the semivolatile suite, not with the
dioxins/furans; therefore, to maintain consistency with the data reporting, it will be discussed
with the semivolatiles.) Dibenzofuran was the only semivolatile retained as a COPC for
sediments. Phenol and 2-methylphenol were recommended for qualitative treatment in the
uncertainty analysis as a result of SQLs frequently exceeding the corresponding sediment
benchmark. The remaining semivolatiles were eliminated as COPCs primarily due to low
toxicity or concentrations similar to background concentrations.

Dibenzofuran was the only semivolatile retained as a COPC for surface soils. Although no soil
benchmark is available for dibenzofuran, it was detected frequently throughout Subreaches 5A
and 5B and was frequently detected in samples collected at the GE facility and in the V2 mile
immediately downstream. Butylbenzyl phthalate and 4-methylphenol were eliminated as soil
COPCs due to low detection frequencies and because both contaminants were not retained as
sediment COPCs.

There were no semivolatiles other than PAHs evaluated in the Tier III assessment for surface
water or fish tissue.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-24

-------
1 B.4.6.1.1 PAHs

2 Sixteen PAHs were evaluated as part of the Tier III sediment assessment. All 16 PAHs were

3 retained as COPCs; however, a majority of these PAHs were detected at concentrations similar

4 to or below those observed in background samples. PAHs were retained as COPCs since recent

5 EPA guidance recommends the quantitative assessment of anthropogenic contaminants that

6 exceed benchmarks, and the discussion of COPCs within background concentrations and their

7 contribution to site risks (EPA 2002).

8 Three PAHs were evaluated in surface soils as part of the Tier III assessment:

9 benzo(k)fluoranthene, benzo(a)pyrene, and pyrene. Benzo(a)pyrene and pyrene were retained as

10 COPCs for surface soils primarily because they had the potential for moderate toxicity.

11 Benzo(k)fluoranthene was eliminated as a surface soil COPC primarily because of its relatively

12 low toxicity potential.

13 Four PAHs were evaluated as COPCs for surface water in the Tier III assessment:

14 benzo(a)anthracene, benzo(a)pyrene, fluoranthene, and pyrene. All four PAHs had a relatively

15 low frequency of detection, but the SQLs for non-detects exceeded corresponding surface water

16 benchmarks. Therefore, these PAHs will be evaluated qualitatively in the ERA.

17 Fish tissue samples were not analyzed for PAHs because of the low concentrations observed in

18 sediment and the high rate of metabolism in fish, making it very unlikely that they would be

19 detected if the analysis were performed.

20 B.4.6.2 Pesticides

21 Six pesticides were evaluated as part of the Tier III sediment assessment: alpha-BHC, beta-BHC,

22 4,4'-DDT; 4,4'-DDD; 4,4'-DDE, and endrin aldehyde. 4,4'-DDT and 4,4'-DDE were retained as

23 sediment COPCs primarily because of their high toxicity and bioaccumulation potential. Alpha-

24 BHC, beta-BHC, and 4,4'-DDD had very low detection frequency, but were recommended for

25 qualitative evaluation in sediments as a result of SQLs frequently exceeding the corresponding

26 benchmark. Endrin aldehyde was eliminated as a COPC primarily because of a low detection

27 frequency and no indication that it could be site related.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-25

-------
1 2,4,5-T was the only pesticide evaluated in surface soils as part of the Tier III assessment. It was

2 eliminated as a soil COPC primarily because of a low detection frequency (i.e., only detected in

3 one sample) and because it was not a sediment COPC.

4 There were no pesticides evaluated in the Tier III assessment for surface water.

5 Eight pesticides were evaluated as part of the Tier III fish tissue assessment: dieldrin, 4,4'-DDE;

6 0,p'-DDT; 4,4'-DDT, heptachlor epoxide, cis-nonachlor, trans-nonachlor, and oxychlordane.

7 Dieldrin was the only pesticide eliminated as a COPC, primarily because it was not detected in

8 any other medium evaluated. The other seven pesticides were tentatively retained as fish tissue

9 COPCs; however, a limited re-analysis of several fish tissue samples indicated that the original

10 analysis results using the GC/ECD method may have over-estimated pesticide concentrations in

11 fish tissue. Attachment B.2 provides an overview of the pesticide re-analysis results and the

12 implications these findings may have on the use of these contaminants as COPCs.

13 B. 4.6.3 Inorganics

14 Fifteen inorganics were evaluated as part of the Tier III sediment assessment. Eleven inorganics

15 were retained as COPCs. In general, inorganics were retained for the following reasons: high

16 detection frequency, detected concentrations greater than background concentrations, and

17 detected concentrations frequently exceeding benchmarks. Four inorganics were eliminated as

18 sediment COPCs: arsenic, nickel, thallium, and zinc. These four inorganics were eliminated

19 primarily because of low toxicity potential and detected concentrations within background

20 concentrations.

21 Nine inorganics were evaluated in surface soils as part of the Tier III assessment. Four

22 inorganics were retained as COPCs: chromium, lead, mercury, and selenium. All four inorganics

23 were retained primarily as a result of the relatively high number of samples that exceeded

24 benchmark values. The five inorganics eliminated as surface soil COPCs were copper, nickel,

25 silver, vanadium, and zinc. The primary reasons for eliminating these inorganics was either

26 because of a low toxicity potential and/or because the inorganic was not a sediment COPC.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-26

-------
1

Four metals were evaluated in surface water as part of the Tier III assessment: copper (dissolved
and total), lead (total), silver (total), and zinc (total). With the exception of silver, which was
recommended for qualitative evaluation because of a high number of SQLs exceeding the
corresponding benchmark, the remaining inorganics were eliminated as surface water COPCs
primarily because of low toxicity potential.

There were no metals evaluated in the Tier III assessment for fish tissue.

B.5 APPROACH BELOW WOODS POND

The approach used to determine the downstream limit of the area of concern for future ecological
risk analysis compared the PCB concentrations observed in sediment at locations downstream of
Woods Pond with benchmarks. PCB concentrations in sediments were selected as an indicator
of the spatial extent of potential ecological risk because sediment serves as a reservoir of PCBs
released from the GE facility and because much of the movement of PCBs in the system is
associated with movement of sediment. The benchmark used for this analysis was a Threshold
Effect Concentration (TEC) of 0.0598 mg PCB/kg sediment (MacDonald et al. 2000). The
benchmark was compared with either the detected concentration or the SQL for surface sediment
data contained in the project as of July 2001. This benchmark is lower than the actual SQLs for
sediment achieved in most analytical programs. Figures depicting the sample locations, total
PCB concentrations, and the location-specific hazard quotients are presented in Figures B-54
through B-74.

The determination of the downstream extent of PCB contamination that warrants future
ecological risk evaluation was based on professional judgment. The objective of this evaluation
was to identify the general area where PCB concentrations in sediments downstream of Woods
Pond were consistently below the conservative sediment benchmark (i.e., an area not expected to
pose ecological risk due to PCBs). This evaluation would establish the downstream extent of the
area of ecological concern.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-27

-------
1	B.6 RESULTS FROM BELOW WOODS POND

2	Between the Massachusetts/Connecticut border and the tidal influenced area of the Housatonic

3	River, much of the data were collected from impoundments. Hazard quotients that were

4	developed based on detected PCB concentrations or SQLs and based on the MacDonald TEC

5	benchmark (0.0598 mg PCB/kg) fall within the 10 to 100 range as far down the river as Lake

6	Housatonic (see Figures B-54 through B-74).

7	Given the magnitude by which the benchmark is exceeded, and the consistency and frequency of

8	exceedances, the potential for ecological risks resulting from elevated PCB concentrations in

9	sediments exists as far downstream as Lake Housatonic (see Figure B-74). Therefore, in the

10	ERA, baseline risk from PCBs will be evaluated from Woods Pond Dam to Derby-Shelton Dam.

11	B.7 REFERENCES

12	Abiola, F.A. 1992. Ecotoxicity of organochloride insecticides: Effects of endosulfan on birds

13	reproduction and evaluation of its induction effects in partridge, Perdix Perdix L. Rev. Vet.

14	Med. 143:443-450. As cited in Sample et al., 1996.

15	Ambrose, A.M., P.S. Larson, J.F. Borzelleca, and G.R. Hennigar, Jr. 1976. Long-term toxicologic

16	assessment of nickel in rats and dogs. J. FoodSci. Tech. 13:181-187. As cited in Sample et al.,

17	1996.

18	Anderson, D.W., R.W. Risebrough, L.A. Woods, Jr., L.R. DeWeese, and W.G. Edgecomb. 1975.

19	Brown pelicans: Improved reproduction off the southern California coast. Science. 190:806-

20	808. As cited in Sample et al., 1996.

21	Arnold, D.L., C.A. Moodie, S.M. Charbonneau, H.C. Grice, P.F. McGuire, F.R. Bryce, B.T Collins,

22	Z.Z. Zawidzka, D.R. Krewski, E.A. Nera, and I.C. Munro. 1985. Long-term toxicity of

23	hexachlorobenzene in the rat and the effect of dietary vitamin A. Fd. Chem. Toxic. 23(9):779-

24	793.

25	Aulerich, R. J., and R.K. Ringer. 1977. Current status of PCB toxicity to mink, and effect on their

26	reproduction. Arch. Environ. Contam. Toxicol. 6:279-292. As cited in Sample et al., 1996.

27	Azar, A., H.J. Trochimowicz, and M.E. Maxwell. 1973. Review of Lead Studies in Animals

28	Carried out at Haskell Laboratory: Two-year Feeding Study and Response to Hemorrhage

29	Study. In: Environmental Health Aspects of Lead: Proceedings, International Symposium, D.

30	Barth et al., eds. Commission of European Communities, pp. 199-210. As cited in Sample et

31	al., 1996.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-28

-------
1	Breslin, W.J., A.B. Liberacki, D.A. Dittenber, and J.F. Quast. 1996. Evaluation of the

2	developmental and reproductive toxicity of chlorpyrifos in the rat. Fundam. Appl. Toxicology.

3	29:119-130.

4	Bush, P.B., J.T. Kiker, R.K. Page, N.H. Booth, and O.J. Fletcher. 1977. Effects of graded levels of

5	toxaphene on poultry residue accumulation, egg production, shell quality and hatchability in

6	white leghorns. J. Agric. FoodChem. 25:928-932. As cited in Eisler and Jacknow, 1985.

7	Cain, B.W., and E.A. Pafford. 1981. Effects of dietary nickel on survival and growth of mallard

8	ducklings. Arch. Environ. Contam. Toxicol. 10:737-745. As cited in Sample et al., 1996.

9	Chakravarty, S., and P. Lahiri. 1986. Effect of lindane on eggshell characteristics and calcium level

10	in the domestic duck. Toxicology. 42:245-258. As cited in Sample et al., 1996.

11	Chu, I., D.C. Villeneuve, V.E. Valli, and V.E. Secours. 1984. Toxicity of 1,2,3,4-, 1,2,4,5-

12	tetrachlorobenzene in the rat: Results of a 90-day feeding study. Drug and Chemical

13	Toxicology. 7(2): 113-127.

14	Clark, O.R., Jr. 1975. Effect of stress on dieldrin toxicity of male redwinged blackbirds

15	(Agelaius Phoeniceus). Bull. Environ. Contam. Toxicol. 14(2):250-256.

16	Conover, W.J. 1980. PracticalNonparametric Statistics. John Wiley & Sons. New York.

17	Dahlgren, R.B., R.L. Linder, and C.W. Carlson. 1972. Polychlorinated biphenyls: Their effects

18	on penned pheasants. Environ. Health Perspect. 1:89-101. As cited in Sample et al., 1996.

19	Dikshith, T.S.S., R.B. Raizada, M.K. Srivastava, and B.S. Kaphalia. 1984. Response of rats to

20	repeated oral administration of endosulfan. Ind. Health. 22:295-304. As cited in Sample et

21	al., 1996.

22	Dourson, M.L. and J.F. Stara. 1983. Regulatory history and experimental support of uncertainty

23	(safety) factors. Reg. Tox. andPharm. 3:224-238.

24	Efroymson, R.A., G.W. Suter II, B.E. Sample, andD.S. Jones. 1997a. Preliminary Remediation

25	Goals for Ecological Endpoints. Oak Ridge National Laboratory for U.S. DOE. ES/ER/TM-

26	162/R2. August 1997.

27	Efroymson, R.A., M.E. Will, G.W. Suter II, and A.C. Wooten. 1997b. Toxicological

28	Benchmarks for Screening Contaminants of Potential Concern for Effects on Terrestrial

29	Plants: 1997 Revision. ES/ER/TM-85/R3. Environmental Restoration Program, Oak Ridge

30	National Laboratory, Oak Ridge, TN.

31	Efroymson, R.A., M.E. Will, and G.W. Suter II. 1997c. Toxicological Benchmarks for

32	Contaminants of Potential Concern for Effects on Soil and Litter Invertebrates and

33	Heterotrophic Process: 1997 Revision. ES/ER/TM-126/R2. Environmental Restoration

34	Program, Oak Ridge National Laboratory, Oak Ridge, TN.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-29

-------
1	Eisler, R. and J. Jacknow. 1985. Toxaphene Hazards to Fish, Wildlife, and Invertebrates: A

2	Synoptic Review. U.S. Fish Wildl. Serv. Biol. Rep. 85(1.4). 26 pp.

3	EPA (U.S. Environmental Protection Agency). 2002. Role of Background in the CERCLA

4	Cleanup Program. Office of Solid Waste and Emergency Response, Office of Emergency

5	and Remedial Response. April 26, 2002. OSWER 9285.6-07.

6	EPA (U.S. Environmental Protection Agency). 1999. National Recommended Water Quality

7	Criteria - Correction. Office of Water. EPA 822-Z-99-001. April 1999.

8	EPA (U.S. Environmental Protection Agency). 1995. Final water quality guidance for the Great

9	Lakes system; final rule. Federal Register 60(56): 15366-15425.

10	EPA (U.S. Environmental Protection Agency). 1993. Wildlife Exposure Factors Handbook.

11	Volume 1. Office of Research and Development. Washington DC. EPA/600/R-93/187a.

12	December 1993.

13	EPA (U.S. Environmental Protection Agency). 1991. Water Quality Criteria Summary. Office

14	of Science and Technology, Health and Ecological Criteria Division, Ecological Risk

15	Assessment Branch (WH-585), Human Risk Assessment Branch (WH-550 D), Washington,

16	D C. May 1, 1991.

17	EPA (U.S. Environmental Protection Agency). 1988. Recommendations for and Documentation

18	of Biological Values for use in Risk Assessment. Environmental Criteria and Assessment

19	Office, Cincinnati, OH. EPA/600/6-87/008. As cited in Sample et al., 1996.

20	Fitzhugh, O.G., A.A. Nelson, and J.P. Frawley. 1950. The chronic toxicities of technical

21	benzene hexachloride and its alpha, beta and gamma isomers. J. Pharmacol. Exp. Ther.

22	100:59-66.

23	Fleming, W.J., M.A. Ross McLane, E. Cromartie. 1982. Endrin decreases screech owl

24	productivity. J. Wildl. Manage. 46:462-468. As cited in Sample et al., 1996.

25	Good, E.E., and G.W Ware. 1969. Effects of insecticides on reproduction in the laboratory

26	mouse, IV. Endrin and dieldrin. Toxicol. Appl. Pharmacol. 14:201-203. As cited in Sample

27	etal., 1996.

28	Green, V.A. 1970. Effects of pesticides on rat and chick embryo. Pp. 183-209. In: Trace

29	Substance Environmental Health 3, Proceedings of the 3rd Annual Conference, University of

30	Missouri. Ed.: D. Hemphill.

31	Hamid, J., A. Sayeed, and H. McFarlane. 1974. The effect of l-(o-chloropheny)-l-(p-

32	chlorophenyl)-2,2-dichloroethane (o,p'-DDD) on the immune response in malnutrition. Br. J.

33	Exp. Path. 55:94-99.

34	Heinz, G.H. 1979. Methyl mercury: Reproductive and Behavioral Effects on Three Generations

35	of Mallard Ducks. J. Wildl. Mgmt. 43:394-401. As cited in Sample et al., 1996.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-30

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

Herin, R.A., J.E. Suggs, E.M. Lores, L.T. Heiderscheit, J.D. Farmer, and D. Prather. 1978.
Correlation of salt gland function with levels of chlorpyrifos in the feed of mallard ducklings.
Pesticide Biochemistry and Physiology. 9:157-164.

Hill, E.F. and M.B. Camardese. 1986. Lethal Dietary Toxicities of Environmental Contaminants
and Pesticides to Coturnix. Fish and Wildlife Technical Report 2. U.S. Department of the
Interior, Fish and Wildlife Service. 147 pp.

Hyde, K.M., J.B. Graves, A.B. Watts, and F.L. Bonner. 1973. Reproductive success of mallard
ducks fed mirex. J. Wildl. Manage. 37(4):479-484.

Ingersoll, C.G., P.S. Haverland, E.L. Brunson, T.J. Canfield, F.J. Dwyer, C.E. Henke, and N.E.
Kemble. 1996. Calculation and evaluation of sediment effect concentrations for the amphipod
Hyalella azteca and the midge Chironomus riparius. J. Great Lakes Res. 22(3):602-623.

Ito, N., H. Nagasaki, M. Arai, S. Sugihara, and S. Makiura. 1973. Histologic and ultrastructural
studies on the hepatocarcinogenicity of benzene hexachloride in mice. J. Natl. Cancer. Inst.
51:817-826.

Jones, D.S., G.W. Suter, and R.N. Hull. 1997. Toxicological Benchmarks for Screening
Contaminants of Potential Concern for Effects on Sediment-Associated Biota: 1997
Revision. Oak Ridge National Laboratory for U.S. DOE. ES/ER/TM-95/R4. November,
1997.

Kennedy, G.L., Jr., J.P. Frawley, and J.C. Calandra. 1973. Multigeneration reproductive effects of
three pesticides. Toxicol. Appl. Pharmacol. 25:589-596. As cited in Sample et al., 1996.

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. 19:81-97.

Longcore, J.R., F.B. Samson, and T.W. Whittendale, Jr. 1971. DDE thins eggshells and lowers
reproductive success of captive black ducks. Bull. Environ. Contam. Toxicol. 6(6):485-490.

MacDonald, D.D., C.G. Ingersoll, and T.A. Berger. 2000. Development and evaluation of
consensus-based sediment quality guidelines for freshwater ecosystems. Arch. Environ.
Contam. Toxicol. 39:20-31.

MacDonald, D.D. 1994. Approach to the Assessment of Sediment Quality of Florida Coastal
Waters. Volume 1 - Development and Evaluation of Sediment Quality Assessment
Guidelines. Report prepared for the Florida Department of Environmental Protection.
Tallahassee, FL. November 1994. As cited in Ingersoll et al., 1996.

MacDonald, D.D., R.S. Carr, F. D. Calder, E.R. Long, and C.G. Ingersoll. 1996. Development
and evaluation of sediment quality guidelines for Florida coastal waters. Ecotoxicology In
press. As cited in Ingersoll et al., 1996.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-31

-------
1	McDonald, M.M. 1991. NTP Report on the Toxicity Studies of 1,2,4,5-Tetrachlorobenzene

2	(CAS No. 95-94-3) in F344/N Rats and B6C3F1 Mice (Feed Studies). NTP Tox 7; NIH

3	Publication No. 91-3126. National Toxicology Program, U.S. Department of Health and

4	Human Services, Public Health Service, National Institutes of Health.

5	Mendenhall, V.M., E.E. Klaas, and M. A.R. McLane. 1983. Breeding success of barn owls (Tyto

6	Alba) fed low levels of DDE and Dieldrin. Arch. Environ. Contam. Toxicol. 12:235-240.

7	As cited in Sample et al., 1996.

8	NCI (National Cancer Institute). 1978. Bioassays of DDT, TDE, and p,p'-DDE for Possible

9	Carcinogenicity CAS No. 50-29-3, 72-54-8, 72-55-9. NCI Carcinogenesis Technical Rep.

10	Series NO. 131, NCI-CG-TR-131. DHEW. 120 pp.

11	NTP (National Toxicity Program). 1991. NTP Report on the Toxicity Studies of

12	Pentachlorobenzene in F344/NRats andB6C3F1 Mice (FeedStudies). NTP Tox 6, Pub. No.

13	(NIH) 91-3125. U.S. Department of Health and Human Services, Public Health Service,

14	National Institutes of Health. 48 pp.

15	Pattee, O.H. 1984. Eggshell thickness and reproduction in American kestrels exposed to chronic

16	dietary lead. Arch Environ. Contam. Toxicol. 13:29-34. As cited in Sample et al., 1996.

17	Persaud, D., R. Jaagumagi, and A. Hayton. 1993. Guidelines for the Protection and Management

18	of Aquatic Sediment Quality in Ontario. Ontario Ministry of Environment and Energy.

19	Queens Printer for Ontario.

20	Sample, B.E., D.M. Opresko, and G.W. Suter II. 1996. Toxicological Benchmarks for Wildlife:

21	1996 Revision. Oak Ridge National Laboratory for U.S. DOE. ES/ER/TM-86/R3. June

22	1996.

23	Stephan, C.E., D.I. Mount, D.J. Hansen, J.H. Gentile, G.A. Chapman, and W.A. Brungs. 1985.

24	Guidelines for Deriving Numerical National Criteria for the Protection of Aquatic

25	Organisms and Their Uses. PB85-227049. National Technical Information Service,

26	Springfield, VA.

27	Stickel, L.F., W.H. Stickel, R.A. Dyrland, and D.L. Hughes. 1983. Oxychlordane, HCS-3260,

28	and nonachlor in birds: Lethal residues and loss rates. J. Toxicol. Environ. Health. 12:611-

29	622. As cited in Sample et al., 1996.

30	Suter, G.W. II. 1993. Ecological Risk Assessment. Lewis Publishers. Boca Raton, FL. ISBN 0-

31	87371-875-5.

32	Suter, G.W. II, and J.B. Mabrey. 1994. Toxicological Benchmarks for Screening Potential

33	Contaminants of Concern for Effects on Aquatic Biota: 1994 Revision. ES/ER/TM-96/R1.

34	Environmental Restoration Program, Oak Ridge National Laboratory, Oak Ridge, TN.

35	Suter, G.W. II, and C.L. Tsao. 1996. Toxicological Benchmarks for Screening Potential

36	Contaminants of Concern for Effects on Aquatic Biota: 1996 Revision. Oak Ridge National

37	Laboratory for U.S. DOE. ES/ER/TM-96/R2. June 1996.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

B-32

-------
1	Treon, J.F., and F.P. Cleveland. 1955. Toxicity of certain chlorinated hydrocarbon insecticides

2	for laboratory animals, with special reference to Aldrin and Dieldrin. Ag. Food Chem.

3	3:402-408. As cited in Sample et al., 1996.

4	Van Velsen, F.L., L.H.J.C. Danse, F.X.R. Van Leeuwen, J.A.M.A. Dormans, and M.J. Van

5	Logten. 1986. The subchronic oral toxicity of the beta-isomer of hexachlorocyclohexane in

6	rats. Fundamental and Applied Toxicology. 6:697-712.

7	Vos, J.G., H.L. Van der Mas, A. Musch, and E. Ram. 1971. Toxicity of hexachlorobenzene in

8	Japanese quail with special reference to porphyria, liver damage, reproduction, and tissue

9	residues. Toxicology and Applied Pharmacology. 18:944-957.

10	Welsh, J.J., T.F.S. Collins, T.N. Black, S.L. Graham, and M.W. O'Donnell Jr. 1987.

11	Teratogenic potential of purified pentachlorophenol and pentachloroanisole in subchronically

12	exposed Sprague-Dawley rats. Fd. Chem. Toxicol. 25(2): 163-172.

13	WESTON (Roy F. Weston, Inc.). 2003, In Preparation. Modeling Framework Design: Modeling

14	Study of PCB Contamination in the Housatonic River. Prepared for U.S. Army Corps of

15	Engineers and U.S. Environmental Protection Agency.

16	WHO (World Health Organization). 1984. Chlordane. Environ. Health Criter. 34. 82 pp. As

17	cited in Sample et al., 1986.

18	Wobeser, G., N.O. Nielson, and B. Schiefer. 1976. Mercury and mink II. Experimental methyl

19	mercury intoxication. Can. J. Comp. Med. 34-45. As cited in Sample et al., 1996.

20

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

-------
ATTACHMENT B.1
TABLES AND FIGURES

(Tables and Figures are presented on CD)

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

-------
Table B-l

Sediment Background Samples
Housatonic River Site, OU 2, Pittsfield, MA







Analyses



Field Sample ID

Depth
Interval (ft)

Study Type

PCBs

Congeners

«

A9 Semi.

A9 Pest.

OP Pest.

.Q
-

3

Metals

Inorg.

VI

0

GS Class

&
o

Total PCBs
(mg/kg)

Washington Mountain Lake

H9-SE000385-0-0000

0-0.5

Leopard / Wood / Bull Frogs

X

X



X

X





X

X

X

X

X

5.56E-01 U

WML-1

H9-SE001259-0-0000

0-0.17

Leopard / Wood / Bull Frogs

X



X



X







X

X

X

X

R

H9-SE001269-0-0000

0-0.17

Leopard / Wood / Bull Frogs

X



X

X

X

X

X

X

X

X

X

X

6.90E-02 U

WML-2

H9-SE001260-0-0000

0-0.17

Leopard / Wood / Bull Frogs

X



X



X







X

X

X

X

R

H9-SE001268-0-0000

0-0.17

Leopard / Wood / Bull Frogs

X



X

X

X

X

X

X

X

X

X

X

1.30E-01 U

WML-3

H9-SE001265-0-0000

0-0.17

Leopard / Wood / Bull Frogs

X



X



X







X

X

X

X

R

H9-SE001270-0-0000

0-0.17

Leopard / Wood / Bull Frogs

X



X

X

X

X

X

X

X

X

X

X

1.10E-01 U

Muddy Pond

H9-SE001279-0-0000

0-0.17

Leopard / Wood / Bull Frogs

X



X

X

X

X

X

X

X

X

X

X

4.00E-02

H9-SEMP0001-0-0000

0-0.5

Discrete River Sampling

X

X

X

X

X





X

X

X

X

X

6.80E-01 U

Three Mile Pond

H9-SE3M0004-0-0000

0-0.5

Discrete River Sampling





X

X

X





X

X

X

X

X



H9-SE3M0008-0-0000

0-0.5

Leopard / Wood / Bull Frogs

X



X

X

X

X

X

X

X

X

X

X

6.90E-02 U

H9-SE3M0009-0-0000

0-0.5

Leopard / Wood / Bull Frogs

X



X

X

X

X

X

X

X

X

X

X

7.10E-02 U

Dalton Site

H0-SE000476-0-0000

0-0.5

Sediment Toxicity



X

X

X

X

X

X

X







X



H0-SE000717-0-0000

0-0.5

Mussel Exposure

X

X

X

X

X

X

X

X

X

X

X

X

5.01E-01 U

H0-SEEC0011-0-0000

0-0.5

Mussel Exposure

X

X

X

X

X





X

X

X

X

X

5.09E-01 U

Center Pond

H0-SEN60003-0-0000

0-0.5

Discrete River Sampling

X



X

X

X

X

X

X

X

X

X

X

6.31E-01 U

Upstream of Unkamet and Landfill

H0-SDN32963-0-0005

0.5-1

Systematic Sampling

X



X

X

X





X

X

X

X

X

2.70E-02 U

Tables B-l through B-4.xls Table B-l

-------
Table B-l

Sediment Background Samples
Housatonic River Site, OU 2, Pittsfield, MA

Field Sample ID

Depth
Interval (ft)

Study Type

Analyses

Total PCBs
(mg/kg)

PCBs

Congeners

«

A9 Semi.

A9 Pest.

OP Pest.

.Q
-

3

Metals

Inorg.

VI

O

GS Class

&
o

H0-SDN3 3001 -0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

2.10E-01 J

H0-SDN33042-0-0000

0-0.5

Systematic Sampling

X



X

X

X

X

X

X

X

X

X

X

2.00E-02 J

H0-SDN3 3121-0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

1.90E-02 U

H0-SDN3 3201 -0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

2.10E-02 U

H0-SDN3 3243 -0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

1.00E+00

H0-SDN3 3281 -0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

2.10E-02 U

H0-SDN44043-0-0000

0-0.5

Systematic Sampling

X



X

X

X

X

X

X

X

X

X

X

5.01E-01 UJ

Analyses:

PCBs = Total PCBs and Aroclors
Congeners = PCB congeners and homologs
D/F = Dioxins/Furans

A9 Semi. = Appendix IX semivolatiles or Semivolatiles
A9 Pest. = Appendix IX pesticides or Pesticides

Herb. = Herbicides

Inorganics = Inorganics (e.g., sulfides)
GS = Grain size

GS Class = Grain size class (e.g., sand)
Org. = Organic (e.g., TOC) or Organic Content

Tables B-l through B-4.xls Table B-l

-------
Table B-2

Surface Water Background Samples
Housatonic River Site, OU 2, Kttsfield, MA

Field Sample ID

Study Type

Analyses

Total PCBs
(tig/L)

PCBs

PCBs Filtered

Congeners

D/F

A9 Vol.

A9 Semi.

PAHs

A9 Pest.

OP Pest.

Herb.

Metals

Metals Filtered

Inorg.

Org.

WML-1

H9-SW000045-0-0A10

Leopard / Wood / Bull Frogs

X





X



X



X

X

X

X



X

X

1.30E-02 U

H9-SW000045-0-0Y01

Leopard / Wood / Bull Frogs

X





X



X



X

X

X

X



X



2.70E-02 U

WML-2

H9-SW000046-0-0A10 | Leopard / Wood / Bull Frogs |x| | |x| | | |x| | | | | | | 1.30E-02 U

WML-3

H9-SW000047-0-0A10 | Leopard / Wood / Bull Frogs |x| | |x| | | |x| | | | | | | 1.30E-02 U

Muddy Pond

H9-SW000049-0-0Y24 | Leopard / Wood / Bull Frogs |x| | |x| |x| |x|x|x|x| |x| | 1.30E-02 U

Dalton Site

H0-SW000020-0-9U24 | Sediment Toxicity | | |X|X| |X| | X | X | X | X | I I I

Upstream of Unkamet and Landfill

080498CT08

Monthly Surface Watei

X

X



X

X

X

X

X

X

X

X

X

X

X

1.30E-02 U

080498CT09

Monthly Surface Watei

X

X



X

X

X

X

X

X

X

X

X

X

X

1.30E-02 U

090198GB03

Monthly Surface Watei

X

X



X

X

X

X

X

X

X

X

X

X

X

2.50E-02 J

090198SB06

Monthly Surface Watei

X

X



X

X

X

X

X

X

X

X

X

X

X

1.20E-02 UJ

092598BS04

Monthly Surface Watei

X

X



X

X

X



X

X

X

X

X

X

X

1.20E-02 UJ

092598GC06

Monthly Surface Watei

X

X



X

X

X



X

X

X

X

X

X

X

1.30E-02 UJ

H0-SW000015-0-0Y01

Leopard / Wood / Bull Frogs

X





X



X



X

X

X

X



X



2.50E-02 U

H0-SW000015-0-8C27

Monthly Surface Water

X

X



X

X

X



X

X

X

X

X

X

X

1.30E-02 UJ

H0-SW000015-0-8D18

Monthly Surface Water

X

X



X



X



X

X

X

X

X

X

X

1.20E-02 UJ

H0-SW000015-0-8N24

Monthly Surface Water

X

X



X

X

X



X

X

X

X

X

X

X

1.20E-02 U

H0-SW000015-0-9A20

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.30E-02 UJ

H0-SW000015-0-9F24

Monthly Surface Water

X

X

X

X



X



X

X



X

X

X

X

1.30E-02 U

H0-SW000015-0-9G31

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.30E-02 U

H0-SW000015-0-9J19

Monthly Surface Water

X

X



X



X



X

X

X

X

X

X

X

1.30E-02 UJ

H0-SW000015-0-9L27

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.20E-02 UJ

H0-SW000015-0-9M23

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.20E-02 UJ

H0-SW000015-0-9S29

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.30E-02 U

H0-SW000015-0-9U24

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.30E-02 U

H0-SW000015-0-9Y27

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.20E-02 U

H0-SW000015-1-8C27

Monthly Surface Water

X

X



X

X

X



X

X

X

X

X

X

X

1.30E-02 UJ

Tables B-l through B-4.xls Table B-2

B-3

-------
Table B-2

Surface Water Background Samples
Housatonic River Site, OU 2, Pittsfield, MA

Field Sample ID

Study Type

Analyses

Total PCBs
(fig/L)

PCBs

PCBs Filtered

Congeners

D/F

A9 Vol.

A9 Semi.

PAHs

A9 Pest.

OP Pest.

Herb.

Metals

Metals Filtered

Inorg.

Org.

H0-SW000015-1-9J19

Monthly Surface Water

X

X



X



X



X

X

X

X

X

X

X

1.30E-02 UJ

H0-SW000016-0-8C27

Monthly Surface Water

X

X



X

X

X



X

X

X

X

X

X

X

1.30E-02 U

H0-SW000016-0-8D18

Monthly Surface Water

X

X



X



X



X

X

X

X

X

X

X

1.20E-02 UJ

H0-SW000016-0-8N24

Monthly Surface Water

X

X



X

X

X



X

X

X

X

X

X

X

1.20E-02 U

H0-SW000016-0-9A20

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.30E-02 UJ

H0-SW000016-0-9F24

Monthly Surface Water

X

X

X

X



X



X

X



X

X

X

X

1.20E-02 U

H0-SW000016-0-9G31

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.20E-02 U

H0-SW000016-0-9J19

Monthly Surface Water

X

X



X



X



X

X

X

X

X

X

X

1.30E-02 UJ

H0-SW000016-0-9L27

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.40E-02 UJ

H0-SW000016-0-9M23

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.20E-02 U

H0-SW000016-0-9S29

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.30E-02 U

H0-SW000016-0-9U24

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

2.00E-01 J

H0-SW000016-0-9Y27

Monthly Surface Water

X

X

X

X



X



X

X

X

X

X

X

X

1.30E-02 U

Analyses:

PCBs = Total PCBs and Aroclors
Filtered PCBs = Dissolved total PCBs and Aroclors
Congeners = PCB congeners and homologs
D/F = Dioxins/Furans

A9 Vol. = Appendix IX volatiles or Volatiles

A9 Semi. = Appendix IX semivolatiles or Semivolatiles
A9 Pest. = Appendix IX pesticides or Pesticides
Herb. = Herbicides

Inorganics = Inorganics (e.g., sulfides)

Org. = Organic (e.g., TOC) or Organic Content

Tables B-l through B-4.xls Table B-2

B-4

-------
Table B-3

Soil Background Samples
Housatonic River Site, OU 2, Pittsfield, MA

Field Sample ID

Depth
Interval (ft)

Study Type

Analyses

Total PCBs
(nig/kg)

PCBs

Congeners

D/F

A9 Semi.

A9 Pest.

OP Pest.

Herb.

Metals

Inorg.

CO

0

GS Class

Org.

H3-F0331003-0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

1.10E-01 J

H3-F0641002-0-0000

0-0.5

Systematic Sampling

X



X

X

X

X

X

X

X

X

X

X

5.80E-01 J

H3-F0849005-0-0010

1 - 1.5

Systematic Sampling

X



X

X

X

X

X

X

X

X

X

X

2.00E-02 U

H3-F1264006-0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

1.20E-01 J

H3-F1264006-1-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

2.40E-01 J

F13-F1366001-0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

2.70E-02

F13-F1366005-0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

2.60E-01

F13-F1367002-0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

4.20E-02 J

F13-F1367005-0-0000

0-0.5

Systematic Sampling

X



X

X

X





X

X

X

X

X

5.40E-01 J

H3-FL000577-0-0005

0.5 - 1

Electric Company

X



X

X

X





X

X

X

X

X

3.18E-02 U

H3-FL000697-0-0005

0.5 - 1

Canoe Meadows

X

X

X

X

X





X

X

X

X

X

2.30E-02 U

H3-FL000724-0-0005

0.5 - 1

Sportsman Club

X



X

X

X

X

X

X

X

X

X

X

2.10E-02 U

H3-FL000726-0-0000

0-0.5

Sportsman Club

X

X

X

X

X





X

X

X

X

X

4.14E-01 J

H3-FL000733-0-0000

0-0.5

Noble Farm

X

X

X

X

X





X

X

X

X

X

3.52E-01 J

H3-FL000738-0-0000

0-0.5

Noble Farm

X

X

X

X

X

X

X

X

X

X

X

X

9.30E-02

H3-FL000747-0-0000

0-0.5

Residential

X

X

X

X

X





X

X

X

X

X

5.62E-01 J

H3-FL000748-0-0000

0-0.5

Residential

X

X

X

X

X





X

X

X

X

X

8.30E-02

H3-FL000753-0-0000

0-0.5

Residential

X

X

X

X

X

X

X

X

X

X

X

X

1.40E-01

H3-FL000763-0-0000

0-0.5

Residential

X

X

X

X

X





X

X

X

X

X

3.40E-02

H3-FL000775-0-0000

0-0.5

Residential

X

X

X

X

X

X

X

X

X

X

X

X

3.10E-01 J

Analyses:

PCBs = Total PCBs and Aroclors
Congeners = PCB congeners and homologs
D/F = Dioxins/Furans

A9 Vol. = Appendix IX volatiles or Volatiles
A9 Semi. = Appendix IX semivolatiles or Semivolatiles
A9 Pest. = Appendix IX pesticides or Pesticides

Herb. = Herbicides

Inorganics = Inorganics (e.g., sulfides)
GS = Grain size

GS Class = Grain size class (e.g., sand)
Org. = Organic (e.g., TOC) or Organic Content

Tables B-l through B-4.xls Table 5-3

-------
Table B-4

Fish Background Sample Identification Numbers and Associated Analyses
Housatonic River Site, OU 2, Pittsfield, MA







Analyses









Xfi

09

£/)
s

gf



1/3
V

Ph

£/)
"c3

£/)
¦e
'E.

Total PCBs

Field Sample ID

Species

Sample Type

U

Q.

o

fj



o\



£

(mg/kg)



East Branch Housatonic River -

Upstream of Newell Street





H0-TFN6BB01-0-8S29

Brown Bullhead

F

X

X

X

X



X

1.21E-01

H0-TFN6BB02-0-8S29

Brown Bullhead

F

X

X

X

X



X

2.95E-02

H0-TFN6BB03-0-8S29

Brown Bullhead

F

X

X

X

X



X

5.25E-02

H0-TFN6BB04-0-8S29

Brown Bullhead

F

X

X

X

X



X

2.99E-02

H0-TFN6BB05-0-8S29

Brown Bullhead

F

X

X

X

X



X

1.50E-01

H0-TFN6LB06-0-8S29

Largemouth Bass

F

X

X

X

X



X

4.25E-02

H0-TFN6 YPO1-0-8S29

Yellow Perch

F

X

X

X

X



X

9.70E-03 J

H0-TFN6YP02-0-8S29

Yellow Perch

F

X

X

X

X



X

1.21E-01

H0-TFN6YP03-0-8S29

Yellow Perch

F

X

X

X

X



X

1.21E-01

H0-TFN6YP04-0-8S29

Yellow Perch

F

X

X



X



X

2.31E-01

H0-TFN6YP05-0-8S29

Yellow Perch

F

X

X



X



X

1.52E-01

H0-TFN6YP06-0-8S29

Yellow Perch

F

X

X

X

X



X

2.97E-01

H0-TFN6YP07-0-8S29

Yellow Perch

F

X

X



X



X

2.14E-01

H0-TFN6YP08-0-8S29

Yellow Perch

F

X

X

X

X



X

3.73E-01

H0-TFN6YP09-0-8S29

Yellow Perch

F

X

X

X

X



X

2.78E-01

H0-TFN6YP10-0-8S29

Yellow Perch

F

X

X



X



X

2.70E-01

H0-TFN6YP11-0-8S29

Yellow Perch

F

X

X

X

X



X

2.88E-01

H0-TFN6YP12-0-8S29

Yellow Perch

F

X

X

X

X



X

3.12E-01

H0-TFN6YP13-0-8S29

Yellow Perch

F

X

X

X

X



X

3.85E-01

H0-TFN6YP14-0-8S29

Yellow Perch

F

X

X

X

X



X

2.29E-01

H0-TFN6YP15-0-8S29

Yellow Perch

F

X

X

X

X



X

2.58E-01

H0-TFN6YP16-0-8S30

Yellow Perch

F

X

X

X

X



X

2.75E-01

H0-TFN6YP17-0-8S30

Yellow Perch

F

X

X

X

X



X

2.45E-01

H0-TFN6YP18-0-8S30

Yellow Perch

F

X

X

X

X



X

2.97E-01

H0-TFN6YP19-0-8S30

Yellow Perch

F

X

X

X

X



X

3.04E-01

HO-TON6BBO1-0-8S29

Brown Bullhead

O

X

X

X

X



X

1.67E-01

H0-TON6BB02-0-8S29

Brown Bullhead

O

X

X

X

X



X

1.15E-01

H0-TON6BB03-0-8S29

Brown Bullhead

o

X

X

X

X



X

7.49E-02

H0-TON6BB04-0-8S29

Brown Bullhead

o

X

X

X

X



X

8.18E-02

H0-TON6BB05-0-8S29

Brown Bullhead

o

X

X

X

X



X

1.15E-01

H0-TON6LB06-0-8S29

Largemouth Bass

o

X

X

X

X



X

4.24E-01

H0-TON6YP01-0-8S29

Yellow Perch

o

X

X

X

X



X

1.66E-01

H0-TON6YP02-0-8S29

Yellow Perch

o

X

X

X

X



X

8.62E-02

H0-TON6YP03-0-8S29

Yellow Perch

o

X

X

X

X



X

1.88E-01

H0-TON6YP04-0-8S29

Yellow Perch

o

X

X



X



X

1.29E-01

H0-TON6YP05-0-8S29

Yellow Perch

o

X

X



X



X

2.60E-01

H0-TON6YP06-0-8S29

Yellow Perch

o

X

X

X

X



X

1.92E-01

H0-TON6YP07-0-8S29

Yellow Perch

o

X

X



X



X

2.00E-01

H0-TON6YP08-0-8S29

Yellow Perch

o

X

X

X

X



X

1.70E-01

H0-TON6YP09-0-8S29

Yellow Perch

o

X

X

X

X



X

2.05E-01

H0-TON6YP10-0-8S29

Yellow Perch

o

X

X



X



X

1.99E-01

HO-TON6YP11-0-8S29

Yellow Perch

o

X

X

X

X



X

1.62E-01

HO-TON6 YP12-0-8S29

Yellow Perch

o

X

X

X

X



X

1.87E-01

H0-TON6YP13-0-8S29

Yellow Perch

o

X

X

X

X



X

1.43E-01

HO-TON6 YP 14-0-8S29

Yellow Perch

o

X

X

X

X



X

1.43E-01

H0-TON6YP15-0-8S29

Yellow Perch

o

X

X

X

X



X

1.12E-01

H0-TON6YP16-0-8S30

Yellow Perch

o

X

X

X

X



X

1.72E-01

H0-TON6YP17-0-8S30

Yellow Perch

o

X

X

X

X



X

2.37E-01

H0-TON6YP18-0-8S30

Yellow Perch

o

X

X

X

X



X

2.47E-01

H0-TON6YP19-0-8S30

Yellow Perch

o

X

X

X

X



X

3.51E-01

H0-TWN6BBC1-0-8S29

Brown Bullhead

CM

X

X

X

X



X

1.28E-01

H0-TWN6BBC2-0-8S29

Brown Bullhead

CM

X

X

X

X



X

1.62E-01

H0-TWN6BBC3-0-8S28

Brown Bullhead

CM

X

X

X

X



X

2.08E-01

H0-TWN6BBC4-0-8S28

Brown Bullhead

CM

X

X

X

X



X

1.20E-01

H0-TWN6BBC5-0-8S28

Brown Bullhead

CM

X

X

X

X



X

1.07E-01

Tables B-l through B-4.xls Table 5-4

-------
Table B-4

Fish Background Sample Identification Numbers and Associated Analyses
Housatonic River Site, OU 2, Pittsfield, MA







Analyses









09

£/)
s

gf



1/3
V

Ph

£/)
"c3

£/)
¦e
'E.

Total PCBs

Field Sample ID

Species

Sample Type

U

Q.

o

fj

s

o\



£

(mg/kg)

H0-TWN6BBC6-0-8S29

Brown Bullhead

CM

X

X

X

X



X

1.11E-01

H0-TWN6BBC7-0-8S29

Brown Bullhead

CM

X

X

X

X



X

1.57E-01

H0-TWN6BBC8-0-8S29

Brown Bullhead

CM

X

X

X

X



X

1.08E-01

H0-TWN6BBC9-0-8S29

Brown Bullhead

CM

X

X

X

X



X

1.17E-01

H0-TWN6CSC1-0-8S29

Common Shiner

CM

X

X

X

X



X

1.58E-01

H0-TWN6FFC6-0-8S30

Fallfish

CM

X

X

X

X



X

2.36E+00

H0-TWN6FFC7-0-8S30

Fallfish

CM

X

X

X

X



X

1.01E+00

F10-TWN6GSC1-0-8S29

Golden Shiner

CM

X

X

X

X



X

1.93E-01

H0-TWN6GSC2-0-8S30

Golden Shiner

CM

X

X

X

X



X

1.33E-01

F10-TWN6LB01-0-8S29

Largemouth Bass

WH

X

X

X

X



X

1.58E-01

H0-TWN6LB01-0-8S30

Largemouth Bass

WH

X

X

X

X



X

1.39E-01

H0-TWN6LB02-0-8S29

Largemouth Bass

WH

X

X



X



X

1.19E-01

H0-TWN6LB02-0-8S30

Largemouth Bass

WH

X

X

X

X



X

3.09E-01

H0-TWN6LB03-0-8S29

Largemouth Bass

WH

X

X

X

X



X

1.41E-01

H0-TWN6LB03-0-8S30

Largemouth Bass

WH

X

X

X

X



X

2.76E-01

H0-TWN6LB04-0-8S29

Largemouth Bass

WH

X

X

X

X



X

1.94E-01

H0-TWN6LB04-0-8S30

Largemouth Bass

WH

X

X



X



X

1.90E-01

H0-TWN6LB05-0-8S29

Largemouth Bass

WH

X

X

X

X



X

1.98E-01

H0-TWN6LB05-0-8S30

Largemouth Bass

WH

X

X

X

X



X

2.52E-01

H0-TWN6LB06-0-8S30

Largemouth Bass

WH

X

X

X

X



X

2.20E-01

H0-TWN6LB07-0-8S29

Largemouth Bass

WH

X

X

X

X



X

2.06E-01

H0-TWN6LB07-0-8S30

Largemouth Bass

WH

X

X

X

X



X

2.62E-01

H0-TWN6LB08-0-8S29

Largemouth Bass

WH

X

X

X

X



X

1.95E-01

H0-TWN6LB08-0-8S30

Largemouth Bass

WH

X

X



X



X

1.64E-01

H0-TWN6LB09-0-8S29

Largemouth Bass

WH

X

X

X

X



X

2.32E-01

H0-TWN6LB09-0-8S30

Largemouth Bass

WH

X

X

X

X



X

2.42E-01

F10-TWN6LB10-0-8S29

Largemouth Bass

WH

X

X

X

X



X

2.08E-01

H0-TWN6LB10-0-8S30

Largemouth Bass

WH

X

X

X

X



X

2.29E-01

H0-TWN6LBC1-0-8S30

Largemouth Bass

CM

X

X

X

X



X

7.46E-02

H0-TWN6NAC1-0-8S29

Bluntnose Minnow

CM

X

X

X

X



X

1.99E-01

H0-TWN6NAC2-0-8S29

Bluntnose Minnow

CM

X

X

X

X



X

2.15E-01

H0-TWN6PSC0-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.63E-01

H0-TWN6PSC1-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.93E-01

H0-TWN6PSC2-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.47E-01

H0-TWN6PSC3-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.50E-01

H0-TWN6PSC4-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.48E-01

H0-TWN6PSC5-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.62E-01

H0-TWN6PSC6-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.50E-01

H0-TWN6PSC7-0-8S29

Pumpkinseed

CM

X

X

X

X



X

2.09E-01

H0-TWN6PSC8-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.43E-01

H0-TWN6PSC9-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.77E-01

F10-TWN6 YPC1-0-8S29

Yellow Perch

CM

X

X

X

X



X

1.46E-01

H0-TWN6YPC2-0-8S29

Yellow Perch

CM

X

X

X

X



X

1.49E-01

H0-TWN6YPC3-0-8S29

Yellow Perch

CM

X

X

X

X



X

1.56E-01

H0-TWN6YPC4-0-8S29

Yellow Perch

CM

X

X

X

X



X

1.40E-01

H0-TWN6YPC5-0-8S30



CM

X

X

X

X



X

1.56E-01

Three-Mile Pond

H9-TF3MBB01-0-8S28

Brown Bullhead

F

X

X

X

X



X

8.40E-03 J

H9-TF3MBB01-0-8S29

Brown Bullhead

F

X

X

X

X



X

1.05E-02 J

H9-TF3MBB02-0-8S29

Brown Bullhead

F

X

X

X

X



X

2.03E-02

H9-TF3MBB03-0-8S29

Brown Bullhead

F

X

X

X

X



X

9.60E-03 J

H9-TF3MBB04-0-8S29

Brown Bullhead

F

X

X

X

X



X

4.70E-03 J

H9-TF3MBB05-0-8S29

Brown Bullhead

F

X

X

X

X



X

1.01E-02 J

H9-TF3MLB01-0-8S28

Largemouth Bass

F

X

X



X



X

9.10E-03 J

H9-TF3MLB01-0-9Y13

Largemouth Bass

F









X

X

NA

H9-TF3MLB02-0-8S28

Largemouth Bass

F

X

X

X

X



X

4.20E-03 J

Tables B-l through B-4.xls Table 5-4

-------
Table B-4

Fish Background Sample Identification Numbers and Associated Analyses
Housatonic River Site, OU 2, Pittsfield, MA







Analyses









Xfi

09

£/)
s

gf



1/3
V

Ph

£/)
"c3

£/)
¦e
'E.

Total PCBs

Field Sample ID

Species

Sample Type

U

Q.

o

fj

s

o\



£

(mg/kg)

H9-TF3MLB02-0-9Y13

Largemouth Bass

F









X

X

NA

H9-TF3MLB03-0-8S28

Largemouth Bass

F

X

X

X

X



X

1.09E-02 J

H9-TF3MLB03-0-9Y13

Largemouth Bass

F









X

X

NA

H9-TF3MLB04-0-8S28

Largemouth Bass

F

X

X

X

X



X

1.54E-02 J

H9-TF3MLB04-0-9Y13

Largemouth Bass

F









X

X

NA

H9-TF3MLB05-0-9Y13

Largemouth Bass

F









X

X

NA

H9-TF3MLB06-0-9Y13

Largemouth Bass

F









X

X

NA

H9-TF3MLB09-0-8S28

Largemouth Bass

F

X

X

X

X



X

1.59E-02 J

H9-TF3MLB12-0-8S28

Largemouth Bass

F

X

X

X

X



X

1.45E-02 J

H9-TF3MLB12-1-8S28

Largemouth Bass

F

X

X

X

X



X

6.70E-03 J

H9-TF3MLB15-0-8S28

Largemouth Bass

F

X

X

X

X



X

3.10E-03 J

H9-TF3MLB19-0-8S28

Largemouth Bass

F

X

X

X

X



X

4.60E-03 J

H9-TF3MPS01-0-8S28

Pumpkinseed

F

X

X

X

X



X

2.40E-03 J

H9-TF3MPS02-0-8S28

Pumpkinseed

F

X

X

X

X



X

2.10E-03 J

H9-TF3MPS03-0-8S28

Pumpkinseed

F

X

X

X

X



X

2.40E-03 J

H9-TF3MPS04-0-8S28

Pumpkinseed

F

X

X

X

X



X

2.60E-03 J

H9-TF3MPS05-0-8S28

Pumpkinseed

F

X

X

X

X



X

4.10E-03 J

H9-TF3MPS06-0-8S28

Pumpkinseed

F

X

X

X

X



X

1.50E-03 J

H9-TF3MPS07-0-8S28

Pumpkinseed

F

X

X

X

X



X

6.00E-03 J

H9-TF3MPS08-0-8S29

Pumpkinseed

F

X

X

X

X



X

5.70E-03 J

H9-TF3MPS09-0-8S29

Pumpkinseed

F

X

X

X

X



X

5.20E-03 J

H9-TF3MPS10-0-8S29

Pumpkinseed

F

X

X

X

X



X

9.80E-03 J

H9-TF3MPS11-0-8S29

Pumpkinseed

F

X

X

X

X



X

9.70E-03 J

H9-TF3MPS12-0-8S29

Pumpkinseed

F

X

X

X

X



X

5.40E-03 J

H9-TF3MYP01-0-8S28

Yellow Perch

F

X

X

X

X



X

1.61E-02 J

H9-TF3MYP02-0-8S28

Yellow Perch

F

X

X

X

X



X

1.60E-02 J

H9-TF3MYP03-0-8S28

Yellow Perch

F

X

X

X

X



X

1.44E-02 J

H9-TF3MYP04-0-8S28

Yellow Perch

F

X

X



X



X

9.20E-03 J

H9-TF3MYP05-0-8S28

Yellow Perch

F

X

X

X

X



X

4.80E-03 J

H9-TF3MYP06-0-8S28

Yellow Perch

F

X

X

X

X



X

3.60E-03 J

H9-TF3MYP07-0-8S28

Yellow Perch

F

X

X

X

X



X

3.60E-03 J

H9-TF3MYP08-0-8S28

Yellow Perch

F

X

X

X

X



X

2.38E-02

H9-TF3MYP09-0-8S28

Yellow Perch

F

X

X

X

X



X

1.84E-02 J

H9-TF3MYP10-0-8S28

Yellow Perch

F

X

X

X

X



X

1.02E-02 J

H9-TF3MYP10-1-8S28

Yellow Perch

F

X

X

X

X



X

6.70E-03 J

H9-TF3MYP12-0-8S29

Yellow Perch

F

X

X

X

X



X

5.50E-03 J

H9-TF3MYP13-0-8S29

Yellow Perch

F

X

X



X



X

1.49E-02 J

H9-TF3MYP14-0-8S29

Yellow Perch

F

X

X

X

X



X

1.66E-02 J

H9-TF3MYP15-0-8S29

Yellow Perch

F

X

X

X

X



X

1.49E-02 J

H9-TF3MYP16-0-8S29

Yellow Perch

F

X

X

X

X



X

7.10E-03 J

H9-TF3MYP17-0-8S29

Yellow Perch

F

X

X

X

X



X

1.22E-02 J

H9-TF3MYP18-0-8S29

Yellow Perch

F

X

X

X

X



X

3.80E-03 J

H9-T03MBB01-0-8S28

Brown Bullhead

O

X

X

X

X



X

2.95E-02

H9-T03MBB01-0-8S29

Brown Bullhead

O

X

X

X

X



X

2.40E-02

H9-TO3MBB02-0-8S29

Brown Bullhead

o

X

X

X

X



X

4.62E-02

H9-TO3MBB03-0-8S29

Brown Bullhead

o

X

X

X

X



X

5.00E-02

H9-TO3MBB04-0-8S29

Brown Bullhead

o

X

X

X

X



X

1.03E-02 J

H9-TO3MBB05-0-8S29

Brown Bullhead

o

X

X

X

X



X

4.37E-02

H9-T03MLB01-0-8S28

Largemouth Bass

o

X

X



X



X

1.19E-01

H9-T03MLB01-0-9Y13

Largemouth Bass

o









X

X

NA

H9-TO3MLB02-0-8S28

Largemouth Bass

o

X

X

X

X



X

1.50E-01

H9-T03MLB02-0-9Y13

Largemouth Bass

o









X

X

NA

H9-TO3MLB03-0-8S28

Largemouth Bass

o

X

X

X

X



X

1.31E-01

H9-T03MLB03-0-9Y13

Largemouth Bass

o









X

X

NA

H9-TO3MLB04-0-8S28

Largemouth Bass

o

X

X

X

X



X

7.93E-02

H9-T03MLB04-0-9Y13

Largemouth Bass

o









X

X

NA

Tables B-l through B-4.xls Table 5-4

-------
Table B-4

Fish Background Sample Identification Numbers and Associated Analyses
Housatonic River Site, OU 2, Pittsfield, MA







Analyses









Xfi

09

£/)
s

gf



1/3
V

Ph

£/)
"c3

£/)
¦e
'E.

Total PCBs

Field Sample ID

Species

Sample Type

U

Q.

o

fj

s

o\



£

(mg/kg)

H9-T03MLB05-0-9Y13

Largemouth Bass

O









X

X

NA

H9-T03MLB06-0-9Y13

Largemouth Bass

O









X

X

NA

H9-T03MLB07-0-9Y13

Largemouth Bass

o

X

X

X

X



X

4.02E-02

H9-T03MLB08-0-9Y13

Largemouth Bass

o

X

X

X

X



X

2.19E-02

H9-TO3MLB09-0-8S28

Largemouth Bass

o

X

X

X

X



X

1.59E-01

H9-T03MLB09-0-9Y13

Largemouth Bass

o

X

X

X

X



X

3.91E-02

H9-T03MLB10-0-9Y13

Largemouth Bass

o

X

X

X

X



X

3.18E-02

H9-TO3MLB11-0-9Y13

Largemouth Bass

o

X

X

X

X



X

2.56E-02

H9-TO3MLB12-0-8S28

Largemouth Bass

o

X

X

X

X



X

1.05E-01

H9-TO3MLB12-0-9Y13

Largemouth Bass

o

X

X

X

X



X

3.48E-02

H9-TO3MLB15-0-8S28

Largemouth Bass

o

X

X

X

X



X

1.30E-01

H9-TO3MLB19-0-8S28

Largemouth Bass

o

X

X

X

X



X

2.66E-01

H9-T03MPS01-0-8S28

Pumpkinseed

o

X

X

X

X



X

1.46E-01

H9-TO3MPS02-0-8S28

Pumpkinseed

o

X

X

X

X



X

6.59E-02

H9-TO3MPS03-0-8S28

Pumpkinseed

o

X

X

X

X



X

8.05E-02

H9-TO3MPS04-0-8S28

Pumpkinseed

o

X

X

X

X



X

6.23 E-02

H9-TO3MPS05-0-8S28

Pumpkinseed

o

X

X

X

X



X

7.83E-02

H9-TO3MPS06-0-8S28

Pumpkinseed

o

X

X

X

X



X

1.14E-01

H9-TO3MPS07-0-8S28

Pumpkinseed

o

X

X

X

X



X

3.43 E-02

H9-TO3MPS08-0-8S29

Pumpkinseed

o

X

X

X

X



X

7.39E-02

H9-TO3MPS09-0-8S29

Pumpkinseed

o

X

X

X

X



X

1.68E-02 J

H9-T03MPS10-0-8S29

Pumpkinseed

o

X

X

X

X



X

3.02E-02

H9-TO3MPS11-0-8S29

Pumpkinseed

o

X

X

X

X



X

7.82E-02

H9-TO3MPS12-0-8S29

Pumpkinseed

o

X

X

X

X



X

7.64E-02

H9-T03MYP01-0-8S28

Yellow Perch

o

X

X

X

X



X

1.80E-01

H9-TO3MYP02-0-8S28

Yellow Perch

o

X

X

X

X



X

1.83E-01

H9-TO3MYP03-0-8S28

Yellow Perch

o

X

X

X

X



X

8.06E-02

H9-TO3MYP04-0-8S28

Yellow Perch

o

X

X



X



X

1.02E-01

H9-TO3MYP05-0-8S28

Yellow Perch

o

X

X

X

X



X

9.51E-02

H9-TO3MYP06-0-8S28

Yellow Perch

o

X

X

X

X



X

8.98E-02

H9-TO3MYP07-0-8S28

Yellow Perch

o

X

X

X

X



X

7.54E-02

H9-TO3MYP08-0-8S28

Yellow Perch

o

X

X



X



X

1.36E-01

H9-TO3MYP09-0-8S28

Yellow Perch

o

X

X

X

X



X

1.63E-01

H9-T03MYP10-0-8S28

Yellow Perch

o

X

X

X

X



X

6.92E-02

H9-TO3MYP12-0-8S29

Yellow Perch

o

X

X

X

X



X

4.04E-02

H9-TO3MYP13-0-8S29

Yellow Perch

o

X

X



X



X

5.10E-02

H9-TO3MYP14-0-8S29

Yellow Perch

o

X

X

X

X



X

4.54E-02

H9-TO3MYP15-0-8S29

Yellow Perch

o

X

X

X

X



X

5.04E-02

H9-TO3MYP16-0-8S29

Yellow Perch

o

X

X

X

X



X

1.66E-02 J

H9-TO3MYP17-0-8S29

Yellow Perch

o

X

X

X

X



X

1.21E-01

H9-TO3MYP18-0-8S29

Yellow Perch

o

X

X

X

X



X

4.31E-02

H9-TW3MGSC1-0-8S28

Golden Shiner

CM

X

X

X

X



X

4.14E-02

H9-TW3MGSC2-0-8S28

Golden Shiner

CM

X

X

X

X



X

6.05E-02

H9-TW3MGSC3-0-8S29

Golden Shiner

CM

X

X

X

X



X

2.86E-02

H9-TW3MGSC4-0-8S29

Golden Shiner

CM

X

X

X

X



X

2.38E-02

H9-TW3MGSC5-0-8S29

Golden Shiner

CM

X

X

X

X



X

2.15E-02

H9-TW3MGSC6-0-8S29

Golden Shiner

CM

X

X

X

X



X

2.44E-02

H9-TW3MLB01-0-8S28

Largemouth Bass

WH

X

X

X

X



X

3.13E-02

H9-TW3MLB02-0-8S28

Largemouth Bass

WH

X

X



X



X

5.20E-02

H9-TW3MLB03-0-8S28

Largemouth Bass

WH

X

X

X

X



X

3.43 E-02

H9-TW3MLB04-0-8S28

Largemouth Bass

WH

X

X

X

X



X

2.45E-02

H9-TW3MLB07-0-8S28

Largemouth Bass

WH

X

X



X



X

3.60E-02

H9-TW3MLB08-0-8S28

Largemouth Bass

WH

X

X

X

X



X

3.57E-02

H9-TW3MLB09-0-8S28

Largemouth Bass

WH

X

X



X



X

2.24E-02

H9-TW3MLBC1-0-8S28

Largemouth Bass

CM

X

X

X

X



X

2.42E-02

H9-TW3MLBC2-0-8S28

Largemouth Bass

CM

X

X

X

X



X

1.64E-02 J

Tables B-l through B-4.xls Table 5-4

-------
Table B-4

Fish Background Sample Identification Numbers and Associated Analyses
Housatonic River Site, OU 2, Pittsfield, MA







Analyses









Xfi

09

£/)
2-

S

gf



1/3
V

Ph

£/)
"c3

Xfi

¦e
'E.

Total PCBs

Field Sample ID

Species

Sample Type

U

Q.

o

fj

s

o\





(mg/kg)

H9-TW3MLBC3-0-8S28

Largemouth Bass

CM

X

X

X

X



X

1.87E-02 J

H9-TW3MLBC4-0-8S28

Largemouth Bass

CM

X

X

X

X



X

3.14E-02 J

H9-TW3MPSC1-0-8S28

Pumpkinseed

CM

X

X

X

X



X

2.34E-02

H9-TW3MPSC2-0-8S28

Pumpkinseed

CM

X

X

X

X



X

2.98E-02

H9-TW3MPSC3-0-8S28

Pumpkinseed

CM

X

X

X

X



X

2.75E-02

H9-TW3MPSC4-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.87E-02 J

H9-TW3MPSC5-0-8S29

Pumpkinseed

CM

X

X

X

X



X

1.97E-02 J

H9-TW3MYPC1-0-8S28

Yellow Perch

CM

X

X

X

X



X

2.08E-02

H9-TW3MYPC2-0-8S28

Yellow Perch

CM

X

X

X

X



X

1.73E-02 J

Analyses

PCBs = Total PCBs and Aroclors
Congeners = PCB congeners and homologs
D/F = Dioxins/Furans

A9 Pest. = Appendix IX pesticides or Pesticides
Metals = Metals (e.g., aluminum and lead)

Sample Types

CM = Composite
F = Fillet
O = Offal
WH = Whole body

Tables B-l through B-4.xls Table 5-4

-------
Table B-5

Sediment Chemistry Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

17 / 22

2.40E-02

3.20E-01

3.60E-01

-

4.50E-01

1.10E-01

4.49E-01

DIBENZOFURAN

15 / 22

3.20E-02

2.00E+00

3.60E-01

-

8.60E-01

2.45E-01

1.83E+00

1,4-DICHLOROBENZENE

16 / 22

2.10E-02

1.20E-01

3.60E-01

-

2.10E+00

7.15E-02

1.86E+00

DI-N-BUTYL PHTHALATE

1 / 22

2.60E-02

2.60E-02

3.60E-01

-

2.10E+00

4.45E-01

1.91E+00

2-METHYLNAPHTHALENE

18 / 22

2.50E-02

1.20E+00

3.60E-01

-

2.10E+00

7.25E-02

1.97E+00

4-METHYLPHENOL

4 / 22

2.90E-02

6.70E-02

3.60E-01

-

2.10E+00

4.30E-01

1.91E+00

N-NITROSO-DI-N-BUTYL AMINE

1 / 22

5.60E-02

5.60E-02

3.60E-01

-

2.10E+00

4.35E-01

1.91E+00

PENTACHLOROBENZENE

7 / 22

2.10E-02

7.00E-02

3.60E-01

-

2.10E+00

4.20E-01

1.91E+00

PAHs











ACENAPHTHENE

16 / 22

3.30E-02

1.70E+00

3.60E-01

-

8.60E-01

2.25E-01

1.57E+00

ACENAPHTHYLENE

20 / 22

2.00E-02

2.30E+00

3.60E-01

-

2.10E+00

8.45E-02

2.27E+00

ANTHRACENE

22 / 22

2.30E-02

1.10E+01

N/A

1.80E-01

9.68E+00

BENZO(A)ANTHRACENE

22 / 22

3.00E-02

1.50E+01

N/A

6.40E-01

1.37E+01

BENZO(B)FLUORANTHENE

22 / 22

2.60E-02

7.30E+00

N/A

7.15E-01

6.97E+00

BENZO(K)FLUORANTHENE

22 / 22

3.00E-02

9.60E+00

N/A

7.30E-01

9.00E+00

BENZO(GHI)PERYLENE

22 / 22

2.20E-02

3.80E+00

N/A

3.10E-01

3.68E+00

BENZO(A)PYRENE

22 / 22

3.10E-02

9.20E+00

N/A

7.50E-01

8.65E+00

CHRYSENE

22 / 22

3.40E-02

1.30E+01

N/A

7.25E-01

1.20E+01

DIBENZO(A,H) ANTHRACENE

20 / 22

4.00E-02

2.30E+00

3.60E-01

-

8.60E-01

1.05E-01

2.10E+00

FLUORANTHENE

22 / 22

5.80E-02

2.00E+01

N/A

1.10E+00

1.85E+01

FLUORENE

19 / 22

3.10E-02

4.00E+00

3.60E-01

-

8.60E-01

1.35E-01

3.53E+00

INDENO( 1,2,3 -C,D)P"YRENE

22 / 22

2.10E-02

4.40E+00

N/A

3.40E-01

4.21E+00

NAPHTHALENE

21 / 22

3.30E-02

4.70E+00

3.60E-01

-

3.60E-01

1.15E-01

4.07E+00

PHENANTHRENE

22 / 22

3.70E-02

2.90E+01

N/A

7.30E-01

2.61E+01

PYRENE

22 / 22

5.80E-02

2.20E+01

N/A

1.20E+00

2.02E+01

TOTAL PAH (USING 0)

24 / 24

3.70E-01

1.59E+02

N/A

8.34E+00

1.36E+02

TOTAL PAH (USING DL)

24 / 24

2.17E+00

1.59E+02

N/A

8.34E+00

1.36E+02

TOTAL PAH (USING HALF DL)

24 / 24

1.27E+00

1.59E+02

N/A

8.34E+00

1.36E+02

N/A = Not Applicable

Tables B-5- B-20.xlsB-5

-------
Table B-5

Sediment Chemistry Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (HIGH) (USING 0)

24 / 24

2.52E-01

8.66E+01

N/A

5.93E+00

7.61E+01

TOTAL PAH (HIGH) (USING DL)

24 / 24

6.12E-01

8.66E+01

N/A

5.93E+00

7.61E+01

TOTAL PAH (HIGH) (USING HALF DL)

24 / 24

4.32E-01

8.66E+01

N/A

5.93E+00

7.61E+01

TOTAL PAH (LOW) (USING 0)

24 / 24

1.18E-01

7.27E+01

N/A

2.47E+00

6.03E+01

TOTAL PAH (LOW) (USING DL)

24 / 24

1.13E+00

7.27E+01

N/A

2.51E+00

6.08E+01

TOTAL PAH (LOW) (USING HALF DL)

24 / 24

7.72E-01

7.27E+01

N/A

2.47E+00

6.06E+01

1,2,4,5-TETRACHLOROBENZENE

3 / 22

2.20E-02

8.30E-02

3.60E-01

-

2.10E+00

4.30E-01

1.91E+00

1,2,4-TRICHLOROBENZENE

10 / 22

2.10E-02

8.40E-02

3.60E-01

-

2.10E+00

3.60E-01

1.91E+00

APPIX PESTICIDES

4,4'-DDD

1 / 20

2.30E-02

2.30E-02

3.80E-03

-

1.10E+00

1.55E-01

1.09E+00

ENDRIN ALDEHYDE

1 / 20

9.00E-01

9.00E-01

5.20E-03

-

1.10E+00

1.55E-01

1.09E+00

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

25 / 25

7.50E-06

2.94E-03

N/A

6.24E-04

2.85E-03

TOTAL DIOXINS (USING DL)

25 / 25

7.79E-06

2.94E-03

N/A

6.24E-04

2.85E-03

TOTAL DIOXINS (USING HALF DL)

25 / 25

7.65E-06

2.94E-03

N/A

6.24E-04

2.85E-03

TOTAL FURANS (USING 0)

25 / 25

4.16E-06

4.47E-03

N/A

7.09E-04

4.19E-03

TOTAL FURANS (USING DL)

25 / 25

4.32E-06

4.47E-03

N/A

7.09E-04

4.19E-03

TOTAL FURANS (USING HALF DL)

25 / 25

4.24E-06

4.47E-03

N/A

7.09E-04

4.19E-03

PCBS

AROCLOR-1248

1 / 400

1.40E+00

1.40E+00

3.60E-02

-

2.51E+01

5.04E-01

5.02E+00

AROCL OR-1254

21 / 400

9.00E-01

4.49E+01

3.60E-02

-

2.51E+01

5.05E-01

5.10E+00

AROCLOR-1260

389 / 400

1.50E-01

2.90E+02

5.01E-01

-

5.32E-01

9.58E+00

7.74E+01

PCB, TOTAL

390 / 400

1.50E-01

2.90E+02

5.01E-01

-

5.32E-01

9.84E+00

8.26E+01

METALS

ANTIMONY

12 / 22

4.00E-01

2.00E+00

3.20E-01

-

9.70E-01

5.60E-01

1.90E+00

ARSENIC

19 / 22

1.40E+00

4.30E+00

1.50E+00

-

5.50E+00

2.20E+00

5.32E+00

BARIUM

22 / 22

8.70E+00

9.57E+01

N/A

1.75E+01

8.93E+01

BERYLLIUM

17 / 22

2.40E-01

7.80E-01

1.80E-01

-

3.20E-01

2.70E-01

7.44E-01

CADMIUM

1 / 22

1.60E+00

1.60E+00

2.00E-02

-

1.20E-01

4.00E-02

1.38E+00

N/A = Not Applicable

Tables B-5- B-20.xlsB-5

-------
Table B-5

Sediment Chemistry Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

CHROMIUM

22 / 22

5.70E+00

8.63E+01

N/A

1.24E+01

8.03E+01

COBALT

22 / 22

3.60E+00

1.46E+01

N/A

6.55E+00

1.41E+01

COPPER

22 / 22

7.60E+00

9.44E+01

N/A

1.71E+01

8.68E+01

LEAD

22 / 22

4.00E+00

3.03E+02

N/A

2.14E+01

2.74E+02

MERCURY

17 / 21

3.00E-02

8.40E-01

2.00E-02

-

2.00E-02

6.00E-02

7.84E-01

NICKEL

20 / 22

9.30E+00

2.64E+01

5.80E+00

-

7.80E+00

1.16E+01

2.52E+01

SILVER

4 / 22

1.40E-01

3.00E-01

1.00E-01

-

2.80E+00

1.80E-01

2.51E+00

THALLIUM

9 / 22

4.90E-01

3.00E+00

3.00E-01

-

1.70E+00

8.05E-01

2.85E+00

TIN

7 / 22

4.10E+00

9.60E+00

1.20E+00

-

1.00E+01

3.80E+00

9.94E+00

VANADIUM

22 / 22

4.90E+00

2.44E+01

N/A

7.90E+00

2.34E+01

ZINC

22 / 22

3.11E+01

2.75E+02

N/A

7.13E+01

2.55E+02

ORGANIC

TOTAL ORGANIC CARBON

308 / 376

1.51E+03

5.37E+05

1.10E+02

-

1.16E+04

7.41E+03

4.50E+04

INORGANICS

AMMONIA AS N

10 / 12

4.40E+00

2.80E+01

2.70E+00

-

2.80E+00

7.25E+00

2.80E+01

PERCENT SOLIDS

298 / 298

0.00E+00

9.85E+01

N/A

7.61E+01

9.12E+01

SULFIDE

2 / 17

6.40E+00

2.03E+01

5.40E+00

-

1.42E+01

1.02E+01

2.03E+01

N/A = Not Applicable

Tables B-5- B-20.xlsB-5

-------
Table B-6

Sediment Chemistry Summary
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 2

6.70E-02

6.70E-02

1.50E+00

-

1.50E+00

7.84E-01

1.50E+00

PAHs











BENZO(A)ANTHRACENE

1 / 2

1.10E-01

1.10E-01

1.50E+00

-

1.50E+00

8.05E-01

1.50E+00

BENZO(B)FLUORANTHENE

1 / 2

1.20E-01

1.20E-01

1.50E+00

-

1.50E+00

8.10E-01

1.50E+00

BENZO(K)FLUORANTHENE

1 / 2

1.90E-01

1.90E-01

1.50E+00

-

1.50E+00

8.45E-01

1.50E+00

BENZO(GHI)PERYLENE

1 / 2

7.00E-02

7.00E-02

1.50E+00

-

1.50E+00

7.85E-01

1.50E+00

BENZO(A)PYRENE

1 / 2

1.20E-01

1.20E-01

1.50E+00

-

1.50E+00

8.10E-01

1.50E+00

CHRYSENE

1 / 2

1.40E-01

1.40E-01

1.50E+00

-

1.50E+00

8.20E-01

1.50E+00

DIBENZO(A,H) ANTHRACENE

1 / 2

8.90E-01

8.90E-01

1.50E+00

-

1.50E+00

1.20E+00

1.50E+00

FLUORANTHENE

1 / 2

3.10E-01

3.10E-01

1.50E+00

-

1.50E+00

9.05E-01

1.50E+00

INDENO(l,2,3-C,D)PYRENE

1 / 2

6.40E-02

6.40E-02

1.50E+00

-

1.50E+00

7.82E-01

1.50E+00

PHENANTHRENE

1 / 2

2.40E-01

2.40E-01

1.50E+00

-

1.50E+00

8.70E-01

1.50E+00

PYRENE

1 / 2

4.80E-01

4.80E-01

1.50E+00

-

1.50E+00

9.90E-01

1.50E+00

TOTAL PAH (USING 0)

2 / 2

0.00E+00

2.73E+00

N/A

1.37E+00

2.73E+00

TOTAL PAH (USING DL)

2 / 2

7.18E+00

2.40E+01

N/A

1.56E+01

2.40E+01

TOTAL PAH (USING HALF DL)

2 / 2

4.96E+00

1.20E+01

N/A

8.48E+00

1.20E+01

TOTAL PAH (HIGH) (USING 0)

2 / 2

0.00E+00

2.18E+00

N/A

1.09E+00

2.18E+00

TOTAL PAH (HIGH) (USING DL)

2 / 2

2.18E+00

1.35E+01

N/A

7.84E+00

1.35E+01

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

2.18E+00

6.75E+00

N/A

4.47E+00

6.75E+00

TOTAL PAH (LOW) (USING 0)

2 / 2

0.00E+00

5.50E-01

N/A

2.75E-01

5.50E-01

TOTAL PAH (LOW) (USING DL)

2 / 2

5.00E+00

1.05E+01

N/A

7.75E+00

1.05E+01

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

2.78E+00

5.25E+00

N/A

4.01E+00

5.25E+00

APP IX PESTICIDES

ENDRIN ALDEHYDE

1 / 2

3.90E-01

3.90E-01

1.50E-02

-

1.50E-02

2.03E-01

3.90E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

2 / 2

4.80E-05

3.09E-04

N/A

1.79E-04

3.09E-04

TOTAL DIOXINS (USING DL)

2 / 2

4.88E-05

3.10E-04

N/A

1.79E-04

3.10E-04

TOTAL DIOXINS (USING HALF DL)

2 / 2

4.84E-05

3.09E-04

N/A

1.79E-04

3.09E-04

N/A = Not Applicable

Tables B-5- B-20.xls B-6

-------
Table B-6

Sediment Chemistry Summary
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL FURANS (USING 0)

2 / 2

3.45E-05

5.10E-04

N/A

2.72E-04

5.10E-04

TOTAL FURANS (USING DL)

2 / 2

3.45E-05

5.10E-04

N/A

2.72E-04

5.10E-04

TOTAL FURANS (USING HALF DL)

2 / 2

3.45E-05

5.10E-04

N/A

2.72E-04

5.10E-04

PCBS

AROCL OR-1254

8 / 24

4.80E-01

1.06E+02

5.03E-01

-

5.06E+00

1.78E+00

8.34E+01

AROCLOR-1260

20 / 24

1.95E+00

1.29E+02

5.01E-01

-

1.81E+00

1.09E+01

1.10E+02

PCB, TOTAL

21 / 24

1.95E+00

2.35E+02

1.06E+00

-

1.81E+00

1.17E+01

1.89E+02

METALS

ARSENIC

2 / 2

3.00E+00

4.20E+00

N/A

3.60E+00

4.20E+00

BARIUM

2 / 2

4.15E+01

8.53E+01

N/A

6.34E+01

8.53E+01

BERYLLIUM

2 / 2

5.90E-01

8.90E-01

N/A

7.40E-01

8.90E-01

CADMIUM

1 / 2

3.30E-01

3.30E-01

9.00E-02

-

9.00E-02

2.10E-01

3.30E-01

CHROMIUM

2 / 2

2.00E+01

2.03E+01

N/A

2.02E+01

2.03E+01

COBALT

2 / 2

6.40E+00

1.17E+01

N/A

9.05E+00

1.17E+01

COPPER

2 / 2

2.84E+01

2.92E+01

N/A

2.88E+01

2.92E+01

LEAD

2 / 2

1.56E+01

4.65E+01

N/A

3.11E+01

4.65E+01

MERCURY

1 / 2

1.70E-01

1.70E-01

1.60E-01

-

1.60E-01

1.65E-01

1.70E-01

NICKEL

2 / 2

1.56E+01

2.36E+01

N/A

1.96E+01

2.36E+01

TIN

1 / 2

2.60E+00

2.60E+00

7.10E+00

-

7.10E+00

4.85E+00

7.10E+00

VANADIUM

2 / 2

2.01E+01

2.23E+01

N/A

2.12E+01

2.23E+01

ZINC

2 / 2

7.83E+01

1.11E+02

N/A

9.47E+01

1.11E+02

ORGANIC

TOTAL ORGANIC CARBON

15 / 15

1.97E+04

3.87E+05

N/A

6.04E+04

3.87E+05

INORGANICS

PERCENT SOLIDS

21 / 21

0.00E+00

8.65E+01



N/A



3.74E+01

8.51E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-6

-------
Table B-7

Sediment Chemistry Summary
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 8

3.90E-02

3.90E-02

4.10E-01

-

1.35E+00

6.95E-01

1.35E+00

DI-N-BUTYL PHTHALATE

1 / 8

6.70E-02

6.70E-02

4.10E-01

-

1.35E+00

6.75E-01

1.35E+00

1,4-DICHLOROBENZENE

3 / 8

4.30E-02

7.00E-02

6.10E-01

-

1.35E+00

6.70E-01

1.35E+00

2-METHYLNAPHTHALENE

1 / 8

6.20E-02

6.20E-02

4.10E-01

-

1.35E+00

6.95E-01

1.35E+00

4-METHYLPHENOL

4 / 8

4.40E-02

1.00E+00

4.10E-01

-

9.40E-01

6.50E-01

1.00E+00

PAHs











ACENAPHTHENE

1 / 8

3.10E-02

3.10E-02

4.10E-01

-

1.35E+00

7.10E-01

1.35E+00

ACENAPHTHYLENE

4 / 8

3.00E-02

1.90E-01

7.30E-01

-

1.35E+00

4.60E-01

1.35E+00

ACETOPHENONE

4 / 8

3.30E-02

7.40E-02

4.10E-01

-

9.40E-01

2.42E-01

9.40E-01

ANTHRACENE

4 / 8

3.60E-02

1.10E-01

7.30E-01

-

1.35E+00

4.20E-01

1.35E+00

BENZO(A)ANTHRACENE

6 / 8

4.30E-02

1.00E+00

1.20E+00

-

1.35E+00

3.40E-01

1.35E+00

BENZO(B)FLUORANTHENE

5 / 8

6.20E-02

1.20E+00

7.30E-01

-

1.35E+00

5.80E-01

1.35E+00

BENZO(K)FLUORANTHENE

5 / 8

7.90E-02

1.20E+00

7.30E-01

-

1.35E+00

6.50E-01

1.35E+00

BENZO(GHI)PERYLENE

5 / 8

7.60E-02

6.50E-01

7.30E-01

-

1.35E+00

5.35E-01

1.35E+00

BENZO(A)PYRENE

5 / 8

7.60E-02

1.40E+00

7.30E-01

-

1.35E+00

6.65E-01

1.40E+00

CHRYSENE

6 / 8

6.00E-02

1.30E+00

1.20E+00

-

1.35E+00

4.70E-01

1.35E+00

DIBENZO(A,H) ANTHRACENE

4 / 8

3.60E-02

2.00E-01

7.30E-01

-

1.35E+00

4.65E-01

1.35E+00

FLUORANTHENE

7 / 8

6.40E-02

1.30E+00

1.35E+00

-

1.35E+00

3.65E-01

1.35E+00

FLUORENE

2 / 8

4.00E-02

4.60E-02

4.10E-01

-

1.35E+00

6.70E-01

1.35E+00

INDENO(l,2,3-C,D)PYRENE

5 / 8

6.80E-02

5.80E-01

7.30E-01

-

1.35E+00

4.90E-01

1.35E+00

NAPHTHALENE

5 / 8

4.60E-02

1.50E-01

7.30E-01

-

1.35E+00

1.50E-01

1.35E+00

PHENANTHRENE

6 / 8

4.70E-02

5.70E-01

1.20E+00

-

1.35E+00

3.75E-01

1.35E+00

PYRENE

6 / 8

7.90E-02

1.70E+00

1.20E+00

-

1.35E+00

7.10E-01

1.70E+00

TOTAL PAH (USING 0)

8 / 8

0.00E+00

1.16E+01

N/A

1.52E+00

1.16E+01

TOTAL PAH (USING DL)

8 / 8

2.98E+00

2.16E+01

N/A

7.66E+00

2.16E+01

TOTAL PAH (USING HALF DL)

8 / 8

2.57E+00

1.16E+01

N/A

5.48E+00

1.16E+01

TOTAL PAH (HIGH) (USING 0)

8 / 8

0.00E+00

9.23E+00

N/A

1.08E+00

9.23E+00

TOTAL PAH (HIGH) (USING DL)

8 / 8

1.50E+00

1.22E+01

N/A

4.61E+00

1.22E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-7

-------
Table B-7

Sediment Chemistry Summary
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA











Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

TOTAL PAH (HIGH) (USING HALF DL)

8 / 8

1.13E+00

9.23E+00

N/A

3.51E+00

9.23E+00

TOTAL PAH (LOW) (USING 0)

8 / 8

0.00E+00

2.40E+00

N/A

4.48E-01

2.40E+00

TOTAL PAH (LOW) (USING DL)

8 / 8

1.48E+00

9.45E+00

N/A

3.10E+00

9.45E+00

TOTAL PAH (LOW) (USING HALF DL)

8 / 8

1.07E+00

4.73E+00

N/A

2.04E+00

4.73E+00

1,2,4-TRICHLOROBENZENE

2 / 8

4.30E-02

5.30E-02

4.10E-01

-

1.35E+00

6.70E-01

1.35E+00

APPIX PESTICIDES

ALPHA-BHC

1 /

10

7.60E-03

7.60E-03

5.10E-03

-

1.50E+00

1.15E-01

1.50E+00

BETA-BHC

1

10

1.60E-02

1.60E-02

5.10E-03

-

1.50E+00

1.15E-01

1.50E+00

4,4'-DDD

2 /

10

2.40E-01

4.80E-01

1.00E-02

-

3.00E+00

2.95E-01

3.00E+00

4,4'-DDE

5 /

11

5.40E-02

2.00E+00

1.00E-02

-

3.00E+00

3.50E-01

3.00E+00

HEPTACHLOR

1 /

10

9.50E-03

9.50E-03

5.10E-03

-

1.50E+00

1.15E-01

1.50E+00

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

12 /

12

0.00E+00

2.27E-02

N/A

2.85E-04

2.27E-02

TOTAL DIOXINS (USING DL)

12 /

12

2.04E-05

2.27E-02

N/A

4.90E-04

2.27E-02

TOTAL DIOXINS (USING HALF DL)

12 /

12

1.99E-05

2.27E-02

N/A

3.34E-04

2.27E-02

TOTAL FURANS (USING 0)

12 /

12

4.23E-06

2.39E-02

N/A

2.98E-04

2.39E-02

TOTAL FURANS (USING DL)

12 /

12

4.45E-06

2.39E-02

N/A

2.98E-04

2.39E-02

TOTAL FURANS (USING HALF DL)

12 /

12

4.34E-06

2.39E-02

N/A

2.98E-04

2.39E-02

PCBS

AROCL OR-1254

23 / 220

1.01E+00

4.70E+01

5.00E-02

-

1.02E+02

1.16E+00

1.01E+01

AROCLOR-1260

178 / 222

9.80E-02

8.74E+02

1.50E-01

-

2.95E+00

7.10E+00

9.29E+01

PCB, TOTAL

185 / 222

9.80E-02

8.74E+02

1.50E-01

-

2.95E+00

7.10E+00

9.49E+01

METALS

ANTIMONY

4 / 8

9.00E-01

2.00E+00

1.30E+00

-

2.20E+00

1.93E+00

2.20E+00

ARSENIC

6 / 8

2.90E+00

9.50E+00

2.35E+00

-

4.70E+00

4.20E+00

9.50E+00

BARIUM

8 / 8

4.43E+01

8.60E+01

N/A

6.78E+01

8.60E+01

BERYLLIUM

8 / 8

3.80E-01

9.50E-01

N/A

5.70E-01

9.50E-01

CADMIUM

4 / 8

3.50E-01

1.50E+00

6.00E-02

-

1.30E-01

2.40E-01

1.50E+00

CHROMIUM

8 / 8

8.40E+00

1.04E+02

N/A

3.10E+01

1.04E+02

N/A = Not Applicable

Tables B-5- B-20.xls B-7

-------
Table B-7

Sediment Chemistry Summary
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

COBALT

8 / 8

6.90E+00

1.29E+01

N/A

1.10E+01

1.29E+01

COPPER

8 / 8

2.16E+01

1.16E+02

N/A

4.55E+01

1.16E+02

LEAD

8 / 8

4.23E+01

2.36E+02

N/A

8.76E+01

2.36E+02

MERCURY

6 / 8

1.80E-01

9.10E-01

6.00E-02

-

6.00E-02

3.30E-01

9.10E-01

NICKEL

8 / 8

1.45E+01

2.55E+01

N/A

1.96E+01

2.55E+01

SELENIUM

1 / 8

1.80E+00

1.80E+00

5.50E-01

-

1.60E+00

8.65E-01

1.80E+00

SILVER

3 / 8

6.20E-01

1.90E+00

2.20E-01

-

7.55E-01

6.55E-01

1.90E+00

THALLIUM

2 / 8

2.40E+00

4.00E+00

7.80E-01

-

2.50E+00

1.93E+00

4.00E+00

TIN

4 / 8

4.40E+00

7.20E+00

1.30E+00

-

1.56E+01

5.15E+00

1.56E+01

VANADIUM

8 / 8

1.38E+01

2.85E+01

N/A

1.81E+01

2.85E+01

ZINC

8 / 8

1.17E+02

2.88E+02

N/A

1.34E+02

2.88E+02

ORGANIC

TOTAL ORGANIC CARBON

187 /197

3.08E+03

8.53E+05

2.93E+03

-

8.05E+05

9.22E+04

5.45E+05

INORGANICS

PERCENT SOLIDS

173 / 173

0.00E+00

9.35E+01

N/A

5.08E+01

7.12E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-7

-------
Table B-8

Sediment Chemistry Summary
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

3 / 6

4.60E-02

6.10E-02

4.80E-01

5.60E-01

2.71E-01

5.60E-01

BUTYLBENZYLPHTHALATE

1 / 6

4.70E-02

4.70E-02

4.40E-01

5.60E-01

4.70E-01

5.60E-01

DIBENZOFURAN

3 / 6

3.00E-02

5.70E-02

4.40E-01

5.10E-01

2.49E-01

5.10E-01

DI-N-BUTYL PHTHALATE

1 / 6

1.60E-01

1.60E-01

4.40E-01

5.60E-01

4.95E-01

5.60E-01

1,4-DICHLOROBENZENE

2 / 6

4.40E-02

5.90E-02

4.40E-01

5.60E-01

4.60E-01

5.60E-01

DIETHYL PHTHALATE

1 / 6

5.00E-02

5.00E-02

4.40E-01

5.20E-01

4.70E-01

5.20E-01

2-METHYLNAPHTHALENE

3 / 6

3.70E-02

6.80E-02

4.40E-01

5.10E-01

2.54E-01

5.10E-01

PENTACHLOROBENZENE

1 / 6

2.70E-02

2.70E-02

4.40E-01

5.60E-01

4.70E-01

5.60E-01

PAHs

ACENAPHTHENE

2 / 6

5.00E-02

5.90E-02

4.40E-01

5.60E-01

4.60E-01

5.60E-01

ACENAPHTHYLENE

4 / 6

2.10E-02

5.70E-02

4.40E-01

4.80E-01

5.35E-02

4.80E-01

ANTHRACENE

4 / 6

7.60E-02

1.40E-01

4.40E-01

4.80E-01

1.20E-01

4.80E-01

BENZO(A)ANTHRACENE

5 / 6

7.20E-02

4.50E-01

4.80E-01

4.80E-01

3.75E-01

4.80E-01

BENZO(B)FLUORANTHENE

5 / 6

8.50E-02

4.00E-01

4.80E-01

4.80E-01

3.35E-01

4.80E-01

BENZO(K)FLUORANTHENE

5 / 6

8.60E-02

4.70E-01

4.80E-01

4.80E-01

4.15E-01

4.80E-01

BENZO(GHI)PERYLENE

5 / 6

7.60E-02

4.20E-01

4.80E-01

4.80E-01

2.10E-01

4.80E-01

BENZO(A)PYRENE

5 / 6

8.60E-02

4.60E-01

4.80E-01

4.80E-01

4.25E-01

4.80E-01

CHRYSENE

5 / 6

1.10E-01

4.90E-01

4.80E-01

4.80E-01

4.55E-01

4.90E-01

DIBENZO(A,H) ANTHRACENE

5 / 6

2.00E-02

7.30E-02

4.80E-01

4.80E-01

5.80E-02

4.80E-01

FLUORANTHENE

6 / 6

2.70E-02

8.40E-01

N/A

4.75E-01

8.40E-01

FLUORENE

4 / 6

3.20E-02

1.50E-01

4.40E-01

4.80E-01

1.12E-01

4.80E-01

INDENO(l,2,3-C,D)PYRENE

5 / 6

6.60E-02

3.30E-01

4.80E-01

4.80E-01

2.15E-01

4.80E-01

NAPHTHALENE

4 / 6

4.40E-02

1.50E-01

4.40E-01

4.80E-01

1.12E-01

4.80E-01

PHENANTHRENE

5 / 6

8.00E-02

8.40E-01

4.80E-01

4.80E-01

4.05E-01

8.40E-01

PYRENE

6 / 6

2.90E-02

7.50E-01

N/A

4.95E-01

7.50E-01

TOTAL PAH (USING 0)

6 / 6

5.60E-02

5.48E+00

N/A

3.65E+00

5.48E+00

TOTAL PAH (USING DL)

6 / 6

3.16E+00

6.78E+00

N/A

4.88E+00

6.78E+00

TOTAL PAH (USING HALF DL)

6 / 6

2.06E+00

5.48E+00

N/A

3.95E+00

5.48E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-8

-------
Table B-8

Sediment Chemistry Summary
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (HIGH) (USING 0)

6 / 6

2.90E-02

3.40E+00

N/A

2.62E+00

3.40E+00

TOTAL PAH (HIGH) (USING DL)

6 / 6

7.41E-01

3.87E+00

N/A

3.32E+00

3.87E+00

TOTAL PAH (HIGH) (USING HALF DL)

6 / 6

7.41E-01

3.40E+00

N/A

2.62E+00

3.40E+00

TOTAL PAH (LOW) (USING 0)

6 / 6

2.70E-02

2.09E+00

N/A

1.03E+00

2.09E+00

TOTAL PAH (LOW) (USING DL)

6 / 6

1.51E+00

2.91E+00

N/A

1.86E+00

2.91E+00

TOTAL PAH (LOW) (USING HALF DL)

6 / 6

1.23E+00

2.09E+00

N/A

1.42E+00

2.09E+00

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

6 / 6

5.01E-05

2.37E-03

N/A

5.73E-04

2.37E-03

TOTAL DIOXINS (USING DL)

6 / 6

5.06E-05

2.37E-03

N/A

5.73E-04

2.37E-03

TOTAL DIOXINS (USING HALF DL)

6 / 6

5.04E-05

2.37E-03

N/A

5.73E-04

2.37E-03

TOTAL FURANS (USING 0)

6 / 6

3.54E-05

1.61E-03

N/A

6.17E-04

1.61E-03

TOTAL FURANS (USING DL)

6 / 6

3.54E-05

1.61E-03

N/A

6.17E-04

1.61E-03

TOTAL FURANS (USING HALF DL)

6 / 6

3.54E-05

1.61E-03

N/A

6.17E-04

1.61E-03

PCBS

AROCLOR-1242

1 / 21

6.50E-01

6.50E-01

1.64E-02

-

3.39E+00

3.41E-01

3.39E+00

AROCL OR-1254

9 / 182

3.90E-01

4.00E+00

1.64E-02

-

5.29E+00

5.03E-01

2.48E+00

AROCLOR-1260

149 / 182

2.62E-02

1.65E+02

5.00E-01

-

1.35E+00

3.32E+00

1.81E+01

PCB, TOTAL

149 / 182

2.62E-02

1.65E+02

5.00E-01

-

1.35E+00

3.34E+00

1.81E+01

METALS

ANTIMONY

3 / 6

3.80E-01

1.80E+00

2.50E-01

-

8.30E-01

5.05E-01

1.80E+00

ARSENIC

6 / 6

9.70E-01

5.50E+00

N/A

1.90E+00

5.50E+00

BARIUM

6 / 6

1.41E+01

6.20E+01

N/A

2.48E+01

6.20E+01

BERYLLIUM

3 / 6

2.00E-01

5.70E-01

2.40E-01

-

3.60E-01

3.55E-01

5.70E-01

CHROMIUM

6 / 6

9.60E+00

1.98E+02

N/A

1.74E+01

1.98E+02

COBALT

6 / 6

5.30E+00

1.13E+01

N/A

6.65E+00

1.13E+01

COPPER

6 / 6

9.50E+00

1.39E+02

N/A

1.59E+01

1.39E+02

LEAD

6 / 6

5.60E+00

1.04E+02

N/A

1.72E+01

1.04E+02

MERCURY

5 / 5

3.00E-02

1.40E+00

N/A

7.00E-02

1.40E+00

NICKEL

6 / 6

9.30E+00

1.90E+01

N/A

1.19E+01

1.90E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-8

-------
Table B-8

Sediment Chemistry Summary
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

SILVER

2 / 6

1.10E-01

8.70E-01

1.10E-01

-

1.00E+00

1.35E-01

1.00E+00

THALLIUM

4 / 6

4.50E-01

2.30E+00

3.90E-01

-

4.20E-01

6.45E-01

2.30E+00

TIN

4 / 6

2.40E+00

5.70E+00

1.40E+00

-

1.26E+01

3.85E+00

1.26E+01

VANADIUM

6 / 6

5.70E+00

1.54E+01

N/A

8.90E+00

1.54E+01

ZINC

6 / 6

4.89E+01

1.93E+02

N/A

6.94E+01

1.93E+02

ORGANIC

TOTAL ORGANIC CARBON

159 / 175

1.62E+03

8.40E+04

1.35E+02

-

2.54E+05

1.10E+04

3.73E+04

INORGANICS

PERCENT SOLIDS

152 / 152

3.50E+01

9.41E+01

N/A

7.00E+01

8.64E+01

SULFIDE

2 / 6

1.03E+01

1.07E+01

1.11E+01

-

1.33E+01

1.15E+01

1.33E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-8

-------
Table B-9

Sediment Chemistry Summary
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 1

5.80E-02 - 5.80E-02

N/A

5.80E-02

5.80E-02

DIBENZOFURAN

1 / 1

3.60E-02 - 3.60E-02

N/A

3.60E-02

3.60E-02

DI-N-BUTYL PHTHALATE

1 / 1

3.00E-02 - 3.00E-02

N/A

3.00E-02

3.00E-02

1,4-DICHLOROBENZENE

1 / 1

6.50E-02 - 6.50E-02

N/A

6.50E-02

6.50E-02

2-METHYLNAPHTHALENE

1 / 1

6.70E-02 - 6.70E-02

N/A

6.70E-02

6.70E-02

4-METHYLPHENOL

1 / 1

3.70E-02 - 3.70E-02

N/A

3.70E-02

3.70E-02

PAHs

ACENAPHTHENE

1 / 1

2.80E-02 - 2.80E-02

N/A

2.80E-02

2.80E-02

ACENAPHTHYLENE

1 / 1

1.10E-01 - 1.10E-01

N/A

1.10E-01

1.10E-01

ANTHRACENE

1 / 1

8.60E-02 - 8.60E-02

N/A

8.60E-02

8.60E-02

BENZO(A)ANTHRACENE

1 / 1

4.80E-01 - 4.80E-01

N/A

4.80E-01

4.80E-01

BENZO(B)FLUORANTHENE

1 / 1

6.50E-01 - 6.50E-01

N/A

6.50E-01

6.50E-01

BENZO(K)FLUORANTHENE

1 / 1

6.80E-01 - 6.80E-01

N/A

6.80E-01

6.80E-01

BENZO(GHI)PERYLENE

1 / 1

2.40E-01 - 2.40E-01

N/A

2.40E-01

2.40E-01

BENZO(A)PYRENE

1 / 1

6.50E-01 - 6.50E-01

N/A

6.50E-01

6.50E-01

CHRYSENE

1 / 1

6.80E-01 - 6.80E-01

N/A

6.80E-01

6.80E-01

DIBENZO(A,H) ANTHRACENE

1 / 1

8.30E-02 - 8.30E-02

N/A

8.30E-02

8.30E-02

FLUORANTHENE

1 / 1

1.10E+00 - 1.10E+00

N/A

1.10E+00

1.10E+00

FLUORENE

1 / 1

4.60E-02 - 4.60E-02

N/A

4.60E-02

4.60E-02

INDENO(l,2,3-C,D)PYRENE

1 / 1

2.20E-01 - 2.20E-01

N/A

2.20E-01

2.20E-01

NAPHTHALENE

1 / 1

1.40E-01 - 1.40E-01

N/A

1.40E-01

1.40E-01

PHENANTHRENE

1 / 1

6.30E-01 - 6.30E-01

N/A

6.30E-01

6.30E-01

PYRENE

1 / 1

1.00E+00 - 1.00E+00

N/A

1.00E+00

1.00E+00

TOTAL PAH (USING 0)

1 / 1

6.82E+00 - 6.82E+00

N/A

6.82E+00

6.82E+00

TOTAL PAH (USING DL)

1 / 1

6.82E+00 - 6.82E+00

N/A

6.82E+00

6.82E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-9

-------
Table B-9

Sediment Chemistry Summary
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (USING HALF DL)

1 / 1

6.82E+00 - 6.82E+00

N/A

6.82E+00

6.82E+00

TOTAL PAH (HIGH) (USING 0)

1 / 1

4.68E+00 - 4.68E+00

N/A

4.68E+00

4.68E+00

TOTAL PAH (HIGH) (USING DL)

1 / 1

4.68E+00 - 4.68E+00

N/A

4.68E+00

4.68E+00

TOTAL PAH (HIGH) (USING HALF DL)

1 / 1

4.68E+00 - 4.68E+00

N/A

4.68E+00

4.68E+00

TOTAL PAH (LOW) (USING 0)

1 / 1

2.14E+00 - 2.14E+00

N/A

2.14E+00

2.14E+00

TOTAL PAH (LOW) (USING DL)

1 / 1

2.14E+00 - 2.14E+00

N/A

2.14E+00

2.14E+00

TOTAL PAH (LOW) (USING HALF DL)

1 / 1

2.14E+00 - 2.14E+00

N/A

2.14E+00

2.14E+00

1,2,4-TRICHLOROBENZENE

1 / 1

3.90E-02 - 3.90E-02

N/A

3.90E-02

3.90E-02

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

1 / 1

7.53E-03 - 7.53E-03

N/A

7.53E-03

7.53E-03

TOTAL DIOXINS (USING DL)

1 / 1

7.53E-03 - 7.53E-03

N/A

7.53E-03

7.53E-03

TOTAL DIOXINS (USING HALF DL)

1 / 1

7.53E-03 - 7.53E-03

N/A

7.53E-03

7.53E-03

TOTAL FURANS (USING 0)

1 / 1

5.60E-03 - 5.60E-03

N/A

5.60E-03

5.60E-03

TOTAL FURANS (USING DL)

1 / 1

5.60E-03 - 5.60E-03

N/A

5.60E-03

5.60E-03

TOTAL FURANS (USING HALF DL)

1 / 1

5.60E-03 - 5.60E-03

N/A

5.60E-03

5.60E-03

PCBS

AROCL OR-1254

4 / 13

6.28E+00 - 2.16E+01

5.05E-01 - 1.02E+01

2.55E+00

2.16E+01

AROCLOR-1260

11 / 13

5.14E+00 - 8.92E+01

9.01E-01 - 1.47E+00

1.99E+01

8.92E+01

PCB, TOTAL

11 / 13

5.14E+00 - 8.92E+01

9.01E-01 - 1.47E+00

2.40E+01

8.92E+01

METALS

ARSENIC

1 / 1

5.80E+00 - 5.80E+00

N/A

5.80E+00

5.80E+00

BARIUM

1 / 1

9.17E+01 - 9.17E+01

N/A

9.17E+01

9.17E+01

BERYLLIUM

1 / 1

7.80E-01 - 7.80E-01

N/A

7.80E-01

7.80E-01

CADMIUM

1 / 1

9.10E-01 - 9.10E-01

N/A

9.10E-01

9.10E-01

CHROMIUM

1 / 1

7.27E+01 - 7.27E+01

N/A

7.27E+01

7.27E+01

COBALT

1 / 1

1.40E+01 - 1.40E+01

N/A

1.40E+01

1.40E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-9

-------
Table B-9

Sediment Chemistry Summary
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

COPPER

1 / 1

6.95E+01 - 6.95E+01

N/A

6.95E+01

6.95E+01

LEAD

1 / 1

1.29E+02 - 1.29E+02

N/A

1.29E+02

1.29E+02

MERCURY

1 / 1

6.10E-01 - 6.10E-01

N/A

6.10E-01

6.10E-01

NICKEL

1 / 1

2.59E+01 - 2.59E+01

N/A

2.59E+01

2.59E+01

SILVER

1 / 1

1.80E+00 - 1.80E+00

N/A

1.80E+00

1.80E+00

THALLIUM

1 / 1

2.30E+00 - 2.30E+00

N/A

2.30E+00

2.30E+00

VANADIUM

1 / 1

2.55E+01 - 2.55E+01

N/A

2.55E+01

2.55E+01

ZINC

1 / 1

2.12E+02 - 2.12E+02

N/A

2.12E+02

2.12E+02

ORGANIC

TOTAL ORGANIC CARBON

6 / 6

3.03E+04 - 1.16E+05

N/A

5.18E+04

1.16E+05

INORGANICS

PERCENT SOLIDS

12 / 12

3.01E+01 - 7.40E+01

N/A

5.56E+01

7.40E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-9

-------
Table B-10

Sediment Chemistry Summary
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

2 / 8

4.40E-02

4.70E-02

5.40E-01

1.60E+00

7.00E-01

1.60E+00

1,4-DICHLOROBENZENE

4 / 8

4.70E-02

8.10E-02

5.40E-01

8.50E-01

3.11E-01

8.50E-01

DIETHYL PHTHALATE

4 / 9

7.00E-02

1.30E-01

5.40E-01

1.60E+00

5.40E-01

1.60E+00

2-METHYLNAPHTHALENE

2 / 8

6.10E-02

6.80E-02

5.40E-01

1.60E+00

7.00E-01

1.60E+00

4-METHYLPHENOL

3 / 8

8.80E-02

5.10E+00

5.40E-01

1.60E+00

8.05E-01

5.10E+00

PHENOL

1 / 8

2.20E+00

2.20E+00

5.40E-01

1.60E+00

8.45E-01

2.20E+00

PAHs

ACENAPHTHYLENE

4 / 8

4.40E-02

1.40E-01

7.70E-01

1.60E+00

4.55E-01

1.60E+00

ANTHRACENE

4 / 8

3.40E-02

7.60E-02

7.70E-01

1.60E+00

4.23E-01

1.60E+00

BENZO(A)ANTHRACENE

7 / 8

6.60E-02

5.10E-01

7.70E-01

7.70E-01

2.25E-01

7.70E-01

BENZO(B)FLUORANTHENE

7 / 8

6.20E-02

5.80E-01

7.70E-01

7.70E-01

2.80E-01

7.70E-01

BENZO(K)FLUORANTHENE

7 / 8

6.60E-02

6.60E-01

7.70E-01

7.70E-01

2.90E-01

7.70E-01

BENZO(GHI)PERYLENE

7 / 8

7.10E-02

6.10E-01

7.70E-01

7.70E-01

1.50E-01

7.70E-01

BENZO(A)PYRENE

7 / 8

7.20E-02

6.50E-01

7.70E-01

7.70E-01

2.40E-01

7.70E-01

CHRYSENE

7 / 8

1.00E-01

6.80E-01

7.70E-01

7.70E-01

2.90E-01

7.70E-01

DIBENZO(A,H) ANTHRACENE

3 / 8

5.00E-02

2.10E-01

6.30E-01

1.60E+00

7.00E-01

1.60E+00

FLUORANTHENE

7 / 8

1.10E-01

8.30E-01

7.70E-01

7.70E-01

4.20E-01

8.30E-01

FLUORENE

1 / 8

3.70E-02

3.70E-02

5.40E-01

1.60E+00

8.05E-01

1.60E+00

INDENO(l,2,3-C,D)PYRENE

7 / 8

6.00E-02

4.90E-01

7.70E-01

7.70E-01

1.30E-01

7.70E-01

NAPHTHALENE

6 / 8

6.40E-02

1.40E-01

7.70E-01

9.10E-01

9.20E-02

9.10E-01

PHENANTHRENE

7 / 8

8.30E-02

4.40E-01

7.70E-01

7.70E-01

2.50E-01

7.70E-01

PYRENE

8 / 9

5.60E-02

1.20E+00

7.70E-01

7.70E-01

3.60E-01

1.20E+00

TOTAL PAH (USING 0)

9 / 9

0.00E+00

7.17E+00

N/A

2.09E+00

7.17E+00

TOTAL PAH (USING DL)

9 / 9

5.60E-02

1.23E+01

N/A

6.31E+00

1.23E+01

TOTAL PAH (USING HALF DL)

9 / 9

5.60E-02

8.02E+00

N/A

4.19E+00

8.02E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-10

-------
Table B-10

Sediment Chemistry Summary
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (HIGH) (USING 0)

9 / 9

0.00E+00

5.59E+00

N/A

1.51E+00

5.59E+00

TOTAL PAH (HIGH) (USING DL)

9 / 9

5.60E-02

6.93E+00

N/A

2.42E+00

6.93E+00

TOTAL PAH (HIGH) (USING HALF DL)

9 / 9

5.60E-02

5.59E+00

N/A

2.11E+00

5.59E+00

TOTAL PAH (LOW) (USING 0)

8 / 8

0.00E+00

1.58E+00

N/A

6.15E-01

1.58E+00

TOTAL PAH (LOW) (USING DL)

8 / 8

1.91E+00

6.86E+00

N/A

3.61E+00

6.86E+00

TOTAL PAH (LOW) (USING HALF DL)

8 / 8

1.28E+00

3.66E+00

N/A

2.34E+00

3.66E+00

1,2,4-TRICHLOROBENZENE

1 / 8

3.90E-02

3.90E-02

5.40E-01

-

1.60E+00

7.60E-01

1.60E+00

APPIX PESTICIDES

4,4'-DDD

1 / 10

9.40E-02

9.40E-02

2.30E-01

-

2.40E+00

7.95E-01

2.40E+00

4,4'-DDT

1 / 8

2.80E+00

2.80E+00

2.30E-01

-

2.40E+00

1.24E+00

2.80E+00

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

12 / 12

5.77E-05

1.98E-02

N/A

7.52E-04

1.98E-02

TOTAL DIOXINS (USING DL)

12 / 12

5.87E-05

1.98E-02

N/A

7.52E-04

1.98E-02

TOTAL DIOXINS (USING HALF DL)

12 / 12

5.82E-05

1.98E-02

N/A

7.52E-04

1.98E-02

TOTAL FURANS (USING 0)

12 / 12

9.91E-05

1.31E-02

N/A

8.63E-04

1.31E-02

TOTAL FURANS (USING DL)

12 / 12

9.91E-05

1.31E-02

N/A

8.63E-04

1.31E-02

TOTAL FURANS (USING HALF DL)

12 / 12

9.91E-05

1.31E-02

N/A

8.63E-04

1.31E-02

PCBS

AROCL OR-1254

25 / 156

6.40E-01

3.73E+01

6.20E-02

-

1.16E+01

1.51E+00

1.19E+01

AROCLOR-1260

139 / 156

9.50E-02

1.36E+02

7.02E-01

-

4.21E+00

1.57E+01

8.88E+01

PCB, TOTAL

141 / 156

9.50E-02

1.36E+02

7.02E-01

-

4.21E+00

1.80E+01

9.07E+01

METALS

ANTIMONY

6 / 9

1.30E+00

2.40E+00

6.00E-01

-

2.10E+00

2.00E+00

2.40E+00

ARSENIC

7 / 9

2.40E+00

7.00E+00

1.80E+00

-

5.20E+00

5.20E+00

7.00E+00

BARIUM

9 / 9

5.02E+01

9.67E+01

N/A

7.89E+01

9.67E+01

BERYLLIUM

9 / 9

3.60E-01

1.00E+00

N/A

7.50E-01

1.00E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-10

-------
Table B-10

Sediment Chemistry Summary
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

CADMIUM

5 / 8

4.60E-01

2.30E+00

4.00E-02

-

1.00E+00

9.75E-01

2.30E+00

CHROMIUM

9 / 9

1.99E+01

1.32E+02

N/A

7.43E+01

1.32E+02

COBALT

9 / 9

7.10E+00

1.66E+01

N/A

1.13E+01

1.66E+01

COPPER

9 / 9

2.66E+01

1.22E+02

N/A

6.66E+01

1.22E+02

LEAD

9 / 9

1.85E+01

1.87E+02

N/A

1.13E+02

1.87E+02

MERCURY

9 / 9

3.90E-01

1.10E+00

N/A

7.50E-01

1.10E+00

NICKEL

9 / 9

1.61E+01

2.79E+01

N/A

2.23E+01

2.79E+01

SELENIUM

1 / 8

1.20E+00

1.20E+00

7.20E-01

-

2.30E+00

8.55E-01

2.30E+00

SILVER

7 / 8

4.60E-01

4.00E+00

2.00E-01

-

2.00E-01

1.04E+00

4.00E+00

THALLIUM

5 / 9

1.40E+00

4.10E+00

7.10E-01

-

2.00E+00

1.70E+00

4.10E+00

TIN

5 / 8

6.80E+00

1.43E+01

4.50E+00

-

1.61E+01

8.90E+00

1.61E+01

VANADIUM

9 / 9

1.28E+01

3.10E+01

N/A

1.91E+01

3.10E+01

ZINC

9 / 9

9.80E+01

2.80E+02

N/A

2.06E+02

2.80E+02

ORGANIC

TOTAL ORGANIC CARBON

128 / 129

1.67E+04

5.73E+05

6.03E+06

-

6.03E+06

1.29E+05

4.38E+05

INORGANICS

PERCENT SOLIDS

120 / 120

0.00E+00

8.35E+01

N/A

4.53E+01

6.68E+01

SULFIDE

1 / 8

6.88E+01

6.88E+01

9.20E+00

-

3.94E+01

1.82E+01

6.88E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-10

-------
Table B-ll

Sediment Chemistry Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

8 / 13

3.00E-02 - 8.60E+00

1.10E-01 - 2.00E+00

4.40E-01

8.60E+00

DIBENZOFURAN

2 / 13

5.80E-02 - 5.00E+00

3.80E-01 - 8.40E-01

7.60E-01

5.00E+00

1,2-DICHLOROBENZENE

2 / 13

8.70E-02 - 6.80E-01

3.80E-01 - 2.00E+00

7.00E-01

2.00E+00

1,3 -DICHLOROBENZENE

7 / 13

3.80E-02 - 2.10E-01

3.80E-01 - 2.00E+00

2.10E-01

2.00E+00

1,4-DICHLOROBENZENE

11 / 13

4.70E-02 - 8.30E-01

3.80E-01 - 4.60E-01

3.80E-01

8.30E-01

METHAPYRILENE

1 / 13

8.20E-01 - 8.20E-01

3.80E-01 - 2.00E+00

7.60E-01

2.00E+00

2-METHYLNAPHTHALENE

8 / 13

3.10E-02 - 2.20E+00

3.80E-01 - 8.20E-01

1.70E-01

2.20E+00

2-METHYLPHENOL (O-CRESOL)

1 / 13

4.30E-02 - 4.30E-02

3.80E-01 - 2.00E+00

7.60E-01

2.00E+00

4-METHYLPHENOL

5 / 13

5.00E-02 - 4.00E-01

3.80E-01 - 2.00E+00

4.60E-01

2.00E+00

PHENOL

1 / 13

2.90E-01 - 2.90E-01

3.80E-01 - 2.00E+00

7.60E-01

2.00E+00

PAHs

ACENAPHTHENE

4 / 13

6.00E-02 - 3.90E+00

3.80E-01 - 8.40E-01

6.20E-01

3.90E+00

ACENAPHTHYLENE

6 / 13

4.00E-02 - 4.30E+00

3.80E-01 - 8.40E-01

4.60E-01

4.30E+00

ANTHRACENE

7 / 13

2.40E-02 - 1.40E+01

4.60E-01 - 8.20E-01

4.60E-01

1.40E+01

BENZO(A)ANTHRACENE

13 / 13

2.50E-02 - 2.00E+01

N/A

2.90E-01

2.00E+01

BENZO(B)FLUORANTHENE

13 / 13

2.40E-02 - 1.40E+01

N/A

3.60E-01

1.40E+01

BENZO(K)FLUORANTHENE

13 / 13

2.80E-02 - 1.20E+01

N/A

3.90E-01

1.20E+01

BENZO(GHI)PERYLENE

13 / 13

3.00E-02 - 4.90E+00

N/A

2.30E-01

4.90E+00

BENZO(A)PYRENE

13 / 13

2.70E-02 - 1.50E+01

N/A

3.20E-01

1.50E+01

CHRYSENE

13 / 13

3.70E-02 - 1.40E+01

N/A

4.10E-01

1.40E+01

DIBENZO(A,H) ANTHRACENE

10 / 13

3.20E-02 - 2.30E+00

3.80E-01 - 7.70E-01

8.50E-02

2.30E+00

FLUORANTHENE

13 / 13

3.70E-02 - 4.00E+01

N/A

5.60E-01

4.00E+01

FLUORENE

5 / 13

4.80E-02 - 1.00E+01

3.80E-01 - 8.20E-01

5.70E-01

1.00E+01

INDENO(l,2,3-C,D)PYRENE

13 / 13

2.60E-02 - 5.00E+00

N/A

2.00E-01

5.00E+00

NAPHTHALENE

11 / 13

3.00E-02 - 6.00E+00

3.80E-01 - 4.60E-01

1.60E-01

6.00E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-ll

-------
Table B-ll

Sediment Chemistry Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PHENANTHRENE

11 / 13

9.60E-02 - 5.40E+01

4.60E-01 - 8.20E-01

4.40E-01

5.40E+01

PYRENE

13 / 13

5.40E-02 - 3.60E+01

N/A

6.30E-01

3.60E+01

TOTAL PAH (USING 0)

13 / 13

2.89E-01 - 2.55E+02

N/A

3.66E+00

2.55E+02

TOTAL PAH (USING DL)

13 / 13

2.71E+00 - 2.55E+02

N/A

5.93E+00

2.55E+02

TOTAL PAH (USING HALF DL)

13 / 13

1.76E+00 - 2.55E+02

N/A

4.78E+00

2.55E+02

TOTAL PAH (HIGH) (USING 0)

13 / 13

2.52E-01 - 1.23E+02

N/A

2.83E+00

1.23E+02

TOTAL PAH (HIGH) (USING DL)

13 / 13

7.12E-01 - 1.23E+02

N/A

3.60E+00

1.23E+02

TOTAL PAH (HIGH) (USING HALF DL)

13 / 13

4.82E-01 - 1.23E+02

N/A

3.22E+00

1.23E+02

TOTAL PAH (LOW) (USING 0)

13 / 13

3.70E-02 - 1.32E+02

N/A

9.04E-01

1.32E+02

TOTAL PAH (LOW) (USING DL)

13 / 13

1.66E+00 - 1.32E+02

N/A

2.80E+00

1.32E+02

TOTAL PAH (LOW) (USING HALF DL)

13 / 13

1.00E+00 - 1.32E+02

N/A

1.98E+00

1.32E+02

1,2,4-TRICHLOROBENZENE

7 / 13

3.50E-02 - 2.70E-01

3.80E-01 - 7.70E-01

2.70E-01

7.70E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

12 / 12

1.28E-05 - 2.74E-02

N/A

7.51E-03

2.74E-02

TOTAL DIOXINS (USING DL)

12 / 12

1.30E-05 - 2.74E-02

N/A

7.51E-03

2.74E-02

TOTAL DIOXINS (USING HALF DL)

12 / 12

1.29E-05 - 2.74E-02

N/A

7.51E-03

2.74E-02

TOTAL FURANS (USING 0)

12 / 12

3.46E-06 - 3.08E-02

N/A

4.47E-03

3.08E-02

TOTAL FURANS (USING DL)

12 / 12

4.18E-06 - 3.08E-02

N/A

4.47E-03

3.08E-02

TOTAL FURANS (USING HALF DL)

12 / 12

3.82E-06 - 3.08E-02

N/A

4.47E-03

3.08E-02

PCBS

AROCLOR-1248

5 / 230

2.15E+00 - 3.50E+01

9.31E-03 - 2.10E+01

5.12E-01

5.76E+00

AROCL OR-1254

21 / 230

1.10E-01 - 6.60E+01

9.31E-03 - 1.13E+01

5.13E-01

1.11E+01

AROCLOR-1260

208 / 230

1.81E-01 - 2.13E+02

5.00E-01 - 1.08E+00

5.97E+00

7.93E+01

PCB, TOTAL

209 / 230

1.81E-01 - 2.13E+02

5.00E-01 - 1.08E+00

6.00E+00

8.92E+01

METALS

ANTIMONY

9 / 13

1.20E+00 - 8.00E+00

3.30E-01 - 8.40E-01

1.90E+00

8.00E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-ll

-------
Table B-ll

Sediment Chemistry Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

ARSENIC

11 /

13

1.40E+00

1.07E+01

1.80E+00

-

1.90E+00

4.70E+00

1.07E+01

BARIUM

13 /

13

1.25E+01

1.72E+02

N/A

1.02E+02

1.72E+02

BERYLLIUM

12 /

13

1.50E-01

8.80E-01

2.80E-01

-

2.80E-01

5.90E-01

8.80E-01

CADMIUM

10 /

13

3.50E-01

1.66E+01

2.00E-02

-

4.00E-02

2.30E+00

1.66E+01

CHROMIUM

13 /

13

5.30E+00

4.59E+02

N/A

1.13E+02

4.59E+02

COBALT

13 /

13

4.70E+00

1.43E+01

N/A

1.15E+01

1.43E+01

COPPER

13 /

13

6.30E+00

2.68E+02

N/A

1.59E+02

2.68E+02

LEAD

13 /

13

5.40E+00

3.08E+02

N/A

1.50E+02

3.08E+02

MERCURY

10 /

13

1.90E-01

2.50E+00

2.00E-02

-

2.00E-02

8.90E-01

2.50E+00

NICKEL

13 /

13

8.60E+00

2.83E+01

N/A

2.28E+01

2.83E+01

SELENIUM

3 /

13

5.50E-01

1.10E+00

2.50E-01

-

7.10E-01

5.90E-01

1.10E+00

SILVER

10

13

3.80E-01

9.10E+00

1.00E-01

-

1.60E-01

3.00E+00

9.10E+00

THALLIUM

7

13

5.70E-01

2.40E+00

4.80E-01

-

1.90E+00

1.10E+00

2.40E+00

TIN

1 /

13

7.70E+00

3.32E+01

6.60E-01

-

1.46E+01

1.46E+01

3.32E+01

VANADIUM

13 /

13

6.40E+00

2.54E+01

N/A

1.51E+01

2.54E+01

ZINC

13 /

13

4.39E+01

9.48E+02

N/A

3.18E+02

9.48E+02

ORGANIC

TOTAL ORGANIC CARBON

234 / 244

2.26E+03

2.50E+05

1.23E+02

-

7.71E+03

2.64E+04

9.35E+04

INORGANICS

AMMONIA AS N

9 / 9

1.01E+01

9.69E+01

N/A

4.10E+01

9.69E+01

CYANIDE

1 / 9

1.40E+00

1.40E+00

8.40E-01

-

1.60E+00

1.40E+00

1.60E+00

PERCENT SOLIDS

193 /

193

0.00E+00

9.84E+01

N/A

6.11E+01

8.77E+01

SULFIDE

8 / 9

2.59E+01

1.22E+02

1.09E+01

-

1.09E+01

6.79E+01

1.22E+02

N/A = Not Applicable

Tables B-5- B-20.xls B-ll

-------
Table B-12

Sediment Chemistry Summary
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

PAHs

CHRYSENE

1 / 2

3.70E-02

3.70E-02

1.50E+00

1.50E+00

7.69E-01

1.50E+00

FLUORANTHENE

1 / 2

6.30E-02

6.30E-02

1.50E+00

1.50E+00

7.82E-01

1.50E+00

PHENANTHRENE

1 / 2

3.40E-02

3.40E-02

1.50E+00

1.50E+00

7.67E-01

1.50E+00

PYRENE

1 / 2

5.60E-02

5.60E-02

1.50E+00

1.50E+00

7.78E-01

1.50E+00

TOTAL PAH (USING 0)

2 / 2

0.00E+00

1.90E-01

N/A

9.50E-02

1.90E-01

TOTAL PAH (USING DL)

2 / 2

7.87E+00

2.40E+01

N/A

1.59E+01

2.40E+01

TOTAL PAH (USING HALF DL)

2 / 2

4.03E+00

1.20E+01

N/A

8.02E+00

1.20E+01

TOTAL PAH (HIGH) (USING 0)

2 / 2

0.00E+00

9.30E-02

N/A

4.65E-02

9.30E-02

TOTAL PAH (HIGH) (USING DL)

2 / 2

4.57E+00

1.35E+01

N/A

9.04E+00

1.35E+01

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

2.33E+00

6.75E+00

N/A

4.54E+00

6.75E+00

TOTAL PAH (LOW) (USING 0)

2 / 2

0.00E+00

9.70E-02

N/A

4.85E-02

9.70E-02

TOTAL PAH (LOW) (USING DL)

2 / 2

3.30E+00

1.05E+01

N/A

6.90E+00

1.05E+01

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

1.70E+00

5.25E+00

N/A

3.47E+00

5.25E+00

APP IX PESTICIDES

4,4'-DDD

1 / 2

8.00E-02

8.00E-02

4.00E-02

4.00E-02

6.00E-02

8.00E-02

4,4'-DDE

1 / 2

1.70E-01

1.70E-01

4.00E-02

4.00E-02

1.05E-01

1.70E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

2 / 2

1.67E-05

4.77E-05

N/A

3.22E-05

4.77E-05

TOTAL DIOXINS (USING DL)

2 / 2

1.75E-05

4.81E-05

N/A

3.28E-05

4.81E-05

TOTAL DIOXINS (USING HALF DL)

2 / 2

1.71E-05

4.79E-05

N/A

3.25E-05

4.79E-05

TOTAL FURANS (USING 0)

2 / 2

1.42E-05

3.19E-05

N/A

2.30E-05

3.19E-05

TOTAL FURANS (USING DL)

2 / 2

1.42E-05

3.19E-05

N/A

2.30E-05

3.19E-05

TOTAL FURANS (USING HALF DL)

2 / 2

1.42E-05

3.19E-05

N/A

2.30E-05

3.19E-05

PCBS

N/A = Not Applicable

Tables B-5- B-20.xls B-12

-------
Table B-12

Sediment Chemistry Summary
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA







Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

AROCL OR-1254

5 / 43

1.50E-01 - 2.89E+01

1.67E-02 - 2.75E+01

7.69E-01

1.00E+02

AROCLOR-1260

37 / 42

1.20E-01 - 2.84E+02

5.04E-01 - 1.43E+00

3.04E+00

1.00E+02

PCB, TOTAL

38 / 43

1.20E-01 - 2.84E+02

5.04E-01 - 1.43E+00

2.66E+00

1.00E+02

METALS









2.06E+01

ARSENIC

2 / 2

5.50E+00 - 1.04E+01

N/A

7.95E+00

1.26E+01

BARIUM

2 / 2

1.12E+02 - 1.33E+02

N/A

1.23E+02

8.85E+01

BERYLLIUM

2 / 2

6.20E-01 - 8.30E-01

N/A

7.25E-01

5.05E+01

CHROMIUM

2 / 2

2.05E+01 - 2.06E+01

N/A

2.06E+01

4.50E-01

COBALT

2 / 2

6.60E+00 - 1.26E+01

N/A

9.60E+00

2.46E+01

COPPER

2 / 2

1.38E+01 - 8.85E+01

N/A

5.12E+01

6.10E+00

LEAD

2 / 2

3.54E+01 - 5.05E+01

N/A

4.30E+01

3.60E+00

MERCURY

2 / 2

1.20E-01 - 4.50E-01

N/A

2.85E-01

2.28E+01

NICKEL

2 / 2

1.91E+01 - 2.46E+01

N/A

2.19E+01

1.63E+02

THALLIUM

2 / 2

5.10E+00 - 6.10E+00

N/A

5.60E+00



TIN

2 / 2

1.70E+00 - 3.60E+00

N/A

2.65E+00

3.05E+05

VANADIUM

2 / 2

1.81E+01 - 2.28E+01

N/A

2.05E+01



ZINC

2 / 2

1.15E+02 - 1.63E+02

N/A

1.39E+02

2.52E+01

ORGANIC









5.12E+01

TOTAL ORGANIC CARBON

31 / 31

1.57E+04 - 3.66E+05

N/A

7.33E+04

3.38E+05

INORGANICS

PERCENT SOLIDS

31 / 31

0.00E+00 - 9.78E+01

N/A

4.31E+01

9.50E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-12

-------
Table B-13

Sediment Chemistry Summary
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

5 /

15

4.70E-02

6.20E-01

8.00E-01

-

2.25E+00

1.00E+00

2.25E+00

DIBENZOFURAN

2 /

14

5.00E-02

1.00E-01

6.40E-01

-

2.25E+00

1.10E+00

2.25E+00

1,4-DICHLOROBENZENE

10 /

16

9.20E-02

2.60E-01

6.40E-01

-

2.25E+00

2.35E-01

2.25E+00

2-METHYLNAPHTHALENE

3 /

15

5.60E-02

2.80E-01

6.40E-01

-

2.40E+00

1.10E+00

2.40E+00

4-METHYLPHENOL

3 /

15

6.70E-02

1.10E+00

6.40E-01

-

2.40E+00

1.10E+00

2.40E+00

PAHs

ACENAPHTHYLENE

3 /

15

1.30E-01

6.10E-01

6.40E-01

-

2.40E+00

1.10E+00

2.40E+00

ACETOPHENONE

2 /

15

3.70E-01

5.10E-01

6.40E-01

-

2.40E+00

1.10E+00

2.40E+00

ANTHRACENE

3 /

14

5.20E-02

8.20E-01

6.40E-01

-

2.25E+00

1.10E+00

2.25E+00

BENZO(A)ANTHRACENE

12 /

15

5.70E-02

7.90E+00

6.40E-01

-

2.25E+00

1.80E-01

7.90E+00

BENZO(B)FLUORANTHENE

11 /

15

1.10E-01

7.80E+00

6.40E-01

-

2.25E+00

2.00E-01

7.80E+00

BENZO(K)FLUORANTHENE

11 /

15

1.10E-01

5.70E+00

6.40E-01

-

2.25E+00

2.30E-01

5.70E+00

BENZO(GHI)PERYLENE

11 /

15

7.20E-02

1.20E+00

6.40E-01

-

2.25E+00

1.90E-01

2.25E+00

BENZO(A)PYRENE

8 /

14

8.40E-02

4.40E+00

6.40E-01

-

2.25E+00

6.70E-01

4.40E+00

CHRYSENE

12

16

7.80E-02

8.20E+00

5.20E-01

-

2.25E+00

2.75E-01

8.20E+00

DIBENZO(A,H) ANTHRACENE

2 /

13

1.40E-01

7.60E-01

6.40E-01

-

2.25E+00

1.10E+00

2.25E+00

FLUORANTHENE

11

14

8.80E-02

1.50E+00

6.40E-01

-

2.25E+00

2.55E-01

2.25E+00

FLUORENE

2 /

14

7.20E-02

8.10E-02

6.40E-01

-

2.25E+00

1.10E+00

2.25E+00

INDENO(l,2,3-C,D)PYRENE

11 /

15

6.60E-02

1.50E+00

6.40E-01

-

2.25E+00

1.50E-01

2.25E+00

NAPHTHALENE

9 /

16

4.30E-02

2.60E-01

6.40E-01

-

2.25E+00

2.15E-01

2.25E+00

PHENANTHRENE

12 /

15

7.20E-02

1.00E+00

6.40E-01

-

2.25E+00

1.80E-01

2.25E+00

PYRENE

13 /

15

6.90E-02

1.50E+00

6.40E-01

-

2.25E+00

3.50E-01

2.25E+00

TOTAL PAH (USING 0)

16 /

16

0.00E+00

4.02E+01

N/A

1.64E+00

4.02E+01

TOTAL PAH (USING DL)

16 /

16

2.07E+00

4.12E+01

N/A

1.02E+01

4.12E+01

TOTAL PAH (USING HALF DL)

16 /

16

2.07E+00

4.07E+01

N/A

6.32E+00

4.07E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-13

-------
Table B-13

Sediment Chemistry Summary
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (HIGH) (USING 0)

16 / 16

0.00E+00 - 3.75E+01

N/A

1.18E+00

3.75E+01

TOTAL PAH (HIGH) (USING DL)

16 / 16

9.00E-01 - 3.75E+01

N/A

3.36E+00

3.75E+01

TOTAL PAH (HIGH) (USING HALF DL)

16 / 16

9.00E-01 - 3.75E+01

N/A

2.43E+00

3.75E+01

TOTAL PAH (LOW) (USING 0)

16 / 16

0.00E+00 - 2.86E+00

N/A

4.00E-01

2.86E+00

TOTAL PAH (LOW) (USING DL)

16 / 16

1.30E-01 - 1.58E+01

N/A

5.02E+00

1.58E+01

TOTAL PAH (LOW) (USING HALF DL)

16 / 16

1.30E-01 - 7.88E+00

N/A

3.12E+00

7.88E+00

1,2,4-TRICHLOROBENZENE

2 / 15

5.90E-02 - 8.40E-02

6.40E-01 - 2.40E+00

1.10E+00

2.40E+00

APPIX PESTICIDES

4,4'-DDT

1 / 14

5.20E-02 - 5.20E-02

1.20E-02 - 4.00E+00

2.20E-01

4.00E+00

HERBICIDES

2,4,5-T

1 / 7

5.20E-02 - 5.20E-02

1.20E-02 - 3.35E-02

2.20E-02

5.20E-02

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

18 / 18

1.42E-05 - 1.34E-02

N/A

1.43E-03

1.34E-02

TOTAL DIOXINS (USING DL)

18 / 18

1.52E-05 - 1.34E-02

N/A

1.44E-03

1.34E-02

TOTAL DIOXINS (USING HALF DL)

18 / 18

1.47E-05 - 1.34E-02

N/A

1.44E-03

1.34E-02

TOTAL FURANS (USING 0)

18 / 18

2.58E-05 - 1.87E-02

N/A

1.78E-03

1.87E-02

TOTAL FURANS (USING DL)

18 / 18

2.58E-05 - 1.87E-02

N/A

1.78E-03

1.87E-02

TOTAL FURANS (USING HALF DL)

18 / 18

2.58E-05 - 1.87E-02

N/A

1.78E-03

1.87E-02

PCBS

AROCLOR-1248

3 / 222

3.40E+00 - 2.70E+01

7.60E-02 - 1.20E+01

9.29E-01

5.63E+00

AROCL OR-1254

38 / 226

3.50E-01 - 3.40E+01

7.60E-02 - 1.06E+01

9.83E-01

1.28E+01

AROCLOR-1260

168 / 226

1.00E-01 - 1.30E+02

1.15E-01 - 2.27E+00

4.11E+00

7.73E+01

PCB, TOTAL

169 / 226

1.00E-01 - 1.60E+02

1.15E-01 - 2.27E+00

4.44E+00

7.87E+01

METALS

ANTIMONY

8 / 14

9.60E-01 - 2.50E+00

5.70E-01 - 3.80E+00

1.85E+00

3.80E+00

ARSENIC

13 / 15

1.90E+00 - 1.44E+01

1.60E+00 - 5.00E+00

7.10E+00

1.44E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-13

-------
Table B-13

Sediment Chemistry Summary
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

BARIUM

16 / 16

2.45E+01 - 2.15E+02

N/A

1.07E+02

2.15E+02

BERYLLIUM

16 / 16

1.80E-01 - 1.60E+00

N/A

8.90E-01

1.60E+00

CADMIUM

13 / 15

5.00E-02 - 7.80E+00

8.00E-02 - 9.00E-02

2.50E+00

7.80E+00

CHROMIUM

16 / 16

2.50E+00 - 2.54E+02

N/A

1.13E+02

2.54E+02

COBALT

16 / 16

3.00E+00 - 2.25E+01

N/A

1.51E+01

2.25E+01

COPPER

16 / 16

6.20E+00 - 2.50E+02

N/A

9.66E+01

2.50E+02

LEAD

16 / 16

2.00E+01 - 2.99E+02

N/A

1.12E+02

2.99E+02

MERCURY

15 / 16

1.00E-01 - 1.70E+00

1.05E-01 - 1.05E-01

1.05E+00

1.70E+00

NICKEL

16 / 16

4.60E+00 - 5.01E+01

N/A

2.49E+01

5.01E+01

SELENIUM

8 / 15

7.40E-01 - 3.90E+00

8.40E-01 - 2.10E+00

1.10E+00

3.90E+00

SILVER

10 / 15

6.60E-01 - 1.01E+01

1.80E-01 - 1.15E+00

2.00E+00

1.01E+01

THALLIUM

7 / 14

1.10E+00 - 7.90E+00

6.60E-01 - 2.35E+00

1.55E+00

7.90E+00

TIN

11 / 15

2.30E+00 - 2.29E+01

1.50E+00 - 1.06E+01

1.21E+01

2.29E+01

VANADIUM

16 / 16

6.60E+00 - 4.10E+01

N/A

2.27E+01

4.10E+01

ZINC

16 / 16

2.44E+01 - 5.20E+02

N/A

3.20E+02

5.20E+02

ORGANIC

TOTAL ORGANIC CARBON

184 / 189

5.69E+03 - 7.80E+05

2.12E+02 - 1.56E+06

1.07E+05

4.01E+05

INORGANICS

PERCENT SOLIDS

142 / 142

0.00E+00 - 9.56E+01

N/A

4.34E+01

8.54E+01

SULFIDE

7 / 13

1.01E+02 - 4.47E+02

9.60E+00 - 5.38E+01

1.01E+02

4.47E+02

N/A = Not Applicable

Tables B-5- B-20.xls B-13

-------
Table B-14

Sediment Chemistry Summary
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

2 / 2

5.30E-02 - 1.40E-01

N/A

9.65E-02

1.40E-01

1,3 -DICHLOROBENZENE

1 / 2

8.80E-02 - 8.80E-02

5.60E-01 - 5.60E-01

3.24E-01

5.60E-01

1,4-DICHLOROBENZENE

2 / 2

8.10E-02 - 4.50E-01

N/A

2.66E-01

4.50E-01

2-METHYLNAPHTHALENE

1 / 2

6.30E-02 - 6.30E-02

5.60E-01 - 5.60E-01

3.12E-01

5.60E-01

PAHs











ANTHRACENE

1 / 2

2.60E-02 - 2.60E-02

1.20E+00 - 1.20E+00

6.13E-01

1.20E+00

BENZO(A)ANTHRACENE

2 / 2

1.50E-01 - 1.90E-01

N/A

1.70E-01

1.90E-01

BENZO(B)FLUORANTHENE

2 / 2

1.70E-01 - 2.90E-01

N/A

2.30E-01

2.90E-01

BENZO(K)FLUORANTHENE

2 / 2

1.70E-01 - 2.60E-01

N/A

2.15E-01

2.60E-01

BENZO(GHI)PERYLENE

2 / 2

1.20E-01 - 2.50E-01

N/A

1.85E-01

2.50E-01

BENZO(A)PYRENE

2 / 2

1.70E-01 - 2.90E-01

N/A

2.30E-01

2.90E-01

CHRYSENE

2 / 2

2.40E-01 - 3.20E-01

N/A

2.80E-01

3.20E-01

DIBENZO(A,H) ANTHRACENE

1 / 2

3.50E-02 - 3.50E-02

1.20E+00 - 1.20E+00

6.18E-01

1.20E+00

FLUORANTHENE

2 / 2

2.60E-01 - 4.60E-01

N/A

3.60E-01

4.60E-01

INDENO(l,2,3-C,D)PYRENE

2 / 2

1.00E-01 - 2.20E-01

N/A

1.60E-01

2.20E-01

NAPHTHALENE

2 / 2

4.00E-02 - 1.30E-01

N/A

8.50E-02

1.30E-01

PHENANTHRENE

2 / 2

7.90E-02 - 3.30E-01

N/A

2.05E-01

3.30E-01

PYRENE

2 / 2

3.40E-01 - 5.40E-01

N/A

4.40E-01

5.40E-01

TOTAL PAH (USING 0)

2 / 2

1.90E+00 - 3.28E+00

N/A

2.59E+00

3.28E+00

TOTAL PAH (USING DL)

2 / 2

3.58E+00 - 9.28E+00

N/A

6.43E+00

9.28E+00

TOTAL PAH (USING HALF DL)

2 / 2

2.74E+00 - 6.28E+00

N/A

4.51E+00

6.28E+00

TOTAL PAH (HIGH) (USING 0)

2 / 2

1.50E+00 - 2.36E+00

N/A

1.93E+00

2.36E+00

TOTAL PAH (HIGH) (USING DL)

2 / 2

1.50E+00 - 3.56E+00

N/A

2.53E+00

3.56E+00

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

1.50E+00 - 2.96E+00

N/A

2.23E+00

2.96E+00

TOTAL PAH (LOW) (USING 0)

2 / 2

4.05E-01 - 9.20E-01

N/A

6.63E-01

9.20E-01

N/A = Not Applicable

Tables B-5- B-20.xls B-14

-------
Table B-14

Sediment Chemistry Summary
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (LOW) (USING DL)

2 / 2

2.09E+00

5.72E+00

N/A

3.90E+00

5.72E+00

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

1.25E+00

3.32E+00

N/A

2.28E+00

3.32E+00

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

2 / 2

1.93E-03

4.91E-03

N/A

3.42E-03

4.91E-03

TOTAL DIOXINS (USING DL)

2 / 2

1.96E-03

4.91E-03

N/A

3.44E-03

4.91E-03

TOTAL DIOXINS (USING HALF DL)

2 / 2

1.95E-03

4.91E-03

N/A

3.43E-03

4.91E-03

TOTAL FURANS (USING 0)

2 / 2

2.35E-03

5.83E-03

N/A

4.09E-03

5.83E-03

TOTAL FURANS (USING DL)

2 / 2

2.35E-03

5.83E-03

N/A

4.09E-03

5.83E-03

TOTAL FURANS (USING HALF DL)

2 / 2

2.35E-03

5.83E-03

N/A

4.09E-03

5.83E-03

PCBS

AROCLOR-1248

4 / 68

2.60E+00

3.87E+01

2.90E-02

-

2.62E+01

8.96E-01

1.28E+01

AROCL OR-1254

4 / 68

4.20E+00

2.99E+01

2.90E-02

-

2.62E+01

9.55E-01

1.10E+01

AROCLOR-1260

64 / 68

3.55E-01

2.05E+02

5.00E-01

-

5.15E-01

1.15E+01

1.06E+02

PCB, TOTAL

64 / 68

3.55E-01

2.05E+02

5.00E-01

-

5.15E-01

1.26E+01

1.06E+02

METALS

ANTIMONY

2 / 2

8.10E-01

2.00E+00

N/A

1.41E+00

2.00E+00

ARSENIC

2 / 2

3.50E+00

6.10E+00

N/A

4.80E+00

6.10E+00

BARIUM

2 / 2

4.48E+01

1.12E+02

N/A

7.84E+01

1.12E+02

BERYLLIUM

2 / 2

4.10E-01

7.80E-01

N/A

5.95E-01

7.80E-01

CADMIUM

2 / 2

7.40E-01

6.00E+00

N/A

3.37E+00

6.00E+00

CHROMIUM

2 / 2

6.74E+01

1.55E+02

N/A

1.11E+02

1.55E+02

COBALT

2 / 2

8.50E+00

1.23E+01

N/A

1.04E+01

1.23E+01

COPPER

2 / 2

6.27E+01

1.76E+02

N/A

1.19E+02

1.76E+02

LEAD

2 / 2

6.61E+01

1.76E+02

N/A

1.21E+02

1.76E+02

MERCURY

2 / 2

3.60E-01

1.10E+00

N/A

7.30E-01

1.10E+00

NICKEL

2 / 2

1.65E+01

2.63E+01

N/A

2.14E+01

2.63E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-14

-------
Table B-14

Sediment Chemistry Summary
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

SILVER

2 / 2

6.80E-01 - 6.50E+00

N/A

3.59E+00

6.50E+00

THALLIUM

2 / 2

1.20E+00 - 1.40E+00

N/A

1.30E+00

1.40E+00

TIN

2 / 2

6.80E+00 - 1.57E+01

N/A

1.13E+01

1.57E+01

VANADIUM

2 / 2

1.16E+01 - 2.08E+01

N/A

1.62E+01

2.08E+01

ZINC

2 / 2

1.67E+02 - 3.73E+02

N/A

2.70E+02

3.73E+02

ORGANIC

TOTAL ORGANIC CARBON

62 / 62

5.02E+03 - 2.84E+05

N/A

5.75E+04

1.40E+05

INORGANICS

PERCENT SOLIDS

43 / 43

1.96E+01 - 8.31E+01

N/A

3.76E+01

7.78E+01

SULFIDE

2 / 2

1.18E+02 - 1.22E+02

N/A

1.20E+02

1.22E+02

N/A = Not Applicable

Tables B-5- B-20.xls B-14

-------
Table B-15

Sediment Chemistry Summary
6ab SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PCBS

AROCLOR-1260

5 / 5

8.31E+00 - 1.39E+02

N/A

4.68E+01

1.39E+02

PCB, TOTAL

5 / 5

8.31E+00 - 1.39E+02

N/A

4.68E+01

1.39E+02

ORGANIC

TOTAL ORGANIC CARBON

4 / 4

6.78E+04 - 1.81E+05

N/A

9.23E+04

1.81E+05

INORGANICS

PERCENT SOLIDS

4 / 4

3.02E+01 - 3.54E+01

N/A

3.15E+01

3.54E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-15

-------
Table B-16

Sediment Chemistry Summary
6ab Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PCBS

AROCL OR-1254

2 / 39

2.20E-01

9.20E+01

5.00E-01

-

1.73E+01

8.27E-01

1.73E+01

AROCLOR-1260

38 / 39

3.64E-01

2.00E+02

8.27E-01

-

8.27E-01

1.08E+01

1.09E+02

PCB, TOTAL

38 / 39

3.64E-01

2.90E+02

8.27E-01

-

8.27E-01

1.08E+01

1.09E+02

ORGANIC

TOTAL ORGANIC CARBON

35 / 35

2.60E+04

1.26E+06

N/A

1.17E+05

4.44E+05

INORGANICS

PERCENT SOLIDS

28 / 28

0.00E+00

6.22E+01



N/A



2.45E+01

5.92E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-16

-------
Table B-17

Sediment Chemistry Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 1

4.90E-02 - 4.90E-02

N/A

4.90E-02

4.90E-02

1,4-DICHLOROBENZENE

1 / 1

1.40E-01 - 1.40E-01

N/A

1.40E-01

1.40E-01

2-METHYLNAPHTHALENE

1 / 1

8.60E-02 - 8.60E-02

N/A

8.60E-02

8.60E-02

4-METHYLPHENOL

1 / 1

8.80E-01 - 8.80E-01

N/A

8.80E-01

8.80E-01

PAHs

ACENAPHTHYLENE

1 / 1

3.80E-02 - 3.80E-02

N/A

3.80E-02

3.80E-02

ANTHRACENE

1 / 1

7.80E-02 - 7.80E-02

N/A

7.80E-02

7.80E-02

BENZO(A)ANTHRACENE

1 / 1

3.00E-01 - 3.00E-01

N/A

3.00E-01

3.00E-01

BENZO(B)FLUORANTHENE

1 / 1

3.40E-01 - 3.40E-01

N/A

3.40E-01

3.40E-01

BENZO(K)FLUORANTHENE

1 / 1

4.00E-01 - 4.00E-01

N/A

4.00E-01

4.00E-01

BENZO(GHI)PERYLENE

1 / 1

2.80E-01 - 2.80E-01

N/A

2.80E-01

2.80E-01

BENZO(A)PYRENE

1 / 1

3.60E-01 - 3.60E-01

N/A

3.60E-01

3.60E-01

CHRYSENE

1 / 1

3.80E-01 - 3.80E-01

N/A

3.80E-01

3.80E-01

DIBENZO(A,H) ANTHRACENE

1 / 1

7.90E-02 - 7.90E-02

N/A

7.90E-02

7.90E-02

FLUORANTHENE

1 / 1

5.20E-01 - 5.20E-01

N/A

5.20E-01

5.20E-01

FLUORENE

1 / 1

4.30E-02 - 4.30E-02

N/A

4.30E-02

4.30E-02

INDENO(l,2,3-C,D)PYRENE

1 / 1

2.70E-01 - 2.70E-01

N/A

2.70E-01

2.70E-01

NAPHTHALENE

1 / 1

1.60E-01 - 1.60E-01

N/A

1.60E-01

1.60E-01

PHENANTHRENE

1 / 1

3.60E-01 - 3.60E-01

N/A

3.60E-01

3.60E-01

PYRENE

1 / 1

5.90E-01 - 5.90E-01

N/A

5.90E-01

5.90E-01

TOTAL PAH (USING 0)

1 / 1

4.20E+00 - 4.20E+00

N/A

4.20E+00

4.20E+00

TOTAL PAH (USING DL)

1 / 1

4.97E+00 - 4.97E+00

N/A

4.97E+00

4.97E+00

TOTAL PAH (USING HALF DL)

1 / 1

4.58E+00 - 4.58E+00

N/A

4.58E+00

4.58E+00

TOTAL PAH (HIGH) (USING 0)

1 / 1

3.00E+00 - 3.00E+00

N/A

3.00E+00

3.00E+00

TOTAL PAH (HIGH) (USING DL)

1 / 1

3.00E+00 - 3.00E+00

N/A

3.00E+00

3.00E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-17

-------
Table B-17

Sediment Chemistry Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (HIGH) (USING HALF DL)

1 / 1

3.00E+00 - 3.00E+00

N/A

3.00E+00

3.00E+00

TOTAL PAH (LOW) (USING 0)

1 / 1

1.20E+00 - 1.20E+00

N/A

1.20E+00

1.20E+00

TOTAL PAH (LOW) (USING DL)

1 / 1

1.97E+00 - 1.97E+00

N/A

1.97E+00

1.97E+00

TOTAL PAH (LOW) (USING HALF DL)

1 / 1

1.58E+00 - 1.58E+00

N/A

1.58E+00

1.58E+00

1,2,4-TRICHLOROBENZENE

1 / 1

6.00E-02 - 6.00E-02

N/A

6.00E-02

6.00E-02

BULK DENSITY

DRY UNIT WEIGHT

1 / 1

3.50E+01 - 3.50E+01

N/A

3.50E+01

3.50E+01

TOTAL UNIT WEIGHT

1 / 1

7.67E+01 - 7.67E+01

N/A

7.67E+01

7.67E+01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

1 / 1

2.26E-03 - 2.26E-03

N/A

2.26E-03

2.26E-03

TOTAL DIOXINS (USING DL)

1 / 1

2.26E-03 - 2.26E-03

N/A

2.26E-03

2.26E-03

TOTAL DIOXINS (USING HALF DL)

1 / 1

2.26E-03 - 2.26E-03

N/A

2.26E-03

2.26E-03

TOTAL FURANS (USING 0)

1 / 1

2.48E-03 - 2.48E-03

N/A

2.48E-03

2.48E-03

TOTAL FURANS (USING DL)

1 / 1

2.48E-03 - 2.48E-03

N/A

2.48E-03

2.48E-03

TOTAL FURANS (USING HALF DL)

1 / 1

2.48E-03 - 2.48E-03

N/A

2.48E-03

2.48E-03

PCBS

AROCLOR-1248

6 / 116

2.57E+00 - 2.00E+01

5.70E-02 - 2.70E+01

1.00E+00

1.90E+01

AROCL OR-1254

19 / 117

7.00E-02 - 7.00E+01

6.60E-02 - 2.60E+01

1.02E+00

2.24E+01

AROCLOR-1260

100 / 117

2.30E-01 - 6.68E+02

5.00E-01 - 2.64E+00

9.46E+00

1.62E+02

PCB, TOTAL

100 / 117

3.00E-01 - 6.68E+02

5.00E-01 - 2.64E+00

1.07E+01

1.91E+02

METALS

ANTIMONY

1 / 1

2.80E+00 - 2.80E+00

N/A

2.80E+00

2.80E+00

ARSENIC

1 / 1

7.10E+00 - 7.10E+00

N/A

7.10E+00

7.10E+00

BARIUM

1 / 1

1.28E+02 - 1.28E+02

N/A

1.28E+02

1.28E+02

BERYLLIUM

1 / 1

8.10E-01 - 8.10E-01

N/A

8.10E-01

8.10E-01

CADMIUM

1 / 1

2.80E+00 - 2.80E+00

N/A

2.80E+00

2.80E+00

N/A = Not Applicable

Tables B-5- B-20.xls B-17

-------
Table B-17

Sediment Chemistry Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

CHROMIUM

1 / 1

1.91E+02 - 1.91E+02

N/A

1.91E+02

1.91E+02

COBALT

1 / 1

1.26E+01 - 1.26E+01

N/A

1.26E+01

1.26E+01

COPPER

1 / 1

1.88E+02 - 1.88E+02

N/A

1.88E+02

1.88E+02

LEAD

1 / 1

2.46E+02 - 2.46E+02

N/A

2.46E+02

2.46E+02

MERCURY

1 / 1

1.70E+00 - 1.70E+00

N/A

1.70E+00

1.70E+00

NICKEL

1 / 1

2.68E+01 - 2.68E+01

N/A

2.68E+01

2.68E+01

SELENIUM

1 / 1

6.90E-01 - 6.90E-01

N/A

6.90E-01

6.90E-01

SILVER

1 / 1

4.60E+00 - 4.60E+00

N/A

4.60E+00

4.60E+00

TIN

1 / 1

2.21E+01 - 2.21E+01

N/A

2.21E+01

2.21E+01

VANADIUM

1 / 1

2.28E+01 - 2.28E+01

N/A

2.28E+01

2.28E+01

ZINC

1 / 1

4.49E+02 - 4.49E+02

N/A

4.49E+02

4.49E+02

ORGANIC

TOTAL ORGANIC CARBON

72 / 72

5.83E+02 - 3.60E+05

N/A

8.00E+04

2.47E+05

INORGANICS

PERCENT SOLIDS

98 / 98

0.00E+00 - 9.39E+01

N/A

3.64E+01

8.50E+01

PERCENT WATER CONTENT

1 / 1

1.19E+02 - 1.19E+02

N/A

1.19E+02

1.19E+02

SULFIDE

1 / 1

4.74E+01 - 4.74E+01

N/A

4.74E+01

4.74E+01

N/A = Not Applicable

Tables B-5- B-20.xls B-17

-------
Table B-18

Sediment Chemistry Summary
Woods Pond Dam to Connecticut PCBs
Housatonic River Site, OU2 - Pittsfield, MA







Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

PCBS

PCB, TOTAL

348 / 496

1.70E-02 - 2.10E+01

1.70E-02 - 8.65E-01

6.20E-01

1.50E+01

Tables B-18 and B-19.xls B-18

-------
Table B-19

Sediment Chemistry Summary
Connecticut PCBs
Housatonic River Site, OU2 - Pittsfield, MA







Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

PCBS

PCB, TOTAL

253 / 344

2.00E-03 - 1.19E+01

0.00E+00 - 4.50E-01

1.90E-01

1.90E+00

Tables B-18 and B-19.xls B-19

-------
Table B-20

Sediment Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

ACETOPHENONE

3 / 20

6.60E-02 - 7.00E-01

3.80E-01 - 2.80E+00

4.55E-01

2.79E+00

BIS(2-ETHYLHEXYL) PHTHALATE

9 / 20

1.90E-02 - 3.20E-01

3.70E-01 - 2.80E+00

3.75E-01

2.79E+00

DIBENZOFURAN

1 / 20

3.40E-02 - 3.40E-02

3.70E-01 - 2.80E+00

4.55E-01

2.79E+00

DI-N-BUTYL PHTHALATE

1 / 20

1.10E-01 - 1.10E-01

3.70E-01 - 2.80E+00

4.55E-01

2.79E+00

1,4-DICHLOROBENZENE

1 / 20

6.20E-02 - 6.20E-02

3.70E-01 - 2.80E+00

4.55E-01

2.79E+00

2-METHYLNAPHTHALENE

3 / 20

2.50E-02 - 6.00E-02

3.70E-01 - 2.80E+00

4.15E-01

2.79E+00

4-METHYLPHENOL

1 / 20

1.60E+00 - 1.60E+00

3.70E-01 - 2.50E+00

4.80E-01

2.49E+00

N-NITROSO-DI-N-BUTYL AMINE

1 / 20

2.90E-02 - 2.90E-02

3.70E-01 - 2.80E+00

4.80E-01

2.79E+00

PAHs

ACENAPHTHENE

1 / 20

6.10E-02 - 6.10E-02

3.70E-01 - 2.80E+00

4.55E-01

2.79E+00

ANTHRACENE

8 / 20

3.30E-02 - 1.00E-01

3.70E-01 - 2.80E+00

4.15E-01

2.79E+00

BENZO(A)ANTHRACENE

11/20

3.00E-02 - 4.40E-01

3.70E-01 - 2.80E+00

3.35E-01

2.79E+00

BENZO(B)FLUORANTHENE

11/20

3.20E-02 - 4.30E-01

3.70E-01 - 2.80E+00

3.75E-01

2.79E+00

BENZO(K)FLUORANTHENE

11/20

3.60E-02 - 4.90E-01

3.70E-01 - 2.80E+00

3.30E-01

2.79E+00

BENZO(GHI)PERYLENE

11/20

2.60E-02 - 2.20E-01

3.70E-01 - 2.80E+00

2.20E-01

2.79E+00

BENZO(A)PYRENE

11/20

3.00E-02 - 4.60E-01

3.70E-01 - 2.80E+00

3.60E-01

2.79E+00

CHRYSENE

12 / 20

2.40E-02 - 5.40E-01

4.30E-01 - 2.80E+00

3.90E-01

2.79E+00

DIBENZO(A,H) ANTHRACENE

8 / 20

1.90E-02 - 8.30E-02

3.70E-01 - 2.80E+00

4.20E-01

2.79E+00

FLUORANTHENE

14 / 20

2.90E-02 - 1.00E+00

4.30E-01 - 2.80E+00

3.05E-01

2.79E+00

FLUORENE

6 / 20

1.80E-02 - 6.20E-02

3.70E-01 - 2.80E+00

4.25E-01

2.79E+00

INDENO(l,2,3-C,D)PYRENE

11/20

2.30E-02 - 2.30E-01

3.70E-01 - 2.80E+00

2.15E-01

2.79E+00

NAPHTHALENE

5 / 20

3.40E-02 - 1.10E-01

3.70E-01 - 2.80E+00

4.05E-01

2.79E+00

PHENANTHRENE

13 / 20

2.20E-02 - 7.50E-01

4.30E-01 - 2.80E+00

2.35E-01

2.79E+00

PYRENE

13 / 20

4.30E-02 - 1.10E+00

4.30E-01 - 2.80E+00

4.50E-01

2.79E+00

APP IX PESTICIDES

4,4'-DDD

3 / 21

3.00E-02 - 4.00E-01

3.80E-03 - 3.60E-01

6.50E-03

3.96E-01

NBA = No Benchmark Available

Tables B-5- B-20.xls B-20

-------
Table B-20

Sediment Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

4,4'-DDE

2 / 20

5.40E-02

5.40E-02

3.80E-03

-

3.60E-01

5.75E-03

3.45E-01

ALPHA-BHC

1 / 20

1.40E-02

1.40E-02

1.90E-03

-

1.80E-01

2.90E-03

1.72E-01

BETA-BHC

2 / 18

2.60E-03

3.50E-02

1.90E-03

-

1.80E-01

2.95E-03

1.80E-01

HERBICIDES

2,4,5-T

1 / 11

6.60E-03

6.60E-03

5.30E-03

-

3.70E-02

1.10E-02

3.70E-02

DIOXINS/FURANS

HPCDD (TOTAL)

19 / 20

1.51E-06

1.03E-03

2.66E-05

-

2.66E-05

2.10E-05

9.97E-04

HPCDF (TOTAL)

17 / 20

3.79E-07

7.05E-04

8.16E-07

-

1.68E-05

1.00E-05

6.89E-04

HXCDD (TOTAL)

15 / 20

9.48E-08

3.65E-04

6.00E-07

-

5.94E-06

5.51E-06

3.53E-04

HXCDF (TOTAL)

18 / 20

6.63E-07

7.69E-04

1.88E-06

-

2.17E-06

8.93E-06

7.55E-04

OCDD

18 / 20

5.31E-06

5.47E-03

5.21E-06

-

1.23E-04

5.64E-05

5.27E-03

OCDF

16 / 20

5.59E-07

6.36E-04

3.29E-06

-

8.77E-06

7.89E-06

6.13E-04

PECDD (TOTAL)

16 / 20

9.97E-08

8.60E-05

3.70E-07

-

1.35E-06

1.34E-06

8.30E-05

PECDF (TOTAL)

18 / 20

6.60E-07

1.23E-03

1.30E-06

-

1.73E-06

4.62E-06

1.20E-03

TCDD (TOTAL)

12 / 20

5.89E-07

4.82E-05

4.12E-08

-

2.83E-06

1.73E-06

4.68E-05

TCDF (TOTAL)

17 / 20

2.26E-07

1.97E-03

3.20E-07

-

1.38E-06

4.75E-06

1.89E-03

PCBS

AROCL OR-1254

2 / 17

4.00E-02

1.60E-01

1.90E-02

-

6.80E-01

1.10E-01

6.80E-01

AROCLOR-1260

2 / 16

4.90E-02

1.00E+00

1.90E-02

-

6.80E-01

9.05E-02

1.00E+00

PCB, TOTAL

4 / 17

2.00E-02

1.00E+00

1.90E-02

-

6.80E-01

1.10E-01

1.00E+00

METALS

ANTIMONY

13 / 20

3.70E-01

4.80E+00

2.40E-01

-

3.20E+00

7.10E-01

4.78E+00

ARSENIC

15 / 20

1.60E+00

5.80E+00

1.00E+00

-

3.90E+00

2.85E+00

5.73E+00

BARIUM

20 / 20

8.60E+00

2.46E+02

N/A

5.54E+01

2.45E+02

BERYLLIUM

19 / 20

2.30E-01

1.90E+00

9.00E-02

-

9.00E-02

4.15E-01

1.85E+00

CADMIUM

6 / 20

1.80E-01

2.10E+00

2.00E-02

-

1.50E+00

8.00E-02

2.07E+00

CHROMIUM

20 / 20

5.30E+00

1.39E+02

N/A

1.13E+01

1.35E+02

NBA = No Benchmark Available

Tables B-5- B-20.xls B-20

-------
Table B-20

Sediment Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

COBALT

20 / 20

4.00E+00

1.55E+01

N/A

6.75E+00

1.53E+01

COPPER

20 / 20

5.80E+00

1.22E+02

N/A

1.32E+01

1.19E+02

LEAD

20 / 20

5.00E+00

2.15E+02

N/A

2.29E+01

2.11E+02

MERCURY

9 / 20

2.00E-02

4.30E-01

2.00E-02

-

1.10E-01

4.00E-02

4.17E-01

NICKEL

20 / 20

3.90E+00

2.10E+01

N/A

1.12E+01

2.10E+01

SELENIUM

1 / 20

9.40E-01

9.40E-01

1.70E-01

-

2.80E+00

4.75E-01

2.78E+00

SILVER

5 / 20

4.80E-01

2.70E+00

8.00E-02

-

1.30E+00

2.15E-01

2.63E+00

THALLIUM

11/20

3.70E-01

3.40E+00

3.90E-01

-

2.70E+00

1.10E+00

3.37E+00

TIN

6 / 20

2.20E+00

1.56E+01

8.50E-01

-

6.90E+00

2.40E+00

1.52E+01

VANADIUM

20 / 20

4.80E+00

3.33E+01

N/A

1.41E+01

3.33E+01

ZINC

20 / 20

2.51E+01

2.04E+02

N/A

5.21E+01

2.04E+02

ORGANIC

TOTAL ORGANIC CARBON

20 / 23

2.54E+03

2.86E+05

1.18E+02

-

1.29E+02

3.53E+04

2.75E+05

INORGANICS

PERCENT SOLIDS

9 / 9

1.26E+01

7.86E+01

N/A

2.36E+01

7.86E+01

SULFIDE

8 / 17

5.60E+00

3.00E+02

6.20E+00

-

4.61E+01

1.72E+01

3.00E+02

NBA = No Benchmark Available

Tables B-5- B-20.xls B-20

-------
Table B-21

Surface Water Chemistry Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median

(HS/L)

95th Percentile

(HS/L)

APPIX VOLATILES

ACETONE

4 / 8

2.50E+00

2.70E+00

2.50E+00

-

4.50E+00

2.60E+00

4.50E+00

CHLOROBENZENE

6 / 10

8.80E-01

1.90E+00

5.00E-01

-

5.00E-01

8.80E-01

1.90E+00

TRICHLOROETHYLENE (TCE)

3 / 10

6.10E-01

7.70E-01

5.00E-01

-

5.00E-01

5.00E-01

7.70E-01

VINYL CHLORIDE

4 / 10

3.60E-01

9.30E-01

5.00E-01

-

5.00E-01

5.00E-01

9.30E-01

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

2 / 32

1.00E+01

1.40E+01

1.00E+01

-

1.10E+01

1.00E+01

1.21E+01

DIETHYL PHTHALATE

2 / 32

6.25E-01

7.00E-01

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

1,4-DICHLOROBENZENE

2 / 32

5.00E-01

5.50E-01

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

PAHs

ACENAPHTHENE

4 / 32

3.90E-02

7.00E-02

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

ACENAPHTHYLENE

1 / 32

1.10E-02

1.10E-02

2.00E-02

-

1.10E+01

1.00E+01

1.10E+01

ACETOPHENONE

1 / 32

2.00E+00

2.00E+00

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

BENZO(B)FLUORANTHENE

1 / 32

1.10E-02

1.10E-02

2.00E-02

-

1.10E+01

1.00E+01

1.10E+01

FLUORANTHENE

4 / 32

1.70E-02

2.80E-02

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

FLUORENE

2 / 32

1.00E-02

1.00E-02

2.00E-02

-

1.10E+01

1.00E+01

1.10E+01

NAPHTHALENE

4 / 32

3.10E-02

1.10E-01

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

PHENANTHRENE

4 / 32

1.20E-02

1.80E-02

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

PYRENE

5 / 32

2.50E-02

5.00E-01

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

TOTAL PAH (USING 0)

32 / 32

0.00E+00

5.00E-01

N/A

0.00E+00

3.48E-01

TOTAL PAH (USING DL)

32 / 32

1.51E+02

1.76E+02

N/A

1.60E+02

1.76E+02

TOTAL PAH (USING HALF DL)

32 / 32

7.55E+01

8.80E+01

N/A

8.00E+01

8.80E+01

TOTAL PAH (HIGH) (USING 0)

32 / 32

0.00E+00

5.00E-01

N/A

0.00E+00

2.02E-01

TOTAL PAH (HIGH) (USING DL)

32 / 32

8.05E+01

9.90E+01

N/A

9.00E+01

9.90E+01

TOTAL PAH (HIGH) (USING HALF DL)

32 / 32

4.05E+01

4.95E+01

N/A

4.50E+01

4.95E+01

TOTAL PAH (LOW) (USING 0)

32 / 32

0.00E+00

2.29E-01

N/A

0.00E+00

1.88E-01

N/A = Not Applicable

Tables B-21- B-30.xls B-21

-------
Table B-21

Surface Water Chemistry Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

TOTAL PAH (LOW) (USING DL)

32 / 32

7.00E+01

7.70E+01

N/A

7.00E+01

7.70E+01

TOTAL PAH (LOW) (USING HALF DL)

32 / 32

3.50E+01

3.85E+01

N/A

3.50E+01

3.85E+01

1,2,4-TRICHLOROBENZENE, FILTERED

22 / 22

7.80E-03

1.10E-01

N/A

4.05E-02

1.10E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

32 / 32

0.00E+00

5.54E-03

N/A

3.60E-06

2.23E-03

TOTAL DIOXINS (USING DL)

32 / 32

1.37E-05

5.54E-03

N/A

2.76E-05

2.24E-03

TOTAL DIOXINS (USING HALF DL)

32 / 32

9.07E-06

5.54E-03

N/A

1.82E-05

2.23E-03

TOTAL FURANS (USING 0)

32 / 32

0.00E+00

1.20E-03

N/A

6.18E-05

8.48E-04

TOTAL FURANS (USING DL)

32 / 32

7.02E-06

1.20E-03

N/A

7.19E-05

8.48E-04

TOTAL FURANS (USING HALF DL)

32 / 32

4.28E-06

1.20E-03

N/A

6.99E-05

8.48E-04

PCBS

AROCLOR-1242

3 / 30

1.45E-02

2.30E-02

1.20E-02

-

6.20E-02

1.30E-02

4.22E-02

AROCL OR-1254

16 / 30

1.40E-02

5.70E-02

1.20E-02

-

6.20E-02

1.75E-02

5.93E-02

AROCLOR-1260

19 / 29

1.30E-02

5.00E-01

1.20E-02

-

1.30E-02

2.80E-02

3.35E-01

PCB, TOTAL

22 / 30

1.30E-02

5.00E-01

1.20E-02

-

1.30E-02

4.30E-02

3.41E-01

PCBS - FILTERED

AROCLOR-1254, DISSOLVED

1 / 30

1.80E-01

1.80E-01

1.20E-02

-

1.30E-01

1.40E-02

1.53E-01

AROCLOR-1260, DISSOLVED

1 / 28

3.30E-02

3.30E-02

1.20E-02

-

3.00E-02

1.40E-02

3.17E-02

TOTAL PCB, DISSOLVED

2 / 30

2.10E-01

2.60E-01

1.20E-02

-

3.00E-02

1.40E-02

2.33E-01

METALS

ARSENIC

2 / 32

3.00E+00

3.50E+00

1.80E+00

-

6.00E+00

2.90E+00

6.00E+00

BARIUM

32 / 32

9.80E+00

3.10E+01

N/A

1.79E+01

2.96E+01

BERYLLIUM

1 / 32

1.50E-01

1.50E-01

1.00E-01

-

4.00E-01

1.15E-01

4.00E-01

CADMIUM

2 / 32

3.00E-01

3.90E-01

2.00E-01

-

9.00E-01

3.30E-01

9.00E-01

CALCIUM

22 / 22

1.26E+04

4.32E+04

N/A

3.40E+04

4.30E+04

CHROMIUM

3 / 32

8.50E-01

4.90E+00

7.00E-01

-

2.90E+00

1.30E+00

3.60E+00

N/A = Not Applicable

Tables B-21- B-30.xls B-21

-------
Table B-21

Surface Water Chemistry Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile

(HS/L)

COBALT

1 / 32

2.30E+00

2.30E+00

1.50E+00

-

3.90E+00

2.30E+00

3.90E+00

COPPER

6 / 32

9.70E-01

6.70E+00

9.00E-01

-

5.30E+00

2.20E+00

5.79E+00

LEAD

4 / 32

1.20E+00

1.09E+01

8.00E-01

-

4.10E+00

1.85E+00

8.69E+00

MAGNESIUM

22 / 22

4.07E+03

1.73E+04

N/A

1.23E+04

1.73E+04

MERCURY

4 / 32

1.00E-01

1.10E-01

1.00E-01

-

1.00E-01

1.00E-01

1.10E-01

NICKEL

1 / 32

4.40E+00

4.40E+00

8.00E-01

-

4.70E+00

2.30E+00

4.51E+00

SELENIUM

1 / 32

4.20E+00

4.20E+00

1.80E+00

-

4.90E+00

3.40E+00

4.90E+00

SILVER

1 / 32

1.30E+00

1.30E+00

9.00E-01

-

3.10E+00

1.40E+00

3.10E+00

VANADIUM

3 / 32

3.00E+00

4.90E+00

1.50E+00

-

4.00E+00

2.40E+00

4.32E+00

ZINC

12 / 32

3.10E+00

3.22E+02

1.70E+00

-

1.22E+01

5.25E+00

1.80E+02

METALS - FILTERED

BARIUM, DISSOLVED

30 / 30

6.80E+00

3.18E+01

N/A

1.56E+01

3.11E+01

BERYLLIUM, DISSOLVED

1 / 30

1.20E-01

1.20E-01

1.00E-01

-

4.00E-01

1.15E-01

4.00E-01

CADMIUM, DISSOLVED

1 / 30

3.10E-01

3.10E-01

2.00E-01

-

9.00E-01

3.00E-01

9.00E-01

CALCIUM, DISSOLVED

20 / 20

1.12E+04

4.28E+04

N/A

3.28E+04

4.27E+04

CHROMIUM, DISSOLVED

1 / 30

1.70E+00

1.70E+00

7.00E-01

-

2.90E+00

1.00E+00

2.90E+00

COBALT, DISSOLVED

2 / 30

2.20E+00

2.20E+00

1.50E+00

-

3.90E+00

2.20E+00

3.90E+00

COPPER, DISSOLVED

3 / 30

1.10E+00

3.00E+00

9.00E-01

-

3.70E+00

1.70E+00

3.70E+00

LEAD, DISSOLVED

1 / 30

1.10E+00

1.10E+00

8.00E-01

-

2.90E+00

1.30E+00

2.85E+00

MAGNESIUM, DISSOLVED

20 / 20

3.81E+03

1.75E+04

N/A

1.17E+04

1.75E+04

MERCURY, FILTERED

3 / 30

1.00E-01

1.00E-01

1.00E-01

-

1.00E-01

1.00E-01

1.00E-01

SELENIUM, DISSOLVED

1 / 30

4.00E+00

4.00E+00

1.80E+00

-

4.90E+00

3.60E+00

4.90E+00

SILVER, DISSOLVED

3 / 30

1.20E+00

2.90E+00

9.00E-01

-

4.00E+00

1.35E+00

3.51E+00

ZINC, DISSOLVED

7 / 30

2.60E+00

1.45E+01

1.00E+00

-

1.18E+01

4.20E+00

1.30E+01

ORGANIC

TOTAL ORGANIC CARBON

16 / 16

2.70E+03

6.90E+03

N/A

4.75E+03

6.90E+03

N/A = Not Applicable

Tables B-21- B-30.xls B-21

-------
Table B-21

Surface Water Chemistry Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

TOTAL ORGANIC CARBON, DISSOLVED

30 / 30

2.70E+03

1.26E+04

N/A

5.03E+03

1.17E+04

INORGANICS

ALKALINITY

30 / 30

3.95E+04

1.70E+05

N/A

1.19E+05

1.70E+05

AMMONIA AS N

30 / 30

4.00E+01

7.75E+02

N/A

1.55E+02

7.61E+02

BIOLOGICAL OXYGEN DEMAND 5 DAY

12 / 32

2.00E+03

3.50E+03

2.00E+03

-

3.00E+03

2.00E+03

3.44E+03

HARDNESS

30 / 30

5.00E+04

1.82E+05

N/A

1.31E+05

1.79E+05

HARDNESS, DISSOLVED

1 / 1

9.00E+04

9.00E+04

N/A

9.00E+04

9.00E+04

NITRATE AND NITRITE AS N

30 / 30

2.00E+01

1.70E+03

N/A

5.28E+02

1.70E+03

NITRITE AS N

19 / 30

5.00E+00

1.25E+02

5.00E+00

-

5.00E+00

8.50E+00

1.05E+02

ORGANIC CARBON, PARTICULATE

3 / 14

5.00E+02

1.50E+03

5.00E+02

-

5.00E+02

5.00E+02

1.50E+03

ORTHOPHOSPHATE AS P

7 / 30

1.00E+01

9.00E+01

1.00E+01

-

1.00E+01

1.00E+01

9.00E+01

PHOSPHATE, AS P

2 / 2

2.00E+01

3.00E+01

N/A

2.50E+01

3.00E+01

PHOSPHATE, TOTAL AS P

25 / 28

1.00E+01

3.10E+02

1.00E+01

-

1.00E+01

2.00E+01

2.74E+02

SULFIDE

7 / 26

5.00E+02

8.00E+02

5.00E+02

-

1.00E+03

8.00E+02

1.00E+03

TKN

29 / 30

2.40E+02

1.10E+03

2.40E+02

-

2.40E+02

7.00E+02

1.05E+03

TOTAL DISSOLVED SOLIDS

30 / 30

8.70E+04

8.13E+05

N/A

2.02E+05

6.12E+05

TOTAL SUSPENDED SOLIDS

30 / 30

8.00E+02

1.27E+05

N/A

3.10E+03

8.42E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-21

-------
Table B-22

Surface Water Chemistry Summary
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median

(HS/L)

95th Percentile
(Jig/L)

ORGANIC

TOTAL ORGANIC CARBON

29 / 29

2.10E+03

7.50E+03

N/A

3.70E+03

7.50E+03

TOTAL ORGANIC CARBON, DISSOLVED

29 / 29

2.80E+03

7.30E+03

N/A

3.90E+03

7.15E+03

INORGANICS

AMMONIA AS N

21 / 29

2.70E+01

1.40E+02

2.00E+01

-

1.00E+02

7.00E+01

1.25E+02

BIOLOGICAL OXYGEN DEMAND 5 DAY

13 / 29

1.00E+03

6.00E+03

2.00E+03

-

3.00E+03

2.00E+03

4.90E+03

HYDROLYZABLE PHOSPHATE, AS P

19 / 26

1.00E+01

1.80E+02

1.00E+01

-

1.00E+01

2.50E+01

1.77E+02

NITRATE AND NITRITE AS N

29 / 29

1.00E+02

1.10E+03

N/A

1.90E+02

1.07E+03

NITRITE AS N

3 / 29

6.00E+00

7.00E+00

5.00E+00

-

2.50E+02

5.00E+00

2.50E+02

ORGANIC CARBON, PARTICULATE

8 / 29

5.00E+02

2.40E+03

5.00E+02

-

5.00E+02

5.00E+02

2.00E+03

ORGANIC PHOSPHATE, AS P

22 / 29

1.00E+01

2.70E+02

1.00E+01

-

5.00E+01

2.00E+01

2.70E+02

ORTHOPHOSPHATE AS P

16 / 29

1.00E+01

7.00E+01

1.00E+01

-

2.50E+02

1.00E+01

2.50E+02

PHOSPHATE, TOTAL AS P

23 / 29

1.00E+01

4.50E+02

1.00E+01

-

5.00E+01

5.00E+01

4.35E+02

TKN

25 / 29

1.80E+02

1.00E+03

2.40E+02

-

2.40E+02

4.20E+02

1.00E+03

TOTAL SUSPENDED SOLIDS

130 / 162

1.00E+03

3.65E+05

5.00E+03

-

5.00E+03

5.60E+03

1.09E+05

N/A = Not Applicable

Tables B-21- B-30.xls B-22

-------
Table B-23

Surface Water Chemistry Summary
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

9 / 9

0.00E+00

7.10E-06

N/A

0.00E+00

7.10E-06

TOTAL DIOXINS (USING DL)

9 / 9

1.76E-05

5.08E-05

N/A

2.43E-05

5.08E-05

TOTAL DIOXINS (USING HALF DL)

9 / 9

8.82E-06

2.67E-05

N/A

1.21E-05

2.67E-05

TOTAL FURANS (USING 0)

9 / 9

0.00E+00

6.84E-05

N/A

0.00E+00

6.84E-05

TOTAL FURANS (USING DL)

9 / 9

9.53E-06

7.26E-05

N/A

1.28E-05

7.26E-05

TOTAL FURANS (USING HALF DL)

9 / 9

4.76E-06

7.05E-05

N/A

6.79E-06

7.05E-05

PCBS

AROCL OR-1254

3 / 8

1.50E-02

2.90E-02

1.30E-02

-

2.70E-02

1.45E-02

2.90E-02

AROCLOR-1260

4 / 8

2.80E-02

1.10E-01

1.30E-02

-

1.40E-02

2.10E-02

1.10E-01

PCB, TOTAL

4 / 8

4.30E-02

1.40E-01

1.30E-02

-

1.40E-02

2.85E-02

1.40E-01

N/A = Not Applicable

Tables B-21- B-30.xls B-23

-------
Table B-24

Surface Water Chemistry Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median

(HS/L)

95th Percentile
(Jig/L)

APPIX VOLATILES

ACETONE

2 / 6

3.70E+00

4.20E+00

2.50E+00

-

5.10E+00

3.95E+00

5.10E+00

BROMODICHLOROMETHANE

4 / 9

5.50E-01

3.20E+00

5.00E-01

-

5.00E-01

5.00E-01

3.20E+00

CHLOROBENZENE

2 / 9

7.80E-01

8.90E-01

5.00E-01

-

5.00E-01

5.00E-01

8.90E-01

CHLOROFORM

5 / 9

2.00E+00

3.80E+00

5.00E-01

-

5.00E-01

2.00E+00

3.80E+00

DIBROMOCHLOROMETHANE

2 / 9

2.00E+00

2.00E+00

5.00E-01

-

5.00E-01

5.00E-01

2.00E+00

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 29

1.00E+00

1.00E+00

1.00E+01

-

1.10E+01

1.00E+01

1.05E+01

PAHs

ACENAPHTHENE

3 / 29

1.80E-02

2.10E-02

1.00E+01

-

1.10E+01

1.00E+01

1.05E+01

BENZO(A)ANTHRACENE

1 / 29

1.10E-02

1.10E-02

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

BENZO(B)FLUORANTHENE

2 / 29

1.50E-02

1.50E-02

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

BENZO(K)FLUORANTHENE

1 / 29

1.20E-02

1.20E-02

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

BENZO(GHI)PERYLENE

2 / 29

1.20E-02

1.50E-02

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

BENZO(A)PYRENE

1 / 29

1.40E-02

1.40E-02

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

CHRYSENE

2 / 29

1.30E-02

5.00E-01

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

FLUORANTHENE

5 / 29

1.50E-02

8.00E-01

1.00E+01

-

1.10E+01

1.00E+01

1.05E+01

INDENO(l,2,3-C,D)PYRENE

1 / 29

1.00E-02

1.00E-02

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

NAPHTHALENE

3 / 29

1.50E-02

2.40E-02

1.00E+01

-

1.10E+01

1.00E+01

1.05E+01

PHENANTHRENE

3 / 29

1.00E-02

6.00E-01

2.00E-02

-

1.10E+01

1.00E+01

1.05E+01

PYRENE

5 / 29

2.00E-02

8.00E-01

1.00E+01

-

1.10E+01

1.00E+01

1.05E+01

TOTAL PAH (USING 0)

29 / 29

0.00E+00

2.70E+00

N/A

0.00E+00

1.95E+00

TOTAL PAH (USING DL)

29 / 29

1.23E+02

1.76E+02

N/A

1.60E+02

1.68E+02

TOTAL PAH (USING HALF DL)

29 / 29

6.27E+01

8.80E+01

N/A

8.00E+01

8.40E+01

TOTAL PAH (HIGH) (USING 0)

29 / 29

0.00E+00

1.30E+00

N/A

0.00E+00

9.50E-01

TOTAL PAH (HIGH) (USING DL)

29 / 29

7.13E+01

9.90E+01

N/A

9.00E+01

9.45E+01

N/A = Not Applicable

Tables B-21- B-30.xls B-24

-------
Table B-24

Surface Water Chemistry Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

TOTAL PAH (HIGH) (USING HALF DL)

29 / 29

3.63E+01

4.95E+01

N/A

4.50E+01

4.73E+01

TOTAL PAH (LOW) (USING 0)

29 / 29

0.00E+00

1.40E+00

N/A

0.00E+00

1.00E+00

TOTAL PAH (LOW) (USING DL)

29 / 29

5.14E+01

7.70E+01

N/A

7.00E+01

7.35E+01

TOTAL PAH (LOW) (USING HALF DL)

29 / 29

2.64E+01

3.85E+01

N/A

3.50E+01

3.68E+01

1,2,4-TRICHLOROBENZENE, FILTERED

24 / 26

4.50E-03

9.00E-02

9.50E-03

-

1.20E-02

1.90E-02

7.99E-02

APPIX PESTICIDES

ENDOSULFANI

1 / 29

1.00E-01

1.00E-01

5.00E-02

-

5.70E-02

5.00E-02

7.85E-02

HERBICIDES

2,4,5-T

1 / 28

1.20E-01

1.20E-01

9.50E-02

-

1.20E-01

9.50E-02

1.20E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

29 / 29

0.00E+00

8.11E-04

N/A

1.60E-05

6.87E-04

TOTAL DIOXINS (USING DL)

29 / 29

1.61E-05

8.30E-04

N/A

4.18E-05

7.05E-04

TOTAL DIOXINS (USING HALF DL)

29 / 29

1.29E-05

8.21E-04

N/A

3.13E-05

6.96E-04

TOTAL FURANS (USING 0)

29 / 29

0.00E+00

1.78E-03

N/A

7.22E-05

1.44E-03

TOTAL FURANS (USING DL)

29 / 29

8.80E-06

1.78E-03

N/A

8.76E-05

1.44E-03

TOTAL FURANS (USING HALF DL)

29 / 29

5.90E-06

1.78E-03

N/A

8.04E-05

1.44E-03

PCBS

AROCLOR-1242

7 / 59

2.70E-02

8.20E-02

1.20E-02

-

1.30E-01

1.30E-02

6.60E-02

AROCL OR-1254

43 / 60

1.30E-02

1.80E-01

1.20E-02

-

6.60E-02

2.50E-02

6.89E-02

AROCLOR-1260

54 / 60

1.40E-02

1.00E+00

1.20E-02

-

3.00E-02

5.85E-02

3.59E-01

PCB, TOTAL

53 / 59

1.40E-02

1.20E+00

1.20E-02

-

3.00E-02

8.70E-02

4.30E-01

PCBS - FILTERED

AROCLOR-1254, DISSOLVED

2 / 60

1.10E-01

2.40E-01

1.20E-02

-

2.70E-02

1.40E-02

2.64E-02

AROCLOR-1260, DISSOLVED

6 / 60

1.50E-02

2.30E-01

1.20E-02

-

1.50E-02

1.40E-02

3.81E-02

TOTAL PCB, DISSOLVED

6 / 60

1.50E-02

3.20E-01

1.20E-02

-

1.50E-02

1.40E-02

2.00E-02

METALS

N/A = Not Applicable

Tables B-21- B-30.xls B-24

-------
Table B-24

Surface Water Chemistry Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

ARSENIC

1 / 29

3.30E+00

3.30E+00

1.80E+00

-

6.00E+00

2.90E+00

6.00E+00

BARIUM

29 / 29

1.17E+01

3.17E+01

N/A

1.64E+01

2.88E+01

BERYLLIUM

1 / 29

1.10E-01

1.10E-01

1.00E-01

-

4.00E-01

1.00E-01

4.00E-01

CADMIUM

1 / 29

4.00E-01

4.00E-01

2.00E-01

-

9.00E-01

3.00E-01

9.00E-01

CALCIUM

20 / 20

1.44E+04

3.94E+04

N/A

3.34E+04

3.94E+04

CHROMIUM

4 / 29

1.50E+00

6.40E+00

7.00E-01

-

2.10E+00

1.00E+00

5.70E+00

COBALT

2 / 29

1.80E+00

2.80E+00

1.50E+00

-

3.90E+00

2.00E+00

3.65E+00

COPPER

17 / 29

1.20E+00

8.60E+00

1.00E+00

-

5.60E+00

3.40E+00

8.35E+00

LEAD

4 / 29

1.20E+00

1.43E+01

8.00E-01

-

3.30E+00

1.70E+00

1.25E+01

MAGNESIUM

20 / 20

6.32E+03

1.61E+04

N/A

1.20E+04

1.61E+04

MERCURY

3 / 29

1.00E-01

1.00E-01

1.00E-01

-

1.00E-01

1.00E-01

1.00E-01

NICKEL

1 / 29

4.40E+00

4.40E+00

1.40E+00

-

4.70E+00

2.30E+00

4.60E+00

SELENIUM

1 / 29

2.90E+00

2.90E+00

1.80E+00

-

4.90E+00

3.60E+00

4.90E+00

SILVER

2 / 29

2.70E+00

2.80E+00

9.00E-01

-

3.10E+00

1.20E+00

2.95E+00

THALLIUM

1 / 29

4.00E+00

4.00E+00

1.90E+00

-

6.50E+00

3.60E+00

6.50E+00

VANADIUM

2 / 29

4.30E+00

5.60E+00

1.50E+00

-

4.00E+00

2.40E+00

4.95E+00

ZINC

11/29

5.50E+00

1.36E+02

5.10E+00

-

1.92E+01

9.40E+00

1.29E+02

METALS - FILTERED

BARIUM, DISSOLVED

29 / 29

7.40E+00

2.16E+01

N/A

1.50E+01

2.12E+01

BERYLLIUM, DISSOLVED

2 / 29

1.00E-01

6.70E-01

1.00E-01

-

4.00E-01

1.00E-01

5.35E-01

CADMIUM, DISSOLVED

1 / 29

2.50E-01

2.50E-01

2.00E-01

-

9.00E-01

3.00E-01

9.00E-01

CALCIUM, DISSOLVED

20 / 20

1.18E+04

4.14E+04

N/A

3.31E+04

4.12E+04

CHROMIUM, DISSOLVED

3 / 29

7.70E-01

3.70E+00

7.00E-01

-

1.60E+00

9.00E-01

2.65E+00

COBALT, DISSOLVED

1 / 29

2.70E+00

2.70E+00

1.50E+00

-

3.90E+00

2.00E+00

3.65E+00

COPPER, DISSOLVED

13 / 29

1.30E+00

1.74E+01

9.00E-01

-

5.00E+00

1.80E+00

1.25E+01

MAGNESIUM, DISSOLVED

20 / 20

4.32E+03

1.62E+04

N/A

1.22E+04

1.62E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-24

-------
Table B-24

Surface Water Chemistry Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

MERCURY, FILTERED

3 / 20

1.00E-01

1.80E-01

1.00E-01

-

1.00E-01

1.00E-01

1.76E-01

ZINC, DISSOLVED

8 / 29

2.10E+00

2.86E+01

2.10E+00

-

2.58E+01

6.10E+00

2.72E+01

ORGANIC

TOTAL ORGANIC CARBON

47 / 47

2.80E+03

1.73E+04

N/A

5.10E+03

1.14E+04

TOTAL ORGANIC CARBON, DISSOLVED

60 / 60

2.70E+03

7.29E+04

N/A

5.10E+03

2.02E+04

INORGANICS

ALKALINITY

59 / 60

4.40E+04

1.44E+05

1.00E+03

-

1.00E+03

1.03E+05

1.30E+05

AMMONIA AS N

60 / 60

4.00E+01

3.00E+03

N/A

1.10E+02

3.80E+02

BIOLOGICAL OXYGEN DEMAND 5 DAY

24 / 62

2.00E+03

8.00E+03

1.00E+03

-

3.00E+03

2.00E+03

4.23E+03

CHEMICAL OXYGEN DEMAND

31 / 31

9.90E+03

4.70E+04

N/A

2.00E+04

4.46E+04

HARDNESS

60 / 60

5.50E+04

1.66E+05

N/A

1.25E+05

1.62E+05

HYDROLYZABLE PHOSPHATE, AS P

27 / 27

4.00E+01

2.60E+02

N/A

1.20E+02

2.44E+02

NITRATE AND NITRITE AS N

59 / 60

1.00E+02

9.80E+03

1.00E+01

-

1.00E+01

2.10E+03

8.99E+03

NITRITE AS N

43 / 60

5.00E+00

9.80E+01

5.00E+00

-

1.00E+03

1.00E+01

1.00E+03

ORGANIC CARBON, PARTICULATE

18 / 45

5.00E+02

1.16E+04

5.00E+02

-

5.00E+02

5.00E+02

6.39E+03

ORGANIC PHOSPHATE, AS P

25 / 31

1.00E+01

2.00E+02

1.00E+01

-

5.00E+01

2.00E+01

2.00E+02

ORTHOPHOSPHATE AS P

56 / 60

1.00E+01

3.20E+02

2.50E+02

-

2.50E+02

9.00E+01

2.50E+02

PHOSPHATE, AS P

2 / 2

5.00E+01

9.00E+01

N/A

7.00E+01

9.00E+01

PHOSPHATE, TOTAL AS P

58 / 58

2.00E+01

4.10E+02

N/A

1.35E+02

3.71E+02

SULFIDE

6 / 25

5.00E+02

8.00E+02

5.00E+02

-

1.00E+03

8.00E+02

1.00E+03

TKN

58 / 60

2.40E+02

2.80E+03

2.40E+02

-

2.40E+03

7.30E+02

1.88E+03

TOTAL DISSOLVED SOLIDS

29 / 29

1.02E+05

3.00E+05

N/A

2.10E+05

2.96E+05

TOTAL SUSPENDED SOLIDS

525 / 568

1.30E+03

3.66E+05

5.00E+03

-

5.00E+03

1.20E+04

9.81E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-24

-------
Table B-25

Surface Water Chemistry Summary
5b Vernal Pool
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

8 / 8

0.00E+00 - 1.13E-04

N/A

0.00E+00

1.13E-04

TOTAL DIOXINS (USING DL)

8 / 8

1.97E-05 - 1.30E-04

N/A

4.54E-05

1.30E-04

TOTAL DIOXINS (USING HALF DL)

8 / 8

9.84E-06 - 1.22E-04

N/A

2.27E-05

1.22E-04

TOTAL FURANS (USING 0)

8 / 8

2.30E-06 - 1.44E-03

N/A

3.41E-05

1.44E-03

TOTAL FURANS (USING DL)

8 / 8

1.46E-05 - 1.44E-03

N/A

3.92E-05

1.44E-03

TOTAL FURANS (USING HALF DL)

8 / 8

8.46E-06 - 1.44E-03

N/A

3.58E-05

1.44E-03

PCBS

AROCL OR-1254

6 / 8

1.90E-02 - 1.10E-01

1.30E-02 - 2.50E-02

3.25E-02

1.10E-01

AROCLOR-1260

8 / 8

3.00E-02 - 4.70E-01

N/A

1.75E-01

4.70E-01

PCB, TOTAL

8 / 8

3.00E-02 - 5.80E-01

N/A

2.15E-01

5.80E-01

METALS

BARIUM

1 / 1

1.44E+01 - 1.44E+01

N/A

1.44E+01

1.44E+01

CALCIUM

1 / 1

1.74E+04 - 1.74E+04

N/A

1.74E+04

1.74E+04

COPPER

1 / 1

4.80E+00 - 4.80E+00

N/A

4.80E+00

4.80E+00

LEAD

1 / 1

6.90E+00 - 6.90E+00

N/A

6.90E+00

6.90E+00

MAGNESIUM

1 / 1

6.49E+03 - 6.49E+03

N/A

6.49E+03

6.49E+03

VANADIUM

1 / 1

1.60E+00 - 1.60E+00

N/A

1.60E+00

1.60E+00

ORGANIC

TOTAL ORGANIC CARBON

1 / 1

9.40E+03 - 9.40E+03

N/A

9.40E+03

9.40E+03

INORGANICS

SULFIDE

1 / 1

1.00E+03 - 1.00E+03

N/A

1.00E+03

1.00E+03

N/A = Not Applicable

Tables B-21- B-30.xls B-25

-------
Table B-26

Surface Water Chemistry Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

APPIX VOLATILES

ACETONE

1 / 3

4.00E+00

4.00E+00

2.50E+00

-

5.60E+00

4.00E+00

5.60E+00

BROMODICHLOROMETHANE

3 / 5

5.70E-01

8.70E-01

5.00E-01

-

5.00E-01

5.70E-01

8.70E-01

CHLOROBENZENE

1 / 5

5.40E-01

5.40E-01

5.00E-01

-

5.00E-01

5.00E-01

5.40E-01

CHLOROFORM

3 / 5

2.00E+00

4.10E+00

5.00E-01

-

5.00E-01

2.00E+00

4.10E+00

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

3 /

16

6.00E-01

1.20E+02

1.00E+01

-

1.10E+01

1.00E+01

1.20E+02

DIETHYL PHTHALATE

1

16

9.00E-01

9.00E-01

1.00E+01

-

1.10E+01

1.00E+01

1.10E+01

PAHs

ACENAPHTHENE

2 /

16

1.10E-02

1.50E-02

1.00E+01

-

2.00E+01

1.00E+01

2.00E+01

BENZO(A)ANTHRACENE

1

16

1.20E-02

1.20E-02

2.00E-02

-

2.00E+01

1.00E+01

2.00E+01

BENZO(B)FLUORANTHENE

2 /

16

1.70E-02

2.10E-02

1.00E+01

-

2.00E+01

1.00E+01

2.00E+01

BENZO(GHI)PERYLENE

1

16

1.10E-02

1.10E-02

2.20E-02

-

2.00E+01

1.00E+01

2.00E+01

BENZO(A)PYRENE

2

16

1.10E-02

1.30E-02

1.00E+01

-

2.00E+01

1.00E+01

2.00E+01

CHRYSENE

2

16

1.30E-02

1.30E-02

1.00E+01

-

2.00E+01

1.00E+01

2.00E+01

FLUORANTHENE

2 /

16

2.50E-02

3.20E-02

1.00E+01

-

2.00E+01

1.00E+01

2.00E+01

INDENO(l,2,3-C,D)PYRENE

1 /

16

1.20E-02

1.20E-02

2.20E-02

-

2.00E+01

1.00E+01

2.00E+01

NAPHTHALENE

1

16

1.50E-02

1.50E-02

2.20E-02

-

2.00E+01

1.00E+01

2.00E+01

PHENANTHRENE

2

16

1.30E-02

1.40E-02

1.00E+01

-

2.00E+01

1.00E+01

2.00E+01

PYRENE

2 /

16

3.20E-02

3.70E-02

1.00E+01

-

2.00E+01

1.00E+01

2.00E+01

TOTAL PAH (USING 0)

16 /

16

0.00E+00

1.70E-01

N/A

0.00E+00

1.70E-01

TOTAL PAH (USING DL)

16 /

16

1.60E+02

3.20E+02

N/A

1.60E+02

3.20E+02

TOTAL PAH (USING HALF DL)

16 /

16

8.00E+01

1.60E+02

N/A

8.00E+01

1.60E+02

TOTAL PAH (HIGH) (USING 0)

16 /

16

0.00E+00

1.01E-01

N/A

0.00E+00

1.01E-01

TOTAL PAH (HIGH) (USING DL)

16 /

16

9.00E+01

1.80E+02

N/A

9.00E+01

1.80E+02

TOTAL PAH (HIGH) (USING HALF DL)

16 /

16

4.50E+01

9.00E+01

N/A

4.50E+01

9.00E+01

N/A = Not Applicable

Tables B-21- B-30.xls B-26

-------
Table B-26

Surface Water Chemistry Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

TOTAL PAH (LOW) (USING 0)

16 / 16

0.00E+00 - 6.90E-02

N/A

0.00E+00

6.90E-02

TOTAL PAH (LOW) (USING DL)

16 / 16

7.00E+01 - 1.40E+02

N/A

7.00E+01

1.40E+02

TOTAL PAH (LOW) (USING HALF DL)

16 / 16

3.50E+01 - 7.00E+01

N/A

3.50E+01

7.00E+01

1,2,4-TRICHLOROBENZENE, FILTERED

10 / 10

5.40E-03 - 5.60E-02

N/A

1.85E-02

5.60E-02

APPIX PESTICIDES

DELTA-BHC

1 / 16

1.10E-01 - 1.10E-01

5.00E-02 - 5.30E-02

5.00E-02

1.10E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

16 / 16

0.00E+00 - 1.01E-04

N/A

1.78E-05

1.01E-04

TOTAL DIOXINS (USING DL)

16 / 16

2.31E-05 - 1.13E-04

N/A

3.66E-05

1.13E-04

TOTAL DIOXINS (USING HALF DL)

16 / 16

1.16E-05 - 1.07E-04

N/A

2.93E-05

1.07E-04

TOTAL FURANS (USING 0)

16 / 16

0.00E+00 - 6.51E-04

N/A

9.06E-05

6.51E-04

TOTAL FURANS (USING DL)

16 / 16

1.79E-05 - 6.51E-04

N/A

1.09E-04

6.51E-04

TOTAL FURANS (USING HALF DL)

16 / 16

1.14E-05 - 6.51E-04

N/A

1.00E-04

6.51E-04

PCBS

AROCLOR-1242

1 / 14

2.40E-02 - 2.40E-02

1.20E-02 - 3.90E-02

1.30E-02

3.90E-02

AROCL OR-1254

7 / 14

2.00E-02 - 5.50E-02

1.20E-02 - 3.90E-02

2.10E-02

5.50E-02

AROCLOR-1260

12 / 14

1.70E-02 - 2.00E-01

1.30E-02 - 1.70E-02

3.45E-02

2.00E-01

PCB, TOTAL

12 / 14

1.70E-02 - 2.00E-01

1.30E-02 - 1.70E-02

5.50E-02

2.00E-01

METALS

ARSENIC

1 / 16

3.10E+00 - 3.10E+00

1.80E+00 - 6.00E+00

3.00E+00

6.00E+00

BARIUM

15 / 16

1.17E+01 - 2.27E+01

2.15E+01 - 2.15E+01

1.76E+01

2.27E+01

CALCIUM

11 / 11

1.91E+04 - 3.98E+04

N/A

3.53E+04

3.98E+04

CHROMIUM

2 / 16

1.70E+00 - 4.30E+00

7.00E-01 - 2.70E+00

1.30E+00

4.30E+00

COPPER

12 / 16

1.10E+00 - 5.70E+00

1.00E+00 - 6.00E+00

3.05E+00

6.00E+00

MAGNESIUM

11 / 11

6.63E+03 - 1.60E+04

N/A

1.28E+04

1.60E+04

MERCURY

3 / 16

1.00E-01 - 2.10E-01

1.00E-01 - 1.00E-01

1.00E-01

2.10E-01

N/A = Not Applicable

Tables B-21- B-30.xls B-26

-------
Table B-26

Surface Water Chemistry Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

TIN

1 / 16

2.30E+00 - 2.30E+00

1.60E+00 - 6.30E+00

2.70E+00

6.30E+00

VANADIUM

2 / 16

3.50E+00 - 4.10E+00

1.50E+00 - 4.00E+00

2.45E+00

4.10E+00

ZINC

5 / 16

4.00E+00 - 5.18E+01

3.50E+00 - 1.92E+01

9.25E+00

5.18E+01

METALS - FILTERED

BARIUM, DISSOLVED

14 / 14

1.00E+01 - 2.12E+01

N/A

1.52E+01

2.12E+01

BERYLLIUM, DISSOLVED

1 / 14

1.10E-01 - 1.10E-01

1.00E-01 - 4.00E-01

1.10E-01

4.00E-01

CALCIUM, DISSOLVED

9 / 9

1.86E+04 - 3.78E+04

N/A

3.30E+04

3.78E+04

CHROMIUM, DISSOLVED

3 / 14

1.50E+00 - 3.90E+00

7.00E-01 - 1.30E+00

9.00E-01

3.90E+00

COBALT, DISSOLVED

1 / 14

2.20E+00 - 2.20E+00

1.50E+00 - 3.90E+00

2.25E+00

3.90E+00

COPPER, DISSOLVED

6 / 14

1.50E+00 - 4.50E+00

9.00E-01 - 6.00E+00

2.15E+00

6.00E+00

LEAD, DISSOLVED

1 / 14

8.40E-01 - 8.40E-01

8.00E-01 - 2.80E+00

1.35E+00

2.80E+00

MAGNESIUM, DISSOLVED

9 / 9

6.37E+03 - 1.53E+04

N/A

1.19E+04

1.53E+04

MERCURY, FILTERED

1 / 9

1.00E-01 - 1.00E-01

1.00E-01 - 1.00E-01

1.00E-01

1.00E-01

SELENIUM, DISSOLVED

1 / 14

3.70E+00 - 3.70E+00

1.80E+00 - 4.90E+00

3.35E+00

4.90E+00

ZINC, DISSOLVED

3 / 14

5.10E+00 - 8.30E+00

3.00E+00 - 1.44E+01

6.45E+00

1.44E+01

ORGANIC

TOTAL ORGANIC CARBON

8 / 8

3.40E+03 - 6.50E+03

N/A

4.25E+03

6.50E+03

TOTAL ORGANIC CARBON, DISSOLVED

14 / 14

2.90E+03 - 1.54E+04

N/A

5.50E+03

1.54E+04

INORGANICS

ALKALINITY

14 / 14

6.10E+04 - 1.40E+05

N/A

1.13E+05

1.40E+05

AMMONIA AS N

14 / 14

6.00E+01 - 1.90E+03

N/A

1.15E+02

1.90E+03

BIOLOGICAL OXYGEN DEMAND 5 DAY

4 / 15

2.10E+03 - 3.00E+03

2.00E+03 - 3.00E+03

2.00E+03

3.00E+03

HARDNESS

14 / 14

7.10E+04 - 1.56E+05

N/A

1.32E+05

1.56E+05

NITRATE AND NITRITE AS N

13 / 14

6.50E+02 - 3.60E+03

1.00E+01 - 1.00E+01

1.70E+03

3.60E+03

NITRITE AS N

10 / 14

8.00E+00 - 5.70E+01

5.00E+00 - 5.00E+00

1.45E+01

5.70E+01

ORGANIC CARBON, PARTICULATE

1 / 7

5.30E+02 - 5.30E+02

5.00E+02 - 5.00E+02

5.00E+02

5.30E+02

N/A = Not Applicable

Tables B-21- B-30.xls B-26

-------
Table B-26

Surface Water Chemistry Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

ORTHOPHOSPHATE AS P

14 / 14

1.00E+01 - 1.60E+02

N/A

7.00E+01

1.60E+02

PHOSPHATE, AS P

1 / 1

5.00E+01 - 5.00E+01

N/A

5.00E+01

5.00E+01

PHOSPHATE, TOTAL AS P

13 / 13

5.00E+01 - 2.00E+02

N/A

1.10E+02

2.00E+02

SULFIDE

2 / 12

5.00E+02 - 6.00E+02

5.00E+02 - 1.00E+03

8.00E+02

1.00E+03

TKN

12 / 14

4.90E+02 - 9.50E+02

2.40E+02 - 2.40E+02

6.85E+02

9.50E+02

TOTAL DISSOLVED SOLIDS

14 / 14

1.25E+05 - 3.07E+05

N/A

2.01E+05

3.07E+05

TOTAL SUSPENDED SOLIDS

14 / 14

1.90E+03 - 2.59E+04

N/A

3.00E+03

2.59E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-26

-------
Table B-27

Surface Water Chemistry Summary
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median

(HS/L)

95th Percentile
(Jig/L)

ORGANIC

TOTAL ORGANIC CARBON

29 / 29

1.70E+03

1.30E+04

N/A

3.20E+03

1.30E+04

TOTAL ORGANIC CARBON, DISSOLVED

29 / 29

2.00E+03

1.15E+04

N/A

3.60E+03

1.11E+04

INORGANICS

AMMONIA AS N

14 / 29

2.00E+01

1.00E+03

2.00E+01

-

1.00E+02

4.00E+01

5.50E+02

BIOLOGICAL OXYGEN DEMAND 5 DAY

9 / 29

1.00E+03

4.00E+03

2.00E+03

-

3.00E+03

2.00E+03

3.60E+03

HYDROLYZABLE PHOSPHATE, AS P

14 / 26

1.00E+01

8.00E+01

1.00E+01

-

1.00E+01

1.00E+01

6.25E+01

NITRATE AND NITRITE AS N

28 / 29

2.00E+01

6.30E+02

1.00E+01

-

1.00E+01

1.30E+02

5.80E+02

NITRITE AS N

1 / 29

6.00E+00

6.00E+00

5.00E+00

-

2.50E+02

5.00E+00

2.50E+02

ORGANIC CARBON, PARTICULATE

5 / 29

5.00E+02

3.60E+03

5.00E+02

-

5.00E+02

5.00E+02

2.55E+03

ORGANIC PHOSPHATE, AS P

15 / 29

1.00E+01

5.00E+01

1.00E+01

-

5.00E+01

1.00E+01

5.00E+01

ORTHOPHOSPHATE AS P

9 / 29

1.00E+01

4.00E+01

1.00E+01

-

2.50E+02

1.00E+01

2.50E+02

PHOSPHATE, TOTAL AS P

17 / 29

1.00E+01

8.00E+01

1.00E+01

-

5.00E+01

2.00E+01

7.50E+01

TKN

15 / 29

1.10E+02

2.80E+03

1.00E+02

-

2.40E+02

2.40E+02

2.00E+03

TOTAL SUSPENDED SOLIDS

129 / 173

5.00E+02

1.02E+05

5.00E+02

-

5.00E+03

4.00E+03

4.16E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-27

-------
Table B-28

Surface Water Chemistry Summary
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

DIOXINS/FURANS

HPCDD (TOTAL)

5 / 12

2.20E-06

2.60E-04

1.69E-06

-

8.63E-06

6.05E-06

2.60E-04

HPCDF (TOTAL)

4 / 12

6.50E-06

2.29E-04

9.78E-07

-

7.50E-06

6.12E-06

2.29E-04

HXCDD (TOTAL)

2 / 12

1.13E-05

1.89E-05

1.49E-06

-

8.70E-06

4.51E-06

1.89E-05

HXCDF (TOTAL)

5 / 12

1.36E-05

3.38E-04

7.02E-07

-

4.61E-06

3.66E-06

3.38E-04

OCDD

4 / 12

1.24E-05

7.71E-04

8.00E-06

-

7.71E-05

2.63E-05

7.71E-04

OCDF

3 / 12

1.01E-05

6.26E-05

3.82E-06

-

2.69E-05

8.85E-06

6.26E-05

PECDD (TOTAL)

1 / 12

1.44E-06

1.44E-06

6.60E-07

-

5.88E-06

2.00E-06

5.88E-06

PECDF (TOTAL)

5 / 12

7.08E-06

3.16E-04

5.84E-07

-

2.69E-06

2.29E-06

3.16E-04

TCDF (TOTAL)

5 / 12

3.44E-05

8.11E-04

1.08E-06

-

5.66E-06

4.98E-06

8.11E-04

PCBS

AROCL OR-1254

7 / 12

1.40E-02

6.70E-02

1.30E-02

-

2.50E-02

1.70E-02

6.70E-02

AROCLOR-1260

6 / 12

2.40E-02

3.40E-01

1.30E-02

-

2.50E-02

2.45E-02

3.40E-01

PCB, TOTAL

7 / 12

1.50E-02

4.10E-01

1.30E-02

-

2.50E-02

3.15E-02

4.10E-01

METALS

BARIUM

3 / 4

7.30E+00

1.99E+01

1.07E+01

-

1.07E+01

1.24E+01

1.99E+01

CALCIUM

4 / 4

1.95E+03

3.75E+04

N/A

1.87E+04

3.75E+04

COPPER

2 / 4

1.10E+00

1.60E+00

1.00E+00

-

9.30E+00

1.35E+00

9.30E+00

LEAD

2 / 4

1.00E+00

1.10E+00

9.00E-01

-

2.30E+00

1.05E+00

2.30E+00

MAGNESIUM

4 / 4

7.26E+02

9.53E+03

N/A

6.46E+03

9.53E+03

VANADIUM

1 / 4

1.90E+00

1.90E+00

1.60E+00

-

2.90E+00

1.75E+00

2.90E+00

ZINC

1 / 4

1.78E+01

1.78E+01

3.30E+00

-

2.05E+01

1.37E+01

2.05E+01

ORGANIC

TOTAL ORGANIC CARBON

3 / 3

2.70E+03

9.00E+03

N/A

4.50E+03

9.00E+03

N/A = Not Applicable

Tables B-21- B-30.xls B-28

-------
Table B-29

Surface Water Chemistry Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile

(HS/L)

APPIX VOLATILES

ACETONE

1 / 2

4.10E+00 - 4.10E+00

4.40E+00 - 4.40E+00

4.25E+00

4.40E+00

BROMODICHLOROMETHANE

2 / 4

5.40E-01 - 1.00E+00

5.00E-01 - 5.00E-01

5.20E-01

1.00E+00

CHLOROFORM

2 / 4

1.40E+00 - 2.20E+00

5.00E-01 - 5.00E-01

9.50E-01

2.20E+00

DIBROMOCHLOROMETHANE

1 / 4

1.00E+00 - 1.00E+00

5.00E-01 - 5.00E-01

5.00E-01

1.00E+00

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 13

7.00E-01 - 7.00E-01

1.00E+01 - 1.10E+01

1.00E+01

1.10E+01

PAHs











BENZO(A)ANTHRACENE

1 / 13

1.00E-02 - 1.00E-02

1.00E+01 - 1.10E+01

1.00E+01

1.10E+01

CHRYSENE

1 / 13

1.20E-02 - 1.20E-02

1.00E+01 - 1.10E+01

1.00E+01

1.10E+01

FLUORANTHENE

1 / 13

2.70E-02 - 2.70E-02

1.00E+01 - 1.10E+01

1.00E+01

1.10E+01

PHENANTHRENE

1 / 13

1.00E-02 - 1.00E-02

1.00E+01 - 1.10E+01

1.00E+01

1.10E+01

PYRENE

1 / 13

2.70E-02 - 2.70E-02

1.00E+01 - 1.10E+01

1.00E+01

1.10E+01

TOTAL PAH (USING 0)

13 / 13

0.00E+00 - 8.60E-02

N/A

0.00E+00

8.60E-02

TOTAL PAH (USING DL)

13 / 13

1.60E+02 - 1.76E+02

N/A

1.60E+02

1.76E+02

TOTAL PAH (USING HALF DL)

13 / 13

8.00E+01 - 8.80E+01

N/A

8.00E+01

8.80E+01

TOTAL PAH (HIGH) (USING 0)

13 / 13

0.00E+00 - 4.90E-02

N/A

0.00E+00

4.90E-02

TOTAL PAH (HIGH) (USING DL)

13 / 13

9.00E+01 - 9.90E+01

N/A

9.00E+01

9.90E+01

TOTAL PAH (HIGH) (USING HALF DL)

13 / 13

4.50E+01 - 4.95E+01

N/A

4.50E+01

4.95E+01

TOTAL PAH (LOW) (USING 0)

13 / 13

0.00E+00 - 3.70E-02

N/A

0.00E+00

3.70E-02

TOTAL PAH (LOW) (USING DL)

13 / 13

7.00E+01 - 7.70E+01

N/A

7.00E+01

7.70E+01

TOTAL PAH (LOW) (USING HALF DL)

13 / 13

3.50E+01 - 3.85E+01

N/A

3.50E+01

3.85E+01

1,2,4-TRICHLOROBENZENE, FILTERED

14 / 15

5.30E-03 - 4.20E-02

3.90E-03 - 3.90E-03

1.50E-02

4.20E-02

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

13 / 13

0.00E+00 - 2.44E-04

N/A

2.00E-05

2.44E-04

TOTAL DIOXINS (USING DL)

13 / 13

1.61E-05 - 2.45E-04

N/A

4.63E-05

2.45E-04

TOTAL DIOXINS (USING HALF DL)

13 / 13

1.04E-05 - 2.45E-04

N/A

3.14E-05

2.45E-04

TOTAL FURANS (USING 0)

13 / 13

0.00E+00 - 4.52E-04

N/A

1.12E-04

4.52E-04

TOTAL FURANS (USING DL)

13 / 13

1.84E-05 - 4.52E-04

N/A

1.19E-04

4.52E-04

N/A = Not Applicable

Tables B-21- B-30.xls B-29

-------
Table B-29

Surface Water Chemistry Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

TOTAL FURANS (USING HALF DL)

13 / 13

9.20E-06

4.52E-04

N/A

1.15E-04

4.52E-04

PCBS

AROCLOR-1242

8 / 40

1.60E-02

4.20E-02

1.20E-02

-

2.60E-02

1.30E-02

2.59E-02

AROCLOR-1248

1 / 40

1.40E-02

1.40E-02

1.20E-02

-

2.70E-02

1.30E-02

2.55E-02

AROCLOR-1254

30 / 40

1.40E-02

5.80E-02

1.20E-02

-

2.60E-02

2.25E-02

5.10E-02

AROCLOR-1260

36 / 39

1.60E-02

1.50E-01

1.20E-02

-

1.40E-02

4.30E-02

1.40E-01

PCB, TOTAL

36 / 40

1.70E-02

2.10E-01

1.20E-02

-

1.40E-02

6.90E-02

2.00E-01

PCBS - FILTERED

AROCLOR-1260, DISSOLVED

1 / 41

1.50E-02

1.50E-02

1.20E-02

-

1.60E-02

1.30E-02

1.50E-02

TOTAL PCB, DISSOLVED

1 / 41

1.50E-02

1.50E-02

1.20E-02

-

1.60E-02

1.30E-02

1.50E-02

METALS

BARIUM

13 / 13

1.14E+01

2.37E+01

N/A

1.49E+01

2.37E+01

CADMIUM

1 / 13

2.10E-01

2.10E-01

2.00E-01

-

9.00E-01

3.00E-01

9.00E-01

CALCIUM

9 / 9

1.86E+04

3.56E+04

N/A

3.14E+04

3.56E+04

CHROMIUM

2 / 13

1.40E+00

1.50E+00

7.00E-01

-

1.30E+00

9.00E-01

1.50E+00

COPPER

7 / 13

1.60E+00

4.40E+00

1.00E+00

-

2.20E+00

1.70E+00

4.40E+00

MAGNESIUM

9 / 9

6.37E+03

1.52E+04

N/A

1.13E+04

1.52E+04

MERCURY

1 / 13

1.00E-01

1.00E-01

1.00E-01

-

1.00E-01

1.00E-01

1.00E-01

SILVER

1 / 13

2.70E+00

2.70E+00

9.00E-01

-

2.50E+00

1.60E+00

2.70E+00

THALLIUM

1 / 13

4.10E+00

4.10E+00

1.90E+00

-

6.50E+00

3.60E+00

6.50E+00

ZINC

3 / 13

4.40E+00

8.60E+02

3.00E+00

-

2.95E+01

8.50E+00

8.60E+02

METALS - FILTERED

BARIUM, DISSOLVED

14 / 14

1.05E+01

2.20E+01

N/A

1.55E+01

2.20E+01

CADMIUM, DISSOLVED

1 / 14

2.80E-01

2.80E-01

2.00E-01

-

9.00E-01

3.00E-01

9.00E-01

CALCIUM, DISSOLVED

9 / 9

1.87E+04

3.57E+04

N/A

3.04E+04

3.57E+04

CHROMIUM, DISSOLVED

1 / 14

3.20E+00

3.20E+00

7.00E-01

-

1.30E+00

1.04E+00

3.20E+00

COPPER, DISSOLVED

5 / 14

1.30E+00

4.50E+00

9.00E-01

-

4.70E+00

2.25E+00

4.70E+00

LEAD, DISSOLVED

1 / 14

1.00E+00

1.00E+00

8.00E-01

-

2.80E+00

1.15E+00

2.80E+00

MAGNESIUM, DISSOLVED

9 / 9

6.27E+03

1.54E+04

N/A

1.08E+04

1.54E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-29

-------
Table B-29

Surface Water Chemistry Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median

(HS/L)

95th Percentile
(Jig/L)

MERCURY, FILTERED

1 / 9

1.00E-01

1.00E-01

1.00E-01

-

1.00E-01

1.00E-01

1.00E-01

NICKEL, DISSOLVED

1 / 14

5.70E+00

5.70E+00

8.00E-01

-

4.00E+00

2.20E+00

5.70E+00

ZINC, DISSOLVED

3 / 14

4.50E+00

3.83E+01

2.20E+00

-

1.10E+01

7.00E+00

3.83E+01

ORGANIC

TOTAL ORGANIC CARBON

35 / 35

3.20E+03

8.10E+03

N/A

5.50E+03

7.70E+03

TOTAL ORGANIC CARBON, DISSOLVED

42 / 42

3.70E+03

7.16E+04

N/A

5.80E+03

2.11E+04

INORGANICS

ALKALINITY

40 / 40

4.60E+04

1.34E+05

N/A

1.03E+05

1.34E+05

AMMONIA AS N

39 / 40

3.00E+01

5.20E+02

2.00E+01

-

2.00E+01

1.45E+02

3.68E+02

BIOLOGICAL OXYGEN DEMAND 5 DAY

13 / 41

2.10E+03

7.00E+03

1.00E+03

-

3.00E+03

2.00E+03

3.76E+03

CHEMICAL OXYGEN DEMAND

27 / 27

1.20E+04

2.40E+04

N/A

1.90E+04

2.32E+04

HARDNESS

40 / 40

5.10E+04

1.58E+05

N/A

1.20E+05

1.54E+05

HYDROLYZABLE PHOSPHATE, AS P

24 / 24

4.00E+01

1.60E+02

N/A

8.50E+01

1.55E+02

NITRATE AND NITRITE AS N

40 / 40

1.80E+02

6.30E+03

N/A

1.80E+03

4.97E+03

NITRITE AS N

31 / 40

6.00E+00

1.10E+02

5.00E+00

-

1.00E+03

2.05E+01

1.00E+03

ORGANIC CARBON, PARTICULATE

8 / 34

5.00E+02

2.90E+03

5.00E+02

-

5.00E+02

5.00E+02

2.30E+03

ORGANIC PHOSPHATE, AS P

21 / 27

1.00E+01

4.00E+01

1.00E+01

-

5.00E+01

2.00E+01

5.00E+01

ORTHOPHOSPHATE AS P

37 / 40

1.00E+01

4.40E+02

2.50E+02

-

2.50E+02

5.50E+01

2.50E+02

PHOSPHATE, AS P

1 / 1

4.00E+01

4.00E+01

N/A

4.00E+01

4.00E+01

PHOSPHATE, TOTAL AS P

39 / 39

4.00E+01

3.00E+02

N/A

1.00E+02

1.90E+02

SULFIDE

1 / 11

5.00E+02

5.00E+02

5.00E+02

-

1.00E+03

8.00E+02

1.00E+03

TKN

39 / 40

2.40E+02

1.60E+03

2.40E+02

-

2.40E+02

6.90E+02

1.10E+03

TOTAL DISSOLVED SOLIDS

13 / 13

1.28E+05

2.67E+05

N/A

1.98E+05

2.67E+05

TOTAL SUSPENDED SOLIDS

570 / 590

1.70E+03

3.70E+05

5.00E+03

-

5.00E+03

5.30E+03

2.35E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-29

-------
Table B-30

Surface Water Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median

(HS/L)

95th Percentile
(Jig/L)

APPIX VOLATILES

ACETONE

1 / 7

3.90E+00

3.90E+00

2.50E+00

6.80E+00

2.50E+00

6.80E+00

CHLOROFORM

5 / 10

7.50E-01

2.00E+00

5.00E-01

5.00E-01

6.25E-01

2.00E+00

TOLUENE

2 / 10

6.00E-01

7.80E-01

5.00E-01

5.00E-01

5.00E-01

7.80E-01

APP IX SEMIVOLATILES

BENZO(B)FLUORANTHENE

1 / 34

1.80E-02

1.80E-02

2.00E-02

1.10E+01

1.00E+01

1.10E+01

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 34

5.10E+01

5.10E+01

1.00E+01

1.10E+01

1.00E+01

2.10E+01

DI-N-OCTYL PHTHALATE

1 / 34

1.80E+00

1.80E+00

1.00E+01

1.10E+01

1.00E+01

1.10E+01

PAHs

CHRYSENE

1 / 34

1.10E-02

1.10E-02

2.00E-02

1.10E+01

1.00E+01

1.10E+01

FLUORANTHENE

4 / 34

1.80E-02

2.60E-02

1.00E+01

1.10E+01

1.00E+01

1.10E+01

FLUORENE

1 / 34

7.00E-01

7.00E-01

2.00E-02

1.10E+01

1.00E+01

1.10E+01

NAPHTHALENE

2 / 34

1.00E-02

1.10E-02

2.10E-02

1.10E+01

1.00E+01

1.10E+01

PHENANTHRENE

4 / 34

1.60E-02

2.10E-02

1.00E+01

1.10E+01

1.00E+01

1.10E+01

PYRENE

4 / 34

2.10E-02

2.90E-02

1.00E+01

1.10E+01

1.00E+01

1.10E+01

2,4,6-TRICHLOROPHENOL

1 / 34

1.00E+00

1.00E+00

1.00E+01

1.10E+01

1.00E+01

1.10E+01

HERBICIDES

2,4,5-T

1 / 29

1.00E-01

1.00E-01

9.50E-02

1.00E-01

9.50E-02

1.00E-01

DIOXINS/FURANS

HPCDD (TOTAL)

7 / 37

1.24E-06

1.66E-04

2.11E-06

1.41E-05

4.90E-06

2.93E-05

HPCDF (TOTAL)

7 / 37

1.94E-06

9.60E-05

1.00E-06

1.88E-05

3.00E-06

2.65E-05

HXCDD (TOTAL)

2 / 37

2.12E-06

1.90E-04

1.00E-06

2.24E-05

4.50E-06

3.92E-05

HXCDF (TOTAL)

1 / 37

3.09E-05

3.09E-05

7.90E-07

2.06E-05

1.97E-06

2.16E-05

OCDD

7 / 37

6.40E-06

2.22E-04

6.40E-06

8.61E-05

1.50E-05

9.97E-05

OCDF

4 / 37

1.76E-06

9.96E-06

1.41E-06

4.24E-05

8.00E-06

4.12E-05

PECDD (TOTAL)

1 / 37

4.18E-05

4.18E-05

1.00E-06

1.14E-05

3.50E-06

1.44E-05

PECDF (TOTAL)

2 / 37

2.00E-06

2.61E-05

7.90E-07

1.07E-05

2.00E-06

1.22E-05

TCDD (TOTAL)

2 / 37

1.96E-06

1.40E-05

6.50E-07

9.35E-06

2.54E-06

9.82E-06

TCDF (TOTAL)

4 / 37

2.10E-06

5.16E-06

2.60E-07

6.02E-06

2.30E-06

5.67E-06

N/A = Not Applicable

Tables B-21- B-30.xls B-30

-------
Table B-30

Surface Water Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

PCBS

AROCL OR-1254

1 / 36

2.50E-02

2.50E-02

1.20E-02

-

6.80E-02

1.30E-02

3.32E-02

AROCLOR-1260

1 / 36

2.00E-01

2.00E-01

1.20E-02

-

2.70E-02

1.30E-02

5.30E-02

PCB, TOTAL

2 / 36

2.50E-02

2.00E-01

1.20E-02

-

2.70E-02

1.30E-02

5.30E-02

PCBS - FILTERED

1,2,4-TRICHLOROBENZENE, FILTERED

2 / 22

4.50E-03

4.90E-03

2.50E-03

-

2.80E-03

2.70E-03

4.84E-03

AROCLOR-1254, DISSOLVED

1 / 30

1.40E-02

1.40E-02

1.20E-02

-

1.40E-02

1.40E-02

1.40E-02

TOTAL PCB, DISSOLVED

1 / 30

1.40E-02

1.40E-02

1.20E-02

-

1.40E-02

1.40E-02

1.40E-02

METALS

BARIUM

33 / 34

9.60E+00

2.76E+01

1.07E+01

-

1.07E+01

1.67E+01

2.66E+01

CALCIUM

23 / 23

6.75E+03

3.66E+04

N/A

2.31E+04

3.65E+04

CHROMIUM

1 / 34

3.00E+00

3.00E+00

7.00E-01

-

2.90E+00

9.50E-01

2.93E+00

COPPER

4 / 34

1.00E+00

1.87E+01

9.00E-01

-

3.70E+00

2.15E+00

7.60E+00

LEAD

4 / 34

1.10E+00

5.27E+01

8.00E-01

-

2.80E+00

1.55E+00

1.53E+01

MAGNESIUM

23 / 23

2.52E+03

1.78E+04

N/A

9.86E+03

1.76E+04

MERCURY

2 / 34

1.00E-01

1.00E-01

1.00E-01

-

1.60E-01

1.00E-01

1.15E-01

NICKEL

1 / 34

3.30E+00

3.30E+00

8.00E-01

-

4.00E+00

2.30E+00

4.00E+00

SILVER

1 / 34

2.30E+00

2.30E+00

9.00E-01

-

3.10E+00

1.50E+00

3.10E+00

TIN

2 / 34

4.70E+00

6.30E+00

1.60E+00

-

6.30E+00

3.40E+00

6.30E+00

VANADIUM

3 / 34

1.80E+00

3.80E+00

1.50E+00

-

4.00E+00

2.40E+00

4.00E+00

ZINC

15 / 34

4.50E+00

1.31E+02

1.50E+00

-

2.59E+01

6.35E+00

8.39E+01

METALS - FILTERED

BARIUM, DISSOLVED

30 / 30

7.60E+00

2.71E+01

N/A

1.57E+01

2.52E+01

CADMIUM, DISSOLVED

1 / 30

2.30E-01

2.30E-01

2.00E-01

-

9.00E-01

3.00E-01

9.00E-01

CALCIUM, DISSOLVED

20 / 20

6.44E+03

3.97E+04

N/A

2.42E+04

3.96E+04

CHROMIUM, DISSOLVED

2 / 30

1.70E+00

1.90E+00

7.00E-01

-

2.90E+00

9.55E-01

2.90E+00

COBALT, DISSOLVED

1 / 30

1.70E+00

1.70E+00

1.50E+00

-

3.90E+00

2.00E+00

3.90E+00

COPPER, DISSOLVED

1 / 30

1.00E+00

1.00E+00

9.00E-01

-

3.70E+00

1.58E+00

3.70E+00

LEAD, DISSOLVED

1 / 30

1.70E+00

1.70E+00

8.00E-01

-

2.80E+00

1.45E+00

2.80E+00

N/A = Not Applicable

Tables B-21- B-30.xls B-30

-------
Table B-30

Surface Water Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (fig/L)

Range of Sample Quantitation
Limits (fig/L)

Median
(Jig/L)

95th Percentile
(Jig/L)

MAGNESIUM, DISSOLVED

20 / 20

2.31E+03

1.78E+04

N/A

9.87E+03

1.77E+04

MERCURY, FILTERED

2 / 20

1.00E-01

1.00E-01

1.00E-01

-

1.00E-01

1.00E-01

1.00E-01

TIN, DISSOLVED

2 / 30

5.30E+00

9.40E+00

1.60E+00

-

6.30E+00

3.40E+00

7.70E+00

VANADIUM, DISSOLVED

1 / 30

1.70E+00

1.70E+00

1.50E+00

-

4.00E+00

2.30E+00

4.00E+00

ZINC, DISSOLVED

8 / 30

1.50E+00

3.41E+01

1.10E+00

-

7.50E+00

4.00E+00

2.18E+01

ORGANIC

CHLOROPHYLL-A, CORRECTED

4 / 4

5.00E-01

1.60E+00

N/A

1.00E+00

1.60E+00

CHLOROPHYLL-A, UNCORRECTED

4 / 4

7.00E-01

2.00E+00

N/A

1.35E+00

2.00E+00

TOTAL ORGANIC CARBON

16 / 16

2.80E+03

1.24E+04

N/A

4.85E+03

1.24E+04

TOTAL ORGANIC CARBON, DISSOLVED

30 / 30

2.80E+03

1.42E+04

N/A

6.05E+03

1.32E+04

INORGANICS

ALKALINITY

30 / 30

1.70E+04

1.98E+05

N/A

1.02E+05

1.86E+05

AMMONIA AS N

28 / 30

2.00E+01

9.70E+02

2.00E+01

-

2.00E+01

5.00E+01

7.83E+02

BIOLOGICAL OXYGEN DEMAND 5 DAY

11/30

2.00E+03

4.10E+03

2.00E+03

-

3.00E+03

2.00E+03

3.99E+03

HARDNESS

30 / 30

2.80E+04

1.60E+05

N/A

1.08E+05

1.59E+05

HARDNESS, DISSOLVED

2 / 2

4.50E+04

5.50E+04

N/A

5.00E+04

5.50E+04

NITRATE AND NITRITE AS N

30 / 30

7.00E+01

3.20E+03

N/A

3.05E+02

2.93E+03

NITRITE AS N

10 / 30

5.00E+00

7.40E+01

5.00E+00

-

5.00E+00

5.00E+00

5.86E+01

ORGANIC CARBON, PARTICULATE

6 / 14

3.20E+02

2.40E+03

5.00E+02

-

5.00E+02

5.00E+02

2.40E+03

ORTHOPHOSPHATE AS P

10 / 30

1.00E+01

2.10E+02

1.00E+01

-

1.00E+01

1.00E+01

1.22E+02

PHOSPHATE, AS P

2 / 2

2.00E+01

2.00E+01

N/A

2.00E+01

2.00E+01

PHOSPHATE, TOTAL AS P

24 / 28

1.00E+01

2.50E+02

1.00E+01

-

2.00E+01

2.50E+01

1.78E+02

SULFIDE

12 / 29

5.50E+02

1.10E+03

5.00E+02

-

1.00E+03

8.00E+02

1.10E+03

TKN

27 / 30

2.60E+02

2.10E+03

2.40E+02

-

2.40E+02

5.45E+02

2.05E+03

TOTAL DISSOLVED SOLIDS

30 / 30

2.90E+04

3.71E+05

N/A

1.72E+05

3.70E+05

TOTAL SUSPENDED SOLIDS

30 / 30

7.00E+02

3.71E+04

N/A

2.35E+03

2.26E+04

N/A = Not Applicable

Tables B-21- B-30.xls B-30

-------
Table B-31

Soil Chemistry Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

2 / 23

8.90E-02

4.40E-01

5.00E-02

-

8.90E-01

4.20E-01

8.86E-01

BUTYLBENZYLPHTHALATE

2 / 23

3.30E-02

3.80E-02

3.50E-01

-

8.90E-01

4.20E-01

8.86E-01

DIBENZOFURAN

9 / 23

3.50E-02

8.20E-02

3.50E-01

-

8.90E-01

3.90E-01

8.86E-01

DI-N-BUTYL PHTHALATE

1 / 23

2.30E-02

2.30E-02

3.50E-01

-

8.90E-01

4.20E-01

8.86E-01

1,3 -DICHLOROBENZENE

1 / 23

2.80E-02

2.80E-02

3.50E-01

-

8.90E-01

4.30E-01

8.86E-01

1,4-DICHLOROBENZENE

8 / 23

4.20E-02

1.70E-01

3.50E-01

-

8.90E-01

4.00E-01

8.86E-01

DIETHYL PHTHALATE

2 / 24

3.30E-02

1.20E-01

3.50E-01

-

8.90E-01

4.20E-01

8.85E-01

2-METHYLNAPHTHALENE

13 / 23

2.60E-02

1.30E-01

3.50E-01

-

8.90E-01

1.00E-01

8.86E-01

PENTACHLOROBENZENE

5 / 23

2.10E-02

5.30E-02

3.50E-01

-

8.90E-01

4.20E-01

8.86E-01

PAHs

ACENAPHTHENE

8 / 23

3.90E-02

9.70E-02

3.50E-01

-

8.90E-01

4.00E-01

8.86E-01

ACENAPHTHYLENE

15 / 23

2.40E-02

3.50E-01

3.50E-01

-

8.90E-01

9.00E-02

8.86E-01

ANTHRACENE

14 / 23

2.80E-02

3.20E-01

3.50E-01

-

8.90E-01

2.10E-01

8.86E-01

BENZO(A)ANTHRACENE

22 / 24

3.20E-02

1.60E+00

3.50E-01

-

8.90E-01

2.70E-01

1.53E+00

BENZO(B)FLUORANTHENE

21 / 24

3.60E-02

1.80E+00

3.50E-01

-

8.90E-01

3.60E-01

1.78E+00

BENZO(K)FLUORANTHENE

21 / 24

3.20E-02

1.30E+00

3.50E-01

-

8.90E-01

4.00E-01

1.25E+00

BENZO(GHI)PERYLENE

20 / 24

3.10E-02

1.00E+00

3.50E-01

-

8.90E-01

3.00E-01

9.73E-01

BENZO(A)PYRENE

22 / 24

2.80E-02

1.60E+00

3.50E-01

-

8.90E-01

2.95E-01

1.48E+00

CHRYSENE

22 / 24

4.20E-02

1.60E+00

3.50E-01

-

8.90E-01

3.55E-01

1.50E+00

DIBENZO(A,H) ANTHRACENE

13 / 23

5.40E-02

2.80E-01

3.50E-01

-

8.90E-01

2.40E-01

8.86E-01

FLUORANTHENE

23 / 24

2.00E-02

2.20E+00

6.70E-02

-

6.70E-02

2.55E-01

2.15E+00

FLUORENE

10 / 23

2.80E-02

2.00E-01

3.50E-01

-

8.90E-01

3.60E-01

8.86E-01

INDENO(l,2,3-C,D)PYRENE

21 / 24

2.90E-02

8.10E-01

3.50E-01

-

8.90E-01

2.00E-01

8.70E-01

NAPHTHALENE

17 / 24

2.20E-02

3.90E-01

3.50E-01

-

8.90E-01

1.55E-01

7.85E-01

PHENANTHRENE

22 / 24

2.70E-02

1.50E+00

3.50E-01

-

8.90E-01

2.70E-01

1.48E+00

PYRENE

24 / 24

2.40E-02

4.30E+00

N/A

3.50E-01

3.90E+00

TOTAL PAH (USING 0)

24 / 24

4.40E-02

1.78E+01

N/A

2.51E+00

1.68E+01

TOTAL PAH (USING DL)

24 / 24

1.29E+00

1.84E+01

N/A

4.85E+00

1.72E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-31

-------
Table B-31

Soil Chemistry Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (USING HALF DL)

24 / 24

1.29E+00

1.81E+01

N/A

3.69E+00

1.70E+01

TOTAL PAH (HIGH) (USING 0)

24 / 24

2.40E-02

1.42E+01

N/A

1.92E+00

1.31E+01

TOTAL PAH (HIGH) (USING DL)

24 / 24

6.85E-01

1.42E+01

N/A

2.58E+00

1.31E+01

TOTAL PAH (HIGH) (USING HALF DL)

24 / 24

4.90E-01

1.42E+01

N/A

2.24E+00

1.31E+01

TOTAL PAH (LOW) (USING 0)

24 / 24

2.00E-02

4.65E+00

N/A

5.50E-01

4.50E+00

TOTAL PAH (LOW) (USING DL)

24 / 24

2.97E-01

5.39E+00

N/A

2.20E+00

5.20E+00

TOTAL PAH (LOW) (USING HALF DL)

24 / 24

2.97E-01

4.65E+00

N/A

1.30E+00

4.50E+00

1,2,4-TRICHLOROBENZENE

9 / 23

2.40E-02

1.60E-01

3.50E-01

-

8.90E-01

3.90E-01

8.86E-01

APPIX PESTICIDES

4,4'-DDT

2 / 18

2.60E-02

2.40E+00

4.00E-03

-

2.40E+00

1.16E-01

2.40E+00

HERBICIDES

2,4,5-T

1 / 1

2.40E-02

2.40E-02

N/A

2.40E-02

2.40E-02

DIOXINS/FURANS

TCDD (TOTAL)

19 / 24

4.99E-07

6.22E-05

1.11E-07

-

1.37E-05

3.24E-06

5.52E-05

TCDF (TOTAL)

24 / 24

5.92E-06

1.60E-03

N/A

2.41E-04

1.51E-03

TOTAL DIOXINS (USING 0)

24 / 24

6.23E-05

7.27E-03

N/A

1.02E-03

7.10E-03

TOTAL DIOXINS (USING DL)

24 / 24

6.23E-05

7.27E-03

N/A

1.02E-03

7.10E-03

TOTAL DIOXINS (USING HALF DL)

24 / 24

6.23E-05

7.27E-03

N/A

1.02E-03

7.10E-03

TOTAL FURANS (USING 0)

24 / 24

7.81E-05

8.04E-03

N/A

1.19E-03

7.81E-03

TOTAL FURANS (USING DL)

24 / 24

7.81E-05

8.04E-03

N/A

1.19E-03

7.81E-03

TOTAL FURANS (USING HALF DL)

24 / 24

7.81E-05

8.04E-03

N/A

1.19E-03

7.81E-03

PCBS

AROCL OR-1254

83 / 827

2.20E-02

2.20E+01

8.80E-03

-

1.03E+01

5.07E-01

5.20E+00

AROCLOR-1260

649 / 827

3.00E-02

2.09E+02

1.80E-02

-

1.50E+00

4.20E+00

5.88E+01

PCB, TOTAL

659 / 827

3.00E-02

2.09E+02

1.80E-02

-

1.50E+00

4.34E+00

6.16E+01

METALS

ANTIMONY

10 / 23

2.80E-01

1.40E+00

3.40E-01

-

1.10E+00

6.60E-01

1.34E+00

ARSENIC

24 / 24

1.60E+00

7.80E+00

N/A

4.60E+00

7.63E+00

BARIUM

24 / 24

1.52E+01

9.69E+01

N/A

4.61E+01

9.67E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-31

-------
Table B-31

Soil Chemistry Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

BERYLLIUM

24 / 24

9.00E-02

1.10E+00

N/A

5.15E-01

1.08E+00

CADMIUM

3 / 23

6.00E-02

3.40E-01

2.00E-02

-

4.40E-01

5.00E-02

4.20E-01

CHROMIUM

24 / 24

8.00E+00

6.02E+01

N/A

1.90E+01

5.67E+01

COBALT

24 / 24

5.40E+00

1.54E+01

N/A

1.06E+01

1.54E+01

COPPER

24 / 24

7.60E+00

7.23E+01

N/A

2.70E+01

7.10E+01

LEAD

24 / 24

8.60E+00

1.27E+02

N/A

3.72E+01

1.24E+02

MERCURY

23 / 24

3.00E-02

4.60E+00

2.00E-02

-

2.00E-02

1.45E-01

3.61E+00

NICKEL

24 / 24

9.90E+00

4.03E+01

N/A

1.79E+01

3.70E+01

SELENIUM

2 / 23

4.00E-01

6.70E-01

1.70E-01

-

8.40E-01

4.00E-01

8.24E-01

SILVER

8 / 24

3.10E-01

9.00E-01

8.00E-02

-

6.10E-01

1.90E-01

8.45E-01

THALLIUM

17 / 24

5.90E-01

5.20E+00

2.80E-01

-

3.10E+00

1.65E+00

4.68E+00

TIN

7 / 23

6.00E-01

9.60E+00

7.50E-01

-

1.08E+01

3.30E+00

1.06E+01

VANADIUM

24 / 24

7.50E+00

3.04E+01

N/A

1.61E+01

3.00E+01

ZINC

24 / 24

4.51E+01

2.03E+02

N/A

9.11E+01

1.97E+02

ORGANIC

TOTAL ORGANIC CARBON

100 / 101

8.89E+02

3.72E+05

1.10E+02

-

1.10E+02

3.91E+04

1.13E+05

INORGANICS

CYANIDE

1 / 23

6.70E-01

6.70E-01

5.40E-01

-

1.60E+00

7.20E-01

1.60E+00

PERCENT SOLIDS

687 / 687

3.00E-02

9.98E+01

N/A

7.61E+01

9.61E+01

SULFIDE

3 / 19

7.10E+00

1.32E+01

5.70E+00

-

1.26E+01

9.40E+00

1.32E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-31

-------
Table B-32

Soil Chemistry Summary
5a Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median
(mg/kg)

95th Percentile
(mg/kg)

APP IX SEMEVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

4 / 10

5.80E-02 - 2.00E-01

1.00E-01 - 2.90E+00

2.80E-01

2.90E+00

BUTYLBENZYLPHTHALATE

2 / 10

4.60E-02 - 7.50E-02

3.60E-01 - 2.90E+00

4.55E-01

2.90E+00

DIBENZOFURAN

8 / 10

3.10E-02 - 8.90E-01

4.90E-02 - 6.70E-01

6.40E-02

8.90E-01

DI-N-BUTYL PHTHALATE

3 / 10

3.00E-02 - 3.60E-02

3.60E-01 - 2.90E+00

4.30E-01

2.90E+00

1.4-DTCHT ,OR OBENZENF,

7 / 10

4.20E-02 - 1.70E-01

4.30E-01 - 6.70E-01

7.95E-02

6.70E-01

DIMETHYL PHTHALATE

1 / 10

4.80E-01 - 4.80E-01

3.60E-01 - 2.90E+00

5.00E-01

2.90E+00

2-METHYLNAPHTHALENE

9 / 10

6.20E-02 - 5.30E-01

6.70E-01 - 6.70E-01

1.35E-01

6.70E-01

PYRIDINE

1 / 10

4.80E-01 - 4.80E-01

3.60E-01 - 2.90E+00

5.00E-01

2.90E+00

PAHs

ACENAPHTHENE

8 / 10

3.70E-02 - 9.10E-01

5.20E-01 - 6.70E-01

1.97E-01

9.10E-01

ACENAPHTHYLENE

9 / 10

1.40E-01 - 2.10E+00

6.70E-01 - 6.70E-01

2.30E-01

2.10E+00

ANTHRACENE

9 / 10

1.00E-01 - 5.30E+00

6.70E-01 - 6.70E-01

2.55E-01

5.30E+00

BENZO(A)ANTHRACENE

10 / 10

4.10E-02 - 1.20E+01

N/A

1.30E+00

1.20E+01

BENZO(B)FLUORANTHENE

10 / 10

5.10E-02 - 1.10E+01

N/A

1.45E+00

1.10E+01

BENZO(K)FLUORANTHENE

9 / 10

6.20E-02 - 1.40E+01

4.80E-01 - 4.80E-01

1.35E+00

1.40E+01

BENZO(GHI)PERYLENE

9 / 10

1.30E-01 - 1.40E+00

6.70E-01 - 6.70E-01

4.70E-01

1.40E+00

BENZO(A)PYRENE

9 / 10

5.70E-01 - 1.10E+01

6.70E-01 - 6.70E-01

1.30E+00

1.10E+01

CHRYSENE

10 / 10

6.50E-02 - 1.30E+01

N/A

1.40E+00

1.30E+01

DIBENZO(A,H)ANTHRACENE

8 / 10

9.40E-02 - 9.40E-01

3.60E-01 - 6.70E-01

2.30E-01

9.40E-01

FLUORANTHENE

10 / 10

7.50E-02 - 2.00E+01

N/A

1.90E+00

2.00E+01

FLUORENE

9 / 10

5.60E-02 - 2.00E+00

6.70E-01 - 6.70E-01

1.45E-01

2.00E+00

INDENO( 1,2,3-C,D)PYRENE

10 / 10

3.30E-02 - 1.90E+00

N/A

4.15E-01

1.90E+00

NAPHTHALENE

9 / 10

1.20E-01 - 9.90E-01

6.70E-01 - 6.70E-01

1.90E-01

9.90E-01

PHENANTHRENE

10 / 10

4.90E-02 - 1.20E+01

N/A

1.05E+00

1.20E+01

PYRENE

10 / 10

8.60E-02 - 1.50E+01

N/A

2.30E+00

1.50E+01

TOTAL PAH (USING 0)

10 / 10

4.62E-01 - 1.23E+02

N/A

1.34E+01

1.23E+02

TOTAL PAH (USING DL)

10 / 10

5.82E+00 - 1.23E+02

N/A

1.37E+01

1.23E+02

TOTAL PAH (USING HALF DL)

10 / 10

3.14E+00 - 1.23E+02

N/A

1.36E+01

1.23E+02

TOTAL PAH (HIGH) (USING 0)

10 / 10

3.38E-01 - 8.02E+01

N/A

9.96E+00

8.02E+01

TOTAL PAH (HIGH) (USING DL)

10 / 10

2.35E+00 - 8.02E+01

N/A

9.96E+00

8.02E+01

NBA = No Benchmark Available

Tables B-31-B-40.xls B-32

-------
Table B-32

Soil Chemistry Summary
5a Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median
(mg/kg)

95th Percentile
(mg/kg)

TOTAL PAH (HIGH) (USING HALF DL)

10 / 10

1.34E+00 - 8.02E+01

N/A

9.96E+00

8.02E+01

TOTAL PAH (LOW) (USING 0)

10 / 10

1.24E-01 - 4.29E+01

N/A

3.57E+00

4.29E+01

TOTAL PAH (LOW) (USING DL)

10 / 10

1.98E+00 - 4.29E+01

N/A

3.74E+00

4.29E+01

TOTAL PAH (LOW) (USING HALF DL)

10 / 10

1.80E+00 - 4.29E+01

N/A

3.61E+00

4.29E+01

1,2,4-TRICHLOROBENZENE

2 / 10

2.30E-02 - 3.00E-02

4.30E-01 - 2.90E+00

5.00E-01

2.90E+00

APP IX PESTICIDES

ENDRIN ALDEHYDE

1 / 9

6.90E-01 - 6.90E-01

5.30E-03 - 2.20E+00

5.60E-01

2.20E+00

DIOXINS/FURANS

TCDD (TOTAL)

7 / 10

9.11E-07 - 4.92E-05

2.78E-07 - 1.35E-05

1.32E-05

4.92E-05

TCDF (TOTAL)

10 / 10

2.18E-05 - 1.67E-03

N/A

4.93E-04

1.67E-03

TOTAL DIOXINS (USING 0)

10 / 10

2.77E-05 - 4.51E-03

N/A

1.68E-03

4.51E-03

TOTAL DIOXINS (USING DL)

10 / 10

2.77E-05 - 4.51E-03

N/A

1.68E-03

4.51E-03

TOTAL DIOXINS (USING HALF DL)

10 / 10

2.77E-05 - 4.51E-03

N/A

1.68E-03

4.51E-03

TOTAL FURANS (USING 0)

10 / 10

4.78E-05 - 6.27E-03

N/A

2.14E-03

6.27E-03

TOTAL FURANS (USING DL)

10 / 10

4.78E-05 - 6.27E-03

N/A

2.14E-03

6.27E-03

TOTAL FURANS (USING HALF DL)

10 / 10

4.78E-05 - 6.27E-03

N/A

2.14E-03

6.27E-03

PCBS

AROCLOR-1248

1 / 128

5.01E+00 - 5.01E+00

6.70E-02 - 1.00E+01

5.05E-01

4.96E+00

AROCLOR-1254

29 / 128

2.59E-01 - 4.92E+01

6.70E-02 - 1.00E+01

5.08E-01

6.82E+00

AROCLOR-1260

112 / 128

4.16E-01 - 1.17E+02

5.02E-01 - 5.96E-01

1.02E+01

4.07E+01

PCB, TOTAL

113 / 128

2.59E-01 - 1.17E+02

5.02E-01 - 5.96E-01

1.11E+01

4.19E+01

METALS

ANTIMONY

7 / 10

6.90E-01 - 1.70E+00

5.50E-01 - 6.70E-01

9.00E-01

1.70E+00

ARSENIC

9 / 10

3.20E+00 - 1.31E+01

2.10E+00 - 2.10E+00

4.20E+00

1.31E+01

BARIUM

10 / 10

2.20E+01 - 9.93E+01

N/A

4.03E+01

9.93E+01

BERYLLIUM

9 / 10

2.10E-01 - 1.30E+00

5.80E-01 - 5.80E-01

5.05E-01

1.30E+00

CHROMIUM

10 / 10

1.37E+01 - 6.86E+01

N/A

2.86E+01

6.86E+01

COBALT

10 / 10

5.20E+00 - 1.88E+01

N/A

8.80E+00

1.88E+01

COPPER

10 / 10

2.02E+01 - 9.89E+01

N/A

3.60E+01

9.89E+01

LEAD

10 / 10

2.23E+01 - 1.28E+02

N/A

5.21E+01

1.28E+02

MERCURY

10 / 10

8.00E-02 - 9.90E-01

N/A

2.10E-01

9.90E-01

NBA = No Benchmark Available

Tables B-31-B-40.xls B-32

-------
Table B-32

Soil Chemistry Summary
5a Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median
(mg/kg)

95th Percentile
(mg/kg)

NICKEL

10 / 10

1.01E+01

2.96E+01

N/A

1.61E+01

2.96E+01

SELENIUM

1 / 10

4.00E-01

4.00E-01

2.40E-01

-

7.30E-01

3.80E-01

7.30E-01

SILVER

5 / 10

2.00E-01

6.60E-01

1.20E-01

-

3.10E-01

2.15E-01

6.60E-01

THALLIUM

4 / 10

1.40E+00

4.00E+00

6.60E-01

-

2.00E+00

1.70E+00

4.00E+00

TIN

5 / 10

5.00E+00

1.74E+01

4.80E+00

-

8.20E+00

6.05E+00

1.74E+01

VANADIUM

10 / 10

8.10E+00

2.06E+01

N/A

1.47E+01

2.06E+01

ZINC

10 / 10

7.58E+01

1.61E+02

N/A

1.19E+02

1.61E+02

ORGANIC

TOTAL ORGANIC CARBON

30 / 30

7.23E+03

7.03E+04

N/A

2.32E+04

5.68E+04

INORGANICS

PERCENT SOLIDS

125 / 125

5.20E+01

9.42E+01



N/A



7.14E+01

8.73E+01

NBA = No Benchmark Available

Tables B-31-B-40.xls B-32

-------
Table B-33

Soil Chemistry Summary
5b Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

12 /

16

2.60E-02

1.10E-01

3.80E-01

-

4.80E-01

7.00E-02

4.80E-01

BUTYLBENZYLPHTHALATE

1 /

16

2.80E-02

2.80E-02

3.80E-01

-

7.20E-01

4.50E-01

7.20E-01

DIBENZOFURAN

9

16

2.10E-02

4.10E-02

4.60E-01

-

7.20E-01

3.65E-02

7.20E-01

DI-N-BUTYL PHTHALATE

2 /

16

3.10E-02

4.50E-02

3.80E-01

-

7.20E-01

4.50E-01

7.20E-01

1,4-DICHLOROBENZENE

10 /

16

2.80E-02

8.90E-02

4.20E-01

-

6.00E-01

6.45E-02

6.00E-01

2-METHYLNAPHTHALENE

13 /

16

3.00E-02

6.40E-02

4.80E-01

-

6.00E-01

5.20E-02

6.00E-01

4-METHYLPHENOL

3

15

2.30E-02

4.40E-02

3.80E-01

-

6.00E-01

4.30E-01

6.00E-01

PHENOL

2 /

16

3.80E-02

4.50E-02

3.80E-01

-

7.20E-01

4.35E-01

7.20E-01

PAHs

ACENAPHTHENE

8 /

16

2.50E-02

4.20E-02

4.20E-01

-

6.00E-01

2.31E-01

6.00E-01

ACENAPHTHYLENE

13 /

16

2.60E-02

1.80E-01

4.90E-01

-

6.00E-01

4.60E-02

6.00E-01

ANTHRACENE

15 /

16

2.30E-02

1.50E-01

6.00E-01

-

6.00E-01

7.45E-02

6.00E-01

BENZO(A)ANTHRACENE

16 /

16

1.30E-01

7.20E-01

N/A

4.25E-01

7.20E-01

BENZO(B)FLUORANTHENE

16 /

16

1.50E-01

1.10E+00

N/A

4.20E-01

1.10E+00

BENZO(K)FLUORANTHENE

16 /

16

1.50E-01

8.60E-01

N/A

4.90E-01

8.60E-01

BENZO(GHI)PERYLENE

16 /

16

1.10E-01

6.20E-01

N/A

2.95E-01

6.20E-01

BENZO(A)PYRENE

16 /

16

1.60E-01

7.70E-01

N/A

4.85E-01

7.70E-01

CHRYSENE

16 /

16

1.80E-01

8.50E-01

N/A

5.40E-01

8.50E-01

DIBENZO(A,H) ANTHRACENE

16 /

16

2.80E-02

1.90E-01

N/A

9.20E-02

1.90E-01

FLUORANTHENE

16 /

16

1.80E-01

1.20E+00

N/A

6.95E-01

1.20E+00

FLUORENE

12 /

16

2.30E-02

6.90E-02

4.60E-01

-

6.00E-01

5.00E-02

6.00E-01

INDENO(l,2,3-C,D)PYRENE

16 /

16

8.80E-02

5.70E-01

N/A

2.85E-01

5.70E-01

NAPHTHALENE

16 /

16

2.90E-02

1.30E-01

N/A

8.25E-02

1.30E-01

PHENANTHRENE

16 /

16

1.10E-01

7.00E-01

N/A

4.50E-01

7.00E-01

PYRENE

16 /

16

2.10E-01

1.30E+00

N/A

7.50E-01

1.30E+00

N/A = Not Applicable

Tables B-31-B-40.xls B-33

-------
Table B-33

Soil Chemistry Summary
5b Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (USING 0)

16 / 16

1.57E+00 - 8.78E+00

N/A

5.22E+00

8.78E+00

TOTAL PAH (USING DL)

16 / 16

2.81E+00 - 8.78E+00

N/A

5.65E+00

8.78E+00

TOTAL PAH (USING HALF DL)

16 / 16

2.33E+00 - 8.78E+00

N/A

5.44E+00

8.78E+00

TOTAL PAH (HIGH) (USING 0)

16 / 16

1.25E+00 - 6.45E+00

N/A

3.77E+00

6.45E+00

TOTAL PAH (HIGH) (USING DL)

16 / 16

1.25E+00 - 6.45E+00

N/A

3.77E+00

6.45E+00

TOTAL PAH (HIGH) (USING HALF DL)

16 / 16

1.25E+00 - 6.45E+00

N/A

3.77E+00

6.45E+00

TOTAL PAH (LOW) (USING 0)

16 / 16

3.19E-01 - 2.33E+00

N/A

1.44E+00

2.33E+00

TOTAL PAH (LOW) (USING DL)

16 / 16

1.18E+00 - 2.72E+00

N/A

1.77E+00

2.72E+00

TOTAL PAH (LOW) (USING HALF DL)

16 / 16

9.82E-01 - 2.33E+00

N/A

1.51E+00

2.33E+00

1,2,4-TRICHLOROBENZENE

7 / 16

1.90E-02 - 3.70E-02

3.80E-01 - 6.00E-01

4.00E-01

6.00E-01

APPIX PESTICIDES

ENDOSULFAN SULFATE

1 / 16

5.20E-02 - 5.20E-02

2.00E-01 - 1.90E+00

4.40E-01

1.90E+00

DIOXINS/FURANS

TCDD (TOTAL)

16 / 16

3.70E-06 - 6.60E-05

N/A

1.38E-05

6.60E-05

TCDF (TOTAL)

16 / 16

1.00E-04 - 3.10E-03

N/A

6.11E-04

3.10E-03

TOTAL DIOXINS (USING 0)

16 / 16

7.98E-04 - 1.72E-02

N/A

2.76E-03

1.72E-02

TOTAL DIOXINS (USING DL)

16 / 16

7.98E-04 - 1.72E-02

N/A

2.76E-03

1.72E-02

TOTAL DIOXINS (USING HALF DL)

16 / 16

7.98E-04 - 1.72E-02

N/A

2.76E-03

1.72E-02

TOTAL FURANS (USING 0)

16 / 16

9.00E-04 - 1.20E-02

N/A

3.19E-03

1.20E-02

TOTAL FURANS (USING DL)

16 / 16

9.00E-04 - 1.20E-02

N/A

3.19E-03

1.20E-02

TOTAL FURANS (USING HALF DL)

16 / 16

9.00E-04 - 1.20E-02

N/A

3.19E-03

1.20E-02

PCBS

AROCLOR-1248

1 / 272

1.02E+01 - 1.02E+01

1.10E-01 - 2.64E+01

7.46E-01

5.12E+00

AROCL OR-1254

23 / 272

1.97E+00 - 2.09E+01

1.10E-01 - 2.64E+01

7.91E-01

7.13E+00

AROCLOR-1260

237 / 273

3.11E-01 - 1.64E+02

5.00E-01 - 1.49E+00

1.31E+01

7.58E+01

PCB, TOTAL

237 / 273

3.11E-01 - 1.64E+02

5.00E-01 - 1.49E+00

1.35E+01

7.61E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-33

-------
Table B-33

Soil Chemistry Summary
5b Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

METALS

ANTIMONY

11 / 16

3.90E-01 - 1.40E+00

7.30E-01 - 1.20E+00

8.15E-01

1.40E+00

ARSENIC

16 / 16

2.10E+00 - 6.70E+00

N/A

3.60E+00

6.70E+00

BARIUM

16 / 16

2.76E+01 - 9.71E+01

N/A

4.18E+01

9.71E+01

BERYLLIUM

16 / 16

9.00E-02 - 8.50E-01

N/A

3.55E-01

8.50E-01

CADMIUM

11 / 16

6.00E-02 - 2.20E+00

2.00E-02 - 4.80E-01

3.40E-01

2.20E+00

CHROMIUM

16 / 16

2.42E+01 - 9.37E+01

N/A

3.78E+01

9.37E+01

COBALT

16 / 16

5.40E+00 - 1.46E+01

N/A

9.00E+00

1.46E+01

COPPER

16 / 16

2.63E+01 - 1.20E+02

N/A

4.03E+01

1.20E+02

LEAD

16 / 16

3.15E+01 - 1.71E+02

N/A

4.78E+01

1.71E+02

MERCURY

16 / 16

1.60E-01 - 6.60E-01

N/A

2.50E-01

6.60E-01

NICKEL

16 / 16

1.13E+01 - 2.63E+01

N/A

1.59E+01

2.63E+01

SELENIUM

2 / 16

4.50E-01 - 4.70E-01

1.90E-01 - 1.00E+00

3.50E-01

1.00E+00

SILVER

11 / 16

2.20E-01 - 4.10E+00

1.10E-01 - 6.00E-01

5.65E-01

4.10E+00

THALLIUM

9 / 16

3.90E-01 - 2.70E+00

3.30E-01 - 1.60E+00

7.15E-01

2.70E+00

TIN

8 / 16

3.40E+00 - 1.33E+01

3.10E+00 - 7.00E+00

4.75E+00

1.33E+01

VANADIUM

16 / 16

7.20E+00 - 2.66E+01

N/A

1.33E+01

2.66E+01

ZINC

16 / 16

9.07E+01 - 2.77E+02

N/A

1.21E+02

2.77E+02

ORGANIC

TOTAL ORGANIC CARBON

44 / 44

5.41E+03 - 1.89E+05

N/A

3.16E+04

1.78E+05

INORGANICS

CYANIDE

2 / 16

8.00E-01 - 3.60E+00

6.30E-01 - 1.20E+00

8.20E-01

3.60E+00

PERCENT SOLIDS

202 / 202

0.00E+00 - 9.65E+01

N/A

6.89E+01

9.04E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-33

-------
Table B-34

Soil Chemistry Summary
5b Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

3 / 4

3.00E-02

7.70E-02

5.80E-01

-

5.80E-01

7.40E-02

5.80E-01

BUTYLBENZYLPHTHALATE

1 / 4

2.60E-02

2.60E-02

4.40E-01

-

5.80E-01

4.70E-01

5.80E-01

DIBENZOFURAN

3 / 4

2.60E-02

4.00E-02

5.80E-01

-

5.80E-01

3.50E-02

5.80E-01

DI-N-BUTYL PHTHALATE

1 / 4

2.10E-02

2.10E-02

4.50E-01

-

5.80E-01

4.75E-01

5.80E-01

1,4-DICHLOROBENZENE

3 / 4

5.20E-02

5.80E-02

5.80E-01

-

5.80E-01

5.55E-02

5.80E-01

2-METHYLNAPHTHALENE

3 / 4

4.60E-02

5.50E-02

5.80E-01

-

5.80E-01

5.40E-02

5.80E-01

PAHs

ACENAPHTHENE

3 / 4

2.40E-02

5.00E-02

5.80E-01

-

5.80E-01

3.85E-02

5.80E-01

ACENAPHTHYLENE

3 / 4

4.50E-02

1.20E-01

5.80E-01

-

5.80E-01

8.95E-02

5.80E-01

ANTHRACENE

4 / 4

2.60E-02

1.00E-01

N/A

8.35E-02

1.00E-01

BENZO(A)ANTHRACENE

4 / 4

1.40E-01

6.60E-01

N/A

4.75E-01

6.60E-01

BENZO(B)FLUORANTHENE

4 / 4

1.20E-01

6.80E-01

N/A

5.80E-01

6.80E-01

BENZO(K)FLUORANTHENE

4 / 4

1.30E-01

8.00E-01

N/A

6.50E-01

8.00E-01

BENZO(GHI)PERYLENE

4 / 4

1.10E-01

4.50E-01

N/A

2.85E-01

4.50E-01

BENZO(A)PYRENE

4 / 4

1.40E-01

6.40E-01

N/A

5.45E-01

6.40E-01

CHRYSENE

4 / 4

1.60E-01

8.30E-01

N/A

5.85E-01

8.30E-01

DIBENZO(A,H) ANTHRACENE

4 / 4

2.70E-02

1.10E-01

N/A

7.40E-02

1.10E-01

FLUORANTHENE

4 / 4

2.40E-01

1.30E+00

N/A

8.80E-01

1.30E+00

FLUORENE

3 / 4

4.10E-02

9.40E-02

5.80E-01

-

5.80E-01

7.20E-02

5.80E-01

INDENO(l,2,3-C,D)PYRENE

4 / 4

1.00E-01

4.10E-01

N/A

2.65E-01

4.10E-01

NAPHTHALENE

4 / 4

2.70E-02

1.40E-01

N/A

9.30E-02

1.40E-01

PHENANTHRENE

4 / 4

1.90E-01

9.30E-01

N/A

5.00E-01

9.30E-01

PYRENE

4 / 4

3.00E-01

1.40E+00

N/A

9.85E-01

1.40E+00

TOTAL PAH (USING 0)

4 / 4

1.71E+00

8.64E+00

N/A

6.16E+00

8.64E+00

TOTAL PAH (USING DL)

4 / 4

3.45E+00

8.64E+00

N/A

6.16E+00

8.64E+00

N/A = Not Applicable

Tables B-31-B-40.xls B-34

-------
Table B-34

Soil Chemistry Summary
5b Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (USING HALF DL)

4 / 4

2.58E+00 - 8.64E+00

N/A

6.16E+00

8.64E+00

TOTAL PAH (HIGH) (USING 0)

4 / 4

1.23E+00 - 5.97E+00

N/A

4.45E+00

5.97E+00

TOTAL PAH (HIGH) (USING DL)

4 / 4

1.23E+00 - 5.97E+00

N/A

4.45E+00

5.97E+00

TOTAL PAH (HIGH) (USING HALF DL)

4 / 4

1.23E+00 - 5.97E+00

N/A

4.45E+00

5.97E+00

TOTAL PAH (LOW) (USING 0)

4 / 4

4.83E-01 - 2.67E+00

N/A

1.71E+00

2.67E+00

TOTAL PAH (LOW) (USING DL)

4 / 4

1.48E+00 - 2.67E+00

N/A

2.08E+00

2.67E+00

TOTAL PAH (LOW) (USING HALF DL)

4 / 4

1.35E+00 - 2.67E+00

N/A

1.71E+00

2.67E+00

1,2,4-TRICHLOROBENZENE

3 / 4

2.40E-02 - 3.10E-02

5.80E-01 - 5.80E-01

2.90E-02

5.80E-01

DIOXINS/FURANS

TCDD (TOTAL)

3 / 4

5.18E-06 - 2.43E-05

1.09E-06 - 1.09E-06

1.37E-05

2.43E-05

TCDF (TOTAL)

4 / 4

8.87E-04 - 1.60E-03

N/A

1.30E-03

1.60E-03

TOTAL DIOXINS (USING 0)

4 / 4

3.56E-03 - 1.33E-02

N/A

6.02E-03

1.33E-02

TOTAL DIOXINS (USING DL)

4 / 4

3.56E-03 - 1.33E-02

N/A

6.02E-03

1.33E-02

TOTAL DIOXINS (USING HALF DL)

4 / 4

3.56E-03 - 1.33E-02

N/A

6.02E-03

1.33E-02

TOTAL FURANS (USING 0)

4 / 4

3.49E-03 - 6.47E-03

N/A

5.32E-03

6.47E-03

TOTAL FURANS (USING DL)

4 / 4

3.49E-03 - 6.47E-03

N/A

5.32E-03

6.47E-03

TOTAL FURANS (USING HALF DL)

4 / 4

3.49E-03 - 6.47E-03

N/A

5.32E-03

6.47E-03

PCBS

AROCL OR-1254

1 / 32

1.50E+00 - 1.50E+00

5.02E-01 - 5.40E+00

7.40E-01

5.24E+00

AROCLOR-1260

28 / 32

1.02E+00 - 4.53E+01

6.33E-01 - 7.74E-01

6.53E+00

4.35E+01

PCB, TOTAL

28 / 32

1.02E+00 - 4.53E+01

6.33E-01 - 7.74E-01

6.53E+00

4.35E+01

METALS

ARSENIC

4 / 4

2.60E+00 - 3.40E+00

N/A

3.20E+00

3.40E+00

BARIUM

4 / 4

3.37E+01 - 5.32E+01

N/A

4.54E+01

5.32E+01

BERYLLIUM

4 / 4

3.60E-01 - 5.00E-01

N/A

4.80E-01

5.00E-01

CADMIUM

4 / 4

4.60E-01 - 1.60E+00

N/A

5.40E-01

1.60E+00

N/A = Not Applicable

Tables B-31-B-40.xls B-34

-------
Table B-34

Soil Chemistry Summary
5b Riverbank
Housatonic River Site, OU2 - Pittsfield, MA







Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

CHROMIUM

4 / 4

3.60E+01 - 6.92E+01

N/A

4.06E+01

6.92E+01

COBALT

4 / 4

7.40E+00 - 9.10E+00

N/A

8.65E+00

9.10E+00

COPPER

4 / 4

3.47E+01 - 6.87E+01

N/A

4.54E+01

6.87E+01

LEAD

4 / 4

4.30E+01 - 9.65E+01

N/A

6.68E+01

9.65E+01

MERCURY

4 / 4

2.50E-01 - 3.70E-01

N/A

3.25E-01

3.70E-01

NICKEL

4 / 4

1.48E+01 - 1.98E+01

N/A

1.69E+01

1.98E+01

SILVER

4 / 4

5.40E-01 - 2.60E+00

N/A

9.75E-01

2.60E+00

THALLIUM

2 / 4

1.70E+00 - 2.30E+00

1.30E+00 - 1.60E+00

1.65E+00

2.30E+00

TIN

1 / 4

7.80E+00 - 7.80E+00

3.70E+00 - 6.40E+00

6.20E+00

7.80E+00

VANADIUM

4 / 4

1.06E+01 - 1.60E+01

N/A

1.47E+01

1.60E+01

ZINC

4 / 4

1.00E+02 - 1.65E+02

N/A

1.32E+02

1.65E+02

ORGANIC

TOTAL ORGANIC CARBON

5 / 5

1.32E+04 - 3.90E+04

N/A

2.32E+04

3.90E+04

INORGANICS

PERCENT SOLIDS

10 / 10

6.33E+01 - 9.24E+01

N/A

7.59E+01

9.24E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-34

-------
Table B-35

Soil Chemistry Summary
5c Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

9 / 18

2.80E-02 - 1.20E+00

4.00E-01 - 1.40E+00

4.53E-01

1.40E+00

BUTYLBENZYLPHTHALATE

1 / 18

2.60E-02 - 2.60E-02

4.00E-01 - 2.40E+00

6.65E-01

2.40E+00

DI-N-BUTYL PHTHALATE

2 / 18

2.80E-02 - 3.50E-02

4.00E-01 - 2.40E+00

7.00E-01

2.40E+00

DIBENZOFURAN

3 / 18

5.00E-02 - 9.20E-02

4.00E-01 - 2.40E+00

5.50E-01

2.40E+00

1,4-DICHLOROBENZENE

9 / 18

2.80E-02 - 1.80E-01

4.00E-01 - 2.40E+00

2.90E-01

2.40E+00

2-METHYLNAPHTHALENE

9 / 18

3.40E-02 - 2.30E-01

4.00E-01 - 2.40E+00

3.15E-01

2.40E+00

4-METHYLPHENOL

4 / 18

3.00E-02 - 8.70E-02

4.00E-01 - 2.40E+00

5.50E-01

2.40E+00

N-NITROSO-DI-N-BUTYL AMINE

1 / 18

4.40E-02 - 4.40E-02

4.00E-01 - 2.40E+00

6.30E-01

2.40E+00

PAHs









ACENAPHTHENE

4 / 18

2.50E-02 - 5.60E-02

4.00E-01 - 2.40E+00

5.50E-01

2.40E+00

ACENAPHTHYLENE

6 / 18

2.40E-02 - 1.50E-01

4.00E-01 - 2.40E+00

5.18E-01

2.40E+00

ANTHRACENE

7 / 18

3.80E-02 - 2.10E-01

4.00E-01 - 2.40E+00

5.03E-01

2.40E+00

BENZO(A)ANTHRACENE

14 / 18

3.90E-02 - 1.20E+00

4.00E-01 - 2.40E+00

2.80E-01

2.40E+00

BENZO(B)FLUORANTHENE

16 / 18

3.20E-02 - 1.60E+00

1.50E+00 - 2.40E+00

2.60E-01

2.40E+00

BENZO(K)FLUORANTHENE

16 / 18

3.10E-02 - 1.40E+00

1.50E+00 - 2.40E+00

3.00E-01

2.40E+00

BENZO(GHI)PERYLENE

14 / 18

3.60E-02 - 1.50E+00

4.00E-01 - 2.40E+00

2.55E-01

2.40E+00

BENZO(A)PYRENE

16 / 18

2.70E-02 - 1.60E+00

1.50E+00 - 2.40E+00

2.95E-01

2.40E+00

CHRYSENE

17 / 18

2.80E-02 - 1.80E+00

2.40E+00 - 2.40E+00

2.80E-01

2.40E+00

DIBENZO(A,H) ANTHRACENE

9 / 18

3.40E-02 - 4.20E-01

4.00E-01 - 2.40E+00

4.10E-01

2.40E+00

FLUORANTHENE

17 / 18

3.00E-02 - 2.40E+00

2.40E+00 - 2.40E+00

3.40E-01

2.40E+00

FLUORENE

7 / 18

2.80E-02 - 1.00E-01

4.00E-01 - 2.40E+00

5.03E-01

2.40E+00

INDENO(l,2,3-C,D)PYRENE

14 / 18

3.50E-02 - 1.40E+00

4.00E-01 - 2.40E+00

2.35E-01

2.40E+00

NAPHTHALENE

11 / 18

4.20E-02 - 3.60E-01

4.00E-01 - 2.40E+00

2.25E-01

2.40E+00

PHENANTHRENE

17 / 18

1.80E-02 - 1.50E+00

2.40E+00 - 2.40E+00

2.70E-01

2.40E+00

PYRENE

17 / 18

3.80E-02 - 3.30E+00

2.40E+00 - 2.40E+00

4.55E-01

3.30E+00

N/A = Not Applicable

Tables B-31-B-40.xls B-35

-------
Table B-35

Soil Chemistry Summary
5c Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (USING 0)

18 / 18

0.00E+00 - 1.90E+01

N/A

2.47E+00

1.90E+01

TOTAL PAH (USING DL)

18 / 18

3.00E+00 - 3.84E+01

N/A

5.63E+00

3.84E+01

TOTAL PAH (USING HALF DL)

18 / 18

2.00E+00 - 1.92E+01

N/A

4.04E+00

1.92E+01

TOTAL PAH (HIGH) (USING 0)

18 / 18

0.00E+00 - 1.42E+01

N/A

1.86E+00

1.42E+01

TOTAL PAH (HIGH) (USING DL)

18 / 18

9.41E-01 - 2.16E+01

N/A

2.60E+00

2.16E+01

TOTAL PAH (HIGH) (USING HALF DL)

18 / 18

6.91E-01 - 1.42E+01

N/A

2.38E+00

1.42E+01

TOTAL PAH (LOW) (USING 0)

18 / 18

0.00E+00 - 4.76E+00

N/A

6.02E-01

4.76E+00

TOTAL PAH (LOW) (USING DL)

18 / 18

7.81E-01 - 1.68E+01

N/A

2.79E+00

1.68E+01

TOTAL PAH (LOW) (USING HALF DL)

18 / 18

7.81E-01 - 8.40E+00

N/A

2.00E+00

8.40E+00

1,2,4-TRICHLOROBENZENE

4 / 18

6.00E-02 - 1.60E-01

4.00E-01 - 2.40E+00

5.35E-01

2.40E+00

DIOXINS/FURANS

TCDD (TOTAL)

18 / 18

2.38E-07 - 1.80E-04

N/A

3.47E-05

1.80E-04

TCDF (TOTAL)

18 / 18

2.78E-05 - 7.90E-03

N/A

1.16E-03

7.90E-03

TOTAL DIOXINS (USING 0)

18 / 18

2.38E-07 - 4.24E-02

N/A

4.20E-03

4.24E-02

TOTAL DIOXINS (USING DL)

18 / 18

1.47E-04 - 4.24E-02

N/A

4.20E-03

4.24E-02

TOTAL DIOXINS (USING HALF DL)

18 / 18

7.37E-05 - 4.24E-02

N/A

4.20E-03

4.24E-02

TOTAL FURANS (USING 0)

18 / 18

1.25E-04 - 6.69E-02

N/A

5.20E-03

6.69E-02

TOTAL FURANS (USING DL)

18 / 18

1.25E-04 - 6.69E-02

N/A

5.20E-03

6.69E-02

TOTAL FURANS (USING HALF DL)

18 / 18

1.25E-04 - 6.69E-02

N/A

5.20E-03

6.69E-02

PCBS

AROCL OR-1254

20 / 672

5.41E-03 - 2.88E+01

1.66E-02 - 2.93E+01

8.73E-01

5.05E+00

AROCLOR-1260

324 / 675

9.83E-03 - 3.34E+02

7.53E-03 - 1.90E+00

1.33E+00

5.64E+01

PCB, TOTAL

331 / 675

7.53E-03 - 3.34E+02

1.69E-02 - 1.90E+00

1.33E+00

5.89E+01

METALS

ANTIMONY

14 / 18

3.60E-01 - 3.30E+00

5.00E-01 - 1.50E+00

1.15E+00

3.30E+00

ARSENIC

18 / 18

1.00E+00 - 1.00E+01

N/A

4.13E+00

1.00E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-35

-------
Table B-35

Soil Chemistry Summary
5c Floodplain
Housatonic River Site, OU2 - Pittsfield, MA







Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

BARIUM

18 / 18

1.97E+01 - 1.48E+02

N/A

8.56E+01

1.48E+02

BERYLLIUM

18 / 18

2.70E-01 - 1.90E+00

N/A

6.05E-01

1.90E+00

CADMIUM

16 / 18

5.00E-02 - 7.20E+00

5.00E-02 - 6.00E-02

1.50E+00

7.20E+00

CHROMIUM

18 / 18

5.30E+00 - 1.90E+02

N/A

5.01E+01

1.90E+02

COBALT

18 / 18

3.80E+00 - 1.77E+01

N/A

1.13E+01

1.77E+01

COPPER

18 / 18

8.90E+00 - 1.78E+02

N/A

6.60E+01

1.78E+02

LEAD

18 / 18

7.80E+00 - 2.41E+02

N/A

7.76E+01

2.41E+02

MERCURY

17 / 18

6.50E-02 - 1.70E+00

2.00E-02 - 2.00E-02

3.60E-01

1.70E+00

NICKEL

18 / 18

3.70E+00 - 3.07E+01

N/A

1.86E+01

3.07E+01

SELENIUM

5 / 18

6.70E-01 - 2.40E+00

2.30E-01 - 1.10E+00

6.00E-01

2.40E+00

SILVER

13 / 18

4.10E-01 - 5.70E+00

1.50E-01 - 6.00E-01

1.55E+00

5.70E+00

THALLIUM

9 / 18

4.30E-01 - 1.50E+00

3.85E-01 - 1.70E+00

8.45E-01

1.70E+00

TIN

11 / 18

5.70E+00 - 2.11E+01

6.40E-01 - 6.60E+00

6.40E+00

2.11E+01

VANADIUM

18 / 18

7.00E+00 - 3.27E+01

N/A

2.22E+01

3.27E+01

ZINC

18 / 18

2.96E+01 - 3.83E+02

N/A

1.75E+02

3.83E+02

INORGANICS

PERCENT SOLIDS

252 / 252

0.00E+00 - 9.75E+01

N/A

5.95E+01

8.94E+01

SULFIDE

2 / 16

4.46E+01 - 9.85E+01

7.30E+00 - 5.48E+01

1.02E+01

9.85E+01

ORGANIC

TOTAL ORGANIC CARBON

98 / 98

4.98E+03 - 4.49E+05

N/A

5.44E+04

2.90E+05

N/A = Not Applicable

Tables B-31-B-40.xls B-35

-------
Table B-36

Soil Chemistry Summary
5c Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 3

3.65E-02 - 3.65E-02

4.60E-01 - 7.20E-01

4.60E-01

7.20E-01

1,4-DICHLOROBENZENE

1 / 3

1.00E-01 - 1.00E-01

4.60E-01 - 6.40E-01

4.60E-01

6.40E-01

2-METHYLNAPHTHALENE

1 / 3

5.90E-02 - 5.90E-02

6.40E-01 - 7.20E-01

6.40E-01

7.20E-01

4-METHYLPHENOL

1 / 3

3.60E-02 - 3.60E-02

4.60E-01 - 6.40E-01

4.60E-01

6.40E-01

PAHs











ACENAPHTHYLENE

1 / 3

1.10E-01 - 1.10E-01

6.40E-01 - 7.20E-01

6.40E-01

7.20E-01

ANTHRACENE

2 / 3

3.40E-02 - 5.20E-02

6.40E-01 - 6.40E-01

5.20E-02

6.40E-01

BENZO(A)ANTHRACENE

3 / 3

1.50E-01 - 2.80E-01

N/A

1.50E-01

2.80E-01

BENZO(B)FLUORANTHENE

3 / 3

1.50E-01 - 3.20E-01

N/A

1.90E-01

3.20E-01

BENZO(K)FLUORANTHENE

3 / 3

2.20E-01 - 4.00E-01

N/A

2.40E-01

4.00E-01

BENZO(GHI)PERYLENE

3 / 3

9.10E-02 - 1.80E-01

N/A

1.50E-01

1.80E-01

BENZO(A)PYRENE

3 / 3

7.50E-02 - 2.40E-01

N/A

2.20E-01

2.40E-01

CHRYSENE

3 / 3

2.20E-01 - 4.20E-01

N/A

2.60E-01

4.20E-01

DIBENZO(A,H) ANTHRACENE

2 / 3

4.00E-02 - 5.20E-02

6.40E-01 - 6.40E-01

5.20E-02

6.40E-01

FLUORANTHENE

3 / 3

2.70E-01 - 3.50E-01

N/A

3.50E-01

3.50E-01

INDENO(l,2,3-C,D)PYRENE

3 / 3

8.25E-02 - 1.60E-01

N/A

1.40E-01

1.60E-01

NAPHTHALENE

3 / 3

5.40E-02 - 1.00E-01

N/A

7.50E-02

1.00E-01

PHENANTHRENE

3 / 3

1.80E-01 - 2.20E-01

N/A

2.10E-01

2.20E-01

PYRENE

3 / 3

3.40E-01 - 5.60E-01

N/A

4.80E-01

5.60E-01

TOTAL PAH (USING 0)

3 / 3

1.79E+00 - 3.30E+00

N/A

2.68E+00

3.30E+00

TOTAL PAH (USING DL)

3 / 3

4.22E+00 - 4.88E+00

N/A

4.84E+00

4.88E+00

TOTAL PAH (USING HALF DL)

3 / 3

3.12E+00 - 3.76E+00

N/A

3.76E+00

3.76E+00

TOTAL PAH (HIGH) (USING 0)

3 / 3

1.29E+00 - 2.48E+00

N/A

2.00E+00

2.48E+00

TOTAL PAH (HIGH) (USING DL)

3 / 3

2.00E+00 - 2.55E+00

N/A

2.48E+00

2.55E+00

TOTAL PAH (HIGH) (USING HALF DL)

3 / 3

1.92E+00 - 2.48E+00

N/A

2.00E+00

2.48E+00

N/A = Not Applicable

Tables B-31-B-40.xls B-36

-------
Table B-36

Soil Chemistry Summary
5c Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL PAH (LOW) (USING 0)

3 / 3

5.04E-01 - 8.22E-01

N/A

6.79E-01

8.22E-01

TOTAL PAH (LOW) (USING DL)

3 / 3

1.74E+00 - 2.94E+00

N/A

2.84E+00

2.94E+00

TOTAL PAH (LOW) (USING HALF DL)

3 / 3

1.28E+00 - 1.76E+00

N/A

1.66E+00

1.76E+00

1,2,4-TRICHLOROBENZENE

1 / 3

3.80E-02 - 3.80E-02

4.60E-01 - 6.40E-01

4.60E-01

6.40E-01

DIOXINS/FURANS

TCDD (TOTAL)

3 / 3

1.09E-06 - 4.28E-05

N/A

1.28E-05

4.28E-05

TCDF (TOTAL)

3 / 3

1.17E-04 - 2.62E-03

N/A

6.10E-04

2.62E-03

TOTAL DIOXINS (USING 0)

3 / 3

2.40E-04 - 7.43E-03

N/A

3.72E-03

7.43E-03

TOTAL DIOXINS (USING DL)

3 / 3

2.40E-04 - 7.43E-03

N/A

3.72E-03

7.43E-03

TOTAL DIOXINS (USING HALF DL)

3 / 3

2.40E-04 - 7.43E-03

N/A

3.72E-03

7.43E-03

TOTAL FURANS (USING 0)

3 / 3

4.17E-04 - 8.55E-03

N/A

3.23E-03

8.55E-03

TOTAL FURANS (USING DL)

3 / 3

4.17E-04 - 8.55E-03

N/A

3.23E-03

8.55E-03

TOTAL FURANS (USING HALF DL)

3 / 3

4.17E-04 - 8.55E-03

N/A

3.23E-03

8.55E-03

PCBS

AROCL OR-1254

1 / 22

9.30E-01 - 9.30E-01

2.60E-01 - 1.02E+01

5.43E-01

1.02E+01

AROCLOR-1260

22 / 22

1.01E+00 - 1.71E+02

N/A

1.10E+01

1.70E+02

PCB, TOTAL

22 / 22

1.01E+00 - 1.71E+02

N/A

1.10E+01

1.70E+02

METALS

ANTIMONY

3 / 3

1.02E+00 - 2.10E+00

N/A

1.60E+00

2.10E+00

ARSENIC

3 / 3

3.30E+00 - 5.70E+00

N/A

5.10E+00

5.70E+00

BARIUM

3 / 3

5.20E+01 - 1.01E+02

N/A

5.25E+01

1.01E+02

BERYLLIUM

2 / 3

5.70E-01 - 9.40E-01

5.85E-01 - 5.85E-01

5.85E-01

9.40E-01

CADMIUM

3 / 3

5.90E-01 - 2.80E+00

N/A

8.00E-01

2.80E+00

CHROMIUM

3 / 3

4.07E+01 - 1.15E+02

N/A

6.80E+01

1.15E+02

COBALT

3 / 3

9.40E+00 - 1.54E+01

N/A

9.90E+00

1.54E+01

COPPER

3 / 3

4.96E+01 - 1.02E+02

N/A

9.00E+01

1.02E+02

N/A = Not Applicable

Tables B-31-B-40.xls B-36

-------
Table B-36

Soil Chemistry Summary
5c Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

LEAD

3 / 3

5.71E+01 - 1.13E+02

N/A

8.37E+01

1.13E+02

MERCURY

3 / 3

3.30E-01 - 7.80E-01

N/A

6.20E-01

7.80E-01

NICKEL

3 / 3

1.53E+01 - 2.79E+01

N/A

1.67E+01

2.79E+01

SILVER

3 / 3

6.80E-01 - 6.30E+00

N/A

9.45E-01

6.30E+00

THALLIUM

2 / 3

6.60E-01 - 1.50E+00

4.40E-01 - 4.40E-01

6.60E-01

1.50E+00

TIN

3 / 3

8.00E+00 - 1.14E+01

N/A

9.50E+00

1.14E+01

VANADIUM

3 / 3

1.42E+01 - 2.64E+01

N/A

1.50E+01

2.64E+01

ZINC

3 / 3

1.42E+02 - 2.89E+02

N/A

1.69E+02

2.89E+02

ORGANIC

TOTAL ORGANIC CARBON

4 / 4

2.00E+04 - 5.54E+04

N/A

3.23E+04

5.54E+04

INORGANICS

PERCENT SOLIDS

17 / 17

0.00E+00 - 7.37E+01

N/A

5.40E+01

7.37E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-36

-------
Table B-37

Soil Chemistry Summary
6ab Floodplain
Housatonic River Site, OU2 - Pittsfield, MA







Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

PCBS

AROCL OR-1254

3 / 67

5.00E-02 - 2.10E+01

2.00E-02 - 1.15E+01

1.15E+00

5.42E+00

AROCLOR-1260

43 / 69

1.10E-01 - 3.21E+02

2.00E-02 - 1.66E+00

4.74E+00

8.96E+01

PCB, TOTAL

43 / 69

1.60E-01 - 3.21E+02

2.00E-02 - 1.66E+00

4.74E+00

8.96E+01

ORGANIC

TOTAL ORGANIC CARBON

4 / 4

9.03E+03 - 3.05E+05

N/A

5.37E+04

3.05E+05

INORGANICS

PERCENT SOLIDS

38 / 38

0.00E+00 - 9.19E+01

N/A

3.80E+01

8.12E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-37

-------
Table B-38

Soil Chemistry Summary
6ab Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PCBS

AROCLOR-1260

2 / 2

2.38E+01 - 2.49E+01

N/A

2.44E+01

2.49E+01

PCB, TOTAL

2 / 2

2.38E+01 - 2.49E+01

N/A

2.44E+01

2.49E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-38

-------
Table B-39

Soil Chemistry Summary
6cd Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median
(mg/kg)

95th Percentile
(mg/kg)

APP IX SEMEVOLATILES

ACETOPHENONE

1 / 7

6.60E-02

6.60E-02

4.00E-01

-

1.00E+00

6.10E-01

1.00E+00

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 7

7.30E-02

7.30E-02

4.00E-01

-

1.00E+00

6.10E-01

1.00E+00

DIBENZOFURAN

1 / 7

1.30E-01

1.30E-01

5.20E-01

-

1.00E+00

7.00E-01

1.00E+00

DI-N-BUTYL PHTHALATE

1 / 7

4.00E-02

4.00E-02

4.00E-01

-

1.00E+00

6.10E-01

1.00E+00

1,4-DICHLOROBENZENE

2 / 7

6.30E-02

6.50E-02

4.00E-01

-

9.70E-01

5.20E-01

9.70E-01

2-METHYLNAPHTHALENE

2 / 7

4.30E-02

2.40E-01

5.20E-01

-

1.00E+00

6.10E-01

1.00E+00

4-METHYLPHENOL

1 / 7

5.00E-02

5.00E-02

4.00E-01

-

1.00E+00

6.10E-01

1.00E+00

4-NITROPHENOL

1 / 7

1.50E+00

1.50E+00

1.00E+00

-

2.60E+00

1.80E+00

2.60E+00

P-PHENYLENEDIAMTNE

1 / 6

6.10E-01

6.10E-01

4.00E-01

-

1.00E+00

6.95E-01

1.00E+00

PAHs









ACENAPHTHYLENE

3 / 7

3.30E-02

2.10E-01

5.20E-01

-

1.00E+00

5.20E-01

1.00E+00

ANTHRACENE

3 / 7

4.40E-02

1.50E-01

5.20E-01

-

1.00E+00

5.20E-01

1.00E+00

BENZO(A)ANTHRACENE

6 / 7

5.10E-02

7.90E-01

7.80E-01

-

7.80E-01

1.90E-01

7.90E-01

BENZO(B)FLUORANTHENE

7 / 7

4.50E-02

1.40E+00

N/A

2.10E-01

1.40E+00

BENZO(K)FLUORANTHENE

7 / 7

4.40E-02

1.80E+00

N/A

2.30E-01

1.80E+00

BENZO(GHI)PERYLENE

6 / 7

4.50E-02

3.40E-01

7.80E-01

-

7.80E-01

1.50E-01

7.80E-01

BENZO(A)PYRENE

3 / 7

2.30E-01

1.10E+00

5.20E-01

-

1.00E+00

7.80E-01

1.10E+00

CHRYSENE

7 / 7

4.30E-02

1.20E+00

N/A

2.80E-01

1.20E+00

DIBENZO(A,H)ANTHRACENE

2 / 7

5.60E-02

6.50E-02

4.00E-01

-

1.00E+00

5.20E-01

1.00E+00

FLUORANTHENE

7 / 7

4.20E-02

8.00E-01

N/A

3.00E-01

8.00E-01

FLUORENE

1 / 7

4.30E-02

4.30E-02

4.00E-01

-

1.00E+00

6.10E-01

1.00E+00

INDENO( 1,2,3-C,D)PYRENE

6 / 7

4.20E-02

3.00E-01

7.80E-01

-

7.80E-01

1.40E-01

7.80E-01

NAPHTHALENE

6 / 7

4.70E-02

1.80E-01

5.20E-01

-

5.20E-01

8.20E-02

5.20E-01

PHENANTHRENE

6 / 7

6.30E-02

7.60E-01

7.80E-01

-

7.80E-01

2.40E-01

7.80E-01

PYRENE

7 / 7

4.50E-02

1.10E+00

N/A

3.50E-01

1.10E+00

TOTAL PAH (USING 0)

7 / 7

3.01E-01

1.01E+01

N/A

2.26E+00

1.01E+01

TOTAL PAH (USING DL)

7 / 7

3.71E+00

1.13E+01

N/A

6.77E+00

1.13E+01

NBA = No Benchmark Available

Tables B-31-B-40.xls B-39

-------
Table B-39

Soil Chemistry Summary
6cd Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median
(mg/kg)

95th Percentile
(mg/kg)

TOTAL PAH (USING HALF DL)

7 / 7

2.40E+00

1.07E+01

N/A

4.20E+00

1.07E+01

TOTAL PAH (HIGH) (USING 0)

7 / 7

1.77E-01

8.03E+00

N/A

1.63E+00

8.03E+00

TOTAL PAH (HIGH) (USING DL)

7 / 7

1.47E+00

8.43E+00

N/A

3.63E+00

8.43E+00

TOTAL PAH (HIGH) (USING HALF DL)

7 / 7

9.51E-01

8.23E+00

N/A

2.13E+00

8.23E+00

TOTAL PAH (LOW) (USING 0)

7 / 7

1.24E-01

2.10E+00

N/A

6.34E-01

2.10E+00

TOTAL PAH (LOW) (USING DL)

7 / 7

1.92E+00

4.63E+00

N/A

2.90E+00

4.63E+00

TOTAL PAH (LOW) (USING HALF DL)

7 / 7

1.31E+00

2.63E+00

N/A

2.07E+00

2.63E+00

DIOXIN S/FURAN S

TCDD (TOTAL)

6 / 7

1.70E-06

5.50E-05

1.38E-05

-

1.38E-05

2.04E-05

5.50E-05

TCDF (TOTAL)

7 / 7

7.12E-05

2.59E-03

N/A

6.42E-04

2.59E-03

TOTAL DIOXINS (USING 0)

7 / 7

1.98E-04

8.49E-03

N/A

1.69E-03

8.49E-03

TOTAL DIOXINS (USING DL)

7 / 7

3.05E-04

8.49E-03

N/A

1.69E-03

8.49E-03

TOTAL DIOXINS (USING HALF DL)

7 / 7

2.51E-04

8.49E-03

N/A

1.69E-03

8.49E-03

TOTAL FURANS (USING 0)

7 / 7

2.25E-04

8.63E-03

N/A

2.96E-03

8.63E-03

TOTAL FURANS (USING DL)

7 / 7

2.25E-04

8.63E-03

N/A

2.96E-03

8.63E-03

TOTAL FURANS (USING HALF DL)

7 / 7

2.25E-04

8.63E-03

N/A

2.96E-03

8.63E-03

PCBS

AROCLOR-1254

4 / 69

7.08E+00

2.20E+01

1.30E-01

-

1.03E+01

5.11E-01

1.11E+01

AROCLOR-1260

43 / 70

3.42E-01

1.10E+02

5.01E-01

-

1.47E+00

1.25E+00

7.66E+01

PCB, TOTAL

43 / 70

1.69E-01

1.30E+02

5.01E-01

-

1.47E+00

1.25E+00

7.66E+01

METALS

ANTIMONY

4 / 7

4.90E-01

1.80E+00

6.80E-01

-

1.80E+00

9.70E-01

1.80E+00

ARSENIC

6 / 7

4.90E+00

7.30E+00

1.02E+01

-

1.02E+01

6.40E+00

1.02E+01

BARIUM

7 / 7

5.70E+01

1.21E+02

N/A

9.45E+01

1.21E+02

BERYLLIUM

7 / 7

4.90E-01

1.00E+00

N/A

7.50E-01

1.00E+00

CADMIUM

5 / 7

1.50E-01

3.90E+00

5.00E-02

-

6.00E-02

6.30E-01

3.90E+00

CHROMIUM

7 / 7

1.14E+01

9.71E+01

N/A

3.14E+01

9.71E+01

COBALT

7 / 7

6.60E+00

1.43E+01

N/A

1.13E+01

1.43E+01

NBA = No Benchmark Available

Tables B-31-B-40.xls B-39

-------
Table B-39

Soil Chemistry Summary
6cd Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median
(mg/kg)

95th Percentile
(mg/kg)

COPPER

7 / 7

3.05E+01

1.26E+02

N/A

4.13E+01

1.26E+02

LEAD

7 / 7

4.37E+01

1.80E+02

N/A

1.03E+02

1.80E+02

MERCURY

7 / 7

7.00E-02

1.10E+00

N/A

4.20E-01

1.10E+00

NICKEL

7 / 7

1.36E+01

2.65E+01

N/A

1.92E+01

2.65E+01

SELENIUM

3 / 7

3.70E-01

1.30E+00

5.20E-01

-

1.00E+00

8.20E-01

1.30E+00

SILVER

6 / 7

2.80E-01

3.50E+00

1.70E-01

-

1.70E-01

1.60E+00

3.50E+00

THALLIUM

5 / 7

1.60E+00

2.40E+00

5.40E-01

-

3.10E+00

2.00E+00

3.10E+00

TIN

4 / 7

6.10E+00

1.42E+01

3.00E+00

-

5.80E+00

6.10E+00

1.42E+01

VANADIUM

7 / 7

1.80E+01

2.70E+01

N/A

2.33E+01

2.70E+01

ZINC

7 / 7

9.58E+01

3.02E+02

N/A

1.92E+02

3.02E+02

INORGANICS

PERCENT SOLIDS

59 / 59

0.00E+00

9.04E+01

N/A

6.47E+01

8.97E+01

ORGANIC

TOTAL ORGANIC CARBON

7 / 7

3.29E+04

1.63E+05



N/A



9.03E+04

1.63E+05

NBA = No Benchmark Available

Tables B-31-B-40.xls B-39

-------
Table B-40

Soil Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX SEMIVOLATILES

BENZYL ALCOHOL

1 / 19

5.90E-01 - 5.90E-01

3.50E-01 - 6.20E-01

4.20E-01

6.20E-01

BIS(2-ETHYLHEXYL) PHTHALATE

5 / 19

3.20E-02 - 5.00E-01

3.50E-01 - 6.20E-01

4.00E-01

6.20E-01

2-METHYLNAPHTHALENE

1 / 19

4.40E-02 - 4.40E-02

3.50E-01 - 6.20E-01

4.10E-01

6.20E-01

PAHs

ACENAPHTHYLENE

2 / 19

2.10E-02 - 4.30E-02

3.60E-01 - 6.20E-01

4.20E-01

6.20E-01

ANTHRACENE

1 / 19

1.90E-02 - 1.90E-02

3.60E-01 - 6.20E-01

4.20E-01

6.20E-01

BENZO(A)ANTHRACENE

10 / 19

2.20E-02 - 1.40E-01

3.60E-01 - 6.20E-01

1.40E-01

6.20E-01

BENZO(B)FLUORANTHENE

10 / 19

3.30E-02 - 1.80E-01

3.60E-01 - 6.20E-01

1.80E-01

6.20E-01

BENZO(K)FLUORANTHENE

10 / 19

3.00E-02 - 2.40E-01

3.60E-01 - 6.20E-01

2.40E-01

6.20E-01

BENZO(GHI)PERYLENE

7 / 19

2.00E-02 - 7.30E-02

3.50E-01 - 6.20E-01

3.80E-01

6.20E-01

BENZO(A)PYRENE

8 / 19

2.90E-02 - 1.40E-01

3.50E-01 - 6.20E-01

3.60E-01

6.20E-01

CHRYSENE

13 / 19

2.10E-02 - 2.30E-01

3.60E-01 - 6.20E-01

6.40E-02

6.20E-01

DIBENZO(A,H) ANTHRACENE

1 / 19

2.40E-02 - 2.40E-02

3.50E-01 - 6.20E-01

4.20E-01

6.20E-01

FLUORANTHENE

15 / 19

1.80E-02 - 2.20E-01

3.60E-01 - 6.20E-01

5.50E-02

6.20E-01

INDENO(l,2,3-C,D)PYRENE

9 / 19

2.00E-02 - 7.20E-02

3.50E-01 - 6.20E-01

3.50E-01

6.20E-01

NAPHTHALENE

1 / 19

4.20E-02 - 4.20E-02

3.50E-01 - 6.20E-01

4.10E-01

6.20E-01

PHENANTHRENE

12 / 19

2.20E-02 - 7.90E-02

4.00E-01 - 6.20E-01

5.90E-02

6.20E-01

PYRENE

15 / 19

2.00E-02 - 3.20E-01

3.60E-01 - 6.20E-01

6.10E-02

6.20E-01

TOTAL PAH

15 / 18

4.50E-02 - 1.74E+00

4.00E-01 - 6.20E-01

4.05E-01

1.74E+00

TOTAL PAH (HIGH MW)

14 / 18

3.40E-02 - 1.40E+00

3.60E-01 - 6.20E-01

3.61E-01

1.40E+00

TOTAL PAH (LOW MW)

15 / 18

2.20E-02 - 3.42E-01

4.00E-01 - 6.20E-01

8.00E-02

6.20E-01

APP IX PESTICIDES

4,4'-DDE

1 / 19

5.90E-02 - 5.90E-02

3.70E-03 - 4.60E-02

8.60E-03

5.90E-02

ALPHA-BHC

1 / 19

2.40E-03 - 2.40E-03

1.90E-03 - 2.30E-02

4.30E-03

2.30E-02

BETA-BHC

2 / 19

5.30E-03 - 8.50E-03

1.90E-03 - 2.30E-02

7.30E-03

2.30E-02

N/A = Not Applicable

Tables B-31-B-40.xls B-40

-------
Table B-40

Soil Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA











Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

HEPTACHLOR

1 /

18

6.00E-03

6.00E-03

1.90E-03

-

2.30E-02

6.65E-03

2.30E-02

HERBICIDES

2,4,5-T

2 / 6

6.00E-03

1.10E-02

5.40E-03

-

6.80E-03

5.90E-03

1.10E-02

2,4,5-TP (SILVEX)

2 / 6

6.00E-03

8.10E-03

5.40E-03

-

6.80E-03

5.90E-03

8.10E-03

DIOXINS/FURANS

TOTAL DIOXINS

19 /

19

9.00E-06

1.83E-03

N/A

1.37E-04

1.83E-03

TOTAL FURANS

19 /

19

2.45E-06

3.23E-04

N/A

1.26E-04

3.23E-04

PCBS

AROCL OR-1254

1 /

19

3.80E-02

3.80E-02

1.80E-02

-

5.08E-01

2.30E-02

5.08E-01

AROCLOR-1260

15 /

19

2.70E-02

5.80E-01

2.00E-02

-

3.18E-02

1.10E-01

5.80E-01

PCB, TOTAL

15 /

19

2.70E-02

5.80E-01

2.00E-02

-

3.18E-02

1.10E-01

5.80E-01

METALS

ANTIMONY

3 /

19

5.60E-01

1.60E+00

2.60E-01

-

8.00E-01

5.90E-01

1.60E+00

ARSENIC

19 /

19

1.30E+00

7.50E+00

N/A

5.00E+00

7.50E+00

BARIUM

19 /

19

1.91E+01

1.04E+02

N/A

3.62E+01

1.04E+02

BERYLLIUM

16 /

19

2.50E-01

9.50E-01

8.00E-02

-

5.60E-01

4.40E-01

9.50E-01

CADMIUM

1 /

19

1.50E-01

1.50E-01

2.00E-02

-

1.20E-01

5.00E-02

1.50E-01

CHROMIUM

19 /

19

5.80E+00

2.17E+01

N/A

1.16E+01

2.17E+01

COBALT

19 /

19

3.90E+00

1.62E+01

N/A

8.80E+00

1.62E+01

COPPER

18 /

19

5.60E+00

3.47E+01

5.90E+00

-

5.90E+00

1.86E+01

3.47E+01

LEAD

19 /

19

7.10E+00

3.92E+01

N/A

1.91E+01

3.92E+01

MERCURY

18 /

19

2.00E-02

5.90E-01

2.00E-02

-

2.00E-02

7.00E-02

5.90E-01

NICKEL

19

19

3.40E+00

2.65E+01

N/A

1.46E+01

2.65E+01

SELENIUM

2 /

19

8.80E-01

1.10E+00

2.10E-01

-

6.00E-01

3.30E-01

1.10E+00

SILVER

1 /

19

4.20E-01

4.20E-01

1.00E-01

-

2.00E-01

1.60E-01

4.20E-01

THALLIUM

12 /

19

6.10E-01

2.40E+00

7.90E-01

-

2.70E+00

1.50E+00

2.70E+00

N/A = Not Applicable

Tables B-31-B-40.xls B-40

-------
Table B-40

Soil Chemistry Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

VANADIUM

19 / 19

1.05E+01 - 2.62E+01

N/A

1.44E+01

2.62E+01

ZINC

19 / 19

3.55E+01 - 1.26E+02

N/A

6.69E+01

1.26E+02

ORGANIC

TOTAL ORGANIC CARBON

18 / 19

1.01E+03 - 4.66E+04

1.22E+02 - 1.22E+02

2.15E+04

4.66E+04

INORGANICS

PERCENT SOLIDS

11 / 11

6.79E+01 - 9.41E+01

N/A

8.33E+01

9.41E+01

N/A = Not Applicable

Tables B-31-B-40.xls B-40

-------
Table B-41

Fish Tissue Chemistry Summary
5a Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APP IX PESTICIDES

ALPHA-BHC

1 / 2

1.10E-04

1.10E-04

1.94E-03

-

1.94E-03

1.03E-03

1.94E-03

BETA-BHC

2 / 2

1.94E-04

6.07E-04

N/A

4.01E-04

6.07E-04

DELTA-BHC

1 / 2

2.32E-02

2.32E-02

1.94E-03

-

1.94E-03

1.26E-02

2.32E-02

CHLORPYRIFOS

2 / 2

1.04E-03

1.25E-03

N/A

1.14E-03

1.25E-03

GAMMA-CHLORDANE

1 / 2

1.07E-03

1.07E-03

1.94E-03

-

1.94E-03

1.51E-03

1.94E-03

0,P'-DDD

2 / 2

1.02E-01

1.46E-01

N/A

1.24E-01

1.46E-01

4,4'-DDD

2 / 2

2.31E-02

4.08E-02

N/A

3.20E-02

4.08E-02

4,4'-DDE

2 / 2

4.74E-02

6.54E-02

N/A

5.64E-02

6.54E-02

0,P'-DDT

2 / 2

1.72E-01

1.83E-01

N/A

1.78E-01

1.83E-01

4,4'-DDT

2 / 2

8.61E-03

1.26E-02

N/A

1.06E-02

1.26E-02

CIS-NONACHLOR

2 / 2

1.30E-01

1.30E-01

N/A

1.30E-01

1.30E-01

TRANS-NONACHLOR

2 / 2

3.81E-03

5.30E-03

N/A

4.55E-03

5.30E-03

DIELDRIN

1 / 2

3.04E-03

3.04E-03

1.94E-03

-

1.94E-03

2.49E-03

3.04E-03

ENDOSULFANII

1 / 2

5.63E-02

5.63E-02

1.99E-03

-

1.99E-03

2.91E-02

5.63E-02

ENDRIN

1 / 2

1.40E-03

1.40E-03

1.94E-03

-

1.94E-03

1.67E-03

1.94E-03

HEPTACHLOR EPOXIDE

1 / 2

8.54E-03

8.54E-03

1.94E-03

-

1.94E-03

5.24E-03

8.54E-03

HEXACHLOROBENZENE

2 / 2

6.80E-03

9.00E-03

N/A

7.90E-03

9.00E-03

OXYCHLORDANE

1 / 2

8.53E-03

8.53E-03

1.99E-03

-

1.99E-03

5.26E-03

8.53E-03

PENTACHLOROAMSOLE

2 / 2

9.72E-04

1.22E-03

N/A

1.09E-03

1.22E-03

PENTACHLOROBENZENE

2 / 2

1.06E-01

1.29E-01

N/A

1.18E-01

1.29E-01

TOXAPHENE

1 / 2

1.94E-02

1.94E-02

1.99E-02

-

1.99E-02

1.97E-02

1.99E-02

1,2,3,4-TETRACHLOROBENZENE

2 / 2

2.06E-01

2.76E-01

N/A

2.41E-01

2.76E-01

1,2,4,5-TETRACHLOROBENZENE

2 / 2

4.60E-02

4.60E-02

N/A

4.60E-02

4.60E-02

DIOXINS/FURANS

N/A = Not applicable.

Tables B-41-B-52.xls B-41

-------
Table B-41

Fish Tissue Chemistry Summary
5a Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL DIOXINS (USING 0)

2 / 2

0.00E+00 - 7.00E-07

N/A

3.50E-07

7.00E-07

TOTAL DIOXINS (USING DL)

2 / 2

3.49E-05 - 6.78E-05

N/A

5.13E-05

6.78E-05

TOTAL DIOXINS (USING HALF DL)

2 / 2

1.78E-05 - 3.39E-05

N/A

2.58E-05

3.39E-05

TOTAL FURANS (USING 0)

2 / 2

1.31E-04 - 1.57E-04

N/A

1.44E-04

1.57E-04

TOTAL FURANS (USING DL)

2 / 2

1.66E-04 - 2.23E-04

N/A

1.94E-04

2.23E-04

TOTAL FURANS (USING HALF DL)

2 / 2

1.48E-04 - 1.90E-04

N/A

1.69E-04

1.90E-04

PCBS

AROCLOR-1254

2 / 2

1.78E+01 - 1.80E+01

N/A

1.79E+01

1.80E+01

AROCLOR-1260

2 / 2

3.18E+01 - 3.30E+01

N/A

3.24E+01

3.30E+01

PCB, TOTAL

2 / 2

5.08E+01 - 5.14E+01

N/A

5.11E+01

5.14E+01

INORGANICS

PERCENT LIPIDS

2 / 2

4.10E+00 - 4.80E+00

N/A

4.45E+00

4.80E+00

N/A = Not applicable.

Tables B-41-B-52.xls B-41

-------
Table B-42

Fish Tissue Chemistry Summary
5a Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

3 / 46

2.50E-05

1.51E-03

1.84E-03

-

2.84E-03

1.97E-03

2.70E-03

0,P'-DDD

46 / 46

1.99E-03

4.34E-01

N/A

2.09E-01

3.62E-01

4,4'-DDD

46 / 46

1.46E-03

1.49E-01

N/A

3.63E-02

1.33E-01

0,P'-DDE

32 / 46

4.60E-05

2.49E-03

1.94E-03

-

2.86E-03

1.83E-03

2.70E-03

4,4'-DDE

46 / 46

1.93E-03

2.07E-01

N/A

8.43E-02

1.73E-01

0,P'-DDT

46 / 46

4.93E-03

7.30E-01

N/A

2.54E-01

6.37E-01

4,4'-DDT

37 / 46

1.42E-03

4.59E-02

1.87E-03

-

2.73E-03

1.88E-03

1.07E-02

ALPHA-BHC

36 / 46

3.20E-05

1.37E-03

1.84E-03

-

1.99E-03

3.38E-04

1.98E-03

BETA-BHC

19 / 46

3.50E-05

3.13E-03

1.87E-03

-

2.86E-03

1.94E-03

2.86E-03

DELTA-BHC

18 / 46

2.04E-03

1.51E-02

1.30E-05

-

2.86E-03

2.01E-03

4.53E-03

GAMMA BHC (LINDANE)

38 / 46

4.50E-05

1.99E-03

1.87E-03

-

2.73E-03

6.57E-04

2.42E-03

ALPHA-CHLORDANE

9 / 46

7.75E-04

1.45E-02

1.84E-03

-

2.84E-03

1.96E-03

1.02E-02

GAMMA-CHLORDANE

40 / 46

8.40E-05

3.71E-03

1.93E-03

-

2.65E-03

1.65E-03

3.23E-03

CHLORPYRIFOS

37 / 46

2.33E-04

2.93E-03

1.55E-04

-

1.94E-03

7.15E-04

1.97E-03

DIELDRIN

45 / 46

1.43E-04

1.37E-02

1.98E-03

-

1.98E-03

3.90E-03

1.14E-02

ENDOSULFANII

42 / 46

1.99E-03

1.94E-01

1.87E-03

-

1.99E-03

1.35E-02

1.70E-01

ENDRIN

24 / 46

4.47E-04

3.99E-03

1.87E-03

-

2.84E-03

1.88E-03

2.80E-03

HEPTACHLOR

19 / 46

5.80E-05

1.66E-03

1.87E-03

-

2.86E-03

1.92E-03

2.80E-03

HEPTACHLOR EPOXIDE

31 / 46

8.70E-05

1.55E-02

1.84E-03

-

2.86E-03

1.97E-03

1.36E-02

HEXACHLOROBENZENE

46 / 46

2.06E-04

1.17E-02

N/A

6.09E-03

1.10E-02

MIREX

16 / 46

1.42E-04

1.51E-03

1.84E-03

-

2.73E-03

1.93E-03

2.42E-03

CIS-NONACHLOR

46 / 46

4.67E-03

4.32E-01

N/A

1.55E-01

3.53E-01

TRANS-NONACHLOR

46 / 46

2.32E-04

3.25E-02

N/A

5.61E-03

1.67E-02

OXYCHLORDANE

20 / 46

1.63E-04

1.47E-02

1.87E-03

-

2.84E-03

1.98E-03

1.33E-02

PENT ACHL ORO ANIS OLE

46 / 46

5.60E-05

3.58E-03

N/A

1.95E-03

3.48E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-42

-------
Table B-42

Fish Tissue Chemistry Summary
5a Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

46 / 46

2.95E-03 - 2.20E-01

N/A

1.00E-01

1.90E-01

1,2,3,4-TETRACHLOROBENZENE

46 / 46

1.99E-03 - 3.85E-01

N/A

1.95E-01

3.81E-01

1,2,4,5-TETRACHLOROBENZENE

45 / 46

1.30E-03 - 6.93E-02

1.87E-03 - 1.87E-03

2.25E-02

6.62E-02

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

36 / 36

0.00E+00 - 1.13E-05

N/A

0.00E+00

8.54E-06

TOTAL DIOXINS (USING DL)

36 / 36

2.23E-05 - 7.21E-05

N/A

3.30E-05

7.18E-05

TOTAL DIOXINS (USING HALF DL)

36 / 36

1.11E-05 - 4.17E-05

N/A

1.67E-05

3.67E-05

TOTAL FURANS (USING 0)

36 / 36

5.06E-06 - 9.96E-04

N/A

1.76E-04

9.25E-04

TOTAL FURANS (USING DL)

36 / 36

5.11E-05 - 1.05E-03

N/A

2.33E-04

9.65E-04

TOTAL FURANS (USING HALF DL)

36 / 36

3.27E-05 - 1.02E-03

N/A

1.94E-04

9.42E-04

PCBS

AROCLOR-1248

6 / 46

9.87E-02 - 9.98E+00

1.86E-02 - 2.84E-02

1.97E-02

3.67E+00

AROCL OR-1254

46 / 46

4.10E-01 - 4.97E+01

N/A

2.24E+01

4.76E+01

AROCLOR-1260

46 / 46

1.64E+00 - 2.98E+02

N/A

6.65E+01

2.01E+02

PCB, TOTAL

46 / 46

2.05E+00 - 3.33E+02

N/A

9.01E+01

2.32E+02

INORGANICS

PERCENT LIPIDS

46 / 46

8.00E-02 - 7.84E+00

N/A

3.80E+00

7.11E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-42

-------
Table B-43

Fish Tissue Chemistry Summary
5a Large Fish
Housatonic River Site, OU2 - Pittsfield, MA









Range of Sample Quantitation

Median

95th Percentile

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Limits (mg/kg)

(mg/kg)

(mg/kg)

APPIX PESTICIDES

ALPHA-BHC

2 / 2

1.65E-04

2.71E-04

N/A

2.18E-04

2.71E-04

BETA-BHC

2 / 2

1.25E-04

6.64E-04

N/A

3.95E-04

6.64E-04

GAMMA BHC (LINDANE)

2 / 2

7.81E-05

5.40E-04

N/A

3.09E-04

5.40E-04

ALPHA-CHLORDANE

2 / 2

2.73E-03

9.41E-03

N/A

6.07E-03

9.41E-03

GAMMA-CHLORDANE

2 / 2

9.00E-04

1.65E-03

N/A

1.27E-03

1.65E-03

CHLORPYRIFOS

2 / 2

6.62E-04

1.03E-03

N/A

8.45E-04

1.03E-03

0,P'-DDD

2 / 2

4.62E-03

2.11E-01

N/A

1.08E-01

2.11E-01

4,4'-DDD

2 / 2

1.62E-03

2.68E-02

N/A

1.42E-02

2.68E-02

4,4'-DDE

2 / 2

5.51E-03

1.05E-01

N/A

5.55E-02

1.05E-01

0,P'-DDT

2 / 2

3.77E-03

2.97E-01

N/A

1.51E-01

2.97E-01

4,4'-DDT

2 / 2

3.84E-03

4.87E-03

N/A

4.36E-03

4.87E-03

DIELDRIN

2 / 2

1.28E-03

6.02E-03

N/A

3.65E-03

6.02E-03

ENDOSULFANII

2 / 2

4.39E-03

6.39E-02

N/A

3.41E-02

6.39E-02

ENDRIN

1 / 2

1.13E-03

1.13E-03

1.94E-03

-

1.94E-03

1.54E-03

1.94E-03

HEPTACHLOR EPOXIDE

1 / 2

1.63E-03

1.63E-03

2.06E-03

-

2.06E-03

1.84E-03

2.06E-03

HEXACHLOROBENZENE

2 / 2

7.72E-04

5.59E-03

N/A

3.18E-03

5.59E-03

MIREX

1 / 2

8.42E-04

8.42E-04

2.06E-03

-

2.06E-03

1.45E-03

2.06E-03

CIS-NONACHLOR

2 / 2

1.21E-02

1.52E-01

N/A

8.23E-02

1.52E-01

TRANS-NONACHLOR

2 / 2

2.97E-03

5.61E-03

N/A

4.29E-03

5.61E-03

OXYCHLORDANE

1 / 2

1.54E-03

1.54E-03

1.94E-03

-

1.94E-03

1.74E-03

1.94E-03

PENT ACHL ORO ANIS OLE

2 / 2

1.27E-04

4.17E-04

N/A

2.72E-04

4.17E-04

PENTACHLOROBENZENE

2 / 2

6.12E-03

7.47E-02

N/A

4.04E-02

7.47E-02

1,2,3,4-TETRACHLOROBENZENE

2 / 2

1.48E-02

1.06E-01

N/A

6.06E-02

1.06E-01

1,2,4,5-TETRACHLOROBENZENE

2 / 2

3.90E-03

2.01E-02

N/A

1.20E-02

2.01E-02

DIOXINS/FURANS

N/A = Not Applicable

Tables B-41-B-52.xls B-43

-------
Table B-43

Fish Tissue Chemistry Summary
5a Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL DIOXINS (USING 0)

2 / 2

0.00E+00 - 7.26E-07

N/A

3.63E-07

7.26E-07

TOTAL DIOXINS (USING DL)

2 / 2

3.47E-05 - 3.57E-05

N/A

3.52E-05

3.57E-05

TOTAL DIOXINS (USING HALF DL)

2 / 2

1.77E-05 - 1.79E-05

N/A

1.78E-05

1.79E-05

TOTAL FURANS (USING 0)

2 / 2

2.86E-05 - 4.94E-05

N/A

3.90E-05

4.94E-05

TOTAL FURANS (USING DL)

2 / 2

5.95E-05 - 8.44E-05

N/A

7.19E-05

8.44E-05

TOTAL FURANS (USING HALF DL)

2 / 2

4.40E-05 - 6.69E-05

N/A

5.55E-05

6.69E-05

PCBS

AROCL OR-1254

2 / 2

1.71E+00 - 2.86E+01

N/A

1.51E+01

2.86E+01

AROCLOR-1260

2 / 2

9.20E+00 - 6.68E+01

N/A

3.80E+01

6.68E+01

PCB, TOTAL

2 / 2

1.09E+01 - 9.54E+01

N/A

5.32E+01

9.54E+01

INORGANICS

PERCENT LIPIDS

2 / 2

6.45E-01 - 2.69E+00

N/A

1.67E+00

2.69E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-43

-------
Table B-44

Fish Tissue Chemistry Summary
5bc Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALPHA-BHC

3 / 6

2.08E-04

2.54E-04

1.88E-03

-

1.95E-03

1.07E-03

1.95E-03

BETA-BHC

3 / 6

1.81E-04

6.88E-04

1.93E-03

-

2.50E-03

1.31E-03

2.50E-03

GAMMA BHC (LINDANE)

6 / 6

5.34E-04

1.84E-03

N/A

6.53E-04

1.84E-03

ALPHA-CHLORDANE

3 / 6

6.90E-04

9.50E-03

1.95E-03

-

2.50E-03

1.96E-03

9.50E-03

GAMMA-CHLORDANE

5 / 6

1.09E-03

2.92E-03

1.88E-03

-

1.88E-03

1.66E-03

2.92E-03

CHLORPYRIFOS

5 / 6

1.66E-04

1.59E-03

1.93E-03

-

1.93E-03

5.87E-04

1.93E-03

0,P'-DDD

6 / 6

4.52E-02

1.26E-01

N/A

9.09E-02

1.26E-01

4,4'-DDD

5 / 6

1.71E-02

6.23E-02

1.97E-03

-

1.97E-03

2.14E-02

6.23E-02

4,4'-DDE

5 / 6

2.62E-02

1.10E-01

1.97E-03

-

1.97E-03

3.97E-02

1.10E-01

0,P'-DDT

6 / 6

5.05E-02

1.55E-01

N/A

7.97E-02

1.55E-01

4,4'-DDT

6 / 6

7.45E-04

5.22E-03

N/A

2.63E-03

5.22E-03

DIELDRIN

5 / 6

6.12E-04

2.69E-03

1.95E-03

-

1.95E-03

1.84E-03

2.69E-03

ENDOSULFANII

5 / 6

4.29E-03

2.05E-02

1.97E-03

-

1.97E-03

1.19E-02

2.05E-02

HEPTACHLOR

1 / 6

8.93E-04

8.93E-04

1.88E-03

-

2.94E-03

1.96E-03

2.94E-03

HEXACHLOROBENZENE

6 / 6

2.60E-03

6.43E-03

N/A

3.99E-03

6.43E-03

MIREX

1 / 6

4.16E-04

4.16E-04

1.88E-03

-

2.50E-03

1.94E-03

2.50E-03

CIS-NONACHLOR

5 / 6

3.22E-02

5.90E-02

1.97E-03

-

1.97E-03

4.30E-02

5.90E-02

TRANS-NONACHLOR

6 / 6

1.91E-03

1.16E-02

N/A

3.96E-03

1.16E-02

OXYCHLORDANE

5 / 6

4.29E-03

1.28E-02

1.97E-03

-

1.97E-03

6.02E-03

1.28E-02

PENT ACHL ORO ANIS OLE

6 / 6

9.62E-04

3.15E-03

N/A

1.42E-03

3.15E-03

PENTACHLOROBENZENE

4 / 6

1.97E-02

4.43E-02

1.97E-03

-

2.94E-03

2.94E-02

4.43E-02

1,2,3,4-TETRACHLOROBENZENE

6 / 6

3.76E-02

3.20E-01

N/A

8.24E-02

3.20E-01

1,2,4,5-TETRACHLOROBENZENE

6 / 6

8.02E-03

6.73E-02

N/A

1.63E-02

6.73E-02

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

6 / 6

0.00E+00

1.50E-06

N/A

0.00E+00

1.50E-06

N/A = Not Applicable

Tables B-41-B-52.xls B-44

-------
Table B-44

Fish Tissue Chemistry Summary
5bc Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

TOTAL DIOXINS (USING DL)

6 / 6

3.44E-05

6.99E-05

N/A

4.76E-05

6.99E-05

TOTAL DIOXINS (USING HALF DL)

6 / 6

1.80E-05

3.50E-05

N/A

2.38E-05

3.50E-05

TOTAL FURANS (USING 0)

6 / 6

1.99E-05

2.75E-04

N/A

7.43E-05

2.75E-04

TOTAL FURANS (USING DL)

6 / 6

9.76E-05

3.64E-04

N/A

1.42E-04

3.64E-04

TOTAL FURANS (USING HALF DL)

6 / 6

5.88E-05

3.19E-04

N/A

1.12E-04

3.19E-04

PCBS

AROCLOR-1248

2 / 6

1.65E+00

3.60E+00

1.95E-02 - 2.94E-02

2.72E-02

3.60E+00

AROCL OR-1254

6 / 6

5.78E+00

1.50E+01

N/A

8.70E+00

1.50E+01

AROCLOR-1260

6 / 6

9.08E+00

2.25E+01

N/A

1.89E+01

2.25E+01

PCB, TOTAL

6 / 6

1.65E+01

3.74E+01

N/A

2.90E+01

3.74E+01

INORGANICS

PERCENT LIPIDS

6 / 6

1.70E+00

4.00E+00

N/A

3.20E+00

4.00E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-44

-------
Table B-45

Fish Tissue Chemistry Summary
5bc Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

7 / 94

5.27E-04

1.57E-03

1.83E-03

-

2.35E-03

1.95E-03

2.27E-03

ALPHA-BHC

56 / 94

5.50E-05

1.68E-03

1.83E-03

-

4.99E-03

1.47E-03

2.09E-03

BETA-BHC

63 / 94

1.46E-05

3.08E-03

1.83E-03

-

2.34E-03

1.38E-03

2.48E-03

DELTA-BHC

37 / 94

1.20E-04

4.05E-03

1.83E-03

-

4.99E-03

1.91E-03

2.27E-03

GAMMA BHC (LINDANE)

94 / 94

6.50E-05

2.87E-03

N/A

8.96E-04

2.32E-03

ALPHA-CHLORDANE

31 / 94

1.18E-03

1.67E-02

1.83E-03

-

4.99E-03

1.95E-03

8.55E-03

GAMMA-CHLORDANE

66 / 94

2.47E-04

8.64E-03

1.83E-03

-

2.28E-03

2.01E-03

4.38E-03

CHLORPYRIFOS

74 / 94

7.29E-05

7.88E-03

1.53E-04

-

4.99E-03

1.14E-03

2.11E-03

0,P'-DDD

94 / 94

1.88E-03

6.00E-01

N/A

1.22E-01

3.36E-01

4,4'-DDD

92 / 94

1.49E-03

3.02E-01

1.96E-03

-

1.96E-03

2.44E-02

7.00E-02

0,P'-DDE

22 / 94

4.64E-04

2.08E-03

1.47E-03

-

4.99E-03

1.95E-03

2.29E-03

4,4'-DDE

91 / 94

4.39E-03

2.60E-01

1.91E-03

-

1.96E-03

6.51E-02

1.87E-01

0,P'-DDT

94 / 94

1.88E-03

2.28E+00

N/A

1.22E-01

3.63E-01

4,4'-DDT

82 / 94

5.96E-04

2.12E-02

1.92E-03

-

4.99E-03

2.98E-03

1.51E-02

DIELDRIN

91 / 94

6.75E-04

3.26E-02

1.93E-03

-

2.25E-03

2.80E-03

2.64E-02

ENDOSULFANII

70 / 94

1.72E-03

1.28E-01

1.86E-03

-

4.99E-03

1.38E-02

4.82E-02

ENDRIN

34 / 94

1.56E-04

3.70E-03

1.83E-03

-

2.34E-03

1.92E-03

2.28E-03

HEPTACHLOR

32 / 94

1.63E-04

3.37E-03

1.83E-03

-

4.99E-03

1.93E-03

2.32E-03

HEPTACHLOR EPOXIDE

15 / 94

1.01E-03

5.24E-03

1.08E-03

-

4.99E-03

1.96E-03

2.43E-03

HEXACHLOROBENZENE

94 / 94

4.59E-04

8.81E-03

N/A

3.69E-03

6.91E-03

MIREX

22 / 94

1.42E-04

6.23E-03

1.83E-03

-

4.99E-03

1.95E-03

4.08E-03

CIS-NONACHLOR

87 / 94

1.71E-03

5.76E-01

1.91E-03

-

1.99E-03

5.34E-02

1.77E-01

TRANS-NONACHLOR

94 / 94

4.93E-04

2.27E-02

N/A

5.19E-03

1.14E-02

OXYCHLORDANE

67 / 94

1.43E-03

1.73E-02

1.86E-03

-

4.99E-03

6.37E-03

1.31E-02

PENT ACHL ORO ANIS OLE

93 / 94

1.80E-04

5.57E-03

2.30E-04

-

2.30E-04

1.70E-03

4.05E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-45

-------
Table B-45

Fish Tissue Chemistry Summary
5bc Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

94 / 94

2.14E-03

1.20E-01

N/A

3.19E-02

8.38E-02

1,2,3,4-TETRACHLOROBENZENE

94 / 94

4.34E-03

3.51E-01

N/A

7.36E-02

2.00E-01

1,2,4,5-TETRACHLOROBENZENE

94 / 94

8.25E-04

9.71E-02

N/A

1.58E-02

6.13E-02

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

66 / 66

0.00E+00

1.51E-05

N/A

1.20E-06

1.22E-05

TOTAL DIOXINS (USING DL)

66 / 66

1.69E-05

3.43E-04

N/A

2.57E-05

5.56E-05

TOTAL DIOXINS (USING HALF DL)

66 / 66

9.15E-06

1.71E-04

N/A

1.38E-05

3.20E-05

TOTAL FURANS (USING 0)

66 / 66

0.00E+00

3.18E-03

N/A

1.77E-04

2.06E-03

TOTAL FURANS (USING DL)

66 / 66

5.09E-05

3.20E-03

N/A

2.25E-04

2.09E-03

TOTAL FURANS (USING HALF DL)

66 / 66

3.70E-05

3.19E-03

N/A

2.10E-04

2.08E-03

PCBS

AROCLOR-1242

1 / 94

1.99E-02

1.99E-02

1.83E-02

-

4.99E-02

1.96E-02

2.31E-02

AROCLOR-1248

46 / 94

1.99E-02

1.12E+01

1.86E-02

-

2.35E-02

2.31E-02

3.57E+00

AROCLOR-1254

93 / 94

9.06E-01

5.03E+01

2.27E-02

-

2.27E-02

1.07E+01

2.41E+01

AROCLOR-1260

94 / 94

1.42E+00

1.04E+02

N/A

2.93E+01

7.88E+01

PCB, TOTAL

94 / 94

2.59E+00

1.92E+02

N/A

3.98E+01

1.08E+02

INORGANICS

PERCENT LIPIDS

94 / 94

2.00E-01

8.55E+00

N/A

3.03E+00

6.62E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-45

-------
Table B-46

Fish Tissue Chemistry Summary
5bc Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

3 / 32

6.89E-04

2.45E-03

1.82E-03

-

2.09E-03

1.94E-03

2.22E-03

ALPHA-BHC

21 / 32

1.45E-04

1.94E-03

1.84E-03

-

2.00E-03

1.36E-03

1.97E-03

BETA-BHC

20 / 32

7.30E-05

2.80E-03

1.68E-03

-

1.98E-03

1.59E-03

2.42E-03

DELTA-BHC

8 / 32

1.02E-04

1.01E-02

1.11E-04

-

2.09E-03

1.92E-03

5.37E-03

GAMMA BHC (LINDANE)

32 / 32

5.23E-04

6.23E-03

N/A

1.48E-03

6.15E-03

ALPHA-CHLORDANE

4 / 32

1.78E-03

1.96E-02

1.82E-03

-

2.00E-03

1.94E-03

1.36E-02

GAMMA-CHLORDANE

31 / 32

1.04E-03

2.35E-02

1.94E-03

-

1.94E-03

3.16E-03

1.82E-02

CHLORPYRIFOS

23 / 32

2.25E-04

4.24E-03

2.77E-04

-

1.98E-03

1.41E-03

3.24E-03

0,P'-DDD

32 / 32

4.67E-02

1.09E+00

N/A

3.61E-01

1.06E+00

4,4'-DDD

32 / 32

1.93E-03

1.64E+00

N/A

4.32E-02

8.04E-01

0,P'-DDE

12 / 32

8.73E-04

7.67E-03

1.89E-03

-

2.09E-03

1.92E-03

4.70E-03

4,4'-DDE

32 / 32

3.61E-03

6.47E-01

N/A

1.07E-01

5.62E-01

0,P'-DDT

32 / 32

3.22E-02

1.19E+00

N/A

3.50E-01

1.19E+00

4,4'-DDT

16 / 32

1.46E-03

5.31E-02

1.82E-03

-

2.01E-03

1.97E-03

4.84E-02

DIELDRIN

30 / 32

1.53E-03

1.53E-02

1.91E-03

-

2.09E-03

3.54E-03

1.43E-02

ENDOSULFANII

17 / 32

4.49E-03

2.44E-01

1.82E-03

-

2.09E-03

5.54E-03

2.01E-01

ENDRIN

7 / 32

8.27E-04

3.24E-02

1.82E-03

-

2.01E-03

1.93E-03

1.27E-02

HEPTACHLOR

5 / 32

2.06E-04

1.45E-03

1.82E-03

-

2.09E-03

1.91E-03

2.04E-03

HEPTACHLOR EPOXIDE

11/32

2.05E-03

8.09E-03

1.89E-03

-

2.09E-03

1.96E-03

7.94E-03

HEXACHLOROBENZENE

32 / 32

2.34E-03

2.18E-02

N/A

5.62E-03

1.73E-02

MIREX

6 / 32

1.83E-04

5.59E-03

1.82E-03

-

2.00E-03

1.93E-03

4.94E-03

CIS-NONACHLOR

31 / 32

2.49E-03

1.16E+00

1.96E-03

-

1.96E-03

1.77E-01

9.34E-01

TRANS-NONACHLOR

30 / 32

3.28E-03

1.10E-01

1.89E-03

-

1.91E-03

7.98E-03

5.95E-02

OXYCHLORDANE

26 / 32

1.53E-03

6.33E-02

1.89E-03

-

2.00E-03

6.20E-03

5.86E-02

PENT ACHL ORO ANIS OLE

31 / 32

4.03E-04

6.43E-03

1.90E-03

-

1.90E-03

1.98E-03

6.16E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-46

-------
Table B-46

Fish Tissue Chemistry Summary
5bc Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

32 / 32

2.49E-02 - 4.21E-01

N/A

7.92E-02

3.45E-01

1,2,3,4-TETRACHLOROBENZENE

32 / 32

5.14E-02 - 1.18E+00

N/A

1.60E-01

9.47E-01

1,2,4,5-TETRACHLOROBENZENE

31 / 32

2.66E-03 - 2.84E-01

1.91E-03 - 1.91E-03

3.32E-02

1.81E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

25 / 25

0.00E+00 - 2.49E-05

N/A

1.80E-06

2.45E-05

TOTAL DIOXINS (USING DL)

25 / 25

2.05E-05 - 3.80E-05

N/A

2.59E-05

3.76E-05

TOTAL DIOXINS (USING HALF DL)

25 / 25

1.10E-05 - 3.14E-05

N/A

1.42E-05

3.10E-05

TOTAL FURANS (USING 0)

25 / 25

7.31E-05 - 1.69E-03

N/A

3.55E-04

1.61E-03

TOTAL FURANS (USING DL)

25 / 25

1.16E-04 - 1.71E-03

N/A

3.90E-04

1.63E-03

TOTAL FURANS (USING HALF DL)

25 / 25

1.00E-04 - 1.70E-03

N/A

3.79E-04

1.62E-03

PCBS

AROCLOR-1242

1 / 32

1.95E-02 - 1.95E-02

1.82E-02 - 2.09E-02

1.94E-02

2.04E-02

AROCLOR-1248

5 / 32

1.95E-02 - 9.07E+00

1.82E-02 - 2.09E-02

1.95E-02

3.67E+00

AROCLOR-1254

32 / 32

4.92E-01 - 1.65E+02

N/A

2.40E+01

1.14E+02

AROCLOR-1260

32 / 32

1.94E+01 - 4.22E+02

N/A

8.06E+01

3.09E+02

PCB, TOTAL

32 / 32

2.96E+01 - 4.24E+02

N/A

1.10E+02

4.17E+02

INORGANICS

PERCENT LIPIDS

32 / 32

2.10E+00 - 2.76E+01

N/A

5.79E+00

2.66E+01

N/A = Not Applicable

Tables B-41-B-52.xls B-46

-------
Table B-47

Fish Tissue Chemistry Summary
6cd Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

1 / 5

1.33E-04

1.33E-04

1.88E-03

-

2.46E-03

1.93E-03

2.46E-03

ALPHA-BHC

4 / 5

8.80E-05

1.71E-04

1.94E-03

-

1.94E-03

1.60E-04

1.94E-03

BETA-BHC

4 / 5

3.60E-05

2.94E-04

2.46E-03

-

2.46E-03

2.17E-04

2.46E-03

DELTA-BHC

1 / 5

9.06E-03

9.06E-03

1.93E-03

-

2.46E-03

1.94E-03

9.06E-03

GAMMA BHC (LINDANE)

5 / 5

1.73E-04

8.16E-04

N/A

5.50E-04

8.16E-04

ALPHA-CHLORDANE

3 / 5

1.62E-03

2.23E-03

1.88E-03

-

1.94E-03

1.94E-03

2.23E-03

GAMMA-CHLORDANE

4 / 5

5.44E-04

1.44E-03

1.94E-03

-

1.94E-03

1.21E-03

1.94E-03

CHLORPYRIFOS

1 / 5

2.69E-04

2.69E-04

1.93E-03

-

2.46E-03

1.94E-03

2.46E-03

0,P'-DDD

5 / 5

4.30E-02

1.05E-01

N/A

6.64E-02

1.05E-01

4,4'-DDD

5 / 5

5.63E-03

1.63E-02

N/A

1.33E-02

1.63E-02

4,4'-DDE

5 / 5

3.00E-02

6.82E-02

N/A

4.68E-02

6.82E-02

0,P'-DDT

5 / 5

6.44E-03

5.95E-02

N/A

5.23E-02

5.95E-02

4,4'-DDT

4 / 5

2.30E-04

3.62E-04

1.94E-03

-

1.94E-03

3.34E-04

1.94E-03

DIELDRIN

5 / 5

8.95E-04

2.48E-03

N/A

1.95E-03

2.48E-03

ENDOSULFANII

4 / 5

8.73E-03

2.33E-02

1.94E-03

-

1.94E-03

1.46E-02

2.33E-02

ENDRIN

3 / 5

1.26E-04

4.71E-04

1.88E-03

-

2.46E-03

4.71E-04

2.46E-03

HEPTACHLOR

2 / 5

1.72E-04

3.63E-04

1.93E-03

-

1.94E-03

1.93E-03

1.94E-03

HEPTACHLOR EPOXIDE

4 / 5

6.46E-03

1.35E-02

1.94E-03

-

1.94E-03

7.62E-03

1.35E-02

HEXACHLOROBENZENE

5 / 5

3.12E-04

2.41E-03

N/A

1.77E-03

2.41E-03

CIS-NONACHLOR

5 / 5

2.03E-02

4.72E-02

N/A

3.98E-02

4.72E-02

TRANS-NONACHLOR

5 / 5

1.02E-03

2.62E-03

N/A

2.16E-03

2.62E-03

PENT ACHL ORO ANIS OLE

4 / 5

9.97E-04

2.08E-03

1.72E-04

-

1.72E-04

1.24E-03

2.08E-03

PENTACHLOROBENZENE

5 / 5

2.58E-03

2.77E-02

N/A

1.86E-02

2.77E-02

1,2,3,4-TETRACHLOROBENZENE

5 / 5

4.75E-03

5.15E-02

N/A

3.94E-02

5.15E-02

1,2,4,5-TETRACHLOROBENZENE

5 / 5

1.58E-03

1.74E-02

N/A

1.48E-02

1.74E-02

N/A = Not Applicable

Tables B-41-B-52.xls B-47

-------
Table B-47

Fish Tissue Chemistry Summary
6cd Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

5 / 5

0.00E+00

3.30E-06

N/A

0.00E+00

3.30E-06

TOTAL DIOXINS (USING DL)

5 / 5

5.54E-05

3.87E-04

N/A

1.35E-04

3.87E-04

TOTAL DIOXINS (USING HALF DL)

5 / 5

2.77E-05

1.94E-04

N/A

6.73E-05

1.94E-04

TOTAL FURANS (USING 0)

5 / 5

5.61E-05

3.85E-04

N/A

5.72E-05

3.85E-04

TOTAL FURANS (USING DL)

5 / 5

1.18E-04

6.67E-04

N/A

3.13E-04

6.67E-04

TOTAL FURANS (USING HALF DL)

5 / 5

8.69E-05

5.26E-04

N/A

2.23E-04

5.26E-04

PCBS

AROCLOR-1248

5 / 5

4.52E-01

2.23E+00

N/A

1.98E+00

2.23E+00

AROCL OR-1254

5 / 5

4.07E+00

1.01E+01

N/A

8.45E+00

1.01E+01

AROCLOR-1260

5 / 5

4.52E+00

1.12E+01

N/A

1.06E+01

1.12E+01

PCB, TOTAL

5 / 5

9.04E+00

2.25E+01

N/A

2.11E+01

2.25E+01

INORGANICS

PERCENT LIPIDS

5 / 5

4.00E-01

3.80E+00

N/A

3.10E+00

3.80E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-47

-------
Table B-48

Fish Tissue Chemistry Summary
6cd Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

18 /

108

2.05E-04

1.66E-03

1.62E-03

-

3.29E-03

1.96E-03

3.26E-03

ALPHA-BHC

80

108

2.80E-05

1.78E-03

1.61E-03

-

3.29E-03

1.16E-03

2.07E-03

BETA-BHC

72 /

108

1.70E-05

7.14E-03

1.39E-03

-

3.27E-03

1.41E-03

2.52E-03

DELTA-BHC

59 /

108

1.55E-04

2.29E-02

1.20E-03

-

3.29E-03

1.98E-03

8.96E-03

GAMMA BHC (LINDANE)

105 /

108

1.22E-04

6.41E-03

1.90E-03

-

1.98E-03

6.55E-04

1.93E-03

ALPHA-CHLORDANE

53 /

108

9.22E-04

2.46E-01

1.32E-03

-

3.29E-03

1.93E-03

8.24E-03

GAMMA-CHLORDANE

76 /

108

1.15E-04

9.42E-02

1.28E-03

-

3.27E-03

1.92E-03

5.04E-03

CHLORPYRIFOS

81 /

108

3.40E-05

3.97E-03

6.96E-05

-

3.29E-03

7.96E-04

2.00E-03

0,P'-DDD

108 /

108

3.42E-02

7.08E-01

N/A

1.50E-01

5.20E-01

4,4'-DDD

108 /

108

2.20E-03

3.04E-01

N/A

2.39E-02

1.09E-01

0,P'-DDE

28 /

108

1.38E-03

5.15E-02

1.62E-03

-

3.29E-03

1.98E-03

1.97E-02

4,4'-DDE

108 /

108

1.78E-03

9.99E-01

N/A

8.95E-02

4.31E-01

0,P'-DDT

108 /

108

2.01E-03

1.19E+00

N/A

1.34E-01

6.13E-01

4,4'-DDT

63 /

108

1.59E-04

1.07E-02

1.23E-03

-

3.29E-03

1.81E-03

4.16E-03

DIELDRIN

99 /

107

4.53E-04

2.71E-01

1.96E-03

-

3.29E-03

3.32E-03

6.56E-02

ENDOSULFANII

78 /

108

2.84E-03

1.34E-01

1.87E-03

-

3.29E-03

1.90E-02

7.75E-02

ENDRIN

30 /

108

1.62E-04

1.77E-03

1.80E-03

-

3.29E-03

1.94E-03

3.26E-03

HEPTACHLOR

39 /

108

7.70E-05

2.85E-02

1.62E-03

-

3.29E-03

1.91E-03

3.29E-03

HEPTACHLOR EPOXIDE

55 /

108

4.97E-04

4.45E-02

1.62E-03

-

3.29E-03

2.02E-03

1.65E-02

HEXACHLOROBENZENE

105

108

1.95E-04

1.16E-02

1.90E-03

-

1.98E-03

2.12E-03

6.83E-03

MIREX

7 /

108

3.80E-05

1.77E-03

1.62E-03

-

3.29E-03

1.97E-03

3.26E-03

CIS-NONACHLOR

107 /

108

1.21E-02

5.35E-01

1.96E-03

-

1.96E-03

6.64E-02

3.51E-01

TRANS-NONACHLOR

108 /

108

5.92E-04

9.65E-02

N/A

3.96E-03

1.80E-02

OXYCHLORDANE

40 /

108

6.68E-04

3.07E-02

1.62E-03

-

3.29E-03

2.00E-03

2.19E-02

PENT ACHL ORO ANIS OLE

103 /

108

1.66E-04

1.13E-02

1.40E-04

-

1.96E-03

1.11E-03

3.28E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-48

-------
Table B-48

Fish Tissue Chemistry Summary
6cd Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

106 / 108

1.37E-03

1.69E-01

1.96E-03

-

3.29E-03

1.59E-02

6.65E-02

TOXAPHENE

1 / 108

2.13E-02

2.13E-02

1.62E-02

-

3.29E-02

1.97E-02

3.13E-02

1,2,3,4-TETRACHLOROBENZENE

106 / 108

2.29E-03

2.83E-01

1.96E-03

-

3.29E-03

3.20E-02

1.60E-01

1,2,4,5-TETRACHLOROBENZENE

108 / 108

1.43E-03

6.49E-02

N/A

1.03E-02

3.64E-02

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

69 / 69

0.00E+00

2.77E-05

N/A

2.79E-06

2.25E-05

TOTAL DIOXINS (USING DL)

69 / 69

7.50E-06

7.50E-05

N/A

2.94E-05

6.09E-05

TOTAL DIOXINS (USING HALF DL)

69 / 69

5.30E-06

3.78E-05

N/A

1.75E-05

3.22E-05

TOTAL FURANS (USING 0)

69 / 69

6.20E-06

9.23E-04

N/A

1.66E-04

6.93E-04

TOTAL FURANS (USING DL)

69 / 69

4.24E-05

9.69E-04

N/A

2.20E-04

8.40E-04

TOTAL FURANS (USING HALF DL)

69 / 69

2.62E-05

9.46E-04

N/A

1.82E-04

7.48E-04

PCBS

AROCLOR-1248

59 / 108

7.27E-02

1.46E+01

1.62E-02

-

3.29E-02

9.19E-02

7.52E+00

AROCL OR-1254

107 / 108

2.69E-01

1.48E+02

2.00E-02

-

2.00E-02

1.32E+01

6.09E+01

AROCLOR-1260

108 / 108

4.84E+00

2.21E+02

N/A

2.47E+01

9.52E+01

PCB, TOTAL

108 / 108

8.27E+00

3.40E+02

N/A

3.68E+01

1.60E+02

INORGANICS

PERCENT LIPIDS

108 / 108

9.26E-02

2.04E+01

N/A

3.65E+00

1.28E+01

N/A = Not Applicable

Tables B-41-B-52.xls B-48

-------
Table B-49

Fish Tissue Chemistry Summary
6cd Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

6 / 36

9.64E-04

2.35E-02

1.81E-03

-

3.26E-03

1.94E-03

6.29E-03

ALPHA-BHC

22 / 36

2.43E-05

1.69E-03

8.60E-05

-

3.26E-03

1.41E-03

2.19E-03

BETA-BHC

21 / 36

1.42E-04

5.48E-03

1.64E-03

-

3.26E-03

1.96E-03

5.20E-03

DELTA-BHC

15 / 36

3.78E-04

1.95E-02

1.63E-03

-

3.26E-03

1.92E-03

1.49E-02

GAMMA BHC (LINDANE)

36 / 36

1.01E-04

5.84E-03

N/A

9.17E-04

5.19E-03

ALPHA-CHLORDANE

8 / 36

1.27E-03

2.43E-02

1.81E-03

-

3.26E-03

1.91E-03

6.42E-03

GAMMA-CHLORDANE

27 / 36

5.24E-04

6.74E-02

1.87E-03

-

1.99E-03

1.99E-03

6.10E-02

CHLORPYRIFOS

24 / 36

2.46E-04

4.68E-03

1.57E-03

-

2.00E-03

1.39E-03

4.49E-03

0,P'-DDD

36 / 36

3.77E-02

1.46E+00

N/A

3.41E-01

1.35E+00

4,4'-DDD

36 / 36

8.04E-03

3.86E-01

N/A

3.78E-02

3.07E-01

0,P'-DDE

22 / 36

1.61E-03

6.44E-02

1.84E-03

-

3.26E-03

2.12E-03

5.46E-02

4,4'-DDE

36 / 36

2.66E-02

1.20E+00

N/A

1.80E-01

8.69E-01

0,P'-DDT

36 / 36

4.79E-02

1.42E+00

N/A

4.38E-01

1.25E+00

4,4'-DDT

15 / 36

5.50E-05

5.05E-02

1.81E-03

-

3.26E-03

1.92E-03

1.39E-02

DIELDRIN

35 / 36

1.72E-03

3.35E-01

1.95E-03

-

1.95E-03

6.05E-03

2.32E-01

ENDOSULFANII

25 / 36

1.77E-03

1.25E-01

1.84E-03

-

3.26E-03

1.89E-02

1.17E-01

ENDRIN

8 / 36

1.27E-04

1.69E-03

1.23E-03

-

3.26E-03

1.91E-03

2.22E-03

HEPTACHLOR

7 / 36

2.30E-04

3.69E-02

1.81E-03

-

3.26E-03

1.93E-03

3.59E-02

HEPTACHLOR EPOXIDE

20 / 36

1.41E-03

4.21E-02

1.81E-03

-

2.04E-03

2.00E-03

4.03E-02

HEXACHLOROBENZENE

36 / 36

5.95E-04

2.63E-02

N/A

3.29E-03

1.69E-02

MIREX

7 / 36

3.80E-04

2.25E-03

1.84E-03

-

3.26E-03

1.93E-03

2.40E-03

CIS-NONACHLOR

34 / 36

2.69E-02

9.71E-01

1.84E-03

-

1.98E-03

1.81E-01

7.81E-01

TRANS-NONACHLOR

35 / 36

1.77E-03

1.26E-01

3.26E-03

-

3.26E-03

7.17E-03

7.30E-02

OXYCHLORDANE

17 / 36

1.50E-03

3.39E-02

1.81E-03

-

3.26E-03

1.99E-03

2.62E-02

PENT ACHL ORO ANIS OLE

35 / 36

1.68E-04

1.06E-02

1.29E-04

-

1.29E-04

9.79E-04

5.81E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-49

-------
Table B-49

Fish Tissue Chemistry Summary
6cd Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

32 / 36

2.01E-03

3.12E-01

1.84E-03

-

1.98E-03

2.15E-02

3.11E-01

TOXAPHENE

1 / 36

3.26E-02

3.26E-02

1.81E-02

-

2.04E-02

1.95E-02

2.22E-02

1,2,3,4-TETRACHLOROBENZENE

32 / 36

8.74E-03

4.57E-01

1.84E-03

-

1.98E-03

4.13E-02

4.51E-01

1,2,4,5-TETRACHLOROBENZENE

33 / 36

3.55E-03

2.27E-01

1.82E-03

-

4.69E-03

1.57E-02

1.37E-01

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

26 / 26

0.00E+00

2.34E-05

N/A

4.07E-06

2.20E-05

TOTAL DIOXINS (USING DL)

26 / 26

1.41E-05

4.90E-05

N/A

2.40E-05

4.69E-05

TOTAL DIOXINS (USING HALF DL)

26 / 26

7.95E-06

3.22E-05

N/A

1.38E-05

3.16E-05

TOTAL FURANS (USING 0)

26 / 26

1.23E-05

1.11E-03

N/A

2.72E-04

9.74E-04

TOTAL FURANS (USING DL)

26 / 26

5.86E-05

1.13E-03

N/A

2.97E-04

9.91E-04

TOTAL FURANS (USING HALF DL)

26 / 26

5.22E-05

1.12E-03

N/A

2.85E-04

9.83E-04

PCBS

AROCLOR-1248

10 / 39

5.40E-01

2.24E+01

1.81E-02

-

1.00E+01

1.98E-02

2.10E+01

AROCL OR-1254

39 / 39

2.51E-01

2.24E+02

N/A

2.90E+01

1.10E+02

AROCLOR-1260

39 / 39

6.48E+00

3.05E+02

N/A

7.80E+01

2.18E+02

PCB, TOTAL

39 / 39

1.08E+01

4.48E+02

N/A

1.08E+02

3.22E+02

METALS

LEAD

3 / 6

8.00E-02

8.00E-02

8.34E-02

-

9.23E-02

8.17E-02

9.23E-02

MERCURY

6 / 6

2.20E-01

5.32E-01

N/A

3.00E-01

5.32E-01

NICKEL

1 / 6

1.42E-01

1.42E-01

1.01E-01

-

1.29E-01

1.20E-01

1.42E-01

INORGANICS

PERCENT LIPIDS

45 / 45

7.51E-01

2.83E+01

N/A

3.72E+00

2.08E+01

N/A = Not Applicable

Tables B-41-B-52.xls B-49

-------
Table B-50

Fish Tissue Chemistry Summary
Background - Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

8 / 20

4.00E-05

2.45E-03

1.97E-03

-

3.32E-03

1.98E-03

3.32E-03

ALPHA-BHC

20 / 20

2.40E-05

3.62E-04

N/A

1.68E-04

3.57E-04

BETA-BHC

12 / 20

5.30E-05

5.36E-03

1.97E-03

-

3.28E-03

3.71E-04

5.25E-03

DELTA-BHC

5 / 20

7.90E-05

3.82E-04

1.92E-03

-

3.28E-03

1.98E-03

3.27E-03

GAMMA BHC (LINDANE)

18 / 20

8.00E-06

1.46E-03

2.46E-03

-

3.13E-03

3.23E-04

3.09E-03

ALPHA-CHLORDANE

19 / 20

9.60E-05

2.71E-03

3.28E-03

-

3.28E-03

1.41E-03

3.25E-03

GAMMA-CHLORDANE

19 / 20

9.10E-05

1.67E-03

3.28E-03

-

3.28E-03

4.26E-04

3.20E-03

CHLORPYRIFOS

12 / 20

7.50E-05

7.05E-04

1.98E-03

-

2.00E-03

6.40E-04

2.00E-03

0,P'-DDD

15 / 20

4.02E-04

1.16E-02

1.98E-03

-

2.32E-03

1.68E-03

1.12E-02

4,4'-DDD

20 / 20

1.04E-03

2.61E-02

N/A

9.45E-03

2.59E-02

0,P'-DDE

2 / 20

5.70E-05

1.00E-04

1.92E-03

-

3.32E-03

1.99E-03

3.32E-03

4,4'-DDE

20 / 20

8.96E-03

4.72E-02

N/A

2.87E-02

4.72E-02

0,P'-DDT

8 / 20

1.00E-06

9.12E-03

1.92E-03

-

3.28E-03

1.97E-03

8.82E-03

4,4'-DDT

19 / 20

7.70E-05

2.13E-03

3.32E-03

-

3.32E-03

5.42E-04

3.26E-03

DIELDRIN

16 / 20

8.00E-06

1.97E-04

1.95E-03

-

3.28E-03

1.11E-04

3.23E-03

ENDOSULFANII

18 / 20

1.10E-05

9.03E-04

1.40E-05

-

4.30E-05

1.53E-04

8.72E-04

ENDRIN

12 / 20

5.00E-05

3.48E-04

1.98E-03

-

3.32E-03

2.89E-04

3.32E-03

HEPTACHLOR

18 / 20

6.20E-05

9.92E-04

1.98E-03

-

3.13E-03

3.55E-04

3.07E-03

HEPTACHLOR EPOXIDE

15 / 20

1.00E-05

4.29E-04

1.95E-03

-

3.22E-03

3.43E-04

3.21E-03

HEXACHLOROBENZENE

20 / 20

1.97E-04

2.97E-03

N/A

1.17E-03

2.96E-03

MIREX

10 / 20

5.50E-05

2.92E-04

1.92E-03

-

3.28E-03

1.10E-03

3.24E-03

CIS-NONACHLOR

20 / 20

5.00E-05

2.30E-03

N/A

1.22E-03

2.30E-03

TRANS-NONACHLOR

20 / 20

1.33E-04

8.77E-03

N/A

3.29E-03

8.77E-03

OXYCHLORDANE

20 / 20

1.43E-04

4.74E-03

N/A

1.83E-03

4.73E-03

PENT ACHL ORO ANIS OLE

20 / 20

1.94E-04

1.44E-03

N/A

7.61E-04

1.43E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-50

-------
Table B-50

Fish Tissue Chemistry Summary
Background - Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

19 / 20

1.57E-04

3.92E-03

2.46E-03

-

2.46E-03

6.23E-04

3.86E-03

1,2,3,4-TETRACHLOROBENZENE

19 / 20

7.70E-05

1.56E-03

1.99E-03

-

1.99E-03

4.03E-04

1.97E-03

1,2,4,5-TETRACHLOROBENZENE

19 / 20

5.10E-05

2.33E-03

1.95E-03

-

1.95E-03

4.65E-04

2.31E-03

DIOXINS/FURANS

1,2,3,4,6,7,8-HPCDF

11/20

6.00E-07

5.40E-06

3.90E-06

-

9.40E-06

4.05E-06

9.35E-06

1,2,3,4,7,8,9-HPCDF

1 / 20

1.20E-06

1.20E-06

3.90E-06

-

9.40E-06

4.85E-06

9.40E-06

1,2,3,4,7,8-HXCDF

1 / 20

1.20E-06

1.20E-06

3.90E-06

-

9.40E-06

4.85E-06

9.40E-06

1,2,3,6,7,8-HXCDF

2 / 20

8.00E-07

9.00E-07

3.90E-06

-

9.40E-06

4.80E-06

9.40E-06

1,2,3,7,8-PECDF

14 / 20

9.00E-07

6.00E-06

3.90E-06

-

7.10E-06

2.90E-06

7.09E-06

2,3,4,6,7,8-HXCDF

1 / 20

4.00E-07

4.00E-07

3.90E-06

-

9.40E-06

4.85E-06

9.40E-06

2,3,4,7,8-PECDF

5 / 20

1.00E-07

3.00E-07

3.90E-06

-

9.40E-06

4.55E-06

9.35E-06

2,3,7,8-TCDD

1 / 20

7.00E-07

7.00E-07

8.00E-07

-

1.90E-06

1.00E-06

1.90E-06

2,3,7,8-TCDF

14 / 20

8.00E-07

2.90E-06

8.00E-07

-

1.40E-06

1.05E-06

2.85E-06

OCDD

1 / 20

1.00E-06

1.00E-06

7.80E-06

-

1.88E-05

9.70E-06

1.88E-05

OCDF

2 / 20

3.00E-07

9.00E-07

7.80E-06

-

1.88E-05

9.50E-06

1.87E-05

PCBS

AROCLOR-1242

12 / 20

1.50E-03

1.62E-02

1.95E-02

-

3.32E-02

1.41E-02

3.32E-02

AROCLOR-1248

16 / 20

2.80E-03

3.87E-02

2.32E-02

-

3.32E-02

1.93E-02

3.84E-02

AROCLOR-1254

19 / 20

3.00E-03

7.73E-02

3.32E-02

-

3.32E-02

4.01E-02

7.72E-02

AROCLOR-1260

20 / 20

2.90E-03

2.36E+00

N/A

6.51E-02

2.26E+00

PCB, TOTAL

20 / 20

1.64E-02

2.36E+00

N/A

1.49E-01

2.26E+00

INORGANICS

PERCENT LIPIDS

20 / 20

1.70E+00

5.70E+00



N/A



3.90E+00

5.70E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-50

-------
Table B-51

Fish Tissue Chemistry Summary
Background - Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

61 / 114

7.00E-06 - 2.33E-03

1.82E-03 - 4.13E-03

1.55E-03

1.99E-03

ALPHA-BHC

106 / 114

1.90E-05 - 1.54E-03

1.01E-04 - 1.98E-03

1.70E-04

1.90E-03

BETA-BHC

90/114

4.80E-05 - 3.85E-03

1.55E-04 - 2.00E-03

1.14E-03

2.19E-03

DELTA-BHC

72/114

2.90E-05 - 1.57E-03

1.83E-03 - 2.39E-03

1.00E-03

1.99E-03

GAMMA BHC (LINDANE)

97/114

1.70E-05 - 3.43E-03

1.83E-03 - 4.13E-03

5.89E-04

2.02E-03

ALPHA-CHLORDANE

92/114

6.00E-05 - 6.31E-03

1.89E-03 - 2.00E-03

1.77E-03

3.89E-03

GAMMA-CHLORDANE

90/114

1.00E-04 - 2.24E-03

1.89E-03 - 2.01E-03

1.09E-03

1.99E-03

CHLORPYRIFOS

66/114

2.50E-05 - 1.59E-03

1.82E-03 - 4.13E-03

1.49E-03

1.99E-03

0,P'-DDD

109 / 114

6.59E-05 - 6.66E-03

1.93E-03 - 1.99E-03

1.21E-03

3.72E-03

4,4'-DDD

114 / 114

3.06E-04 - 3.11E-02

N/A

6.87E-03

2.25E-02

0,P'-DDE

56/114

4.20E-05 - 1.81E-03

1.83E-03 - 4.13E-03

1.86E-03

2.00E-03

4,4'-DDE

114 / 114

1.88E-03 - 1.67E-01

N/A

3.05E-02

9.28E-02

0,P'-DDT

39 / 114

1.00E-06 - 2.40E-03

1.83E-03 - 4.13E-03

1.94E-03

2.00E-03

4,4'-DDT

110 / 114

4.80E-05 - 6.53E-03

1.85E-03 - 1.98E-03

8.15E-04

3.07E-03

DIELDRIN

100 / 114

7.00E-06 - 1.78E-03

1.85E-03 - 1.99E-03

5.60E-04

1.97E-03

ENDOSULFANII

101 / 114

1.20E-05 - 1.56E-03

1.50E-05 - 2.00E-03

2.43E-04

1.95E-03

ENDRIN

58 / 114

1.80E-05 - 2.22E-03

1.84E-03 - 4.13E-03

1.85E-03

1.99E-03

HEPTACHLOR

96/114

2.30E-05 - 1.65E-03

1.85E-03 - 2.01E-03

5.53E-04

1.98E-03

HEPTACHLOR EPOXIDE

53 / 114

5.00E-06 - 9.86E-03

1.44E-03 - 4.13E-03

1.91E-03

2.54E-03

HEXACHLOROBENZENE

113 / 114

3.90E-05 - 1.85E-02

1.98E-03 - 1.98E-03

1.41E-03

3.72E-03

MIREX

63/114

2.70E-05 - 1.57E-03

1.83E-03 - 4.13E-03

1.38E-03

1.99E-03

OXYCHLORDANE

101 / 114

6.36E-05 - 5.32E-03

1.92E-03 - 1.99E-03

1.84E-03

3.65E-03

CIS-NONACHLOR

106 / 114

3.99E-05 - 4.59E-03

1.89E-03 - 2.00E-03

1.51E-03

3.76E-03

TRANS-NONACHLOR

112 / 114

3.40E-05 - 1.53E-02

1.97E-03 - 2.00E-03

4.10E-03

1.17E-02

PENT ACHL ORO ANIS OLE

109 / 114

7.62E-05 - 1.71E-03

1.43E-03 - 1.99E-03

7.15E-04

1.62E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-51

-------
Table B-51

Fish Tissue Chemistry Summary
Background - Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

104 / 114

2.50E-05 - 5.46E-03

9.74E-04 - 2.00E-03

1.09E-03

3.30E-03

1,2,3,4-TETRACHLOROBENZENE

104 / 114

4.30E-05 - 2.19E-03

1.33E-03 - 1.98E-03

6.24E-04

1.96E-03

1,2,4,5-TETRACHLOROBENZENE

101 / 114

4.90E-05 - 8.23E-03

1.36E-03 - 1.96E-03

9.33E-04

4.29E-03

DIOXINS/FURANS

1,2,3,4,6,7,8-HPCDD

13 / 101

5.00E-07 - 4.80E-06

2.90E-06 - 1.72E-05

4.42E-06

4.93E-06

1,2,3,4,6,7,8-HPCDF

53 / 101

4.00E-07 - 8.07E-06

3.60E-06 - 1.72E-05

4.09E-06

5.92E-06

1,2,3,4,7,8,9-HPCDF

7 / 101

2.00E-07 - 5.18E-06

2.90E-06 - 1.72E-05

4.47E-06

5.17E-06

1,2,3,4,7,8-HXCDD

1 / 101

1.50E-06 - 1.50E-06

2.90E-06 - 1.72E-05

4.53E-06

5.52E-06

1,2,3,4,7,8-HXCDF

7 / 101

8.00E-07 - 4.00E-06

2.90E-06 - 7.89E-06

4.47E-06

5.08E-06

1,2,3,6,7,8-HXCDD

8 / 101

8.00E-07 - 4.57E-06

2.90E-06 - 1.72E-05

4.47E-06

5.08E-06

1,2,3,6,7,8-HXCDF

5 / 101

3.00E-07 - 3.93E-06

2.90E-06 - 1.72E-05

4.47E-06

5.00E-06

1,2,3,7,8,9-HXCDD

6 / 101

5.00E-07 - 4.44E-06

2.90E-06 - 1.72E-05

4.44E-06

4.99E-06

1,2,3,7,8,9-HXCDF

4 / 101

1.60E-06 - 4.41E-06

2.90E-06 - 1.72E-05

4.50E-06

5.00E-06

1,2,3,7,8-PECDD

4 / 101

5.00E-07 - 4.00E-06

2.90E-06 - 1.72E-05

4.52E-06

5.08E-06

1,2,3,7,8-PECDF

64 / 101

2.27E-07 - 1.06E-05

2.55E-06 - 5.00E-06

3.43E-06

5.84E-06

2,3,4,6,7,8-HXCDF

7 / 101

2.07E-06 - 4.43E-06

2.90E-06 - 1.72E-05

4.47E-06

5.00E-06

2,3,4,7,8-PECDF

38 / 101

3.00E-07 - 3.80E-06

2.58E-06 - 1.72E-05

3.70E-06

4.90E-06

2,3,7,8-TCDD

6 / 101

3.00E-07 - 6.36E-07

6.00E-07 - 3.40E-06

9.00E-07

1.08E-06

2,3,7,8-TCDF

79 / 101

3.68E-07 - 6.03E-06

6.00E-07 - 6.74E-06

1.00E-06

3.55E-06

OCDD

16 / 101

8.00E-07 - 1.26E-05

5.90E-06 - 3.44E-05

8.85E-06

1.02E-05

OCDF

9 / 101

2.00E-07 - 1.12E-05

5.90E-06 - 3.44E-05

9.00E-06

1.02E-05

PCBS

AROCLOR-1242

61 / 114

2.50E-03 - 5.19E-02

1.82E-02 - 4.13E-02

1.94E-02

3.12E-02

AROCLOR-1248

88/114

1.20E-03 - 7.58E-02

1.85E-02 - 4.13E-02

1.94E-02

4.66E-02

AROCLOR-1254

103 / 114

1.60E-03 - 1.23E-01

1.91E-02 - 1.99E-02

3.75E-02

1.08E-01

AROCLOR-1260

108 / 114

1.40E-03 - 1.01E+00

1.96E-02 - 1.98E-02

5.39E-02

1.50E-01

N/A = Not Applicable

Tables B-41-B-52.xls B-51

-------
Table B-51

Fish Tissue Chemistry Summary
Background - Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PCB, TOTAL

114 / 114

9.06E-03 - 1.01E+00

N/A

1.16E-01

2.66E-01

METALS

MERCURY

1 / 1

1.94E-01 - 1.94E-01

N/A

1.94E-01

1.94E-01

INORGANICS

PERCENT LIPIDS

115 / 115

5.32E-01 - 6.14E+00

N/A

2.56E+00

4.80E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-51

-------
Table B-52

Fish Tissue Chemistry Summary
Background - Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

APPIX PESTICIDES

ALDRIN

10 /

15

3.70E-05

1.65E-03

1.85E-03

-

1.98E-03

1.46E-03

1.98E-03

ALPHA-BHC

10 /

15

1.60E-04

1.56E-03

2.00E-05

-

8.80E-05

2.07E-04

1.56E-03

BETA-BHC

12 /

15

2.28E-04

1.67E-03

1.85E-03

-

1.93E-03

1.51E-03

1.93E-03

DELTA-BHC

8 /

15

7.93E-05

1.64E-03

1.85E-03

-

1.98E-03

1.64E-03

1.98E-03

GAMMA BHC (LINDANE)

12 /

15

4.10E-05

1.55E-03

1.85E-03

-

1.92E-03

1.04E-03

1.92E-03

ALPHA-CHLORDANE

6 /

15

5.29E-04

1.63E-03

1.85E-03

-

1.99E-03

1.91E-03

1.99E-03

GAMMA-CHLORDANE

3 /

15

7.71E-04

1.47E-03

1.85E-03

-

1.98E-03

1.94E-03

1.98E-03

CHLORPYRIFOS

14 /

15

4.16E-04

1.78E-03

1.98E-03

-

1.98E-03

1.07E-03

1.98E-03

0,P'-DDD

13 /

15

9.56E-05

2.49E-03

1.98E-03

-

1.98E-03

5.92E-04

2.49E-03

4,4'-DDD

15

15

1.29E-03

7.78E-03

N/A

3.91E-03

7.78E-03

0,P'-DDE

7 /

15

8.60E-05

1.66E-03

1.85E-03

-

1.99E-03

1.85E-03

1.99E-03

4,4'-DDE

15 /

15

2.08E-02

1.14E-01

N/A

5.18E-02

1.14E-01

0,P'-DDT

6 /

15

7.10E-05

1.45E-03

1.90E-03

-

1.99E-03

1.94E-03

1.99E-03

4,4'-DDT

15 /

15

4.30E-05

1.06E-03

N/A

3.02E-04

1.06E-03

DIELDRIN

15 /

15

7.80E-05

1.56E-03

N/A

4.65E-04

1.56E-03

ENDOSULFANII

10 /

15

7.65E-05

1.53E-03

1.85E-03

-

1.98E-03

1.25E-03

1.98E-03

ENDRIN

5 /

15

3.65E-04

1.72E-03

1.85E-03

-

1.99E-03

1.92E-03

1.99E-03

HEPTACHLOR

9 /

15

1.62E-04

1.59E-03

1.85E-03

-

1.97E-03

1.44E-03

1.97E-03

HEPTACHLOR EPOXIDE

3 /

15

9.28E-04

2.31E-03

1.85E-03

-

1.99E-03

1.95E-03

2.31E-03

HEXACHLOROBENZENE

11 /

15

2.50E-05

1.38E-02

1.92E-03

-

1.98E-03

1.50E-03

1.38E-02

MIREX

5 /

15

3.00E-05

6.43E-04

1.90E-03

-

1.99E-03

1.93E-03

1.99E-03

CIS-NONACHLOR

15 /

15

1.05E-04

1.37E-03

N/A

3.31E-04

1.37E-03

TRANS-NONACHLOR

15 /

15

1.41E-04

2.69E-03

N/A

6.80E-04

2.69E-03

OXYCHLORDANE

9 /

15

3.35E-04

1.58E-03

1.85E-03

-

1.97E-03

1.28E-03

1.97E-03

PENT ACHL ORO ANIS OLE

9 /

15

1.92E-04

1.50E-03

5.00E-05

-

1.98E-03

7.64E-04

1.98E-03

N/A = Not Applicable

Tables B-41-B-52.xls B-52

-------
Table B-52

Fish Tissue Chemistry Summary
Background - Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

PENTACHLOROBENZENE

9 / 15

1.64E-04

8.98E-03

3.84E-04

-

1.90E-03

1.61E-03

8.98E-03

1,2,3,4-TETRACHLOROBENZENE

13 / 15

6.07E-04

3.55E-03

2.96E-04

-

3.83E-04

1.14E-03

3.55E-03

1,2,4,5-TETRACHLOROBENZENE

10 / 15

2.43E-04

5.29E-03

9.66E-04

-

1.95E-03

1.95E-03

5.29E-03

DIOXINS/FURANS

1,2,3,4,6,7,8-HPCDD

2 / 9

3.81E-06

4.16E-06

3.94E-06

-

4.85E-06

4.62E-06

4.85E-06

1,2,3,4,6,7,8-HPCDF

8 / 9

9.32E-07

4.32E-06

4.33E-06

-

4.33E-06

1.65E-06

4.33E-06

1,2,3,4,7,8,9-HPCDF

1 / 9

3.71E-06

3.71E-06

3.94E-06

-

4.85E-06

4.64E-06

4.85E-06

1,2,3,4,7,8-HXCDF

2 / 9

3.15E-06

3.28E-06

4.33E-06

-

4.85E-06

4.64E-06

4.85E-06

1,2,3,6,7,8-HXCDD

1 / 9

3.77E-06

3.77E-06

3.94E-06

-

4.87E-06

4.64E-06

4.87E-06

1,2,3,6,7,8-HXCDF

2 / 9

3.14E-06

3.15E-06

4.33E-06

-

4.85E-06

4.64E-06

4.85E-06

1,2,3,7,8,9-HXCDF

3 / 9

3.17E-06

3.95E-06

4.33E-06

-

4.85E-06

4.62E-06

4.85E-06

1,2,3,7,8-PECDD

2 / 9

3.77E-06

3.90E-06

3.94E-06

-

4.87E-06

4.62E-06

4.87E-06

1,2,3,7,8-PECDF

9 / 9

2.72E-07

3.56E-06

N/A

1.64E-06

3.56E-06

2,3,4,6,7,8-HXCDF

2 / 9

3.11E-06

3.21E-06

4.33E-06

-

4.85E-06

4.64E-06

4.85E-06

2,3,4,7,8-PECDF

4 / 9

1.00E-07

3.61E-06

4.62E-06

-

4.87E-06

4.62E-06

4.87E-06

2,3,7,8-TCDD

2 / 9

3.18E-07

7.53E-07

8.00E-07

-

1.00E-06

9.00E-07

1.00E-06

2,3,7,8-TCDF

7 / 9

4.43E-07

1.28E-06

8.00E-07

-

1.00E-06

7.18E-07

1.28E-06

OCDD

3 / 9

4.56E-06

9.68E-06

7.91E-06

-

9.69E-06

9.27E-06

9.69E-06

OCDF

4 / 9

4.96E-06

9.11E-06

9.27E-06

-

9.69E-06

9.27E-06

9.69E-06

PCBS

AROCLOR-1242

10 / 15

4.94E-03

9.14E-02

1.85E-02

-

1.97E-02

1.85E-02

9.14E-02

AROCLOR-1248

13 / 15

1.30E-03

3.91E-02

1.97E-02

-

2.00E-02

1.41E-02

3.91E-02

AROCLOR-1254

15 / 15

5.96E-03

5.77E-02

N/A

2.21E-02

5.77E-02

AROCLOR-1260

15 / 15

4.22E-03

3.79E-02

N/A

1.91E-02

3.79E-02

PCB, TOTAL

15 / 15

2.11E-02

1.92E-01

N/A

4.04E-02

1.92E-01

METALS

N/A = Not Applicable

Tables B-41-B-52.xls B-52

-------
Table B-52

Fish Tissue Chemistry Summary
Background - Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Range of Detects (mg/kg)

Range of Sample Quantitation
Limits (mg/kg)

Median

(mg/kg)

95th Percentile

(mg/kg)

MERCURY

5 / 5

2.14E-01 - 3.05E-01

N/A

2.52E-01

3.05E-01

NICKEL

2 / 5

1.03E-01 - 1.88E-01

1.18E-01 - 1.23E-01

1.21E-01

1.88E-01

INORGANICS

PERCENT LIPIDS

20 / 20

5.00E-01 - 3.98E+00

N/A

1.63E+00

3.94E+00

N/A = Not Applicable

Tables B-41-B-52.xls B-52

-------
Table B-53

Pre-ERA Sediment Benchmarks
Housatonic River Site, OU2, Pittsfield, MA



MacDonald et al.a

OMEEb

Ingersoll et al.c



Selected Benchmarks6



(mg/kg)

(mg/kg)

(mg/kg)

Sediment PRGsd

(mg/kg)



Chemical

TECs

PECs

LEL

SEL

Lowest

Higher

(mg/kg)

Low

High

Basis

Semivolatiles

Acetophenone

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

Bis(2-ethylhexyl)phthalate

NBA

NBA

NBA

NBA

NBA

NBA

2.70E+00

2.70E+00

NBA

PRG

Butyl benzylphthalate

NBA

NBA

NBA

NBA

NBA

NBA

1.10E+01

1.10E+01

NBA

PRG

Dibenzofuran

NBA

NBA

NBA

NBA

NBA

NBA

4.20E-01

4.20E-01

NBA

PRG

1,3-Dichlorobenzene

NBA

NBA

NBA

NBA

NBA

NBA

1.70E+00

1.70E+00

NBA

PRG

1,4-Dichlorobenzene

NBA

NBA

NBA

NBA

NBA

NBA

3.50E-01

3.50E-01

NBA

PRG

Diethyl phthalate

NBA

NBA

NBA

NBA

NBA

NBA

6.10E-01

6.10E-01

NBA

PRG

Di-n-butyl phthalate

NBA

NBA

NBA

NBA

NBA

NBA

2.40E+02

2.40E+02

NBA

PRG

Methapyrilene

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

4-Methylphenol

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

PAHs

Acenaphthene

NBA

NBA

NBA

NBA

NBA

NBA

8.90E-02

8.90E-02

NBA

PRG

Aeenaphthylene

NBA

NBA

NBA

NBA

NBA

NBA

1.30E-01

1.30E-01

NBA

PRG

Anthracene

5.72E-02

8.45E-01

2.20E-01

3.7/37 h

1.00E-02 1

1.40E-01 k

2.50E-01

5.72E-02

8.45E-01

MacDonald

Benzo(a)anthracene

1.08E-01

1.05E+00

3.20E-01

14.8/148 h

1.60E-02 J

2.80E-01 m

6.90E-01

1.08E-01

1.05E+00

MacDonald

Benzo(b)fluoranthene

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

Benzo(k)fluoranthene

NBA

NBA

2.40E-01

13.4/134 h

NBA

NBA

NBA

2.40E-01

13.4/134

OMEE

Benzo(ghi)perylene

NBA

NBA

1.70E-01

3.2/32 h

1.30E-02 k

2.80E-01 1

6.30E+00

1.30E-02

2.80E-01

Ingersoll et al.

Benzo(a)pyrene

1.50E-01

1.45E+00

3.70E-01

14.4/144 h

3.20E-02 J

3.20E-01 m

3.94E-01

1.50E-01

1.45E+00

MacDonald

Chrysene

1.66E-01

1.29E+00

3.40E-01

4.6/46 h

2.70E-02 J

4.10E-01 m

8.50E-01

1.66E-01

1.29E+00

MacDonald

Dibenzo(a,h)anthracene

3.30E-02

NBA

6.00E-02

1.3/13 h

1.00E-02 1

NBA

2.82E-02

3.30E-02

NBA

MacDonald

Fluoranthene

4.23E-01

2.23E+00

7.50E-01

10.2/102 h

3.10E-02 J

3.20E-01 m

8.34E-01

4.23E-01

2.23E+00

MacDonald

Fluorene

7.74E-02

5.36E-01

1.90E-01

1.6/16 h

1.00E-02 1

1.40E-01 1

1.40E-01

7.74E-02

5.36E-01

MacDonald

Indeno( 1,2,3-c,d)pyrene

NBA

NBA

2.00E-01

3.2/32 h

1.70E-02 J

2.40E-01 m

8.37E-01

1.70E-02

2.40E-01

Ingersoll et al.

2-Methylnaphthalene

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

Naphthalene

1.76E-01

5.61E-01

NBA

NBA

1.30E-02 k

9.80E-02 1

3.90E-01

1.76E-01

5.61E-01

MacDonald

Tables B-53 - B-55.xlsB-53

-------
Table B-53

Pre-ERA Sediment Benchmarks
Housatonic River Site, OU2, Pittsfield, MA



MacDonald et al.a

OMEEb

Ingersoll et al.c



Selected Benchmarks6



(m?

/kg)

(mg

/kg)

(mg

/kg)

Sediment PRGsd

(mg/kg)



Chemical

TECs

PECs

LEL

SEL

Lowest

Higher

(mg/kg)

Low

High

Basis

Phenanthrene

2.04E-01

1.17E+00

5.60E-01

9.5/95 h

1.90E-02 J

4.10E-01 m

5.40E-01

2.04E-01

1.17E+00

MacDonald

Pyrene

1.95E-01

1.52E+00

4.90E-01

8.5/85 h

4.00E-02 k

3.50E-01 1

1.40E+00

1.95E-01

1.52E+00

MacDonald

Total PAH

1.61E+00

2.28E+01

4.00E+00

100/1000 h

2.40E-01 k

2.20E+00 1

1.37E+01

1.61E+00

2.28E+01

MacDonald

Total PAH (High MW)

NBA

NBA

NBA

NBA

1.70E-01 k

1.70E+00 1

4.35E+00

1.70E-01

1.70E+00

Ingersoll et al.

Total PAH (Low MW)

NBA

NBA

NBA

NBA

7.60E-02 J

1.20E+00 m

3.37E+00

7.60E-02

1.20E+00

Ingersoll et al.

Pentachlorobenzene

NBA

NBA

NBA

NBA

NBA

NBA

7.00E-01

7.00E-01

NBA

PRO

Phenol

NBA

NBA

NBA

NBA

NBA

NBA

3.20E-02

3.20E-02

NBA

PRO

1,2,4,5-Tetrachlorobenzene

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

1,2,4-Trichlorobenzene

NBA

NBA

NBA

NBA

NBA

NBA

9.70E+00

9.70E+00

NBA

PRO

Pesticides

4,4'-DDD

4.88E-03

2.80E-02

8.00E-03

0.06/0.6 h

NBA

NBA

7.80E-03

4.88E-03

2.80E-02

MacDonald

4,4'-DDE

3.16E-03

3.13E-02

5.00E-03

0.19/1.9 h

NBA

NBA

2.70E-02

3.16E-03

3.13E-02

MacDonald

4,4'-DDT

4.16E-03

6.29E-02

8.00E-03

0.71/7.1 h

NBA

NBA

5.20E-02

4.16E-03

6.29E-02

MacDonald

alpha-BHC

NBA

NBA

6.00E-03

0.1/1 h

NBA

NBA

NBA

6.00E-03

0.1/1

OMEE

beta-BHC

NBA

NBA

5.00E-03

0.21/2.1 h

NBA

NBA

NBA

5.00E-03

0.21/2.1

OMEE

Endrin aldehyde

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

Heptaehlor

NBA

NBA

3.00E-04 f

NBA

NBA

NBA

1.30E+01

3.00E-04

NBA

OMEE

Herbicides

2,4,5-T

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

Dioxins/Furans

Dioxin (Total)

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

Furan (Total)

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

PCBs

Total PCBs

5.98E-02

6.76E-01

7.00E-02

5.3/53 h

3.20E-02 J

2.40E-01 m

1.80E-01

5.98E-02

6.76E-01

MacDonald

Metals

Antimony

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

Arsenic

9.79E+00

3.30E+01

6.00E+00

3.30E+01

1.10E+01 J

4.80E+01 m

4.20E+01

9.79E+00

3.30E+01

MacDonald

Tables B-53 - B-55.xlsB-53

-------
Table B-53

Pre-ERA Sediment Benchmarks
Housatonic River Site, OU2, Pittsfield, MA

MacDonald et al.a

OMEEb

Ingersoll et al.c

Selected Benchmarks6

(mg

/kg)

(mg

/kg)

(mg

/kg)

Sediment PRGsd

(mg/kg)

Chemical

TECs

PECs

LEL

SEL

Lowest

Higher

(mg/kg)

Low

High

Basis

Barium

NBA

...

Beryllium

NBA

...

Cadmium

9.90E-01

4.98E+00

6.00E-01

1.00E+01

5.80E-01 J

3.20E+00 m

4.20E+00

9.90E-01

4.98E+00

MacDonald

Chromium

4.34E+01

1.11E+02

2.60E+01

1.10E+02

3.60E+01 J

1.20E+02 m

1.59E+02

4.34E+01

1.11E+02

MacDonald

Cobalt

NBA

5.00E+01 g

NBA

5.00E+01

NBA

OMEE

Copper

3.16E+01

1.49E+02

1.60E+01

1.10E+02

2.80E+01 J

1.00E+02 m

7.77E+01

3.16E+01

1.49E+02

MacDonald

Lead

3.58E+01

1.28E+02

3.10E+01

2.50E+02

3.70E+01 J

8.20E+01 m

1.10E+02

3.58E+01

1.28E+02

MacDonald

Mercury

1.80E-01

1.06E+00

2.00E-01

2.00E+00

NBA

7.00E-01

1.80E-01

1.06E+00

MacDonald

Nickel

2.27E+01

4.86E+01

1.60E+01

7.50E+01

2.00E+01 J

3.30E+01 m

3.85E+01

2.27E+01

4.86E+01

MacDonald

Selenium

NBA

...

Silver

NBA

5.00E-01 g

NBA

1.80E+00

5.00E-01

NBA

OMEE

Thallium

NBA

...

Tin

NBA

...

Vanadium

NBA

...

Zinc

1.21E+02

4.59E+02

1.20E+02

8.20E+02

9.80E+01 J

5.40E+02 m

2.70E+02

1.21E+02

4.59E+02

MacDonald

Inorganics

Ammonia

NBA

1.00E+02 g

NBA

1.00E+02

NBA

OMEE

Sulfide

NBA

...

LEL = Lowest Effect Level

NBA = No benchmark available.

PEC = Probable Effect Concentration

SEL = Severe Effect Level

TEC = Threshold Effect Concentration

a MacDonald et al., 2000.

b Ontario Ministry of the Environment and Energy. 1996. Guidelines for the Protection and Management of Aquatic Sediment Quality in Ontaric

Tables B-53 - B-55.xlsB-53

-------
Table B-53

Pre-ERA Sediment Benchmarks
Housatonic River Site, OU2, Pittsfield, MA



MacDonald et al.a

OMEEb

Ingersoll et al.c



Selected Benchmarks6



(mg/kg)

(mg/kg)

(mg/kg)

Sediment PRGsd

(mg/kg)



Chemical

TECs | PECs

LEL | SEL

Lowest | Higher

(mg/kg)

Low | High

Basis

c Ingersoll et al. 1996. Calculation and Evaluation of Sediment Effect Concentrations for the Amphipod Hyalella azteca and the Midge Chironomus riparius. Great Lakes Research 22(3):602-623.
Lowest value = lowest reliable benchmark available. High value = high-end benchmark that corresponds to the lowest value (e.g., ERM if an ERL was the lowest value; PEL if a TEL was the lowest
value).

d Efroymson et al. 1997. Preliminary Remediation Goals for Ecological Endpoints. Unless otherwise notec

e For low value, MacDonald et al. value preferentially selected, followed by the lowest of the rest. The high value is the high-end benchmark associated with the selected low-end benchmark (e.g., if tl
TEC was selected, the PEC is presented; if LEL then SEL; if ERL then ERM).
f Value represents NEL (no effects level)

8 Open Water Disposal Guidelines, equivalent to LELs in terms of management decisions
h Values calculated assuming 1% TOC and 10% TOC, respectively.

1 Effects Range-Low (ERL) and Threshold Effect Level (TEL).

' TEL.
kERL.

'Effects Range-Median (ERM).

"Probable Effect Level (PEL).

"Jones et al. 1997. Toxicological Benchmarks for Screening Contaminants of Potential Concern for Effects on Sediment-Associated Biota: 1997 Revision. EqP-derived sediment quaility benchmarks
for nonionic organic chemicals. Based on secondary chronic value.

Tables B-53 - B-55.xlsB-53

-------
Table B-54

Pre-ERA Surface Water Benchmarks
Housatonic River Project, OU2, Pittsfield, MA



AWQCa

ORNL Tier IIb

Canadian Water

Literature-based value







Freshwater Chronic

Freshwater Chronic

Quality Guidelines0

Value

Endpoint;



Selected Value11

Chemical

0*g/L)

0*g/L)

0*g/L)

0*g/L)

Organism

Reference

0*g/L)

Basis

Volatiles

Acetone

NBA

1.50E+03

NBA

ND

—

—

1.50E+03

Tier II

Bromodichloromethane

NBA

NBA

NBA

2.40E+05

24 hr EC50 (proliferation
restricted to one-half blank);
Tetrahvmena pvriformis

Yoshioka et al., 1997

2.40E+05

Yoshioka et al., 1997

Chlorobenzene

NBA

6.40E+01

1.30E+00

ND

—

—

6.40E+01

Tier II

Chloroform

NBA

2.80E+01

1.80E+00

ND

—

—

2.80E+01

Tier II

Dibromochloromethane

NBA

NBA

NBA

3.40E+04

5-day LC50; embryo lifestage of
Cyprinus carpio

Mattice et al., 1981

3.40E+04

Mattice et al., 1981

Trichloroethylene (TCE)

NBA

4.70E+01

2.10E+01

ND

—

—

4.70E+01

Tier II

Vinyl chloride

NBA

NBA

NBA



—







Semivolatiles

Bis(2-ethylhexyl) phthalate

NBA

3.00E+00

1.60E+01

ND

—

—

3.00E+00

Tier II

1,4-Dichlorobenzene

NBA

1.50E+01

2.60E+01

ND

—

—

1.50E+01

Tier II

Diethyl phthalate

NBA

2.10E+02

NBA

ND

—

—

2.10E+02

Tier II

PAHs

Acenaphthene

NBA

NBA

5.80E+00

ND

—

—

5.80E+00

Canadian

Benzo(a)anthracene

NBA

2.70E-02

1.80E-02

ND

—

—

2.70E-02

Tier II

Benzo(b)fluoranthene

NBA

NBA

NBA

4.20E+00

24 hr EC50 (immobilization in
the presence of UV light);
Daphnia magna

Wernersson and
Dave, 1997

4.20E+00

Wernersson and Dave, 1997

Benzo(g,h,I)perylene

NBA

NBA

NBA



—







Benzo(a)pyrene

NBA

1.40E-02

1.50E-02

ND

—

—

1.40E-02

Tier II

Chrysene

NBA

NBA

NBA

2.00E+03

Effect Concentration (Inhibition
of growth); Lemna

Krylov et al., 1997

2.00E+03

Krylov etal., 1997

Fluoranthene

NBA

NBA

4.00E-02

ND

—

—

4.00E-02

Canadian

Fluorene

NBA

3.90E+00

3.00E+00

ND

—

—

3.90E+00

Tier II

Indeno( 1,2,3-cd)pyrene

NBA

NBA

NBA



—







Naphthalene

NBA

1.20E+01

1.10E+00

ND

—

—

1.20E+01

Tier II

Phenanthrene

6.30E+00

NBA

4.00E-01

ND

—

—

6.30E+00

AWQC

Tables B-53 - B-55.xlsB-54

-------
Table B-54

Pre-ERA Surface Water Benchmarks
Housatonic River Project, OU2, Pittsfield, MA



AWQCa

ORNL Tier IIb

Canadian Water

Literature-based value







Freshwater Chronic

Freshwater Chronic

Quality Guidelines0

Value

Endpoint;



Selected Valued

Chemical

0*g/L)

(^g/L)

0*g/L)

(^g/L)

Organism

Reference

(Mg/L)

Basis

Pyrene

NBA

NBA

2.50E-02

ND

—

—

2.50E-02

Canadian

Total PAH

NBA

NBA

NBA

NBA

—

—

NBA

—

Total PAH (High MW)

NBA

NBA

NBA

NBA

—

—

NBA

—

Total PAH (Low MW)

NBA

NBA

NBA

NBA

—

—

NBA

—

1,2,4-Trichlorobenzene, Filtered

NBA

1.10E+02

2.40E+01

ND

—

—

1.10E+02

Tier II

Pesticides

delta-BHC

NBA

2.20E+00

NBA

ND

—

—

2.20E+00

Tier II

gamma-BHC

NBA

NBA

1.00E-02

ND

—

—

1.00E-02

Canadian

Dioxins/Furans

Dioxin (Total)

<1.00E-05 £g

NBA

NBA

NBA

—



<1.00E-05

AWQC

Furan (Total)

NBA

NBA

NBA

NBA

—

—

NBA

—

PCBs

Total PCBs

1.40E-02

1.40E-01

NBA

ND

—

—

1.40E-02

AWQC

Total PCBs, Dissolved

1.40E-02

1.40E-01

NBA

ND

—

—

1.40E-02

AWQC

Metals, totals

Arsenic

1.50E+02

3.10E+00

5.00E+00

ND

—

—

1.50E+02

AWQC

Barium

NBA

4.00E+00

NBA

ND

—

—

4.00E+00

Tier II

Beryllium

NBA

6.60E-01

NBA

ND

—

—

6.60E-01

Tier II

Cadmium

1.52E-01 h

NBA

1.70E-02

ND

—

—

1.52E-01

AWQC

Calcium

NBA

NBA

NBA

NE

—

—

NE

—

Chromium

1.14E+01 w

NBA

1.00E+00 1

ND

—

—

1.14E+01

AWQC

Cobalt

NBA

2.30E+01

NBA

ND

—

—

2.30E+01

Tier II

Copper

5.16E+00 h

NBA

2.00E+00 h

ND

—

—

5.16E+00

AWQC

Lead

1.32E+00 h

NBA

1.00E+00 h

ND

—

—

1.32E+00

AWQC

Magnesium

NBA

NBA

NBA

NE

—

—

NE

—

Mercury

9.06E-01

2.80E-03

1.00E-01

ND

—

—

9.06E-01

AWQC

Nickel

2.90E+01 h

NBA

2.50E+01 h

ND

—

—

2.90E+01

AWQC

Selenium

5.00E+00

NBA

1.00E+00

ND

___

___

5.00E+00

AWQC

Tables B-53 - B-55.xlsB-54

-------
Table B-54

Pre-ERA Surface Water Benchmarks
Housatonic River Project, OU2, Pittsfield, MA



AWQCa

ORNL Tier IIb

Canadian Water

Literature-based value







Freshwater Chronic

Freshwater Chronic

Quality Guidelines0

Value

Endpoint;



Selected Value11

Chemical

0*g/L)

0*g/L)

0*g/L)

0*g/L)

Organism

Reference

0*g/L)

Basis

Silver

NBA

3.60E-01

1.00E-01

ND

—

—

3.60E-01

Tier II

Thallium

NBA

1.20E+01

8.00E-01

ND

—

—

1.20E+01

Tier II

Tin

NBA

7.30E+01

NBA

ND

—

—

7.30E+01

Tier II

Vanadium

NBA

2.00E+01

NBA

ND

—

—

2.00E+01

Tier II

Zinc

6.66E+01 h

NBA

3.00E+01

ND

—

—

6.66E+01

AWQC

Metals, dissolved

Barium, Dissolved

NBA

NBA

NBA

NBA

—



NBA

—

Beryllium, Dissolved

NBA

NBA

NBA

NBA

—



NBA

—

Cadmium, Dissolved

1.52E-01 h

NBA

NBA

ND

—

—

1.52E-01

AWQC

Calcium, Dissolved

NBA

NBA

NBA

NBA

—



NBA

—

Chromium, Dissolved

1.10E+01 1

NBA

NBA

ND

—

—

1.10E+01

AWQC

Cobalt, Dissolved

NBA

NBA

NBA

NBA

—



NBA

—

Copper, Dissolved

4.95E+00 h

NBA

NBA

ND

—

—

4.95E+00

AWQC

Lead, Dissolved

1.17E+00 h

NBA

NBA

ND

—

—

1.17E+00

AWQC

Magnesium, Dissolved

NBA

NBA

NBA

NBA

—



NBA

—

Mercury, Filtered

7.70E-01

NBA

NBA

ND

—

—

7.70E-01

AWQC

Nickel, Dissolved

2.89E+01 h

NBA

NBA

ND

—

—

2.89E+01

AWQC

Selenium, Dissolved

4.61E+00

NBA

NBA

ND

—

—

4.61E+00

AWQC

Silver, Dissolved

NBA

NBA

NBA

NBA

—



NBA

—

Zinc, Dissolved

6.57E+01 h

NBA

NBA

ND

—

—

6.57E+01

AWQC

Inorganics

Ammonia As N

3.70E+00 j

NBA

NBA

ND

—

—

3.70E+00

AWQC

Sulfide

2.00E+00

NBA

NBA

ND

—

—

2.00E+00

AWQC

NBA = No benchmark available.

ND = Not determined. Value not searched for if AWQC or Tier II value available or for dissolved metals or inorganics.
a U.S. EPA. 1999. National Recommended Water Quality Criteria - Correction. Office of Water. EPA 822-Z-99-001. April 1999

Tables B-53 - B-55.xlsB-54

-------
Table B-54

Pre-ERA Surface Water Benchmarks
Housatonic River Project, OU2, Pittsfield, MA



AWQCa

ORNL Tier IIb

Canadian Water

Literature-based value







Freshwater Chronic

Freshwater Chronic

Quality Guidelines0

Value

Endpoint;



Selected Value11

Chemical

0*g/L)

0*g/L)

0*g/L)

0*g/L)

Organism

Reference

0*g/L)

Basis

Suter, G. W. II, and C. L. Tsao. 1996. Toxicological Benchmarks for Screening of Potential Contaminants of Concern for Effects on Aquatic Biota on Oak Ridge Reservation: 1996 Revision. Oak Ridge National Laboratory,
Oak Ridge, TN. 104 pp, ES/ER/TM-96/R2

c Canadian Water Quality Guidelines for the Protection of Aquatic Life, CCME. 1999
d AWQC preferentially selected, followed by Tier II value, Canadian value, then literature value
e Proposed criterion; U.S. EPA. 1991. Water Quality Criteria Summary
fLOEL value; U.S. EPA. 1991. Water Quality Criteria Summary
B2,3,7,8-TCDD value; U.S. EPA. 1991. Water Quality Criteria Summary
h Hardness dependent value. Calculated assuming a hardness of 50 mg/L (site minimum)

1 Chromium VI value.

1 pH dependent value. Calculated assuming a pH of 4.6 (site minimum)

Mattice, J.S., S.C. Tsai, M.B. Burch, and J.J. Beauchamp. 1981. "Toxicity of trihalomethanes to common carp embryos." Trans. Amer. Fish. Soc. 110:261-269.

Krylov, S.N., X. Huang, L.F. Zeiler, D.G. Dixon, and B.M. Greenberg. 1997. "Mechanistic quantitative structure-activity relationship model for the photoinduced toxicity of polycyclic aromatic hydrocarbons: I. Physical
model based on chemical kinetics in a two-compartment system." Environ. Toxicol. Chem. 16(ll):2283-2295.

Wernersson, A.S. and G. Dave. 1997. "Phototoxicity identification by solid phase extraction and photoinduced toxicity toDaphnia magna ." Arch. Environ. Contam. Toxicol. 32:268-273.

Yoshioka, Y., Y. Ose, and T. Sato. 1985. "Testing for the toxicity of chemicals vaTetrahymena pyriformis . Sci. Total Environ . 43:149-157.

Tables B-53 - B-55.xlsB-54

-------
Table B-55

Pre-ERA Soil Benchmarks
Housatonic River Project, OU2, Pittsfield, MA

Chemical

Wildlife Soil
PRGsa
(mg/kg)

Phytotoxicityb
(mg/kg)

Earthworm0
(mg/kg)

Microorganisms
and Microbial
Processes0
(mg/kg)

Canadian Soil
Quality
Guidelines'1
(mg/kg)

Lowest Value

(mg/kg)

Basis

Acetone

NBA

NBA

NBA

NBA

3.50E+00

3.50E+00

Ontario SQG

2-Butanone

NBA

NBA

NBA

NBA

2.70E-01

2.70E-01

Ontario SQG

Carbon tetrachloride

NBA

NBA

NBA

1.00E+03

ND

1.00E+03

Microorganisms and Microbial Processes

Chlorobenzene

NBA

NBA

4.00E+01

NBA

ND

4.00E+01

Earthworm

Dibromochloromethane

NBA

NBA

NBA

NBA

9.00E-02

9.00E-02

Ontario SQG

1,1 -Dichloroethylene

NBA

NBA

NBA

NBA

2.40E-03

2.40E-03

Ontario SQG

Ethylbenzene

NBA

NBA

NBA

NBA

1.00E-01

1.00E-01

Ontario SQG

Methylene chloride

NBA

NBA

NBA

NBA

1.10E+00

1.10E+00

Ontario SQG

2-Methyl-2-pentanone

NBA

NBA

NBA

NBA

4.80E-01

4.80E-01

Ontario SQG

Toluene

NBA

2.00E+02

NBA

NBA

ND

2.00E+02

Phytotoxicity

1,1,1,2-T etrachloroethane

NBA

NBA

NBA

NBA

1.90E-02

1.90E-02

Ontario SQG

1,1,2,2-T etrachloroethane

NBA

NBA

NBA

NBA

1.00E-02

1.00E-02

Ontario SQG

1,1,1 ,-T richloroethane

NBA

NBA

NBA

NBA

2.60E+01

2.60E+01

Ontario SQG

1,1,2,-Trichloroethane

NBA

NBA

NBA

NBA

2.80E-01

2.80E-01

Ontario SQG

T etrachloroethylene

NBA

NBA

NBA

NBA

1.00E-01

1.00E-01

Canadian SQG

T richloroethy lene

NBA

NBA

NBA

NBA

1.00E-01

1.00E-01

Canadian SQG

Vinyl chloride

NBA

NBA

NBA

NBA

3.00E-03

3.00E-03

Ontario SQG

Xylene

NBA

NBA

NBA

NBA

1.00E-01

1.00E-01

Canadian SQG

Semivolatiles

Acetophenone

NBA

NBA

NBA

NBA

NBA

NBA

...

Acrylonitrile

NBA

NBA

NBA

1.00E+03

ND

1.00E+03

Microorganisms and Microbial Processes

Benzyl alcohol

NBA

NBA

NBA

NBA

NBA

NBA

...

Bis(2-chloroethyl)ether

NBA

NBA

NBA

NBA

6.60E-01

6.60E-01

Ontario SQG

Bis(2-chloroisopropyl)ether

NBA

NBA

NBA

NBA

6.60E-01

6.60E-01

Ontario SQG

Bis(2-ethylhexyl)phthalate

NBA

NBA

NBA

NBA

1.00E+02

1.00E+02

Ontario SQG

Bromodichloromethane

NBA

NBA

NBA

NBA

1.20E-01

1.20E-01

Ontario SQG

Tables B-53 - B-55.xlsB-55

-------
Table B-55

Pre-ERA Soil Benchmarks
Housatonic River Project, OU2, Pittsfield, MA

Chemical

Wildlife Soil
PRGsa
(mg/kg)

Phytotoxicityb
(mg/kg)

Earthworm0
(mg/kg)

Microorganisms
and Microbial
Processes0
(mg/kg)

Canadian Soil
Quality
Guidelines'1
(mg/kg)

Lowest Value

(mg/kg)

Basis

Bromoform

NBA

NBA

NBA

NBA

1.10E-01

1.10E-01

Ontario SQG

Bromomethane

NBA

NBA

NBA

NBA

6.10E-02

6.10E-02

Ontario SQG

Butylbenzylphthalate

NBA

NBA

NBA

NBA

NBA

NBA

...

2-Chlorophenol

NBA

NBA

NBA

NBA

1.00E-01

1.00E-01

Ontario SQG

Dibenzofuran

NBA

NBA

NBA

NBA

NBA

NBA

...

Di-n-butyl phthalate

NBA

2.00E+02

NBA

NBA

ND

2.00E+02

Phytotoxicity

1,2-Dichlorobenzene

NBA

NBA

NBA

NBA

8.80E-01

8.80E-01

Ontario SQG

1,3-Dichlorobenzene

NBA

NBA

NBA

NBA

3.00E+01

3.00E+01

Ontario SQG

1,4-Dichlorobenzene

NBA

NBA

2.00E+01

NBA

ND

2.00E+01

Earthworm

3,3 '-Dichlorobenzidine

NBA

NBA

NBA

NBA

1.30E+00

1.30E+00

Ontario SQG

1,1 -Dichloroethane

NBA

NBA

NBA

NBA

3.00E+00

3.00E+00

Ontario SQG

1,2-Dichloroethane

NBA

NBA

NBA

NBA

2.20E-02

2.20E-02

Ontario SQG

2,4-Dichlorophenol

NBA

NBA

NBA

NBA

3.00E-01

3.00E-01

Ontario SQG

1,2-Dichloropropane

NBA

NBA

7.00E+02

7.00E+02

ND

7.00E+02

Earthworm

Diethylphthalate

NBA

1.00E+02

NBA

NBA

ND

1.00E+02

Phytotoxicity

Dimethylphthalate

NBA

NBA

2.00E+02

NBA

ND

2.00E+02

Earthworm

2,4-Dimethylphenol

NBA

NBA

NBA

NBA

9.40E-01

9.40E-01

Ontario SQG

2,4-Dinitrophenol

NBA

2.00E+01

NBA

NBA

ND

2.00E+01

Phytotoxicity

2,4-Dinitrotoluene

NBA

NBA

NBA

NBA

6.60E-01

6.60E-01

Ontario SQG

Hexachlorobenzene

NBA

NBA

NBA

1.00E+03

ND

1.00E+03

Microorganisms and Microbial Processes

Hexachlorobutadiene

NBA

NBA

NBA

NBA

3.80E-01

3.80E-01

Ontario SQG

Hexachlorocyclopentadiene

NBA

1.00E+01

NBA

NBA

ND

1.00E+01

Phytotoxicity

Hexachloroethane

NBA

NBA

NBA

NBA

3.80E+00

3.80E+00

Ontario SQG

4-Methylphenol

NBA

NBA

NBA

NBA

NBA

NBA

...

Nitrobenzene

NBA

NBA

4.00E+01

1.00E+03

ND

4.00E+01

Earthworm

4-Nitrophenol

7.00E+00

NBA

7.00E+00

NBA

ND

7.00E+00

PRG

Tables B-53 - B-55.xlsB-55

-------
Table B-55

Pre-ERA Soil Benchmarks
Housatonic River Project, OU2, Pittsfield, MA

Chemical

Wildlife Soil
PRGsa
(mg/kg)

Phytotoxicityb
(mg/kg)

Earthworm0
(mg/kg)

Microorganisms
and Microbial
Processes0
(mg/kg)

Canadian Soil
Quality
Guidelines'1
(mg/kg)

Lowest Value

(mg/kg)

Basis

N-Nitrosodiphenylamine

NBA

NBA

2.00E+01

NBA

ND

2.00E+01

Earthworm

PAHs

Acenaphthene

NBA

2.00E+01

NBA

NBA

ND

2.00E+01

Phytotoxicity

Acenaphthylene

NBA

NBA

NBA

NBA

1.00E+02

1.00E+02

Ontario SQG

Anthracene

NBA

NBA

NBA

NBA

2.80E+01

2.80E+01

Ontario SQG

Benzo(a)anthracene

NBA

NBA

NBA

NBA

6.60E+00

6.60E+00

Ontario SQG

Benzo(b)fluoranthene

NBA

NBA

NBA

NBA

1.20E+01

1.20E+01

Ontario SQG

Benzo(k)fluoranthene

NBA

NBA

NBA

NBA

1.20E+01

1.20E+01

Ontario SQG

Benzo(ghi)perylene

NBA

NBA

NBA

NBA

4.00E+01

4.00E+01

Ontario SQG

Benzo(a)pyrene

NBA

NBA

NBA

NBA

7.00E-01 f

7.00E-01

Canadian SQG

Chrysene

NBA

NBA

NBA

NBA

1.20E+01

1.20E+01

Ontario SQG

Dibenzo(a,h)anthracene

NBA

NBA

NBA

NBA

1.20E+00

1.20E+00

Ontario SQG

Fluoranthene

NBA

NBA

NBA

NBA

4.00E+01

4.00E+01

Ontario SQG

Fluorene

NBA

NBA

3.00E+01

NBA

ND

3.00E+01

Earthworm

Indeno(l ,2,3-c,d)pyrene

NBA

NBA

NBA

NBA

1.20E+01

1.20E+01

Ontario SQG

2-Methylnaphthalene

NBA

NBA

NBA

NBA

1.20E+00

1.20E+00

Ontario SQG

Naphthalene

NBA

NBA

NBA

NBA

4.60E+00

4.60E+00

Ontario SQG

Phenanthrene

NBA

NBA

NBA

NBA

4.00E+01

4.00E+01

Ontario SQG

Pyrene

NBA

NBA

NBA

NBA

1.30E+00

1.30E+00

Ontario SQG

Total PAH

NBA

NBA

NBA

NBA

NBA

NBA

...

Total PAH (High MW)

NBA

NBA

NBA

NBA

NBA

NBA

...

Total PAH (Low MW)

NBA

NBA

NBA

NBA

NBA

NBA

...

Pentachlorobenzene

NBA

NBA

2.00E+01

NBA

ND

2.00E+01

Earthworm

Pentachlorophenol

NBA

3.00E+00

6.00E+00

4.00E+02

ND

3.00E+00

Phytotoxicity

Phenol

NBA

7.00E+01

3.00E+01

1.00E+02

ND

3.00E+01

Earthworm

p-Phenylenediamine

NBA

NBA

NBA

NBA

NBA

NBA

—

Tables B-53 - B-55.xlsB-55

-------
Table B-55

Pre-ERA Soil Benchmarks
Housatonic River Project, OU2, Pittsfield, MA

Chemical

Wildlife Soil
PRGsa
(mg/kg)

Phytotoxicityb
(mg/kg)

Earthworm0
(mg/kg)

Microorganisms
and Microbial
Processes0
(mg/kg)

Canadian Soil
Quality
Guidelines'1
(mg/kg)

Lowest Value

(mg/kg)

Basis

Pyridine

NBA

NBA

NBA

NBA

NBA

NBA

...

Styrene

NBA

3.00E+02

NBA

NBA

ND

3.00E+02

Phytotoxicity

1,2,3,4-Tetrachlorobenzene

NBA

NBA

1.00E+01

NBA

ND

1.00E+01

Earthworm

1,2,3-Trichlorobenzene

NBA

NBA

2.00E+01

NBA

ND

2.00E+01

Earthworm

1,2,4-Trichlorobenzene

NBA

NBA

2.00E+01

NBA

ND

2.00E+01

Earthworm

2,4,5-Trichlorophenol

NBA

4.00E+00

9.00E+00

NBA

ND

4.00E+00

Phytotoxicity

2,4,6-Trichlorophenol

NBA

NBA

1.00E+01

NBA

ND

1.00E+01

Earthworm

Pesticides

Aldrin

NBA

NBA

NBA

NBA

5.00E-02

5.00E-02

Ontario SQG

beta-BHC

NBA

NBA

NBA

NBA

NBA

NBA

...

gamma-BHC

NBA

NBA

NBA

NBA

4.10E-01

4.10E-01

Ontario SQG

Chlordane

NBA

NBA

NBA

NBA

2.90E-01

2.90E-01

Ontario SQG

4,4'-DDT

NBA

NBA

NBA

NBA

7.00E-01

7.00E-01

Canadian SQG

Dieldrin

NBA

NBA

NBA

NBA

5.00E-02

5.00E-02

Ontario SQG

Endosulfan

NBA

NBA

NBA

NBA

1.80E-01

1.80E-01

Ontario SQG

Endosulfan sulfate

NBA

NBA

NBA

NBA

NBA

NBA

...

Endrin

NBA

NBA

NBA

NBA

5.00E-02

5.00E-02

Ontario SQG

Endrin aldehyde

NBA

NBA

NBA

NBA

NBA

NBA

...

Heptachlor

NBA

NBA

NBA

NBA

8.40E-02

8.40E-02

Ontario SQG

Heptachlor epoxide

NBA

NBA

NBA

NBA

6.00E-02

6.00E-02

Ontario SQG

Methoxychlor

NBA

NBA

NBA

NBA

4.00E+00

4.00E+00

Ontario SQG

2,4,5-T

NBA

NBA

NBA

NBA

NBA

NBA

...

2,4,5-TP (Silvex)

NBA

NBA

NBA

NBA

NBA

NBA

...

Dioxins/Furans

TCDD (Total)

3.15E-06

NBA

NBA

NBA

ND

3.15E-06

PRG

TCDF (Total)

8.40E-04

NBA

NBA

NBA

ND

8.40E-04

PRG

Tables B-53 - B-55.xlsB-55

-------
Table B-55

Pre-ERA Soil Benchmarks
Housatonic River Project, OU2, Pittsfield, MA

Chemical

Wildlife Soil
PRGsa
(mg/kg)

Phytotoxicityb
(mg/kg)

Earthworm0
(mg/kg)

Microorganisms
and Microbial
Processes0
(mg/kg)

Canadian Soil
Quality
Guidelines'1
(mg/kg)

Lowest Value

(mg/kg)

Basis

PCBs

Total PCBs

3.71E-01

4.00E+01

NBA

NBA

ND

3.71E-01

PRO

Metals

Aluminum

NBA

5.00E+01

NBA

6.00E+02

ND

5.00E+01

Phytotoxicity

Antimony

NBA

5.00E+00

NBA

NBA

ND

5.00E+00

Phytotoxicity

Arsenic

9.90E+00

1.00E+01

6.00E+01

1.00E+02

ND

9.90E+00

PRO

Barium

2.83E+02

5.00E+02

NBA

3.00E+03

ND

2.83E+02

PRO

Beryllium

NBA

1.00E+01

NBA

NBA

ND

1.00E+01

Phytotoxicity

Boron

NBA

5.00E-01

NBA

2.00E+01

ND

5.00E-01

Phytotoxicity

Cadmium

4.20E+00

4.00E+00

2.00E+01

2.00E+01

ND

4.00E+00

Phytotoxicity

Chromium

1.61E+01

1.00E+00

4.00E-01

1.00E+01

ND

4.00E-01

Earthworm

Cobalt

NBA

2.00E+01

NBA

1.00E+03

ND

2.00E+01

Phytotoxicity

Copper

3.70E+02

1.00E+02

6.00E+01

1.00E+02

ND

6.00E+01

Earthworm

Iron

NBA

NBA

NBA

2.00E+02

ND

2.00E+02

Microorganisms and Microbial Processes

Lead

4.05E+01

5.00E+01

5.00E+02

9.00E+02

ND

4.05E+01

PRO

Manganese

NBA

5.00E+02

NBA

1.00E+02

ND

1.00E+02

Microorganisms and Microbial Processes

Mercury

5.10E-04

3.00E-01

1.00E-01

3.00E+01

ND

5.10E-04

PRO

Nickel

1.21E+02

3.00E+01

2.00E+02

9.00E+01

ND

3.00E+01

Phytotoxicity

Selenium

2.10E-01

1.00E+00

7.00E+01

1.00E+02

ND

2.10E-01

PRO

Silver

NBA

2.00E+00

NBA

5.00E+01

ND

2.00E+00

Phytotoxicity

Thallium

2.10E+00

1.00E+00

NBA

NBA

ND

1.00E+00

Phytotoxicity

Tin

NBA

5.00E+01

NBA

2.00E+03

ND

5.00E+01

Phytotoxicity

Vanadium

5.50E+01

2.00E+00

NBA

2.00E+01

ND

2.00E+00

Phytotoxicity

Zinc

8.50E+00

5.00E+01

2.00E+02

1.00E+02

ND

8.50E+00

PRO

Inorganics

Cyanide

NBA

NBA

NBA

NBA

9.00E-01

9.00E-01

Canadian SQG

Tables B-53 - B-55.xlsB-55

-------
Table B-55

Pre-ERA Soil Benchmarks
Housatonic River Project, OU2, Pittsfield, MA

Chemical

Wildlife Soil
PRGsa
(mg/kg)

Phytotoxicityb
(mg/kg)

Earthworm0
(mg/kg)

Microorganisms
and Microbial
Processes0
(mg/kg)

Canadian Soil
Quality
Guidelines'1
(mg/kg)

Lowest Value

(mg/kg)

Basis

Sulfide

NBA

—

NBA = No benchmark available.

ND = Not determined. Value not searched for if ORNL value available.

a Efroymson, R.A., G.W. Suter II, B.E. Sample, and D.S. Jones. 1997a. Preliminary Remediation Goals for Ecological Endpoints . Oak Ridge National Laboratory for U.S. DOE.
ES/ER/TM-162/R2. August 1997.

b Efroymson, R.A., M.E. Will, G.W. Suter II, and A.C. Wooten. 1997b. Toxicological Benchmarks for Screening Contaminants of Potential Concern for Effects on Terrestrial Plants:
1997 Revision. ES/ER/TM-85/R3. Environmental Restoration Program, Oak Ridge National Laboratory, Oak Ridge, TN.

0 Efroymson, R.A., M.E. Will, and G.W. Suter II. 1997c. Toxicological Benchmarks for Contaminants of Potential Concern for Effects on Soil and Litter Invertebrates and Heterotrophic
Process: 1997Revision. ES/ER/TM-126/R2. Environmental Restoration Program, Oak Ridge National Laboratory, Oak Ridge, TN.

d Canadian Soil Quality Guidelines for the Protection of Environmental and Human Health . 2001. Lowest ecologically-based value from 4 available land uses: Agricultural,
Residential/parkland, Commercial, and Industrial; unless otherwise noted.

e Guideline for Use at Contaminated Sites in Ontario. 1996. Lower soil quality guideline for soil remediation criteria of agricultural land or residential/parkland land use for a potable
ground wtaer condition. Values are the lower of human health and ecological; therefore, should be protective of ecological receptors via direct contact.

Provisional CCME guideline

Tables B-53 - B-55.xlsB-55

-------
Table B-56

Fish Concentration Benchmarks for Piscivores
Housatonic River Site, OU2, Pittsfield Massachusetts



NOAEL-Based







Benchmark





Chemical

(mg/kg)



Receptor Basis

Semivolatiles

Hexachlorobenzene

2.76E-01

Kingfisher

Pentachloroanisole

8.76E+01

Mink

Pentachlorobenzene

1.75E+01

Mink

1,2,3,4-Tetrachlorobenzene

2.99E+02

Mink

1,2,4,5-Tetrachlorobenzene

1.53E+01

Mink

Pesticides

Aldrin

3.95E-02

Kingfisher

alpha-BHC

1.82E+01

Mink

beta-BHC

1.31E-01

Mink

delta-BHC

2.37E+02

Mink

gamma-BHC (Lindane)

3.95E+00

Kingfisher

alpha-Chlordane

4.22E+00

a

Kingfisher

gamma-Chlordane

4.22E+00

a

Kingfisher

Chlorpyrifos

3.55E-02

Kingfisher

o,p'-DDD

8.31E+00

b

Kingfisher

4,4'-DDD

8.31E+00

b

Kingfisher

o,p'-DDE

9.87E-02

c

Kingfisher

4,4'-DDE

9.87E-02

c

Kingfisher

o,p'-DDT

5.53E-03

d

Kingfisher

4,4'-DDT

5.53E-03

d

Kingfisher

Dieldrin

1.46E-01

Mink

Endosulfan II

1.09E+00

Mink

Endrin

2.04E-02

Kingfisher

Heptachlor

1.82E-01

Mink

Heptachlor Epoxide

NBA

—

Mirex

1.31E+00

Mink

cis-Nonachlor

NBA

—

trans-Nonachlor

NBA

—

Oxychlordane

NBA

—

Toxaphene

4.14E-01

Kingfisher

Dioxins/Furans

Total Dioxins

NBA

—

Total Furans

NBA

—

PCBs

Total PCBs

3.55E-01



Kingfisher

Metals

Lead

7.60E+00

Kingfisher

Mercury

1.26E-02

Kingfisher

Nickel

1.53E+02

Kingfisher

NBA = No benchmark available.
aChlordane value.

"TlDD value.

CDDE value.

dDDT and Metabolites value.

Tables B-56, B-58, B-59 valued.xlsB-56

-------
Table B-57

Secondary Sources to be Reviewed for the Identification of Primary Articles

for Toxicity Reference Values

Literature Search

The Dialog Information Retrieval Service will be accessed to perform a comprehensive literature search for avian
and mammalian toxicity data. The databases to be searched include:

Secondary Sources

The secondary sources listed as follows will be reviewed for studies relevant to the development of TRVs:

¦	Agency for Toxic Substances and Disease Registry (ATSDR) Toxicological Profiles

¦	Great Lakes Water Quality Initiative Criteria Documents for the Protection of Wildlife (Proposed) (EPA, 99-
0116)

¦	Toxicological Benchmarks for Wildlife (99-0719)

¦	USFWS Biological Reports, Contaminant Hazard Reviews

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

Biosis Previews
CA Search
EM Base

Life Sciences Collection

Medline
Toxline
SciSearch

Database Searches
The databases listed as follows will be accessed via various internet sites.

Hazardous Substances Data Bank (HSDB)

Integrated Risk Information System (IRIS)

Registry of Toxic Effects of Chemical Substances (RTECS)

-------
Table B-58

Mammalian Toxicity Reference Values
Housatonic River Site, OU2, Pittsfield Massachusetts

Chemical

Receptor (strain)

Sex

Exposure Mode

Duration Category

Dose (m

l/kg-day)

Endpoints

Reference

ORNL
Benchmark Dose
fmg/kg-davt

Study
NOAEL
(mg/kg-day)

Chronic NOAEL
(mg/kg-day)

Fish
Concentration
(mg/kg)

No Effect

Effect

NOAEL

LOAEL

Semivolatiles

Hexachlorobenzene

rat (Sprague-Dawley)

both

Food

developmental





4.00E-01

2.00E+00

severe chronic nephrosis in F1 males; decrease in
viability of pups; significant increase in parathyroid
adenomas in males; phaeochromocytoma of adrenals in
females (F1 generation); and neoplastic liver nodules in
F1 females

Arnold et al., 1985, 99-
0673

4.00E-01

4.00E-01

4.00E-01

2.92E+00

Pentachloroanisole

rat (Sprague-Dawley)

female

Food

chronic





1.20E+01

4.10E+01

embryolethality; significantly less food consumption,
body weight gain significant lower, gravid uterus weigl
significantly reduced and decrease in the number of

Welsh etal., 1987, 99-0727

1.20E+01

1.20E+01

1.20E+01

8.76E+01

Pentachlorobenzene

rat (F344/N)

male

Food

chronic





2.40E+00

6.70E+00

tubular dilation, hyaline droplets, tubular cortex
degeneration and chronic inflammation of the kidneys

NTP, 1991, 99-0716

2.40E+00

2.40E+00

2.40E+00

1.75E+01

1,2,3,4-Tetrachlorobenzene

rat (Spraaue-Dawlev

female

Food

chronic

4.10E+01







no effects

Chu et al., 1984,99-0680

4.10E+01

4.10E+01

4.10E+01

2.99E+02

1,2,4,5-Tetrachlorobenzene

rat (F344/N)

male

Food

chronic





2.10E+00

7.10E+00

renal lesions, abnormal hyaline droplet accumulation,
dilation of tubules, mineralization of medullary
collecting tubules, renal cortical tubular regeneration;
significant increase in serum albumen; significant
increase in relative liver weights

McDonald, 1991, 99-0713

2.10E+00

2.10E+00

2.10E+00

1.53E+01

Pesticides

Aldrin

rat

not reported

Food

developmental





2.00E-01

1.00E+00

offspring mortality, reduced number of Utters

Treon and Cleveland, 1955,
99-0724

2.00E-01

2.00E-01

2.00E-01

1.46E+00

alpha-BHC

rat (Wistar)

both

Food

chronic





2.50E+00

5.00E+00

slight liver damage

Fitzhugh et al., 1950, 99-
0695

2.50E+00

2.50E+00

2.50E+00

1.82E+01

beta-BHC

rat (Wistar)

both

Food

chronic



1.80E-01





centrilobular liver effects in males; enzyme induction;
small number of males showed slight centrilobular

Van Velsen et al., 1986,99-
0725

1.80E-01

1.80E-02

1.80E-02

1.31E-01

delta-BHC

mouse

male

Food

chronic





3.25E+01

6.50E+01

very slight increase in liver weiah

Ito et al., 1973, 99-0705

3.25E+01

3.25E+01

3.25E+01

2.37E+02

gamma-BHC (Lindane)

rat (Wistar)

both

Food

chronic





2.50E+00

5.00E+00

slight kidney damage; liver enlargement;
microscopicallv. verv slight dama2i

Fitzhugh et al., 1950, 99-
0695

2.50E+00

2.50E+00

2.50E+00

1.82E+01

alpha-Chlordane

mouse

not reported

Food

chronic





4.58E+00

9.16E+00

decreased viability and reducted abundance of offspring

WHO, 1984, 99-0728

4.58E+00

4.58E+00

4.58E+00

3.34E+01

gamma-Chlordane

mouse

not reported

Food

chronic





4.58E+00

9.16E+00

decreased viability and reducted abundance of offspring

WHO, 1984, 99-0728

4.58E+00

4.58E+00

4.58E+00

3.34E+01

Chlorpyrifos

rat (Sprague-Dawley)

both

Food

developmental





1.00E+00

5.00E+00

decrease in body weights and accompanied by decrease
in survival of F1 offspring

Breslinet al., 1996, 99-
0676

1.00E+00

1.00E+00

1.00E+00

7.30E+00

o,p'-DDD

rat (Sprague-Dawley)

male

Gavage

subchronic



1.21E+02





decrease in body weight; organ weight reduction;
minimal changes in the liver; thymus attrophy and
suppression of the immune response of the thymus and
spleen; moderate reduction in size of spleen;moderate

Hamidet al., 1974, 99-0699

1.21E+02

1.21E+01

1.21E+00

8.83E+00

4,4'-DDDa

rat (Sprague-Dawley)

male

Gavage

subchronic



1.21E+02





decrease in body weight; organ weight reduction;
minimal changes in the liver; thymus attrophy and
suppression of the immune response of the thymus and
spleen; moderate reduction in size of spleen;moderate

Hamidet al., 1974, 99-0699

1.21E+02

1.21E+01

1.21E+00

8.83E+00

o.p'-DDE

rat

not reported

Food

chronic



2.19E+01





mortality associated with tumor arowtl

NCI. 1978. 99-0715

2.19E+01

2.19E+00

2.19E+00

1.59E+01

4,4'-DDE

rat

not reported

Food

chronic



2.19E+01





mortality associated with tumor arowtl

NCI, 1978, 99-0715

2.19E+01

2.19E+00

2.19E+00

1.59E+01

o.p'-DDT

rat

not reported

Food

chronic





8.00E-01

4.00E+00

reduced number of vouna produce*

EPA. 1988. 99-0694

8.00E-01

8.00E-01

8.00E-01

5.84E+00

4,4'-DDT

rat

not reported

Food

chronic





8.00E-01

4.00E+00

reduced number of youna produce*

EPA, 1988, 99-0694

8.00E-01

8.00E-01

8.00E-01

5.84E+00

Dieldrin

rat

not reported

Food

chronic



2.00E-01





reduced number of pregnancies

Treon and Cleveland, 1955,
99-0724

2.00E-01

2.00E-02

2.00E-02

1.46E-01

Endosulfan II

rat

female

Gavage

subchronic

1.50E+00







reproductive effects and blood chemistry

Dikshith et al., 1984, 99-
0684

1.50E+00

1.50E+00

1.50E-01

1.09E+00

Endiin

mouse

not reported

Food

chronic



9.20E-01





reduced parental survival, litter size, and number of
vouna

Good and Ware, 1969, 99-
0697

9.20E-01

9.20E-02

9.20E-02

6.72E-01

Heptachlor

rat (Sprague-Dawley)

female

Food

chronic



2.50E-01





increased numbers of resorptions; decrease in
conceptions (zero pregnancies in F2 generation);
survival of F1 embryos low (19/25 surviving for 21

Green, 1970, 99-0698

2.50E-01

2.50E-02

2.50E-02

1.82E-01

Heptachlor Epoxide





















NBA

NBA

NBA

NBA

Mirex

rat (Sherman)

both

Food

developmental



1.80E+00





increase in incidence of cataracts in F1 offspring;
decrease in offspring viabiliu



1.80E+00

1.80E-01

1.80E-01

1.31E+00

cis-Nonachloi





















NBA

NBA

NBA

NBA

trans-Nonachlor





















NBA

NBA

NBA

NBA

Oxvchlordane





















NBA

NBA

NBA

NBA

Tables B-56, B-58, B-59 valued.xlsB-58

-------
Table B-58

Mammalian Toxicity Reference Values
Housatonic River Site, OU2, Pittsfield Massachusetts

Chemical

Receptor (strain)

Sex

Exposure Mode

Duration Category

Dose (m

l/kg-day)

Endpoints

Reference

ORNL
Benchmark Dose

Study
NOAEL

Chronic NOAEL

Fish
Concentration

No Effect

Effect

NOAEL

LOAEL

Toxaphene

rat

not reported

Food

chronic

8.00E+00







reproduction

Kennedy et al., 1973, 99-
0707

8.00E+00

8.00E+00

8.00E+00

5.84E+01

Dioxins/Furans

Total Dioxins





















NBA

NBA

NBA

NBA

Total Furans





















NBA

NBA

NBA

NBA

PCBs

Total PCBsb

mink

not reported

Food

chronic





1.37E-01

6.85E-01

reduced number of offspring born alive

Aulerich and Ringer, 1977,
99-0674

1.37E-01

1.37E-01

1.37E-01

1.00E+00

Metals

Lead

rat

not reported

Food

chronic





8.00E+00

8.00E+01

reduced offspring weights and kidney damage in young

Azaret al., 1973, 99-0675

8.00E+00

8.00E+00

8.00E+00

5.84E+01

Mercury1

mink

not reported

Food

subchronic





1.50E-01

2.47E-01

mortality, weight loss, behavioral abnormalities

Wobeser et al., 1976, 99-
0729

1.50E-01

1.50E-01

1.50E-02

1.09E-01

Nickel

rat

not reported

Food

developmental





4.00E+01

8.00E+01

reduced offspring body weights

Ambrose et al., 1976, 99-
0671 (as cited in Sample et
a1 19961

4.00E+01

4.00E+01

4.00E+01

2.92E+02

NBA = No benchmark available

a o,p'-DDD value.

Aroclor 1254 value.
cMethylmercury value.

Tables B-56, B-58, B-59 valued.xlsB-58

-------
Table B-59

Avian Reference Toxicity Values
Housatonic River Site, OU2, Pittsfield Massachusetts











Dose (mg/kg-day)





ORNL





Fish

Chemical

Receptor (sceintific name)

Sex

Exposure Modi

Duration Category

No Effect Dose

Effect Dose

NOAEL

LOAEL

LD50/LC50

Endpoints

Authors

Benchmark Dose
(mg/kg-day)

NOAEL
(mg/kg-dav)

Chronic NOAEL
(mg/kg-dav)

Concentration
(mg/kg)

Semivolatiles

Hexachlorobenzene

Japanese quail (Coturnix
coturnix iavomca}

both

Food

chronic





1.40E-01

7.20E-01



increase in liver weight, slight liver damage, and
enlaraed fecal excretion of coproporphvrin

Vos et al., 1971,99-0726

1.40E-01

1.40E-01

1.40E-01

2.76E-01

Pentachloroanisole























NBA

NBA

NBA

NBA

Pentachlorobenzene























NBA

NBA

NBA

NBA

1.2.3.4-Tetrachlorobenzene























NBA

NBA

NBA

NBA

1.2.4.5-Tetrachlorobenzene























NBA

NBA

NBA

NBA

Pesticides



Aldrin

red-winged blackbird
(Aselaius Dhoeniceus 1

male

Food

subchronic



2.00E+00







weight loss

Clark, 1975, 99-0681

2.00E+00

2.00E-01

2.00E-02

3.95E-02

alpha-BHC























NBA

NBA

NBA

NBA

beta-BHC























NBA

NBA

NBA

NBA

delta-BHC























NBA

NBA

NBA

NBA

gamma-BHC (Lindane)

mallard (Anas platyrhyncos)

not reported

Gavage

chronic



2.00E+01







reduced eggshell tickness, fewer eggs laid and longer
itme intervals between eaas

Chakravarty and Lahiri,
1986. 99-0679

2.00E+01

2.00E+00

2.00E+00

3.95E+00

alpha-Chlordane

red-winged blackbird
(Aaelaius ohoeniceus)

not reported

Food

chronic





2.14E+00

1.07E+01



26% mortality

Stickel et al., 1983, 99-
0721

2.14E+00

2.14E+00

2.14E+00

4.22E+00

gamma-Chlordane

red-winged blackbird
(Aaelaius ohoeniceus")

not reported

Food

chronic





2.14E+00

1.07E+01



26% mortality

Stickel etal., 1983, 99-
0721

2.14E+00

2.14E+00

2.14E+00

4.22E+00

Chlorpyrifos

mallard (Anas
platyrhyncos)

both

Food

acute



1.80E+01







significant decrease in body weight; significant
decrease in brain chholinesterase ; significant
decresase in salt gland weight, function, and

Herin et al., 1978, 99-0701

1.80E+01

1.80E+00

1.80E-02

3.55E-02

o,p'-DDDa

Japanese quail (Coturnix
coturnix iavomca 1

unknown

Food

acute





4.21E+02

5.55E+02

7.78E+02

mortality

Hill and Camardese, 1986,
99-0702

4.21E+02

4.21E+02

4.21E+00

8.31E+00

4,4'-DDD

Japanese quail (Coturnix
coturnix iavomca 1

unknown

Food

acute





4.21E+02

5.55E+02

7.78E+02

mortality

Hill and Camardese, 1986,
99-0702

4.21E+02

4.21E+02

4.21E+00

8.31E+00

o,p'-DDEb

black duck (Anas rubripes)

both

Food

developmental



5.00E-01







signficant eggshell thinning and cracking; embryonic
mortality; significantly lower duckling survival at 21
davs

Longcore et al., 1971, 99-
0709

5.00E-01

5.00E-02

5.00E-02

9.87E-02

4,4'-DDE

black duck (Anas rubripes)

both

Food

developmental



5.00E-01







signficant eggshell thinning and cracking; embryonic
mortality; significantly lower duckling survival at 21
davs

Longcore et al., 1971, 99-
0709

5.00E-01

5.00E-02

5.00E-02

9.87E-02

o,p'-DDT

brown pelican

not reported

Food

chronic



2.80E-02







fledgling rate 30% below that needed to maintain a
stable Dooulation

Anderson et al., 1975, 99-
0672

2.80E-02

2.80E-03

2.80E-03

5.53E-03

4,4'-DDT

brown pelican

not reported

Food

chronic



2.80E-02







fledgling rate 30% below that needed to maintain a
stable Dooulation

Anderson et al., 1975, 99-
0672

2.80E-02

2.80E-03

2.80E-03

5.53E-03

Dieldrin

barn owl

not reported

Food

chronic

7.70E-02









significant reduction in egshell thickness, no
significant effect on number of eggs laid/pair, number
of eggs hatched/pair, percent eggs broken, embryo or

Mendenhall et al., 1983 ,
99-0714

7.70E-02

7.70E-02

7.70E-02

1.52E-01

Endosulfan II

arav partridae

unknown

Food

chronic

1.00E+01









no effects on reproduction

Abiola. 1992. 99-0670

1.00E+01

1.00E+01

1.00E+01

1.97E+01

Endrin

screech owl

unknown

Food

chronic



1.04E-01







reduced egg production and hatching success

Fleming et al., 1982, 99-
0696

1.04E-01

1.04E-02

1.04E-02

2.04E-02

Heptachlor

Japanese quail (Coturnix
coturnix iavomca 1

unknown

Food

acute





1.10E+01

1.60E+01

1.70E+01

mortality

Hill and Camardese, 1986,
99-0702

1.10E+01

1.10E+01

1.10E-01

2.17E-01

Heptachlor Epoxide























NBA

NBA

NBA

NBA

Mirex

mallard (Anas platyrhyncos)

both

Food

developmental



1.00E+01







reduced duckling survival

Hyde etal., 1973,99-0703

1.00E+01

1.00E+00

1.00E+00

1.97E+00

cis-Nonachlor























NBA

NBA

NBA

NBA

trans-Nonachlor























NBA

NBA

NBA

NBA

Oxvchlordane























NBA

NBA

NBA

NBA

Toxaphene

chicken

unknown

Food

chronic

2.10E-01









everv measured parameter

Bush etal.. 1977. 99-0677

2.10E-01

2.10E-01

2.10E-01

4.14E-01

Dioxins/Furans



Total Dioxins























NBA

NBA

NBA

NBA

Total Furans























NBA

NBA

NBA

NBA

PCBs





















NBA

NBA

NBA

NBA

Total PCBsc

ring-necked pheasant

unknown

Capsule

chronic



1.80E+00







significantly reduced egg hatchability

Dahlgren et al., 1972, 99-
0683

1.80E+00

1.80E-01

1.80E-01

3.55E-01

Metals































Lead

American kestrel

unknown

Food

chronic

3.85E+00









no effects on reproduction

Pattee. 1984. 99-0717

3.85E+00

3.85E+00

3.85E+00

7.60E+00

Mercury

mallard (Anas platyrhyncos)

unknown

Food

chronic



6.40E-02







fewer eggs and ducklings produced

Heinz, 1979, 99-0700

6.40E-02

6.40E-03

6.40E-03

1.26E-02

Nickel

mallard (Anas platyrhyncos)

unknown

Food

chronic





7.74E+01

1.07E+02



reduced growth and 70% mortality of ducklings

Cain and Pafford, 1981, 99
0678

7.74E+01

7.74E+01

7.74E+01

1.53E+02

NBA =No benchmark available.

0.506756756756757 food factor for kingfisher (Sample et al., 1996; Table B.l)
a 4,4'-DDD value.
b 4,4'-DDE value.
c Aroclor 1254 value.

Methylmercury value.

Tables B-56, B-58, B-59 valued.xlsB-59

-------
Table B-60

Sediment Comparison to Benchmark Summary - Low
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 17

—

—

—

—

DIBENZOFURAN

2 / 15

2

—

—

—

1,4-DICHLOROBENZENE

0 / 16

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PENTACHLOROBENZENE

0 / 7

—

—

—

—

PAHs

ACENAPHTHENE

8 / 16

7

1

—

—

ACENAPHTHYLENE

5 / 20

4

1

—

—

ANTHRACENE

17 / 22

12

4

1

—

BENZO(A)ANTHRACENE

21 / 22

15

5

1

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

20 / 22

18

2

—

—

BENZO(GHI)PERYLENE

22 / 22

3

16

3

—

BENZO(A)PYRENE

20 / 22

15

5

—

—

CHRYSENE

21 / 22

15

6

—

—

DIBENZO(A,H) ANTHRACENE

20 / 20

16

4

—

—

FLUORANTHENE

19 / 22

15

4

—

—

FLUORENE

14 / 19

11

3

—

—

INDENO( 1,2,3 -C,D)P"YRENE

22 / 22

6

14

2

—

NAPHTHALENE

7 / 21

6

1

—

—

PHENANTHRENE

20 / 22

15

4

1

—

PYRENE

21 / 22

14

6

1

—

TOTAL PAH (USING 0)

22 / 24

15

7

—

—

NBA = No benchmark available.

Tables B-60 - B-87.xls B-60

-------
Table B-60

Sediment Comparison to Benchmark Summary - Low
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING DL)

24 / 24

17

7

—

—

TOTAL PAH (USING HALF DL)

23 / 24

16

7

—

—

TOTAL PAH (HIGH) (USING 0)

24 / 24

2

17

5

—

TOTAL PAH (HIGH) (USING DL)

24 / 24

2

17

5

—

TOTAL PAH (HIGH) (USING HALF DL)

24 / 24

2

17

5

—

TOTAL PAH (LOW) (USING 0)

24 / 24

2

16

6

—

TOTAL PAH (LOW) (USING DL)

24 / 24

—

18

6

—

TOTAL PAH (LOW) (USING HALF DL)

24 / 24

—

18

6

—

1,2,4,5-TETRACHLOROBENZENE

NBA

—

—

—

—

1,2,4-TRICHLOROBENZENE

0 / 10

—

—

—

—

APP IX PESTICIDES

4,4'-DDD

1 / 1

1

—

—

—

ENDRIN ALDEHYDE

NBA

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

405 / 405

12

104

260

29

METALS

NBA = No benchmark available.

Tables B-60 - B-87.xls B-60

-------
Table B-60

Sediment Comparison to Benchmark Summary - Low
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 19

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CHROMIUM

2 / 22

2

—

—

—

COBALT

0 / 22

—

—

—

—

COPPER

3 / 22

3

—

—

—

LEAD

6 / 22

6

—

—

—

MERCURY

3 / 17

3

—

—

—

NICKEL

1 / 20

1

—

—

—

SILVER

0 / 4

—

—

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

4 / 22

4

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

AMMONIA AS N

0 / 11

—

—

—

—

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-60 - B-87.xls B-60

-------
Table B-61

Sediment Comparison to Benchmark Summary - High*
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

—

DIBENZOFURAN

NBA

—

—

—

—

1,4-DICHLOROBENZENE

NBA

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PENTACHLOROBENZENE

NBA

—

—

—

—

PAHs

ACENAPHTHENE

NBA

—

—

—

—

ACENAPHTHYLENE

NBA

—

—

—

—

ANTHRACENE

4 / 22

3

1

—

—

BENZO(A)ANTHRACENE

6 / 22

5

1

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 22

—

—

—

—

BENZO(GHI)PERYLENE

13 / 22

11

2

—

—

BENZO(A)PYRENE

5 / 22

5

—

—

—

CHRYSENE

6 / 22

5

1

—

—

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

—

FLUORANTHENE

6 / 22

6

—

—

—

FLUORENE

4 / 19

4

—

—

—

INDENO( 1,2,3 -C,D)PYRENE

14 / 22

12

2

—

—

NAPHTHALENE

1 / 21

1

—

—

—

PHENANTHRENE

7 / 22

6

1

—

—

PYRENE

8 / 22

7

1

—

—

TOTAL PAH (USING 0)

6 / 24

6

—

—

—

NBA = No benchmark available.

Tables B-60-B-87.xls B-61

-------
Table B-61

Sediment Comparison to Benchmark Summary - High*
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING DL)

6 / 24

6

—

—

—

TOTAL PAH (USING HALF DL)

6 / 24

6

—

—

—

TOTAL PAH (HIGH) (USING 0)

22 / 24

17

5

—

—

TOTAL PAH (HIGH) (USING DL)

22 / 24

17

5

—

—

TOTAL PAH (HIGH) (USING HALF DL)

22 / 24

17

5

—

—

TOTAL PAH (LOW) (USING 0)

18 / 24

14

4

—

—

TOTAL PAH (LOW) (USING DL)

23 / 24

19

4

—

—

TOTAL PAH (LOW) (USING HALF DL)

21 / 24

17

4

—

—

1,2,4,5-TETRACHLOROBENZENE

NBA

—

—

—

—

1,2,4-TRICHLOROBENZENE

NBA

—

—

—

—

APP IX PESTICIDES

4,4'-DDD

0 / 1

—

—

—

—

ENDRIN ALDEHYDE

NBA

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

393 / 405

126

243

24

—

METALS

NBA = No benchmark available.

Tables B-60-B-87.xls B-61

-------
Table B-61

Sediment Comparison to Benchmark Summary - High*
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 19

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CHROMIUM

0 / 22

—

—

—

—

COBALT

NBA

—

—

—

—

COPPER

0 / 22

—

—

—

—

LEAD

1 / 22

1

—

—

—

MERCURY

0 / 17

—

—

—

—

NICKEL

0 / 20

—

—

—

—

SILVER

NBA

—

—

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

0 / 22

—

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

AMMONIA AS N

NBA

—

—

—

—

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-60-B-87.xls B-61

-------
Table B-62

Sediment Comparison to Benchmark Summary - Low
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 1

—

—

—

—

PAHs

BENZO(A)ANTHRACENE

1 / 1

1

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 1

—

—

—

—

BENZO(GHI)PERYLENE

1 / 1

1

—

—

—

BENZO(A)PYRENE

0 / 1

—

—

—

—

CHRYSENE

0 / 1

—

—

—

—

DIBENZO(A,H) ANTHRACENE

1 / 1

—

1

—

—

FLUORANTHENE

0 / 1

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

1 / 1

1

—

—

—

PHENANTHRENE

1 / 1

1

—

—

—

PYRENE

1 / 1

1

—

—

—

TOTAL PAH (USING 0)

1 / 2

1

—

—

—

TOTAL PAH (USING DL)

2 / 2

1

1

—

—

TOTAL PAH (USING HALF DL)

2 / 2

2

—

—

—

TOTAL PAH (HIGH) (USING 0)

1 / 2

—

1

—

—

TOTAL PAH (HIGH) (USING DL)

2 / 2

—

2

—

—

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

—

2

—

—

TOTAL PAH (LOW) (USING 0)

1 / 2

1

—

—

—

TOTAL PAH (LOW) (USING DL)

2 / 2

—

1

1

—

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

—

2

—

—

APP IX PESTICIDES

ENDRIN ALDEHYDE

NBA

—

—

—

—

NBA = No benchmarks available.

Tables B-60 - B-87.xls B-62

-------
Table B-62

Sediment Comparison to Benchmark Summary - Low
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

22 / 22

—

5

16

1

METALS

ARSENIC

0 / 2

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

0 / 1

—

—

—

—

CHROMIUM

0 / 2

—

—

—

—

COBALT

0 / 2

—

—

—

—

COPPER

0 / 2

—

—

—

—

LEAD

1 / 2

1

—

—

—

MERCURY

0 / 1

—

—

—

—

NICKEL

1 / 2

1

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

0 / 2

—

—

—

—

NBA = No benchmarks available.

Tables B-60 - B-87.xls B-62

-------
Table B-62

Sediment Comparison to Benchmark Summary - Low
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmarks available.

Tables B-60 - B-87.xls B-62

-------
Table B-63

Sediment Comparison to Benchmark Summary - High*
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

—

PAHs

BENZO(A)ANTHRACENE

0 / 1

—

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 1

—

—

—

—

BENZO(GHI)PERYLENE

0 / 1

—

—

—

—

BENZO(A)PYRENE

0 / 1

—

—

—

—

CHRYSENE

0 / 1

—

—

—

—

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

—

FLUORANTHENE

0 / 1

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

0 / 1

—

—

—

—

PHENANTHRENE

0 / 1

—

—

—

—

PYRENE

0 / 1

—

—

—

—

TOTAL PAH (USING 0)

0 / 2

—

—

—

—

TOTAL PAH (USING DL)

1 / 2

1

—

—

—

TOTAL PAH (USING HALF DL)

0 / 2

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

1 / 2

1

—

—

—

TOTAL PAH (HIGH) (USING DL)

2 / 2

2

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

2

—

—

—

TOTAL PAH (LOW) (USING 0)

0 / 2

—

—

—

—

TOTAL PAH (LOW) (USING DL)

2 / 2

2

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

2

—

—

—

APP IX PESTICIDES

ENDRIN ALDEHYDE

NBA

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-63

-------
Table B-63

Sediment Comparison to Benchmark Summary - High*
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

22 / 22

6

15

1

—

METALS

ARSENIC

0 / 2

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

0 / 1

—

—

—

—

CHROMIUM

0 / 2

—

—

—

—

COBALT

NBA

—

—

—

—

COPPER

0 / 2

—

—

—

—

LEAD

0 / 2

—

—

—

—

MERCURY

0 / 1

—

—

—

—

NICKEL

0 / 2

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

0 / 2

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-63

-------
Table B-63

Sediment Comparison to Benchmark Summary - High*
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-63

-------
Table B-64

Sediment Comparison to Benchmark Summary - Low
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

ACETOPHENONE

NBA

—

—

—

...

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 1

...

...

...

...

DI-N-BUTYL PHTHALATE

0 / 1

—

—

—

...

1,4-DICHLOROBENZENE

0 / 3

—

—

—

...

2-METHYLNAPHTHALENE

NBA

—

—

—

...

4-METHYLPHENOL

NBA

—

—

—

...

PAHs











ACENAPHTHENE

0 / 1

—

—

—

...

ACENAPHTHYLENE

1 / 4

1

—

—

...

ANTHRACENE

2 / 4

2

—

—

...

BENZO(A)ANTHRACENE

4 / 6

4

—

—

...

BENZO(B)FLUORANTHENE

NBA

—

—

—

...

BENZO(K)FLUORANTHENE

2 / 5

2

—

—

...

BENZO(GHI)PERYLENE

5 / 5

2

3

—

...

BENZO(A)PYRENE

4 / 5

4

—

—

...

CHRYSENE

4 / 6

4

—

—

...

DIBENZO(A,H) ANTHRACENE

4 / 4

4

—

—

...

FLUORANTHENE

2 / 7

2

—

—

...

FLUORENE

0 / 2

...

...

—

...

INDENO( 1,2,3 -C,D)PYRENE

5 / 5

3

2

—

...

NAPHTHALENE

0 / 5

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-64

-------
Table B-64

Sediment Comparison to Benchmark Summary - Low
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PHENANTHRENE

4 / 6

4

—

—

—

PYRENE

4 / 6

4

...

...

...

TOTAL PAH (USING 0)

4 / 9

4

—

—

...

TOTAL PAH (USING DL)

9 / 9

6

3

—

...

TOTAL PAH (USING HALF DL)

9 / 9

9

—

—

...

TOTAL PAH (HIGH) (USING 0)

6 / 9

3

3

—

...

TOTAL PAH (HIGH) (USING DL)

9 / 9

2

7

—

...

TOTAL PAH (HIGH) (USING HALF DL)

9 / 9

2

7

—

...

TOTAL PAH (LOW) (USING 0)

6 / 9

3

3

—

...

TOTAL PAH (LOW) (USING DL)

9 / 9

—

7

2

...

TOTAL PAH (LOW) (USING HALF DL)

9 / 9

—

9

—

...

1,2.4-TRICHLOROBENZENE

0 / 2

—

...

...

...

APP IX PESTICIDES

ALPHA-BHC

1 / 1

1

—

—

...

BETA-BHC

1 / 1

1

...

...

...

4.4'-DDD

2 / 2

—

2

—

...

4.4'-DDE

5 / 5

—

2

3

...

HEPTACHLOR

1 / 1

—

1

—

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-64

-------
Table B-64

Sediment Comparison to Benchmark Summary - Low
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

...

...

...

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

189 / 189

7

60

97

25

METALS

ANTIMONY

NBA

—

—

—

...

ARSENIC

0 / 6

—

—

—

...

BARIUM

NBA

—

—

—

...

BERYLLIUM

NBA

—

—

—

...

CADMIUM

2 / 4

2

—

—

...

CHROMIUM

1 / 9

1

—

—

...

COBALT

0 / 9

...

...

...

...

COPPER

5 / 9

5

—

—

...

LEAD

8 / 9

8

...

...

...

MERCURY

6 / 6

6

—

—

...

NICKEL

3 / 9

3

...

...

...

SELENIUM

NBA

—

—

—

...

SILVER

3 / 3

3

...

...

...

THALLIUM

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-64

-------
Table B-64

Sediment Comparison to Benchmark Summary - Low
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TIN

NBA

—

—

—

—

VANADIUM

NBA

...

...

...

...

ZINC

6 / 9

6

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

INORGANICS

PERCENT SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-64

-------
Table B-65

Sediment Comparison to Benchmark Summary - High*
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

ACETOPHENONE

NBA

—

—

—

...

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

...

...

...

...

1,4-DICHLOROBENZENE

NBA

—

—

—

...

DI-N-BUTYL PHTHALATE

NBA

—

—

—

...

2-METHYLNAPHTHALENE

NBA

—

—

—

...

4-METHYLPHENOL

NBA

—

—

—

...

PAHs











ACENAPHTHENE

NBA

—

—

—

...

ACENAPHTHYLENE

NBA

—

—

—

...

ANTHRACENE

0 / 4

—

—

—

...

BENZO(A)ANTHRACENE

0 / 6

—

—

—

...

BENZO(B)FLUORANTHENE

NBA

—

—

—

...

BENZO(K)FLUORANTHENE

0 / 5

—

—

—

...

BENZO(GHI)PERYLENE

2 / 5

2

—

—

...

BENZO(A)PYRENE

0 / 5

—

—

—

...

CHRYSENE

1 / 6

1

—

—

...

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

...

FLUORANTHENE

0 / 7

—

—

—

...

FLUORENE

0 / 2

...

...

—

...

INDENO( 1,2,3 -C,D)PYRENE

2 / 5

2

—

—

...

NAPHTHALENE

0 / 5

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-65

-------
Table B-65

Sediment Comparison to Benchmark Summary - High*
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PHENANTHRENE

0 / 6

—

—

—

—

PYRENE

1 / 6

1

...

...

...

TOTAL PAH (USING 0)

0 / 9

—

—

—

...

TOTAL PAH (USING DL)

0 / 9

—

—

—

...

TOTAL PAH (USING HALF DL)

0 / 9

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

3 / 9

3

—

—

...

TOTAL PAH (HIGH) (USING DL)

7 / 9

7

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

7 / 9

7

—

—

...

TOTAL PAH (LOW) (USING 0)

2 / 9

2

—

—

...

TOTAL PAH (LOW) (USING DL)

9 / 9

9

—

—

...

TOTAL PAH (LOW) (USING HALF DL)

8 / 9

8

—

—

...

1,2.4-TRICHLOROBENZENE

NBA

—

—

—

...

APP IX PESTICIDES

ALPHA-BHC

0 / 1

—

—

—

...

BETA-BHC

0 / 1

—

...

...

...

4.4'-DDD

2 / 2

1

1

—

...

4.4'-DDE

5 / 5

2

3

—

...

HEPTACHLOR

NBA

—

—

—

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-65

-------
Table B-65

Sediment Comparison to Benchmark Summary - High*
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

...

...

...

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

180 / 189

64

93

22

1

METALS

ANTIMONY

NBA

—

—

—

...

ARSENIC

0 / 6

—

—

—

...

BARIUM

NBA

—

—

—

...

BERYLLIUM

NBA

—

—

—

...

CADMIUM

0 / 4

—

—

—

...

CHROMIUM

0 / 9

—

—

—

...

COBALT

NBA

—

—

—

...

COPPER

0 / 9

—

—

—

...

LEAD

1 / 9

1

...

...

...

MERCURY

0 / 6

—

—

—

...

NICKEL

0 / 9

...

...

...

...

SELENIUM

NBA

—

—

—

...

SILVER

NBA

—

—

—

...

THALLIUM

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-65

-------
Table B-65

Sediment Comparison to Benchmark Summary - High*
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TIN

NBA

—

—

—

—

VANADIUM

NBA

...

...

...

...

ZINC

0 / 9

...

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

INORGANICS

PERCENT SOLIDS

NBA

...

...

...

...

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-65

-------
Table B-66

Sediment Comparison to Benchmark Summary - Low
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 3

—

—

—

—

BUTYLBENZYLPHTHALATE

0 / 1

—

—

—

—

DIBENZOFURAN

0 / 3

—

—

—

—

DI-N-BUTYL PHTHALATE

0 / 1

—

—

—

—

1.4-DICHLOROBENZENE

0 / 2

—

—

—

—

DIETHYL PHTHALATE

0 / 1

—

—

—

—

2-METHYLNAPHTHALENE

NBA

---

---

---

—

PENTACHLOROBENZENE

0 / 1

—

—

—

—

PAHs

ACENAPHTHENE

0 / 2







—

ACENAPHTHYLENE

0 / 4

—

—

—

—

ANTHRACENE

4 / 4

4





—

BENZO(A)ANTHRACENE

4 / 5

4

—

—

—

BENZO(B)FLUORANTHENE

NBA







—

BENZO(K)FLUORANTHENE

4 / 5

4

—

—

—

BENZO(GHI)PERYLENE

5 / 5

2

3

—

—

BENZO(A)PYRENE

4 / 5

4

---

---

—

CHRYSENE

4 / 5

4





—

DIBENZO(A,H) ANTHRACENE

3 / 5

3

—

—

—

FLUORANTHENE

3 / 6

3

—

—

—

FLUORENE

1 / 4

1

—

—

—

INDENO( 1,2,3 -C,D)P YRENE

5 / 5

2

3



—

NAPHTHALENE

0 / 4

—

—

—

—

PHENANTHRENE

4 / 5

4

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-66

-------
Table B-66

Sediment Comparison to Benchmark Summary - Low
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PYRENE

4 / 6

4

—

---

—

TOTAL PAH (USING 0)

4 / 6

4

—

—

—

TOTAL PAH (USING DL)

6 / 6

6

—

—

—

TOTAL PAH (USING HALF DL)

6 / 6

6

—

—

—

TOTAL PAH (HIGH) (USING 0)

5 / 6

1

4

—

—

TOTAL PAH (HIGH) (USING DL)

6 / 6

1

5

—

—

TOTAL PAH (HIGH) (USING HALF DL)

6 / 6

1

5

—

—

TOTAL PAH (LOW) (USING 0)

5 / 6

1

4

—

—

TOTAL PAH (LOW) (USING DL)

6 / 6

—

6

—

—

TOTAL PAH (LOW) (USING HALF DL)

6 / 6

—

6

—

—

DIOXINS/FURANS

1.2,3,4,6,7,8-HPCDD

NBA







—

1,2,3,4,6,7,8-HPCDF

NBA

---

---

---

—

1,2,3,4,7,8,9-HPCDF

NBA







—

1,2,3,4,7,8-HXCDD

NBA

---

---

---

—

1,2,3,4,7,8-HXCDF

NBA







—

1,2,3,6,7,8-HXCDD

NBA

---

---

---

—

1,2,3,6,7,8-HXCDF

NBA







—

1,2,3,7,8,9-HXCDD

NBA

---

---

---

—

1,2,3,7,8,9-HXCDF

NBA







—

1,2,3,7,8-PECDD

NBA

---

---

---

—

1,2,3,7,8-PECDF

NBA







—

2,3,4,6,7,8-HXCDF

NBA

---

---

---

—

2,3,4,7,8-PECDF

NBA







—

2,3,7,8-TCDD

NBA

—

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-66

-------
Table B-66

Sediment Comparison to Benchmark Summary - Low
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

2,3,7,8-TCDF

NBA

---

---

---

—

HPCDD (TOTAL)

NBA







—

HPCDF (TOTAL)

NBA

---

---

---

—

HXCDD (TOTAL)

NBA







—

HXCDF (TOTAL)

NBA

---

---

---

—

OCDD

NBA







—

OCDF

NBA

---

---

---

—

PECDD (TOTAL)

NBA







—

PECDF (TOTAL)

NBA

---

---

---

—

TCDD (TOTAL)

NBA







—

TCDF (TOTAL)

NBA

---

---

---

—

TEQ 2,3,7,8-TCDD (EPA)

NBA







—

TEQ 2,3,7,8-TCDD (MADEP)

NBA

---

---

---

—

TOTAL DIOXINS (USING 0)

NBA







—

TOTAL DIOXINS (USING DL)

NBA

---

---

---

—

TOTAL DIOXINS (USING HALF DL)

NBA







—

TOTAL FURANS (USING 0)

NBA

---

---

---

—

TOTAL FURANS (USING DL)

NBA







—

TOTAL FURANS (USING HALF DL)

NBA

---

---

---

—

GRAIN SIZE

SIZE (01) 075.0 MM

NBA

---

---

---

—

SIZE (02) 050.0 MM

NBA







—

SIZE (03) 037.5 MM

NBA

---

---

---

—

SIZE (04) 025.0 MM

NBA







—

SIZE (05) 019.0 MM

NBA

—

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-66

-------
Table B-66

Sediment Comparison to Benchmark Summary - Low
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

SIZE (06) 012.7 MM

NBA

---

---

---

—

SIZE (07) 009.5 MM

NBA







—

SIZE (08) 004.75 MM

NBA

---

---

---

—

SIZE (09) 002.0 MM

NBA







—

SIZE (10) 850.0 UM

NBA

---

---

---

—

SIZE (11) 425.0 UM

NBA







—

SIZE (12) 250.0 UM

NBA

---

---

---

—

SIZE (13) 180.0 UM

NBA







—

SIZE (14) 150.0 UM

NBA

---

---

---

—

SIZE (15) 075.0 UM

NBA







—

SIZE HYDRO (16) -49.0 UM

NBA

---

---

---

—

SIZE HYDRO (17) -37.0 UM

NBA







—

SIZE HYDRO (18) -23.0 UM

NBA

---

---

---

—

SIZE HYDRO (19) -18.0 UM

NBA







—

SIZE HYDRO (20) -13.0 UM

NBA

---

---

---

—

SIZE HYDRO (21) -09.0 UM

NBA







—

SIZE HYDRO (22) -07.0 UM

NBA

---

---

---

—

SIZE HYDRO (23) -04.0 UM

NBA







—

SIZE HYDRO (24) -03.0 UM

NBA

---

---

---

—

SIZE HYDRO (25) -01.4 UM

NBA







—

PCBS

AROCLOR-1260

NBA







—

PCB, TOTAL

154 / 156

14

88

51

1

METALS

ANTIMONY

NBA

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-66

-------
Table B-66

Sediment Comparison to Benchmark Summary - Low
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

0 / 6

---

---

---

—

BARIUM

NBA







—

BERYLLIUM

NBA

---

---

---

—

CHROMIUM

1 / 6

1





—

COBALT

0 / 6

—

—

—

—

COPPER

2 / 6

2





—

LEAD

2 / 6

2

—

—

—

MERCURY

1 / 5

1





—

NICKEL

0 / 6

—

—

—

—

SILVER

1 / 2

1

—

—

—

THALLIUM

NBA

---

---

---

—

TIN

NBA







—

VANADIUM

NBA

---

---

---

—

ZINC

2 / 6

2

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

INORGANICS

<63UM

NBA







—

>250UM

NBA

---

---

---

—

63-250UM

NBA







—

PERCENT SOLIDS

NBA

---

---

---

—

SULFIDE

NBA







—

TOTAL WEIGHT

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-66

-------
Table B-67

Sediment Comparison to Benchmark Summary - High*
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

---

---

---

—

BUTYLBENZYLPHTHALATE

NBA







—

DIBENZOFURAN

NBA

---

---

---

—

DI-N-BUTYL PHTHALATE

NBA







—

1.4-DICHLOROBENZENE

NBA

---

---

---

—

DIETHYL PHTHALATE

NBA







—

2-METHYLNAPHTHALENE

NBA

---

---

---

—

PENTACHLOROBENZENE

NBA







—

PAHs

ACENAPHTHENE

NBA







—

ACENAPHTHYLENE

NBA

---

---

---

—

ANTHRACENE

0 / 4







—

BENZO(A)ANTHRACENE

0 / 5

—

—

—

—

BENZO(B)FLUORANTHENE

NBA







—

BENZO(K)FLUORANTHENE

0 / 5

—

—

—

—

BENZO(GHI)PERYLENE

1 / 5

1

—

—

—

BENZO(A)PYRENE

0 / 5

---

---

---

—

CHRYSENE

0 / 5







—

DIBENZO(A,H) ANTHRACENE

NBA

---

---

---

—

FLUORANTHENE

0 / 6

—

—

—

—

FLUORENE

0 / 4

—

—

—

—

INDENO( 1,2,3 -C,D)P YRENE

1 / 5

1





—

NAPHTHALENE

0 / 4

—

—

—

—

PHENANTHRENE

0 / 5

—

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-67

-------
Table B-67

Sediment Comparison to Benchmark Summary - High*
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PYRENE

0 / 6

—

—

---

—

TOTAL PAH (USING 0)

0 / 6

—

—

—

—

TOTAL PAH (USING DL)

0 / 6

—

—

—

—

TOTAL PAH (USING HALF DL)

0 / 6

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

4 / 6

4

—

—

—

TOTAL PAH (HIGH) (USING DL)

5 / 6

5

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

5 / 6

5

—

—

—

TOTAL PAH (LOW) (USING 0)

2 / 6

2

—

—

—

TOTAL PAH (LOW) (USING DL)

6 / 6

6

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

6 / 6

6

—

—

—

DIOXINS/FURANS

1.2,3,4,6,7,8-HPCDD

NBA







—

1,2,3,4,6,7,8-HPCDF

NBA

---

---

---

—

1,2,3,4,7,8,9-HPCDF

NBA







—

1,2,3,4,7,8-HXCDD

NBA

---

---

---

—

1,2,3,4,7,8-HXCDF

NBA







—

1,2,3,6,7,8-HXCDD

NBA

---

---

---

—

1,2,3,6,7,8-HXCDF

NBA







—

1,2,3,7,8,9-HXCDD

NBA

---

---

---

—

1,2,3,7,8,9-HXCDF

NBA







—

1,2,3,7,8-PECDD

NBA

---

---

---

—

1,2,3,7,8-PECDF

NBA







—

2,3,4,6,7,8-HXCDF

NBA

---

---

---

—

2,3,4,7,8-PECDF

NBA







—

2,3,7,8-TCDD

NBA

—

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-67

-------
Table B-67

Sediment Comparison to Benchmark Summary - High*
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

2,3,7,8-TCDF

NBA

---

---

---

—

HPCDD (TOTAL)

NBA







—

HPCDF (TOTAL)

NBA

---

---

---

—

HXCDD (TOTAL)

NBA







—

HXCDF (TOTAL)

NBA

---

---

---

—

OCDD

NBA







—

OCDF

NBA

---

---

---

—

PECDD (TOTAL)

NBA







—

PECDF (TOTAL)

NBA

---

---

---

—

TCDD (TOTAL)

NBA







—

TCDF (TOTAL)

NBA

---

---

---

—

TEQ 2,3,7,8-TCDD (EPA)

NBA







—

TEQ 2,3,7,8-TCDD (MADEP)

NBA

---

---

---

—

TOTAL DIOXINS (USING 0)

NBA







—

TOTAL DIOXINS (USING DL)

NBA

---

---

---

—

TOTAL DIOXINS (USING HALF DL)

NBA







—

TOTAL FURANS (USING 0)

NBA

---

---

---

—

TOTAL FURANS (USING DL)

NBA







—

TOTAL FURANS (USING HALF DL)

NBA

---

---

---

—

GRAIN SIZE

SIZE (01) 075.0 MM

NBA

---

---

---

—

SIZE (02) 050.0 MM

NBA







—

SIZE (03) 037.5 MM

NBA

---

---

---

—

SIZE (04) 025.0 MM

NBA







—

SIZE (05) 019.0 MM

NBA

—

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-67

-------
Table B-67

Sediment Comparison to Benchmark Summary - High*
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

SIZE (06) 012.7 MM

NBA

---

---

---

—

SIZE (07) 009.5 MM

NBA







—

SIZE (08) 004.75 MM

NBA

---

---

---

—

SIZE (09) 002.0 MM

NBA







—

SIZE (10) 850.0 UM

NBA

---

---

---

—

SIZE (11) 425.0 UM

NBA







—

SIZE (12) 250.0 UM

NBA

---

---

---

—

SIZE (13) 180.0 UM

NBA







—

SIZE (14) 150.0 UM

NBA

---

---

---

—

SIZE (15) 075.0 UM

NBA







—

SIZE HYDRO (16) -49.0 UM

NBA

---

---

---

—

SIZE HYDRO (17) -37.0 UM

NBA







—

SIZE HYDRO (18) -23.0 UM

NBA

---

---

---

—

SIZE HYDRO (19) -18.0 UM

NBA







—

SIZE HYDRO (20) -13.0 UM

NBA

---

---

---

—

SIZE HYDRO (21) -09.0 UM

NBA







—

SIZE HYDRO (22) -07.0 UM

NBA

---

---

---

—

SIZE HYDRO (23) -04.0 UM

NBA







—

SIZE HYDRO (24) -03.0 UM

NBA

---

---

---

—

SIZE HYDRO (25) -01.4 UM

NBA







—

PCBS

AROCLOR-1260

NBA







—

PCB, TOTAL

138 / 156

94

43

1

—

METALS

ANTIMONY

NBA

—

—

—

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-67

-------
Table B-67

Sediment Comparison to Benchmark Summary - High*
5b Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

0 / 6

---

---

---

—

BARIUM

NBA







—

BERYLLIUM

NBA

---

---

---

—

CHROMIUM

1 / 6

1





—

COBALT

NBA

---

---

---

—

COPPER

0 / 6







—

LEAD

0 / 6

—

—

—

—

MERCURY

1 / 5

1





—

NICKEL

0 / 6

—

—

—

—

SILVER

NBA







—

THALLIUM

NBA

---

---

---

—

TIN

NBA







—

VANADIUM

NBA

---

---

---

—

ZINC

0 / 6

—

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

INORGANICS

<63UM

NBA







—

>250UM

NBA

---

---

---

—

63-250UM

NBA







—

PERCENT SOLIDS

NBA

---

---

---

—

SULFIDE

NBA







—

TOTAL WEIGHT

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-67

-------
Table B-68

Sediment Comparison to Benchmark Summary - Low
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 1

—

—

—

...

DIBENZOFURAN

0 / 1

...

...

...

...

1,4-DICHLOROBENZENE

0 / 1

—

—

—

...

DI-N-BUTYL PHTHALATE

0 / 1

—

—

—

...

2-METHYLNAPHTHALENE

NBA

—

—

—

...

4-METHYLPHENOL

NBA

—

—

—

...

PAHs











ACENAPHTHENE

0 / 1

—

—

—

...

ACENAPHTHYLENE

0 / 1

—

—

—

...

ANTHRACENE

1 / 1

1

—

—

...

BENZO(A)ANTHRACENE

1 / 1

1

—

—

...

BENZO(B)FLUORANTHENE

NBA

—

—

—

...

BENZO(K)FLUORANTHENE

1 / 1

1

—

—

...

BENZO(GHI)PERYLENE

1 / 1

—

1

—

...

BENZO(A)PYRENE

1 / 1

1

—

—

...

CHRYSENE

1 / 1

1

—

—

...

DIBENZO(A,H) ANTHRACENE

1 / 1

1

—

—

...

FLUORANTHENE

1 / 1

1

—

—

...

FLUORENE

0 / 1

...

...

—

...

INDENO( 1,2,3 -C,D)PYRENE

1 / 1

—

1

—

...

NAPHTHALENE

0 / 1

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-68

-------
Table B-68

Sediment Comparison to Benchmark Summary - Low
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PHENANTHRENE

1 / 1

1

—

—

—

PYRENE

1 / 1

1

...

...

...

TOTAL PAH (USING 0)

1 / 1

1

—

—

...

TOTAL PAH (USING DL)

1 / 1

1

—

—

...

TOTAL PAH (USING HALF DL)

1 / 1

1

—

—

...

TOTAL PAH (HIGH) (USING 0)

1 / 1

—

1

—

...

TOTAL PAH (HIGH) (USING DL)

1 / 1

—

1

—

...

TOTAL PAH (HIGH) (USING HALF DL)

1 / 1

—

1

—

...

TOTAL PAH (LOW) (USING 0)

1 / 1

—

1

—

...

TOTAL PAH (LOW) (USING DL)

1 / 1

—

1

—

...

TOTAL PAH (LOW) (USING HALF DL)

1 / 1

—

1

—

...

1,2.4-TRICHLOROBENZENE

0 / 1

—

...

...

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-68

-------
Table B-68

Sediment Comparison to Benchmark Summary - Low
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCB, TOTAL

13 / 13

—

2

7

4

METALS

ARSENIC

0 / 1

—

—

—

...

BARIUM

NBA

...

...

...

...

BERYLLIUM

NBA

—

—

—

...

CADMIUM

0 / 1

—

—

—

...

CHROMIUM

1 / 1

1

—

—

...

COBALT

0 / 1

...

...

...

...

COPPER

1 / 1

1

—

—

...

LEAD

1 / 1

1

...

...

...

MERCURY

1 / 1

1

—

—

...

NICKEL

1 / 1

1

...

...

...

SILVER

1 / 1

1

...

...

...

THALLIUM

NBA

—

—

—

...

VANADIUM

NBA

—

—

—

...

ZINC

1 / 1

1

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

INORGANICS

PERCENT SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-68

-------
Table B-69

Sediment Comparison to Benchmark Summary - High*
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

...

DIBENZOFURAN

NBA

...

...

...

...

1,4-DICHLOROBENZENE

NBA

—

—

—

...

DI-N-BUTYL PHTHALATE

NBA

—

—

—

...

2-METHYLNAPHTHALENE

NBA

—

—

—

...

4-METHYLPHENOL

NBA

—

—

—

...

PAHs











ACENAPHTHENE

NBA

—

—

—

...

ACENAPHTHYLENE

NBA

—

—

—

...

ANTHRACENE

0 / 1

—

—

—

...

BENZO(A)ANTHRACENE

0 / 1

—

—

—

...

BENZO(B)FLUORANTHENE

NBA

—

—

—

...

BENZO(K)FLUORANTHENE

0 / 1

—

—

—

...

BENZO(GHI)PERYLENE

0 / 1

—

—

—

...

BENZO(A)PYRENE

0 / 1

—

—

—

...

CHRYSENE

0 / 1

—

—

—

...

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

...

FLUORANTHENE

0 / 1

—

—

—

...

FLUORENE

0 / 1

...

...

—

...

INDENO( 1,2,3 -C,D)PYRENE

0 / 1

—

—

—

...

NAPHTHALENE

0 / 1

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-69

-------
Table B-69

Sediment Comparison to Benchmark Summary - High*
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PHENANTHRENE

0 / 1

—

—

—

—

PYRENE

0 / 1

...

...

...

...

TOTAL PAH (USING 0)

0 / 1

—

—

—

...

TOTAL PAH (USING DL)

0 / 1

—

—

—

...

TOTAL PAH (USING HALF DL)

0 / 1

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

1 /1

1

—

—

...

TOTAL PAH (HIGH) (USING DL)

1 /1

1

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

1 /1

1

—

—

...

TOTAL PAH (LOW) (USING 0)

1 /1

1

—

—

...

TOTAL PAH (LOW) (USING DL)

1 /1

1

—

—

...

TOTAL PAH (LOW) (USING HALF DL)

1 /1

1

—

—

...

1,2.4-TRICHLOROBENZENE

NBA

—

—

—

...

DIOXINS/FURANS











TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS











AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-69

-------
Table B-69

Sediment Comparison to Benchmark Summary - High*
5b SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCB, TOTAL

13 / 13

3

7

3

—

METALS











ARSENIC

0 / 1

—

—

—

...

BARIUM

NBA

...

...

...

...

BERYLLIUM

NBA

—

—

—

...

CADMIUM

0 / 1

—

—

—

...

CHROMIUM

0 / 1

—

—

—

...

COBALT

NBA

—

—

—

...

COPPER

0 / 1

—

—

—

...

LEAD

1 / 1

1

...

...

...

MERCURY

0 / 1

—

—

—

...

NICKEL

0 / 1

...

...

...

...

SILVER

NBA

—

—

—

...

THALLIUM

NBA

—

—

—

...

VANADIUM

NBA

—

—

—

...

ZINC

0 / 1

...

...

...

...

ORGANIC











TOTAL ORGANIC CARBON

NBA

—

—

—

...

INORGANICS











PERCENT SOLIDS

NBA

...

...

...

...

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-69

-------
Table B-70

Sediment Comparison to Benchmark Summary - Low
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 2

—

—

—

—

1,4-DICHLOROBENZENE

0 / 4

—

—

—

—

DIETHYL PHTHALATE

0 / 4

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PHENOL

1 / 1

—

1

—

—

PAHs











ACENAPHTHYLENE

1 / 4

1

—

—

—

ANTHRACENE

2 / 4

2

—

—

—

BENZO(A)ANTHRACENE

6 / 7

6

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

4 / 7

4

—

—

—

BENZO(GHI)PERYLENE

7 / 7

3

4

—

—

BENZO(A)PYRENE

5 / 7

5

—

—

—

CHRYSENE

6 / 7

6

—

—

—

DIBENZO(A,H) ANTHRACENE

3 / 3

3

—

—

—

FLUORANTHENE

3 / 7

3

—

—

—

FLUORENE

0 / 1

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

7 / 7

5

2

—

—

NAPHTHALENE

0 / 6

—

—

—

—

PHENANTHRENE

4 / 7

4

—

—

—

PYRENE

6 / 8

6

—

—

—

TOTAL PAH (USING 0)

6 / 9

6

—

—

—

TOTAL PAH (USING DL)

8 / 9

8

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-70

-------
Table B-70

Sediment Comparison to Benchmark Summary - Low
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING HALF DL)

8 / 9

8

—

—

—

TOTAL PAH (HIGH) (USING 0)

7 / 9

3

4

—

—

TOTAL PAH (HIGH) (USING DL)

8 / 9

1

7

—

—

TOTAL PAH (HIGH) (USING HALF DL)

8 / 9

1

7

—

—

TOTAL PAH (LOW) (USING 0)

7 / 8

4

3

—

—

TOTAL PAH (LOW) (USING DL)

8 / 8

—

8

—

—

TOTAL PAH (LOW) (USING HALF DL)

8 / 8

—

8

—

—

1,2,4-TRICHLOROBENZENE

0 / 1

—

—

—

—

APP IX PESTICIDES

4,4'-DDD

1 / 1

—

1

—

—

4,4'-DDT

1 / 1

—

—

1

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

144 / 144

3

27

97

17

METALS

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 7

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-70

-------
Table B-70

Sediment Comparison to Benchmark Summary - Low
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

3 / 5

3

—

—

—

CHROMIUM

8 / 9

8

—

—

—

COBALT

0 / 9

—

—

—

—

COPPER

8 / 9

8

—

—

—

LEAD

8 / 9

8

—

—

—

MERCURY

9 / 9

9

—

—

—

NICKEL

4 / 9

4

—

—

—

SELENIUM

NBA

—

—

—

—

SILVER

6 / 7

6

—

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

8 / 9

8

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-70

-------
Table B-71

Sediment Comparison to Benchmark Summary - High*
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

—

DIETHYL PHTHALATE

NBA

—

—

—

—

1,4-DICHLOROBENZENE

NBA

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PHENOL

NBA

—

—

—

—

PAHs











ACENAPHTHYLENE

NBA

—

—

—

—

ANTHRACENE

0 / 4

—

—

—

—

BENZO(A)ANTHRACENE

0 / 7

—

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 7

—

—

—

—

BENZO(GHI)PERYLENE

2 / 7

2

—

—

—

BENZO(A)PYRENE

0 / 7

—

—

—

—

CHRYSENE

0 / 7

—

—

—

—

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

—

FLUORANTHENE

0 / 7

—

—

—

—

FLUORENE

0 / 1

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

2 / 7

2

—

—

—

NAPHTHALENE

0 / 6

—

—

—

—

PHENANTHRENE

0 / 7

—

—

—

—

PYRENE

0 / 8

—

—

—

—

TOTAL PAH (USING 0)

0 / 9

—

—

—

—

TOTAL PAH (USING DL)

0 / 9

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-71

-------
Table B-71

Sediment Comparison to Benchmark Summary - High*
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING HALF DL)

0 / 9

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

4 / 9

4

—

—

—

TOTAL PAH (HIGH) (USING DL)

7 / 9

7

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

7 / 9

7

—

—

—

TOTAL PAH (LOW) (USING 0)

2 / 8

2

—

—

—

TOTAL PAH (LOW) (USING DL)

8 / 8

8

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

8 / 8

8

—

—

—

1,2,4-TRICHLOROBENZENE

NBA

—

—

—

—

APP IX PESTICIDES

4,4'-DDD

1 / 1

1

—

—

—

4,4'-DDT

1 / 1

—

1

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

141 / 144

33

94

14

—

METALS

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 7

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-71

-------
Table B-71

Sediment Comparison to Benchmark Summary - High*
5b Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

0 / 5

—

—

—

—

CHROMIUM

1 / 9

1

—

—

—

COBALT

NBA

—

—

—

—

COPPER

0 / 9

—

—

—

—

LEAD

2 / 9

2

—

—

—

MERCURY

1 / 9

1

—

—

—

NICKEL

0 / 9

—

—

—

—

SELENIUM

NBA

—

—

—

—

SILVER

NBA

—

—

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

0 / 9

—

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-71

-------
Table B-72

Sediment Comparison to Benchmark Summary - Low
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 8

1

—

—

—

DIBENZOFURAN

1 / 2

—

1

—

—

1,2-DICHLOROBENZENE

1 / 2

1

—

—

—

1,3 -DICHLOROBENZENE

0 / 7

—

—

—

—

1,4-DICHLOROBENZENE

5 / 11

5

—

—

—

METHAPYRILENE

NBA

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

2-METHYLPHENOL (O-CRESOL)

1 / 1

1

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PHENOL

1 / 1

1

—

—

—

PAHs

ACENAPHTHENE

1 / 4

—

1

—

—

ACENAPHTHYLENE

1 / 6

—

1

—

—

ANTHRACENE

5 / 7

4

—

1

—

BENZO(A)ANTHRACENE

11 / 13

10

—

1

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

10 / 13

9

1

—

—

BENZO(GHI)PERYLENE

13 / 13

3

9

1

—

BENZO(A)PYRENE

11 / 13

10

—

1

—

CHRYSENE

11 / 13

10

1

—

—

DIBENZO(A,H) ANTHRACENE

9 / 10

8

1

—

—

FLUORANTHENE

7 / 13

6

1

—

—

FLUORENE

1 / 5

—

—

1

—

INDENO( 1,2,3 -C,D)PYRENE

13 / 13

4

8

1

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-72

-------
Table B-72

Sediment Comparison to Benchmark Summary - Low
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

NAPHTHALENE

4 / 11

3

1

—

—

PHENANTHRENE

7 / 11

6

—

1

—

PYRENE

11 / 13

9

1

1

—

TOTAL PAH (USING 0)

11 / 13

10

—

1

—

TOTAL PAH (USING DL)

13 / 13

12

—

1

—

TOTAL PAH (USING HALF DL)

13 / 13

12

—

1

—

TOTAL PAH (HIGH) (USING 0)

13 / 13

3

9

1

—

TOTAL PAH (HIGH) (USING DL)

13 / 13

3

9

1

—

TOTAL PAH (HIGH) (USING HALF DL)

13 / 13

3

9

1

—

TOTAL PAH (LOW) (USING 0)

12 / 13

3

8

—

1

TOTAL PAH (LOW) (USING DL)

13 / 13

—

12

—

1

TOTAL PAH (LOW) (USING HALF DL)

13 / 13

—

12

—

1

1,2,4-TRICHLOROBENZENE

0 / 7

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

218 / 218

16

80

103

19

METALS

N/A = Not Applicable

Tables B-60 - B-87.xls B-72

-------
Table B-72

Sediment Comparison to Benchmark Summary - Low
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ANTIMONY

NBA

—

—

—

—

ARSENIC

1 / 11

1

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

8 / 10

7

1

—

—

CHROMIUM

8 / 13

7

1

—

—

COBALT

0 / 13

—

—

—

—

COPPER

10 / 13

10

—

—

—

LEAD

9 / 13

9

—

—

—

MERCURY

10 / 10

8

2

—

—

NICKEL

7 / 13

7

—

—

—

SELENIUM

NBA

—

—

—

—

SILVER

9 / 10

4

5

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

8 / 13

8

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

AMMONIA AS N

0 / 9

—

—

—

—

CYANIDE

1 / 1

—

1

—

—

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-72

-------
Table B-73

Sediment Comparison to Benchmark Summary - High*
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BI S(2-ETH YLHEX YL) PHTHALATE

NBA

—

—

—

—

DIBENZOFURAN

NBA

—

—

—

—

1,2-DICHLOROBENZENE

NBA

—

—

—

—

1,3 -DICHLOROBENZENE

NBA

—

—

—

—

1,4-DICHLOROBENZENE

NBA

—

—

—

—

METHAPYRILENE

NBA

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

2-METHYLPHENOL (O-CRESOL)

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PHENOL

NBA

—

—

—

—

PAHs

ACENAPHTHENE

NBA

—

—

—

—

ACENAPHTHYLENE

NBA

—

—

—

—

ANTHRACENE

1 / 7

—

1

—

—

BENZO(A)ANTHRACENE

1 / 13

—

1

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 13

—

—

—

—

BENZO(GHI)PERYLENE

4 / 13

3

1

—

—

BENZO(A)PYRENE

1 / 13

—

1

—

—

CHRYSENE

1 / 13

—

1

—

—

DIBENZO(A,H)ANTHRACENE

NBA

—

—

—

—

FLUORANTHENE

1 / 13

—

1

—

—

FLUORENE

1 / 5

—

1

—

—

INDENO(l ,2,3-C,D)PYRENE

5 / 13

4

1

—

—

NAPHTHALENE

1 / 11

—

1

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-73

-------
Table B-73

Sediment Comparison to Benchmark Summary - High*
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PHENANTHRENE

1 / 11

—

1

—

—

PYRENE

2 / 13

1

1

—

—

TOTAL PAH (USING 0)

1 / 13

—

1

—

—

TOTAL PAH (USING DL)

1 / 13

—

1

—

—

TOTAL PAH (USING HALF DL)

1 / 13

—

1

—

—

TOTAL PAH (HIGH) (USING 0)

10 / 13

9

1

—

—

TOTAL PAH (HIGH) (USING DL)

10 / 13

9

1

—

—

TOTAL PAH (HIGH) (USING HALF DL)

10 / 13

9

1

—

—

TOTAL PAH (LOW) (USING 0)

6 / 13

5

—

1

—

TOTAL PAH (LOW) (USING DL)

13 / 13

12

—

1

—

TOTAL PAH (LOW) (USING HALF DL)

12 / 13

11

—

1

—

1,2,4-TRICHLOROBENZENE

NBA

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

201 / 218

92

92

17

—

METALS

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 11

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-73

-------
Table B-73

Sediment Comparison to Benchmark Summary - High*
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

BARIUM

NBA

—

—

...

...

BERYLLIUM

NBA

—

—

...

...

CADMIUM

5 / 10

5

—

...

...

CHROMIUM

7 / 13

7

—

...

...

COBALT

NBA

—

—

...

...

COPPER

7 / 13

7

—

...

...

LEAD

7 / 13

7

...

...

...

MERCURY

6 / 10

6

...

...

...

NICKEL

0 / 13

—

...

...

...

SELENIUM

NBA

—

...

...

...

SILVER

NBA

—

...

...

...

THALLIUM

NBA

—

...

...

...

TIN

NBA

—

...

...

...

VANADIUM

NBA

—

...

...

...

ZINC

3 / 13

3

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

...

...

...

INORGANICS

AMMONIA AS N

NBA

—

...

...

...

CYANIDE

NBA

—

...

...

...

PERCENT SOLIDS

NBA

—

...

...

...

SULFIDE

NBA

—

...

...

...

*SEL based on 1% TOC where applicable. No concentration exceeded the benchmarks @10% TOC.
Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-73

-------
Table B-74

Sediment Comparison to Benchmark Summary - Low
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES



PAHs

CHRYSENE

0 / 1

—

—

—

...

FLUORANTHENE

0 / 1

—

—

—

...

PHENANTHRENE

0 / 1

—

—

—

...

PYRENE

0 / 1

—

—

—

...

TOTAL PAH (USING 0)

0 / 2

—

—

—

...

TOTAL PAH (USING DL)

2 / 2

1

1

—

...

TOTAL PAH (USING HALF DL)

2 / 2

2

—

—

...

TOTAL PAH (HIGH) (USING 0)

0 / 2

—

—

—

...

TOTAL PAH (HIGH) (USING DL)

2 / 2

—

2

—

...

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

—

2

—

...

TOTAL PAH (LOW) (USING 0)

1 / 2

1

—

—

...

TOTAL PAH (LOW) (USING DL)

2 / 2

—

1

1

...

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

—

2

—

...

APP IX PESTICIDES

4,4'-DDD

1 / 1

—

1

—

...

4,4'-DDE

1 / 1

—

1

—

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

38 / 39

4

14

18

2

METALS

NBA = No Benchmark Available	Tables B-60 - B-87.xls B-74

-------
Table B-74

Sediment Comparison to Benchmark Summary - Low
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

1 / 2

1

...

...

...

BARIUM

NBA

—

...

...

...

BERYLLIUM

NBA

—

...

...

...

CHROMIUM

0 / 2

—

...

...

...

COBALT

0 / 2

—

...

...

...

COPPER

1 / 2

1

...

...

...

LEAD

1 / 2

1

...

...

...

MERCURY

1 / 2

1

...

...

...

NICKEL

1 / 2

1

...

...

...

THALLIUM

NBA

—

...

...

...

TIN

NBA

—

...

...

...

VANADIUM

NBA

—

...

...

...

ZINC

1 / 2

1

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

...

...

...

INORGANICS

PERCENT SOLIDS

NBA

—

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-74

-------
Table B-75

Sediment Comparison to Benchmark Summary - High*
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

PAHs

CHRYSENE

0 / 1

—

—

—

...

FLUORANTHENE

0 / 1

—

—

—

...

PHENANTHRENE

0 / 1

—

—

—

...

PYRENE

0 / 1

—

—

—

...

TOTAL PAH (USING 0)

0 / 2

—

—

—

...

TOTAL PAH (USING DL)

1 / 2

1

—

—

...

TOTAL PAH (USING HALF DL)

0 / 2

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

0 / 2

—

—

—

...

TOTAL PAH (HIGH) (USING DL)

2 / 2

2

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

2

—

—

...

TOTAL PAH (LOW) (USING 0)

0 / 2

—

—

—

...

TOTAL PAH (LOW) (USING DL)

2 / 2

2

—

—

...

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

2

—

—

...

APP IX PESTICIDES

4,4'-DDD

1 / 1

1

—

—

...

4,4'-DDE

1 / 1

1

—

—

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

32 / 39

14

16

2

...

METALS

NBA = No Benchmark Available	Tables B-60 - B-87.xls B-75

-------
Table B-75

Sediment Comparison to Benchmark Summary - High*
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

0 / 2

...

...

...

...

BARIUM

NBA

...

...

...

...

BERYLLIUM

NBA

...

...

...

...

CHROMIUM

0 / 2

...

...

...

...

COBALT

NBA

...

...

...

...

COPPER

0 / 2

...

...

...

...

LEAD

0 / 2

...

...

...

...

MERCURY

0 / 2

...

...

...

...

NICKEL

0 / 2

...

...

...

...

THALLIUM

NBA

...

...

...

...

TIN

NBA

...

...

...

...

VANADIUM

NBA

...

...

...

...

ZINC

0 / 2

...

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

...

...

...

...

INORGANICS

PERCENT SOLIDS

NBA

...

...

...

...

*SEL based on 1% TOC where applicable. No concentration exceeded the benchmarks @10% TOC.
Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-75

-------
Table B-76

Sediment Comparison to Benchmark Summary - Low
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 5

—

—

—

—

DIBENZOFURAN

0 / 2

—

—

—

—

1,4-DICHLOROBENZENE

0 / 10

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PAHs

ACENAPHTHYLENE

3 / 3

3

—

—

—

ACETOPHENONE

NBA

—

—

—

—

ANTHRACENE

2 / 3

1

1

—

—

BENZO(A)ANTHRACENE

10 / 12

9

1

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

3 / 11

2

1

—

—

BENZO(GHI)PERYLENE

11 / 11

3

8

—

—

BENZO(A)PYRENE

5 / 8

4

1

—

—

CHRYSENE

10 / 12

9

1

—

—

DIBENZO(A,H) ANTHRACENE

2 / 2

1

1

—

—

FLUORANTHENE

2 / 11

2

—

—

—

FLUORENE

1 / 2

1

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

11 / 11

8

3

—

—

NAPHTHALENE

1 / 9

1

—

—

—

PHENANTHRENE

3 / 12

3

—

—

—

PYRENE

11 / 13

11

—

—

—

TOTAL PAH (USING 0)

8 / 17

7

1

—

—

TOTAL PAH (USING DL)

17 / 17

13

4

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-76

-------
Table B-76

Sediment Comparison to Benchmark Summary - Low
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING HALF DL)

17 / 17

15

2

—

—

TOTAL PAH (HIGH) (USING 0)

13 / 17

9

3

1

—

TOTAL PAH (HIGH) (USING DL)

17 / 17

2

12

3

—

TOTAL PAH (HIGH) (USING HALF DL)

17 / 17

3

13

1

—

TOTAL PAH (LOW) (USING 0)

13 / 17

10

3

—

—

TOTAL PAH (LOW) (USING DL)

17 / 17

1

10

6

—

TOTAL PAH (LOW) (USING HALF DL)

17 / 17

1

15

1

—

1,2,4-TRICHLOROBENZENE

0 / 2

—

—

—

—

APP IX PESTICIDES

4,4'-DDT

1 / 1

—

1

—

—

HERBICIDES

2,4,5-T

NBA

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

176 / 176

8

61

92

15

METALS

ANTIMONY

NBA

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-76

-------
Table B-76

Sediment Comparison to Benchmark Summary - Low
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

3 / 13

3

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

10 / 14

10

—

—

—

CHROMIUM

12 / 16

12

—

—

—

COBALT

0 / 16

—

—

—

—

COPPER

13 / 17

13

—

—

—

LEAD

14 / 17

14

—

—

—

MERCURY

13 / 15

13

—

—

—

NICKEL

10 / 17

10

—

—

—

SELENIUM

NBA

—

—

—

—

SILVER

10 / 10

8

2

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

13 / 17

13

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-76

-------
Table B-77

Sediment Comparison to Benchmark Summary - High*
5c Vernal Pool
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

—

DIBENZOFURAN

NBA

—

—

—

—

1,4-DICHLOROBENZENE

NBA

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PAHs

ACENAPHTHYLENE

NBA

—

—

—

—

ACETOPHENONE

NBA

—

—

—

—

ANTHRACENE

0 / 3

—

—

—

—

BENZO(A)ANTHRACENE

1 / 12

1

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 11

—

—

—

—

BENZO(GHI)PERYLENE

2 / 11

2

—

—

—

BENZO(A)PYRENE

1 / 8

1

—

—

—

CHRYSENE

1 / 12

1

—

—

—

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

—

FLUORANTHENE

0 / 11

—

—

—

—

FLUORENE

0 / 2

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

2 / 11

2

—

—

—

NAPHTHALENE

0 / 9

—

—

—

—

PHENANTHRENE

0 / 12

—

—

—

—

PYRENE

0 / 13

—

—

—

—

TOTAL PAH (USING 0)

1 / 17

1

—

—

—

TOTAL PAH (USING DL)

3 / 17

3

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-77

-------
Table B-77

Sediment Comparison to Benchmark Summary - High*
5c Vernal Pool
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING HALF DL)

1 / 17

1

—

—

—

TOTAL PAH (HIGH) (USING 0)

4 / 17

3

1

—

—

TOTAL PAH (HIGH) (USING DL)

15 / 17

12

3

—

—

TOTAL PAH (HIGH) (USING HALF DL)

14 / 17

13

1

—

—

TOTAL PAH (LOW) (USING 0)

3 / 17

3

—

—

—

TOTAL PAH (LOW) (USING DL)

16 / 17

14

2

—

—

TOTAL PAH (LOW) (USING HALF DL)

16 / 17

16

—

—

—

1,2,4-TRICHLOROBENZENE

NBA

—

—

—

—

APP IX PESTICIDES

4,4'-DDT

0 / 1

—

—

—

—

HERBICIDES

2,4,5-T

NBA

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

166 / 176

65

87

14

—

METALS

ANTIMONY

NBA

—

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-77

-------
Table B-77

Sediment Comparison to Benchmark Summary - High*
5c Vernal Pool
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

0 / 13

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

3 / 14

3

—

—

—

CHROMIUM

8 / 16

8

—

—

—

COBALT

NBA

—

—

—

—

COPPER

7 / 17

7

—

—

—

LEAD

7 / 17

7

—

—

—

MERCURY

8 / 15

8

—

—

—

NICKEL

1 / 17

1

—

—

—

SELENIUM

NBA

—

—

—

—

SILVER

NBA

—

—

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

3 / 17

3

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentration exceeded the benchmarks @10% TOC.
Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-77

-------
Table B-78

Sediment Comparison to Benchmark Summary - Low
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 2

—

—

—

—

1,3 -DICHLOROBENZENE

0 / 1

—

—

—

—

1,4-DICHLOROBENZENE

1 / 2

1

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

PAHs

ANTHRACENE

0 / 1

—

—

—

—

BENZO(A)ANTHRACENE

2 / 2

2

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

1 / 2

1

—

—

—

BENZO(GHI)PERYLENE

2 / 2

1

1

—

—

BENZO(A)PYRENE

2 / 2

2

—

—

—

CHRYSENE

2 / 2

2

—

—

—

DIBENZO(A,H) ANTHRACENE

1 / 1

1

—

—

—

FLUORANTHENE

1 / 2

1

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

2 / 2

1

1

—

—

NAPHTHALENE

0 / 2

—

—

—

—

PHENANTHRENE

1 / 2

1

—

—

—

PYRENE

2 / 2

2

—

—

—

TOTAL PAH (USING 0)

2 / 2

2

—

—

—

TOTAL PAH (USING DL)

2 / 2

2

—

—

—

TOTAL PAH (USING HALF DL)

2 / 2

2

—

—

—

TOTAL PAH (HIGH) (USING 0)

2 / 2

1

1

—

—

TOTAL PAH (HIGH) (USING DL)

2 / 2

1

1

—

—

TOTAL PAH (HIGH) (USING HALF DL)

2 / 2

1

1

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-78

-------
Table B-78

Sediment Comparison to Benchmark Summary - Low
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (LOW) (USING 0)

2 / 2

1

1

—

—

TOTAL PAH (LOW) (USING DL)

2 / 2

—

2

—

—

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

—

2

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

66 / 66

1

19

37

9

METALS

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 2

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

1 / 2

1

—

—

—

CHROMIUM

2 / 2

2

—

—

—

COBALT

0 / 2

—

—

—

—

COPPER

2 / 2

2

—

—

—

LEAD

2 / 2

2

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-78

-------
Table B-78

Sediment Comparison to Benchmark Summary - Low
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

MERCURY

2 / 2

2

—

—

—

NICKEL

1 / 2

1

—

—

—

SILVER

2 / 2

1

1

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

2 / 2

2

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-78

-------
Table B-79

Sediment Comparison to Benchmark Summary - High*
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES











BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

—

1,3 -DICHLOROBENZENE

NBA

—

—

—

—

1,4-DICHLOROBENZENE

NBA

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

PAHs

ANTHRACENE

0 / 1

—

—

—

—

BENZO(A)ANTHRACENE

0 / 2

—

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 2

—

—

—

—

BENZO(GHI)PERYLENE

0 / 2

—

—

—

—

BENZO(A)PYRENE

0 / 2

—

—

—

—

CHRYSENE

0 / 2

—

—

—

—

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

—

FLUORANTHENE

0 / 2

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

0 / 2

—

—

—

—

NAPHTHALENE

0 / 2

—

—

—

—

PHENANTHRENE

0 / 2

—

—

—

—

PYRENE

0 / 2

—

—

—

—

TOTAL PAH (USING 0)

0 / 2

—

—

—

—

TOTAL PAH (USING DL)

0 / 2

—

—

—

—

TOTAL PAH (USING HALF DL)

0 / 2

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

1 / 2

1

—

—

—

TOTAL PAH (HIGH) (USING DL)

1 / 2

1

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

1 / 2

1

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-79

-------
Table B-79

Sediment Comparison to Benchmark Summary - High*
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (LOW) (USING 0)

0 / 2

—

—

—

—

TOTAL PAH (LOW) (USING DL)

2 / 2

2

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

2 / 2

2

—

—

—

DIOXINS/FURANS











TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS











AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

64 / 66

20

35

9

—

METALS











ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 2

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

1 / 2

1

—

—

—

CHROMIUM

1 / 2

1

—

—

—

COBALT

NBA

—

—

—

—

COPPER

1 / 2

1

—

—

—

LEAD

1 / 2

1

—

—

—

N/A = Not Applicable

Tables B-60 - B-87.xls B-79

-------
Table B-79

Sediment Comparison to Benchmark Summary - High*
6ab Main Channel
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

MERCURY

1 / 2

1

—

—

—

NICKEL

0 / 2

—

—

—

—

SILVER

NBA

—

—

—

—

THALLIUM

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

0 / 2

—

—

—

—

ORGANIC











TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS











PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-79

-------
Table B-80

Sediment Comparison to Benchmark Summary - Low
6ab SCOX

Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCBS

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

6 / 6

—

—

4

2

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60 - B-87.xls B-80

-------
Table B-81

Sediment Comparison to Benchmark Summary - High*
6ab SCOX

Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCBS

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

6 / 6

—

4

2

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

N/A = Not Applicable

Tables B-60-B-87.xls B-81

-------
Table B-82

Sediment Comparison to Benchmark Summary - Low
6ab Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

38 / 38

1

13

18

6

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-60 - B-87.xlsB-82

-------
Table B-83

Sediment Comparison to Benchmark Summary - High*
6ab Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

37 / 38

14

17

6

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

TOTAL WEIGHT

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-60 - B-87.xlsB-83

-------
Table B-84

Sediment Comparison to Benchmark Summary - Low
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 1

—

—

—

—

1,4-DICHLOROBENZENE

0 / 1

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PAHs

ACENAPHTHYLENE

0 / 1

—

—

—

—

ANTHRACENE

1 / 1

1

—

—

—

BENZO(A)ANTHRACENE

1 / 1

1

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

1 / 1

1

—

—

—

BENZO(GHI)PERYLENE

1 / 1

—

1

—

—

BENZO(A)PYRENE

1 / 1

1

—

—

—

CHRYSENE

1 / 1

1

—

—

—

DIBENZO(A,H) ANTHRACENE

1 / 1

1

—

—

—

FLUORANTHENE

1 / 1

1

—

—

—

FLUORENE

0 / 1

—

—

—

—

INDENO(l,2,3-C,D)PYRENE

1 / 1

—

1

—

—

NAPHTHALENE

0 / 1

—

—

—

—

PHENANTHRENE

1 / 1

1

—

—

—

PYRENE

1 / 1

1

—

—

—

TOTAL PAH (USING 0)

1 / 1

1

—

—

—

TOTAL PAH (USING DL)

1 / 1

1

—

—

—

TOTAL PAH (USING HALF DL)

1 / 1

1

—

—

—

TOTAL PAH (HIGH) (USING 0)

1 / 1

—

1

—

—

NBA = No benchmark available.

Tables B-60 - B-87.xls B-84

-------
Table B-84

Sediment Comparison to Benchmark Summary - Low
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (HIGH) (USING DL)

1 / 1

—

1

—

—

TOTAL PAH (HIGH) (USING HALF DL)

1 / 1

—

1

—

—

TOTAL PAH (LOW) (USING 0)

1 / 1

—

1

—

—

TOTAL PAH (LOW) (USING DL)

1 / 1

—

1

—

—

TOTAL PAH (LOW) (USING HALF DL)

1 / 1

—

1

—

—

1,2,4-TRICHLOROBENZENE

0 / 1

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

102 / 102

5

22

55

20

METALS

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 1

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

1 / 1

1

—

—

—

CHROMIUM

1 / 1

1

—

—

—

NBA = No benchmark available.

Tables B-60 - B-87.xls B-84

-------
Table B-84

Sediment Comparison to Benchmark Summary - Low
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

COBALT

0 / 1

—

—

—

—

COPPER

1 / 1

1

—

—

—

LEAD

1 / 1

1

—

—

—

MERCURY

1 / 1

1

—

—

—

NICKEL

1 / 1

1

—

—

—

SELENIUM

NBA

—

—

—

—

SILVER

1 / 1

1

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

1 / 1

1

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

PERCENT WATER CONTENT

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-60 - B-87.xls B-84

-------
Table B-85

Sediment Comparison to Benchmark Summary - High*
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

—

1,4-DICHLOROBENZENE

NBA

—

—

—

—

2-METHYLNAPHTHALENE

NBA

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PAHs

ACENAPHTHYLENE

NBA

—

—

—

—

ANTHRACENE

0 / 1

—

—

—

—

BENZO(A)ANTHRACENE

0 / 1

—

—

—

—

BENZO(B)FLUORANTHENE

NBA

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 1

—

—

—

—

BENZO(GHI)PERYLENE

1 / 1

1

—

—

—

BENZO(A)PYRENE

0 / 1

—

—

—

—

CHRYSENE

0 / 1

—

—

—

—

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

—

FLUORANTHENE

0 / 1

—

—

—

—

FLUORENE

0 / 1

—

—

—

—

INDENO(l,2,3-C,D)PYRENE

1 / 1

1

—

—

—

NAPHTHALENE

0 / 1

—

—

—

—

PHENANTHRENE

0 / 1

—

—

—

—

PYRENE

0 / 1

—

—

—

—

TOTAL PAH (USING 0)

0 / 1

—

—

—

—

TOTAL PAH (USING DL)

0 / 1

—

—

—

—

TOTAL PAH (USING HALF DL)

0 / 1

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

1 / 1

1

—

—

—

NBA = No benchmark available.

Tables B-60 - B-87.xls B-85

-------
Table B-85

Sediment Comparison to Benchmark Summary - High*
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (HIGH) (USING DL)

1 / 1

1

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

1 / 1

1

—

—

—

TOTAL PAH (LOW) (USING 0)

0 / 1

—

—

—

—

TOTAL PAH (LOW) (USING DL)

1 / 1

1

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

1 / 1

1

—

—

—

1,2,4-TRICHLOROBENZENE

NBA

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

97 / 102

26

52

19

—

METALS

ANTIMONY

NBA

—

—

—

—

ARSENIC

0 / 1

—

—

—

—

BARIUM

NBA

—

—

—

—

BERYLLIUM

NBA

—

—

—

—

CADMIUM

0 / 1

—

—

—

—

CHROMIUM

1 / 1

1

—

—

—

NBA = No benchmark available.

Tables B-60 - B-87.xls B-85

-------
Table B-85

Sediment Comparison to Benchmark Summary - High*
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

COBALT

NBA

—

—

—

—

COPPER

1 / 1

1

—

—

—

LEAD

1 / 1

1

—

—

—

MERCURY

1 / 1

1

—

—

—

NICKEL

0 / 1

—

—

—

—

SELENIUM

NBA

—

—

—

—

SILVER

NBA

—

—

—

—

TIN

NBA

—

—

—

—

VANADIUM

NBA

—

—

—

—

ZINC

0 / 1

—

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

PERCENT WATER CONTENT

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

NBA = No benchmark available.	Tables B-60 - B-87.xls B-85

-------
Table B-86

Sediment Comparison to Benchmark Summary - Low
Background
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALPHA-BHC

1 / 1

1

—

—

...

BETA-BHC

1 / 2

1

...

...

...

4,4'-DDD

3 / 3

2

1

—

...

4,4'-DDE

2 / 2

—

2

—

...

APP IX SEMIVOLATILES

ACETOPHENONE

NBA

...

...

...

...

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 9

—

—

—

...

CHRYSENE

6 / 12

6

—

—

...

DIBENZOFURAN

0 / 1

—

—

—

...

1,4-DICHLOROBENZENE

0 / 1

—

—

—

...

2-METHYLNAPHTHALENE

NBA

—

—

—

...

4-METHYLPHENOL

NBA

—

—

—

...

N-NITROSO-DI-N-BUTYL AMINE

NBA

—

—

—

...

PAHs

ACENAPHTHENE

0 / 1

—

—

—

...

ANTHRACENE

2 / 8

2

—

—

...

BENZO(A)ANTHRACENE

8 / 11

8

—

—

...

BENZO(B)FLUORANTHENE

NBA

—

—

—

...

BENZO(K)FLUORANTHENE

3 / 11

3

—

—

...

BENZO(GHI)PERYLENE

11 / 11

8

3

...

...

BENZO(A)PYRENE

6 / 11

6

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-86

-------
Table B-86

Sediment Comparison to Benchmark Summary - Low
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIBENZO(A,H) ANTHRACENE

3 / 8

3

—

—

—

DI-N-BUTYL PHTHALATE

0 / 1

...

...

...

...

FLUORANTHENE

2 / 14

2

—

—

...

FLUORENE

0 / 6

...

...

—

...

INDENO( 1,2,3 -C,D)PYRENE

11 / 11

8

3

—

...

NAPHTHALENE

0 / 5

—

—

—

...

PHENANTHRENE

6 / 13

6

—

—

...

PYRENE

9 / 13

9

...

—

...

DIOXINS/FURANS

HPCDD (TOTAL)

NBA

—

—

—

...

HPCDF (TOTAL)

NBA

—

—

—

...

HXCDD (TOTAL)

NBA

—

—

—

...

HXCDF (TOTAL)

NBA

—

—

—

...

OCDD

NBA

—

—

—

...

OCDF

NBA

—

—

—

...

PECDD (TOTAL)

NBA

—

—

—

...

PECDF (TOTAL)

NBA

—

—

—

...

TCDD (TOTAL)

NBA

—

—

—

...

TCDF (TOTAL)

NBA

—

—

—

...

HERBICIDES

2,4,5-T

NBA

—

—

—

...

PCBS

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-86

-------
Table B-86

Sediment Comparison to Benchmark Summary - Low
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

...

...

...

...

PCB, TOTAL

2 / 4

1

1

...

...

METALS

ANTIMONY

NBA

—

—

—

...

ARSENIC

0 / 15

—

—

—

...

BARIUM

NBA

—

—

—

...

BERYLLIUM

NBA

—

—

—

...

CADMIUM

3 / 6

3

—

—

...

CHROMIUM

2 / 20

2

—

—

...

COBALT

0 / 20

...

...

...

...

COPPER

4 / 20

4

—

—

...

LEAD

8 / 20

8

...

...

...

MERCURY

1 / 9

1

—

—

...

NICKEL

0 / 20

...

...

...

...

SELENIUM

NBA

—

—

—

...

SILVER

4 / 5

4

...

...

...

THALLIUM

NBA

—

—

—

...

TIN

NBA

—

—

—

...

VANADIUM

NBA

—

—

—

...

ZINC

3 / 20

3

...

...

...

ORGANIC

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-86

-------
Table B-86

Sediment Comparison to Benchmark Summary - Low
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

...

SULFIDE

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-86

-------
Table B-87

Sediment Comparison to Benchmark Summary - High*
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALPHA-BHC

0 / 1

—

—

—

...

BETA-BHC

0 / 2

...

...

...

...

4,4'-DDD

3 / 3

2

1

—

...

4,4'-DDE

2 / 2

2

—

—

...

APP IX SEMIVOLATILES

ACETOPHENONE

NBA

...

...

...

...

BIS(2-ETHYLHEXYL) PHTHALATE

NBA

—

—

—

...

CHRYSENE

0 / 12

—

—

—

...

DIBENZOFURAN

NBA

—

—

—

...

1,4-DICHLOROBENZENE

NBA

—

—

—

...

2-METHYLNAPHTHALENE

NBA

—

—

—

...

4-METHYLPHENOL

NBA

—

—

—

...

N-NITROSO-DI-N-BUTYL AMINE

NBA

—

—

—

...

PAHs

ACENAPHTHENE

NBA

—

—

—

...

ANTHRACENE

0 / 8

—

—

—

...

BENZO(A)ANTHRACENE

0 / 11

—

—

—

...

BENZO(B)FLUORANTHENE

NBA

—

—

—

...

BENZO(K)FLUORANTHENE

0 / 11

—

—

—

...

BENZO(GHI)PERYLENE

0 / 11

—

—

—

...

BENZO(A)PYRENE

0 / 11

...

...

...

...

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-87

-------
Table B-87

Sediment Comparison to Benchmark Summary - High*
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIBENZO(A,H) ANTHRACENE

NBA

—

—

—

—

DI-N-BUTYL PHTHALATE

NBA

...

...

...

...

FLUORANTHENE

0 / 14

—

—

—

...

FLUORENE

0 / 6

...

...

—

...

INDENO( 1,2,3 -C,D)PYRENE

0 / 11

—

—

—

...

NAPHTHALENE

0 / 5

—

—

—

...

PHENANTHRENE

0 / 13

—

—

—

...

PYRENE

0 / 13

...

...

—

...

DIOXINS/FURANS

HPCDD (TOTAL)

NBA

—

—

—

...

HPCDF (TOTAL)

NBA

—

—

—

...

HXCDD (TOTAL)

NBA

—

—

—

...

HXCDF (TOTAL)

NBA

—

—

—

...

OCDD

NBA

—

—

—

...

OCDF

NBA

—

—

—

...

PECDD (TOTAL)

NBA

—

—

—

...

PECDF (TOTAL)

NBA

—

—

—

...

TCDD (TOTAL)

NBA

—

—

—

...

TCDF (TOTAL)

NBA

—

—

—

...

HERBICIDES

2,4,5-T

NBA

—

—

—

...

PCBS

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-87

-------
Table B-87

Sediment Comparison to Benchmark Summary - High*
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

...

...

...

...

PCB, TOTAL

1 / 4

1

...

...

...

METALS

ANTIMONY

NBA

—

—

—

...

ARSENIC

0 / 15

—

—

—

...

BARIUM

NBA

—

—

—

...

BERYLLIUM

NBA

—

—

—

...

CADMIUM

0 / 6

—

—

—

...

CHROMIUM

1 / 20

1

—

—

...

COBALT

NBA

—

—

—

...

COPPER

0 / 20

—

—

—

...

LEAD

2 / 20

2

...

...

...

MERCURY

0 / 9

—

—

—

...

NICKEL

0 / 20

...

...

...

...

SELENIUM

NBA

—

—

—

...

SILVER

NBA

—

—

—

...

THALLIUM

NBA

—

—

—

...

TIN

NBA

—

—

—

...

VANADIUM

NBA

—

—

—

...

ZINC

0 / 20

...

...

...

...

ORGANIC

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-87

-------
Table B-87

Sediment Comparison to Benchmark Summary - High*
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

...

SULFIDE

NBA

...

...

...

...

*SEL based on 1% TOC where applicable. No concentrations exceeded the benchmarks @ 10% TOC.
Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-60 - B-87.xls B-87

-------
Table B-88

Surface Water Comparison to Benchmark Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX VOLATILES

ACETONE

0 / 4

—

...

...

...

CHLOROBENZENE

0 / 6

...

...

...

...

TRICHLOROETHYLENE (TCE)

0 / 3

...

...

...

...

VINYL CHLORIDE

NBA

—

—

—

...

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

2 / 2

2

...

...

...

1,4-DICHLOROBENZENE

0 / 3

—

—

—

...

DIETHYL PHTHALATE

0 / 3

—

—

—

...

PAHs











ACENAPHTHENE

0 / 4

...

...

...

...

FLUORANTHENE

0 / 4

—

—

—

...

FLUORENE

0 / 2

...

...

—

...

NAPHTHALENE

0 / 4

—

—

—

...

PHENANTHRENE

0 / 4

—

—

—

...

PYRENE

5 / 5

4

1

...

...

TOTAL PAH (USING 0)

NBA

—

—

—

...

TOTAL PAH (USING DL)

NBA

—

—

—

...

TOTAL PAH (USING HALF DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-88

-------
Table B-88

Surface Water Comparison to Benchmark Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING DL)

NBA

...

...

...

...

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

...

1.2.4-TRICHLOROBENZENE, FILTERED

0 / 25

...

...

...

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1242

NBA

—

—

—

...

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

25 / 26

22

3

...

...

PCBS - FILTERED

TOTAL PCB, DISSOLVED

2 / 2

—

2

—

...

METALS

ARSENIC

0 / 2

—

—

—

...

BARIUM

39 / 39

39

—

—

...

CADMIUM

2 / 2

2

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-88

-------
Table B-88

Surface Water Comparison to Benchmark Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

CALCIUM

NBA

—

—

—

—

CHROMIUM

0 / 3

...

...

...

...

COPPER

1 / 7

1

—

—

...

LEAD

3 / 4

3

...

...

...

MAGNESIUM

NBA

—

—

—

...

MERCURY

0 / 5

—

—

—

...

SELENIUM

0 / 2

...

...

—

...

VANADIUM

0 / 3

—

—

—

...

ZINC

2 / 14

2

...

...

...

METALS - FILTERED

BARIUM, DISSOLVED

NBA

—

—

—

...

BERYLLIUM, DISSOLVED

NBA

—

—

—

...

CALCIUM, DISSOLVED

NBA

—

—

—

...

COBALT, DISSOLVED

NBA

—

—

—

...

COPPER, DISSOLVED

0 / 3

—

—

—

...

MAGNESIUM, DISSOLVED

NBA

—

—

—

...

MERCURY, FILTERED

0 / 4

...

...

...

...

SILVER, DISSOLVED

NBA

—

—

—

...

ZINC, DISSOLVED

0 / 8

—

—

—

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

TOTAL ORGANIC CARBON, DISSOLVED

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-88

-------
Table B-88

Surface Water Comparison to Benchmark Summary
5a Main Channel and Aggrading Bars
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

INORGANICS

ALKALINITY

NBA

...

...

...

...

AMMONIA AS N

37 / 37

—

30

7

...

BIOLOGICAL OXYGEN DEMAND 5 DAY

NBA

...

...

...

...

HARDNESS

NBA

—

—

—

...

HARDNESS, DISSOLVED

NBA

—

—

—

...

NITRATE AND NITRITE AS N

NBA

—

—

—

...

NITRITE AS N

NBA

—

—

—

...

ORGANIC CARBON, PARTICULATE

NBA

—

—

—

...

ORTHOPHOSPHATE AS P

NBA

—

—

—

...

PHOSPHATE, AS P

NBA

—

—

—

...

PHOSPHATE, TOTAL AS P

NBA

—

—

—

...

SULFIDE

7 / 7

—

—

7

...

TKN

NBA

—

—

—

...

TOTAL DISSOLVED SOLIDS

NBA

—

—

—

...

TOTAL SUSPENDED SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-88

-------
Table B-89

Surface Water Comparison to Benchmark Summary
5a SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

TOTAL ORGANIC CARBON, DISSOLVED

NBA

...

...

...

...

INORGANICS

AMMONIA AS N

21 / 21

2

19

...

...

BIOLOGICAL OXYGEN DEMAND 5 DAY

NBA

—

—

—

...

HYDROLYZABLE PHOSPHATE, AS P

NBA

...

...

...

...

NITRATE AND NITRITE AS N

NBA

—

—

—

...

NITRITE AS N

NBA

—

—

—

...

ORGANIC CARBON, PARTICULATE

NBA

—

—

—

...

ORGANIC PHOSPHATE, AS P

NBA

—

—

—

...

ORTHOPHOSPHATE AS P

NBA

—

—

—

...

PHOSPHATE, TOTAL AS P

NBA

—

—

—

...

TKN

NBA

—

—

—

...

TOTAL SUSPENDED SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available	Tables B-88 - B-97.xls B-89

-------
Table B-90

Surface Water Comparison to Benchmark Summary
5a Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

...

...

...

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

5 / 5

4

1

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-90

-------
Table B-91

Surface Water Comparison to Benchmark Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX VOLATILES

ACETONE

0 / 2

—

...

...

...

BROMODICHLOROMETHANE

0 / 4

...

...

...

...

CHLOROBENZENE

0 / 2

...

...

...

...

CHLOROFORM

0 / 5

—

—

—

...

DIBROMOCHLOROMETHANE

0 / 2

—

—

—

...

APP IX SEMIVOLATILES

PAHs











ACENAPHTHENE

0 / 3

—

—

—

...

BENZO(B)FLUORANTHENE

0 / 2

—

—

—

...

BENZO(GHI)PERYLENE

NBA

—

—

—

...

CHRYSENE

0 / 2

—

—

—

...

FLUORANTHENE

2 / 5

—

2

—

...

NAPHTHALENE

0 / 3

—

—

—

...

PHENANTHRENE

0 / 3

—

—

—

...

PYRENE

4 / 5

2

2

...

...

TOTAL PAH (USING 0)

NBA

—

—

—

...

TOTAL PAH (USING DL)

NBA

—

—

—

...

TOTAL PAH (USING HALF DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-91

-------
Table B-91

Surface Water Comparison to Benchmark Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING DL)

NBA

...

...

...

...

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1242

NBA

—

—

—

...

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

53 / 53

33

20

...

...

PCBS - FILTERED

AROCLOR-1260, DISSOLVED

NBA

—

—

—

...

TOTAL PCB, DISSOLVED

6 / 6

4

2

—

...

1.2.4-TRICHLOROBENZENE, FILTERED

0 / 24

...

...

...

...

METALS

BARIUM

29 / 29

29

—

—

...

CALCIUM

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-91

-------
Table B-91

Surface Water Comparison to Benchmark Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

CHROMIUM

0 / 4

—

—

—

—

COBALT

0 / 2

...

...

...

...

COPPER

9 / 17

9

—

—

...

LEAD

3 / 4

2

1

...

...

MAGNESIUM

NBA

—

—

—

...

MERCURY

0 / 3

—

—

—

...

SILVER

2 / 2

2

...

...

...

VANADIUM

0 / 2

—

—

—

...

ZINC

2 / 11

2

...

...

...

METALS - FILTERED

BARIUM, DISSOLVED

NBA

—

—

—

...

BERYLLIUM, DISSOLVED

NBA

—

—

—

...

CALCIUM, DISSOLVED

NBA

—

—

—

...

CHROMIUM, DISSOLVED

0 / 3

—

—

—

...

COPPER, DISSOLVED

6 / 13

6

—

—

...

MAGNESIUM, DISSOLVED

NBA

—

—

—

...

MERCURY, FILTERED

0 / 3

...

...

...

...

ZINC, DISSOLVED

0 / 8

—

—

—

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

TOTAL ORGANIC CARBON, DISSOLVED

NBA

—

—

—

...

INORGANICS

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-91

-------
Table B-91

Surface Water Comparison to Benchmark Summary
5b Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ALKALINITY

NBA

—

—

—

—

AMMONIA AS N

60 / 60

...

56

4

...

BIOLOGICAL OXYGEN DEMAND 5 DAY

NBA

—

—

—

...

CHEMICAL OXYGEN DEMAND

NBA

—

—

—

...

HARDNESS

NBA

—

—

—

...

HYDROLYZABLE PHOSPHATE, AS P

NBA

—

—

—

...

NITRATE AND NITRITE AS N

NBA

—

—

—

...

NITRITE AS N

NBA

—

—

—

...

ORGANIC CARBON, PARTICULATE

NBA

—

—

—

...

ORGANIC PHOSPHATE, AS P

NBA

—

—

—

...

ORTHOPHOSPHATE AS P

NBA

—

—

—

...

PHOSPHATE, AS P

NBA

—

—

—

...

PHOSPHATE, TOTAL AS P

NBA

—

—

—

...

SULFIDE

6 / 6

—

—

6

...

TKN

NBA

—

—

—

...

TOTAL DISSOLVED SOLIDS

NBA

—

—

—

...

TOTAL SUSPENDED SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-91

-------
Table B-92

Surface Water Comparison to Benchmark Summary
5b Vernal Pool
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIOXINS/FURANS

1,2,3,4,6,7,8-HPCDD

NBA

—

—

—

—

1,2,3,4,6,7,8-HPCDF

NBA

—

—

—

—

1,2,3,4,7,8-HXCDF

NBA

—

—

—

—

1,2,3,7,8-PECDF

NBA

—

—

—

—

2,3,4,6,7,8-HXCDF

NBA

—

—

—

—

2,3,7,8-TCDF

NBA

—

—

—

—

HPCDD (TOTAL)

NBA

—

—

—

—

HPCDF (TOTAL)

NBA

—

—

—

—

HXCDF (TOTAL)

NBA

—

—

—

—

OCDD

NBA

—

—

—

—

OCDF

NBA

—

—

—

—

PECDF (TOTAL)

NBA

—

—

—

—

TCDF (TOTAL)

NBA

—

—

—

—

TEQ 2,3,7,8-TCDD (EPA)

NBA

—

—

—

—

TEQ 2,3,7,8-TCDD (MADEP)

NBA

—

—

—

—

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-88 - B-97.xls B-92

-------
Table B-92

Surface Water Comparison to Benchmark Summary
5b Vernal Pool
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCB, TOTAL

8 / 8

3

5

—

—

METALS

BARIUM

1 / 1

1

—

—

—

CALCIUM

NBA

—

—

—

—

COPPER

0 / 1

—

—

—

—

LEAD

1 / 1

1

—

—

—

MAGNESIUM

NBA

—

—

—

—

VANADIUM

0 / 1

—

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

SULFIDE

1 / 1

—

—

1

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-88 - B-97.xls B-92

-------
Table B-93

Surface Water Comparison to Benchmark Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX VOLATILES

ACETONE

0 / 1

—

...

...

...

BROMODICHLOROMETHANE

0 / 3

...

...

...

...

CHLOROBENZENE

0 / 1

...

...

...

...

CHLOROFORM

0 / 3

—

—

—

...

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 3

...

1

...

...

DIETHYL PHTHALATE

0 / 1

—

—

—

...

PAHs











ACENAPHTHENE

0 / 2

—

—

—

...

BENZO(A)ANTHRACENE

0 / 1

...

...

...

...

BENZO(B)FLUORANTHENE

0 / 2

—

—

—

...

BENZO(GHI)PERYLENE

NBA

—

—

—

...

BENZO(A)PYRENE

0 / 2

—

—

—

...

CHRYSENE

0 / 2

—

—

—

...

FLUORANTHENE

0 / 2

—

—

—

...

INDENO(l,2,3-C,D)PYRENE

NBA

—

—

—

...

NAPHTHALENE

0 / 1

—

—

—

...

PHENANTHRENE

0 / 2

—

—

—

...

PYRENE

2 / 2

2

...

...

...

TOTAL PAH (USING 0)

NBA

—

—

—

...

TOTAL PAH (USING DL)

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-93

-------
Table B-93

Surface Water Comparison to Benchmark Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

NBA

...

...

...

...

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING DL)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

...

1.2.4-TRICHLOROBENZENE, FILTERED

0 / 10

...

...

...

...

APP IX PESTICIDES

DELTA-BHC

0 / 1

...

...

...

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1242

NBA

—

—

—

...

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

12 / 12

10

2

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-93

-------
Table B-93

Surface Water Comparison to Benchmark Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

METALS

ARSENIC

0 / 1

...

...

...

...

BARIUM

15 / 15

15

—

—

...

CALCIUM

NBA

...

...

...

...

CHROMIUM

0 / 2

—

—

—

...

COPPER

2 / 12

2

—

—

...

MAGNESIUM

NBA

—

—

—

...

MERCURY

0 / 3

—

—

—

...

TIN

0 / 1

...

...

...

...

VANADIUM

0 / 2

—

—

—

...

ZINC

0 / 5

...

...

...

...

METALS - FILTERED

BARIUM, DISSOLVED

NBA

—

—

—

...

BERYLLIUM, DISSOLVED

NBA

—

—

—

...

CALCIUM, DISSOLVED

NBA

—

—

—

...

CHROMIUM, DISSOLVED

0 / 3

—

—

—

...

COBALT, DISSOLVED

NBA

—

—

—

...

COPPER, DISSOLVED

0 / 6

—

—

—

...

LEAD, DISSOLVED

0 / 1

—

—

—

...

MAGNESIUM, DISSOLVED

NBA

—

—

—

...

MERCURY, FILTERED

0 / 1

...

...

...

...

SELENIUM, DISSOLVED

0 / 1

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-93

-------
Table B-93

Surface Water Comparison to Benchmark Summary
5c Main Channel
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ZINC, DISSOLVED

0 / 3

—

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

TOTAL ORGANIC CARBON, DISSOLVED

NBA

...

...

...

...

INORGANICS

ALKALINITY

NBA

...

...

...

...

AMMONIA AS N

14 / 14

—

12

2

...

BIOLOGICAL OXYGEN DEMAND 5 DAY

NBA

...

...

...

...

HARDNESS

NBA

—

—

—

...

NITRATE AND NITRITE AS N

NBA

—

—

—

...

NITRITE AS N

NBA

—

—

—

...

ORGANIC CARBON, PARTICULATE

NBA

—

—

—

...

ORTHOPHOSPHATE AS P

NBA

—

—

—

...

PHOSPHATE, AS P

NBA

—

—

—

...

PHOSPHATE, TOTAL AS P

NBA

—

—

—

...

SULFIDE

2 / 2

—

—

2

...

TKN

NBA

—

—

—

...

TOTAL DISSOLVED SOLIDS

NBA

—

—

—

...

TOTAL SUSPENDED SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-93

-------
Table B-94

Surface Water Comparison to Benchmark Summary
5c SCOX

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

TOTAL ORGANIC CARBON, DISSOLVED

NBA

...

...

...

...

INORGANICS

AMMONIA AS N

14 / 14

2

11

1

...

BIOLOGICAL OXYGEN DEMAND 5 DAY

NBA

—

—

—

...

HYDROLYZABLE PHOSPHATE, AS P

NBA

...

...

...

...

NITRATE AND NITRITE AS N

NBA

—

—

—

...

ORGANIC CARBON, PARTICULATE

NBA

—

—

—

...

ORGANIC PHOSPHATE, AS P

NBA

—

—

—

...

ORTHOPHOSPHATE AS P

NBA

—

—

—

...

PHOSPHATE, TOTAL AS P

NBA

—

—

—

...

TKN

NBA

—

—

—

...

TOTAL SUSPENDED SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-94

-------
Table B-95

Surface Water Comparison to Benchmark Summary
5c Vernal Pools
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIOXINS/FURANS

TCDF (TOTAL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

...

...

...

...

PCB, TOTAL

7 / 7

5

2

...

...

METALS

BARIUM

3 / 3

3

—

—

...

CALCIUM

NBA

...

...

...

...

COPPER

0 / 2

—

—

—

...

LEAD

0 / 2

...

...

...

...

MAGNESIUM

NBA

—

—

—

...

VANADIUM

0 / 1

—

—

—

...

ZINC

0 / 1

...

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-95

-------
Table B-96

Surface Water Comparison to Benchmark Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX VOLATILES

ACETONE

1 / 2

—

...

—

...

BROMODICHLOROMETHANE

2 / 4

...

...

...

...

CHLOROFORM

2 / 4

—

—

—

...

DIBROMOCHLOROMETHANE

1 / 4

—

—

—

...

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

1 / 13

...

...

...

...

PAHs











BENZO(A)ANTHRACENE

1 / 13

...

...

...

...

CHRYSENE

1 / 13

—

—

—

...

FLUORANTHENE

1 / 13

...

...

...

...

PHENANTHRENE

1 / 13

—

—

—

...

PYRENE

1 / 13

1

—

—

...

TOTAL PAH (USING 0)

13 / 13

—

—

—

...

TOTAL PAH (USING DL)

13 / 13

—

—

—

...

TOTAL PAH (USING HALF DL)

13 / 13

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

13 / 13

—

—

—

...

TOTAL PAH (HIGH) (USING DL)

13 / 13

—

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

13 / 13

—

—

—

...

TOTAL PAH (LOW) (USING 0)

13 / 13

—

—

—

...

TOTAL PAH (LOW) (USING DL)

13 / 13

—

—

—

...

TOTAL PAH (LOW) (USING HALF DL)

13 / 13

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-96

-------
Table B-96

Surface Water Comparison to Benchmark Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

1.2.4-TRICHLOROBENZENE, FILTERED

14 / 15

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

13 / 13

—

—

—

...

TOTAL DIOXINS (USING DL)

13 / 13

...

...

...

...

TOTAL DIOXINS (USING HALF DL)

13 / 13

—

—

—

...

TOTAL FURANS (USING 0)

13 / 13

—

—

—

...

TOTAL FURANS (USING DL)

13 / 13

—

—

—

...

TOTAL FURANS (USING HALF DL)

13 / 13

—

—

—

...

PCBS

AROCLOR-1242

8 / 40

—

—

—

...

AROCLOR-1248

1 / 40

**

**

**

**

AROCLOR-1254

30 / 40

—

—

—

...

AROCLOR-1260

36 / 39

—

—

—

...

PCB, TOTAL

36 / 40

32

4

...

...

PCBS - FILTERED

AROCLOR-1260, DISSOLVED

1 / 41

**

**

**

**

TOTAL PCB, DISSOLVED

1 / 41

**

**

**

**

METALS

BARIUM

13 / 13

13

—

—

...

CADMIUM

1 / 13

1

—

—

...

CALCIUM

9 / 9

—

—

—

...

CHROMIUM

2 / 13

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-96

-------
Table B-96

Surface Water Comparison to Benchmark Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

COPPER

7 / 13

—

—

—

—

MAGNESIUM

9 / 9

...

...

...

...

MERCURY

1 / 13

—

—

—

...

SILVER

1 / 13

1

...

—

...

THALLIUM

1 / 13

—

—

—

...

ZINC

3 / 13

...

1

—

...

METALS - FILTERED

BARIUM, DISSOLVED

14 / 14

—

—

—

...

CADMIUM, DISSOLVED

1 / 14

1

—

—

...

CALCIUM, DISSOLVED

9 / 9

—

—

—

...

CHROMIUM, DISSOLVED

1 / 14

—

—

—

...

COPPER, DISSOLVED

5 / 14

—

—

—

...

LEAD, DISSOLVED

1 / 14

—

—

—

...

MAGNESIUM, DISSOLVED

9 / 9

—

—

—

...

MERCURY, FILTERED

1 / 9

...

...

...

...

NICKEL, DISSOLVED

1 / 14

—

—

—

...

ZINC, DISSOLVED

3 / 14

—

—

—

...

ORGANIC

TOTAL ORGANIC CARBON

35 / 35

—

—

—

...

TOTAL ORGANIC CARBON, DISSOLVED

42 / 42

—

—

—

...

INORGANICS

ALKALINITY

40 / 40

...

...

...

...

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-96

-------
Table B-96

Surface Water Comparison to Benchmark Summary
6cd Pond

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detection

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

AMMONIA AS N

39 / 40

1

36

2

—

BIOLOGICAL OXYGEN DEMAND 5 DAY

13 / 41

...

...

...

...

CHEMICAL OXYGEN DEMAND

27 / 27

—

—

—

...

HARDNESS

40 / 40

—

—

—

...

HYDROLYZABLE PHOSPHATE, AS P

24 / 24

—

—

—

...

NITRATE AND NITRITE AS N

40 / 40

—

—

—

...

NITRITE AS N

31 / 40

—

—

—

...

ORGANIC CARBON, PARTICULATE

8 / 34

—

—

—

...

ORGANIC PHOSPHATE, AS P

21 / 27

—

—

—

...

ORTHOPHOSPHATE AS P

37 / 40

—

—

—

...

PHOSPHATE, AS P

1 / 1

—

—

—

...

PHOSPHATE, TOTAL AS P

39 / 39

—

—

—

...

SULFIDE

1 / 11

—

—

1

...

TKN

39 / 40

—

—

—

...

TOTAL DISSOLVED SOLIDS

13 / 13

—

—

—

...

TOTAL SUSPENDED SOLIDS

570 / 590

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-88 - B-97.xls B-96

-------
Table B-97

Surface Water Comparison to Benchmark Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX VOLATILES

ACETONE

0 / 1

—

—

—

—

CHLOROFORM

0 / 6

—

—

—

—

TOLUENE

NBA

—

—

—

—

APP IX SEMIVOLATILES

PAHs

FLUORANTHENE

0 / 4

—

—

—

—

NAPHTHALENE

0 / 2

—

—

—

—

PHENANTHRENE

0 / 4

—

—

—

—

PYRENE

2 / 4

2

—

—

—

1,2,4-TRICHLOROBENZENE, FILTERED

0 / 2

—

—

—

—

DIOXINS/FURANS

HPCDD (TOTAL)

NBA

—

—

—

—

HPCDF (TOTAL)

NBA

—

—

—

—

HXCDD (TOTAL)

NBA

—

—

—

—

OCDD

NBA

—

—

—

—

OCDF

NBA

—

—

—

—

PECDF (TOTAL)

NBA

—

—

—

—

TCDD (TOTAL)

NBA

—

—

—

—

TCDF (TOTAL)

NBA

—

—

—

—

PCBS

PCB, TOTAL

2 / 2

1

1

—

—

METALS

BARIUM

35 / 35

35

—

—

—

CALCIUM

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-88 - B-97.xls B-97

-------
Table B-97

Surface Water Comparison to Benchmark Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

COPPER

1 / 4

1

—

—

—

LEAD

2 / 4

1

1

—

—

MAGNESIUM

NBA

—

—

—

—

MERCURY

0 / 2

—

—

—

—

TIN

0 / 2

—

—

—

—

VANADIUM

0 / 3

—

—

—

—

ZINC

2 / 16

2

—

—

—

METALS - FILTERED

BARIUM, DISSOLVED

NBA

—

—

—

—

CALCIUM, DISSOLVED

NBA

—

—

—

—

CHROMIUM, DISSOLVED

0 / 2

—

—

—

—

MAGNESIUM, DISSOLVED

NBA

—

—

—

—

MERCURY, FILTERED

0 / 2

—

—

—

—

TIN, DISSOLVED

NBA

—

—

—

—

VANADIUM, DISSOLVED

NBA

—

—

—

—

ZINC, DISSOLVED

0 / 9

—

—

—

—

ORGANIC

CHLOROPHYLL-A, CORRECTED

NBA

—

—

—

—

CHLOROPHYLL-A, UNCORRECTED

NBA

—

—

—

—

TOTAL ORGANIC CARBON

NBA

—

—

—

—

TOTAL ORGANIC CARBON, DISSOLVED

NBA

—

—

—

—

INORGANICS

ALKALINITY

NBA

—

—

—

—

AMMONIA AS N

30 / 30

4

23

3

—

BIOLOGICAL OXYGEN DEMAND 5 DAY

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-88 - B-97.xls B-97

-------
Table B-97

Surface Water Comparison to Benchmark Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

HARDNESS

NBA

—

—

—

—

HARDNESS, DISSOLVED

NBA

—

—

—

—

NITRATE AND NITRITE AS N

NBA

—

—

—

—

NITRITE AS N

NBA

—

—

—

—

ORGANIC CARBON, PARTICULATE

NBA

—

—

—

—

ORTHOPHOSPHATE AS P

NBA

—

—

—

—

PHOSPHATE, AS P

NBA

—

—

—

—

PHOSPHATE, TOTAL AS P

NBA

—

—

—

—

SULFIDE

13 / 13

—

—

13

—

TKN

NBA

—

—

—

—

TOTAL DISSOLVED SOLIDS

NBA

—

—

—

—

TOTAL SUSPENDED SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-88 - B-97.xls B-97

-------
Table B-98

Soil Comparison to Benchmark Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 2

—

—

—

—

BUTYLBENZYLPHTHALATE

NBA

—

—

—

—

DIBENZOFURAN

NBA

—

—

—

—

1,4-DICHLOROBENZENE

0 / 8

—

—

—

—

DIETHYL PHTHALATE

0 / 2

—

—

—

—

2-METHYLNAPHTHALENE

0 / 13

—

—

—

—

PENTACHLOROBENZENE

0 / 5

—

—

—

—

PAHs









ACENAPHTHENE

0 / 8

—

—

—

—

ACENAPHTHYLENE

0 / 15

—

—

—

—

ANTHRACENE

0 / 14

—

—

—

—

BENZO(A)ANTHRACENE

0 / 22

—

—

—

—

BENZO(B)FLUORANTHENE

0 / 21

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 21

—

—

—

—

BENZO(GHI)PERYLENE

0 / 20

—

—

—

—

BENZO(A)PYRENE

7 / 22

7

—

—

—

CHRYSENE

0 / 22

—

—

—

—

DIBENZO(A,H) ANTHRACENE

0 / 13

—

—

—

—

FLUORANTHENE

0 / 23

—

—

—

—

FLUORENE

0 / 10

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

0 / 21

—

—

—

—

NAPHTHALENE

0 / 17

—

—

—

—

PHENANTHRENE

0 / 22

—

—

—

—

PYRENE

8 / 24

8

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-98

-------
Table B-98

Soil Comparison to Benchmark Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING 0)

NBA

—

—

—

—

TOTAL PAH (USING DL)

NBA

—

—

—

—

TOTAL PAH (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

—

1,2,4-TRICHLOROBENZENE

0 / 9

—

—

—

—

APP IX PESTICIDES

4,4'-DDT

1 / 2

1

—

—

—

HERBICIDES

2,4,5-T

NBA

—

—

—

—

DIOXINS/FURANS

TCDD (TOTAL)

12 / 19

10

2

—

—

TCDF (TOTAL)

4 / 24

4

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

646 / 679

198

348

100

—

METALS

ANTIMONY

0 / 10

—

—

—

—

ARSENIC

0 / 24

—

—

—

—

BARIUM

0 / 24

—

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-98

-------
Table B-98

Soil Comparison to Benchmark Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

BERYLLIUM

0 / 24

—

—

—

—

CADMIUM

0 / 3

—

—

—

—

CHROMIUM

24 / 24

—

22

2

—

COBALT

0 / 24

—

—

—

—

COPPER

2 / 24

2

—

—

—

LEAD

11/24

11

—

—

—

MERCURY

23 / 23

—

4

16

3

NICKEL

1 / 24

1

—

—

—

SELENIUM

2 / 2

2

—

—

—

SILVER

0 / 8

—

—

—

—

THALLIUM

10 / 17

10

—

—

—

TIN

0 / 7

—

—

—

—

VANADIUM

24 / 24

18

6

—

—

ZINC

24 / 24

8

16

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-98 - B-107.xls B-98

-------
Table B-99

Soil Comparison to Benchmark Summary
5 a Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APPIX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 4

...

...

...

...

BUTYLBENZYLPHTHALATE

NBA

...

...

...

...

DIBENZOFURAN

NBA

...

...

...

...

DI-N-BUTYL PHTHALATE

0 / 3

...

...

...

...

1,4-DICHLOROBENZENE

0 / 7

...

...

...

...

DIMETHYL PHTHALATE

0 / 1

...

...

...

...

2-METHYLNAPHTHALENE

0 / 9

...

...

...

...

PAHs











ACENAPHTHENE

0 / 8

...

...

...

...

ACENAPHTHYLENE

0 / 9

...

...

...

...

ANTHRACENE

0 / 9

...

...

...

...

BENZO(A)ANTHRACENE

1 / 10

1

...

...

...

BENZO(B)FLUORANTHENE

0 / 10

...

...

...

...

BENZO(K)FLUORANTHENE

1 / 9

1

...

...

...

BENZO(GHI)PERYLENE

0 / 9

...

...

...

...

BENZO(A )P YRENE

8 / 9

7

1

...

...

CHRYSENE

1 / 10

1

...

...

...

DIBENZO(A,H)ANTHRACENE

0 / 8

...

...

...

...

FLUORANTHENE

0 / 10

...

...

...

...

FLUORENE

0 / 9

...

...

...

...

INDENO(l ,2,3-C,D)PYRENE

0 / 10

...

...

...

...

NAPHTHALENE

0 / 9

...

...

...

...

PHENANTHRENE

0 / 10

...

...

...

...

PYRENE

7 / 10

6

1

...

...

TOTAL PAH (USING 0)

NBA

...

...

...

...

TOTAL PAH (USING DL)

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-99

-------
Table B-99

Soil Comparison to Benchmark Summary
5 a Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (USING HALF DL)

NBA

—

...

...

...

TOTAL PAH (HIGH) (USING 0)

NBA

...

...

...

...

TOTAL PAH (HIGH) (USING DL)

NBA

...

...

...

...

TOTAL PAH (HIGH) (USING HALF DL)

NBA

...

...

...

...

TOTAL PAH (LOW) (USING 0)

NBA

...

...

...

...

TOTAL PAH (LOW) (USING DL)

NBA

...

...

...

...

TOTAL PAH (LOW) (USING HALF DL)

NBA

...

...

...

...

PYRIDINE

NBA

...

...

...

...

1,2,4-TRICHLOROBENZENE

0 / 2

...

...

...

...

APPIX PESTICIDES

ENDRIN ALDEHYDE

NBA

...

...

...

...

DIOXIN S/FURAN S

TCDD (TOTAL)

6 / 7

4

2

...

...

TCDF (TOTAL)

2 / 10

2

...

...

...

TOTAL DIOXINS (USING 0)

NBA

...

...

...

...

TOTAL DIOXINS (USING DL)

NBA

...

...

...

...

TOTAL DIOXINS (USING HALF DL)

NBA

...

...

...

...

TOTAL FURANS (USING 0)

NBA

...

...

...

...

TOTAL FURANS (USING DL)

NBA

...

...

...

...

TOTAL FURANS (USING HALF DL)

NBA

...

...

...

...

PCBS

AROCLOR-1248



AROCLOR-1254

NBA

...

...

...

...

AROCLOR-1260

NBA

...

...

...

...

PCB, TOTAL

115 / 116

18

84

13

...

METALS

ANTIMONY

0 / 7

...

...

...

...

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-99

-------
Table B-99

Soil Comparison to Benchmark Summary
5 a Riverbank
Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

1 / 9

1

...

...

...

BARIUM

0 / 10

...

...

...

...

BERYLLIUM

0 / 9

...

...

...

...

CHROMIUM

10 / 10

...

8

2

...

COBALT

0 / 10

...

...

...

...

COPPER

2 / 10

2

...

...

...

LEAD

8 / 10

8

...

...

...

MERCURY

10 / 10

...

...

9

1

NICKEL

0 / 10

...

...

...

...

SELENIUM

1 / 1

1

...

...

...

SILVER

0 / 5

...

...

...

...

THALLIUM

4 / 4

4

...

...

...

TIN

0 / 5

...

...

...

...

VANADIUM

10 / 10

9

1

...

...

ZINC

10 / 10

1

9

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

...

...

...

...

INORGANICS

PERCENT SOLIDS

NBA

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-99

-------
Table B-100

Soil Comparison to Benchmark Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 12

—

—

—

—

BUTYLBENZYLPHTHALATE

NBA

—

—

—

—

DIBENZOFURAN

NBA

—

—

—

—

DI-N-BUTYL PHTHALATE

0 / 2

—

—

—

—

1,4-DICHLOROBENZENE

0 / 10

—

—

—

—

2-METHYLNAPHTHALENE

0 / 13

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PHENOL

0 / 2

—

—

—

—

PAHs











ACENAPHTHENE

0 / 8

—

—

—

—

ACENAPHTHYLENE

0 / 13

—

—

—

—

ANTHRACENE

0 / 15

—

—

—

—

BENZO(A)ANTHRACENE

0 / 16

—

—

—

—

BENZO(B)FLUORANTHENE

0 / 16

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 16

—

—

—

—

BENZO(GHI)PERYLENE

0 / 16

—

—

—

—

BENZO(A)PYRENE

2 / 16

2

—

—

—

CHRYSENE

0 / 16

—

—

—

—

DIBENZO(A,H) ANTHRACENE

0 / 16

—

—

—

—

FLUORANTHENE

0 / 16

—

—

—

—

FLUORENE

0 / 12

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

0 / 16

—

—

—

—

NAPHTHALENE

0 / 16

—

—

—

—

PHENANTHRENE

0 / 16

—

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-100

-------
Table B-100

Soil Comparison to Benchmark Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PYRENE

1 / 16

1

—

—

—

TOTAL PAH (USING 0)

NBA

—

—

—

—

TOTAL PAH (USING DL)

NBA

—

—

—

—

TOTAL PAH (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

—

1,2,4-TRICHLOROBENZENE

0 / 7

—

—

—

—

APP IX PESTICIDES

ENDOSULFAN SULFATE

NBA

—

—

—

—

DIOXINS/FURANS

TCDD (TOTAL)

16 / 16

14

2

—

—

TCDF (TOTAL)

4 / 16

4

—

—

—

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-100

-------
Table B-100

Soil Comparison to Benchmark Summary
5a Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCB, TOTAL

239 / 240

41

147

51

—

METALS

ANTIMONY

0 / 11

—

—

—

—

ARSENIC

0 / 16

—

—

—

—

BARIUM

0 / 16

—

—

—

—

BERYLLIUM

0 / 16

—

—

—

—

CADMIUM

0 / 11

—

—

—

—

CHROMIUM

16 / 16

—

10

6

—

COBALT

0 / 16

—

—

—

—

COPPER

1 / 16

1

—

—

—

LEAD

14 / 16

14

—

—

—

MERCURY

16 / 16

—

—

15

1

NICKEL

0 / 16

—

—

—

—

SELENIUM

2 / 2

2

—

—

—

SILVER

1 / 11

1

—

—

—

THALLIUM

5 / 9

5

—

—

—

TIN

0 / 8

—

—

—

—

VANADIUM

16 / 16

15

1

—

—

ZINC

16 / 16

—

16

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

CYANIDE

1 / 2

1

—

—

—

PERCENT SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-98 - B-107.xls B-100

-------
Table B-101

Soil Comparison to Benchmark Summary

5b Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 3

—

—

—

...

BUTYLBENZYLPHTHALATE

NBA

...

...

...

...

DIBENZOFURAN

NBA

—

—

—

...

DI-N-BUTYL PHTHALATE

0 / 1

—

—

—

...

1,4-DICHLOROBENZENE

0 / 3

—

—

—

...

2-METHYLNAPHTHALENE

0 / 3

—

—

—

...

PAHs











ACENAPHTHENE

0 / 3

—

—

—

...

ACENAPHTHYLENE

0 / 3

—

—

—

...

ANTHRACENE

0 / 4

—

—

—

...

BENZO(A)ANTHRACENE

0 / 4

—

—

—

...

BENZO(B)FLUORANTHENE

0 / 4

—

—

—

...

BENZO(K)FLUORANTHENE

0 / 4

—

—

—

...

BENZO(GHI)PERYLENE

0 / 4

—

—

—

...

BENZO(A)PYRENE

0 / 4

—

—

—

...

CHRYSENE

0 / 4

—

—

—

...

DIBENZO(A,H) ANTHRACENE

0 / 4

—

—

—

...

FLUORANTHENE

0 / 4

—

—

—

...

FLUORENE

0 / 3

...

...

—

...

INDENO(l,2,3-C,D)PYRENE

0 / 4

—

—

—

...

NAPHTHALENE

0 / 4

...

...

...

...

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-101

-------
Table B-101

Soil Comparison to Benchmark Summary

5b Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PHENANTHRENE

0 / 4

—

—

—

—

PYRENE

1 / 4

1

...

...

...

TOTAL PAH (USING 0)

NBA

—

—

—

...

TOTAL PAH (USING DL)

NBA

—

—

—

...

TOTAL PAH (USING HALF DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING DL)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

...

1,2.4-TRICHLOROBENZENE

0 / 3

—

...

...

...

DIOXINS/FURANS

TCDD (TOTAL)

3 / 3

3

...

...

...

TCDF (TOTAL)

4 / 4

4

...

...

...

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-101

-------
Table B-101

Soil Comparison to Benchmark Summary

5b Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

30 / 30

10

15

5

...

METALS

ARSENIC

0 / 4

...

...

...

...

BARIUM

0 / 4

—

—

—

...

BERYLLIUM

0 / 4

...

...

...

...

CADMIUM

0 / 4

—

—

—

...

CHROMIUM

4 / 4

—

2

2

...

COBALT

0 / 4

...

...

...

...

COPPER

1 / 4

1

—

—

...

LEAD

4 / 4

4

...

...

...

MERCURY

4 / 4

—

—

4

...

NICKEL

0 / 4

...

...

...

...

SILVER

1 / 4

1

...

...

...

THALLIUM

2 / 2

2

...

—

...

TIN

0 / 1

—

—

—

—

VANADIUM

4 / 4

4

—

—

...

ZINC

4 / 4

...

4

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

...

INORGANICS

PERCENT SOLIDS

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-101

-------
Table B-102

Soil Comparison to Benchmark Summary
5c Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 9

—

—

—

—

BUTYLBENZYLPHTHALATE

NBA

—

—

—

—

DIBENZOFURAN

NBA

—

—

—

—

DI-N-BUTYL PHTHALATE

0 / 2

—

—

—

—

1,4-DICHLOROBENZENE

0 / 9

—

—

—

—

2-METHYLNAPHTHALENE

0 / 9

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

N-NITROSO-DI-N-BUTYL AMINE

NBA

—

—

—

—

PAHs









ACENAPHTHENE

0 / 4

—

—

—

—

ACENAPHTHYLENE

0 / 6

—

—

—

—

ANTHRACENE

0 / 7

—

—

—

—

BENZO(A)ANTHRACENE

0 / 15

—

—

—

—

BENZO(B)FLUORANTHENE

0 / 17

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 17

—

—

—

—

BENZO(GHI)PERYLENE

0 / 15

—

—

—

—

BENZO(A)PYRENE

2 / 17

2

—

—

—

CHRYSENE

0 / 18

—

—

—

—

DIBENZO(A,H) ANTHRACENE

0 / 9

—

—

—

—

FLUORANTHENE

0 / 18

—

—

—

—

FLUORENE

0 / 7

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

0 / 15

—

—

—

—

NAPHTHALENE

0 / 11

—

—

—

—

PHENANTHRENE

0 / 18

—

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-102

-------
Table B-102

Soil Comparison to Benchmark Summary
5c Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PYRENE

3 / 18

3

—

—

—

TOTAL PAH (USING 0)

NBA

—

—

—

—

TOTAL PAH (USING DL)

NBA

—

—

—

—

TOTAL PAH (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

—

1,2,4-TRICHLOROBENZENE

0 / 4

—

—

—

—

DIOXINS/FURANS

TCDD (TOTAL)

15 / 18

6

9

—

—

TCDF (TOTAL)

10 / 19

10

—

—

—

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

317 / 342

102

155

60

—

METALS

ANTIMONY

0 / 14

—

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-102

-------
Table B-102

Soil Comparison to Benchmark Summary
5c Floodplain
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

ARSENIC

1 / 19

1

—

—

—

BARIUM

0 / 19

—

—

—

—

BERYLLIUM

0 / 19

—

—

—

—

CADMIUM

1 / 17

1

—

—

—

CHROMIUM

19 / 19

—

7

12

—

COBALT

0 / 19

—

—

—

—

COPPER

10 / 19

10

—

—

—

LEAD

14 / 19

14

—

—

—

MERCURY

18 / 18

—

—

10

8

NICKEL

2 / 19

2

—

—

—

SELENIUM

5 / 5

3

2

—

—

SILVER

7 / 14

7

—

—

—

THALLIUM

3 / 9

3

—

—

—

TIN

0 / 11

—

—

—

—

VANADIUM

19 / 19

7

12

—

—

ZINC

19 / 19

3

16

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

SULFIDE

NBA

—

—

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-98 - B-107.xls B-102

-------
Table B-103

Soil Comparison to Benchmark Summary
5c Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 2

—

—

—

—

1,4-DICHLOROBENZENE

0 / 1

—

—

—

—

2-METHYLNAPHTHALENE

0 / 1

—

—

—

—

4-METHYLPHENOL

NBA

—

—

—

—

PAHs

ACENAPHTHYLENE

0 / 1

—

—

—

—

ANTHRACENE

0 / 2

—

—

—

—

BENZO(A)ANTHRACENE

0 / 4

—

—

—

—

BENZO(B)FLUORANTHENE

0 / 4

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 4

—

—

—

—

BENZO(GHI)PERYLENE

0 / 4

—

—

—

—

BENZO(A)PYRENE

0 / 3

—

—

—

—

CHRYSENE

0 / 4

—

—

—

—

DIBENZO(A,H) ANTHRACENE

0 / 2

—

—

—

—

FLUORANTHENE

0 / 4

—

—

—

—

INDENO( 1,2,3 -C,D)P"YRENE

0 / 4

—

—

—

—

NAPHTHALENE

0 / 4

—

—

—

—

PHENANTHRENE

0 / 4

—

—

—

—

PYRENE

0 / 4

—

—

—

—

TOTAL PAH (USING 0)

NBA

—

—

—

—

TOTAL PAH (USING DL)

NBA

—

—

—

—

TOTAL PAH (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

—

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-103

-------
Table B-103

Soil Comparison to Benchmark Summary
5c Riverbank
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL PAH (HIGH) (USING HALF DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING DL)

NBA

—

—

—

—

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

—

1,2,4-TRICHLOROBENZENE

0 / 1

—

—

—

—

DIOXINS/FURANS

TCDD (TOTAL)

2 / 4

1

1

—

—

TCDF (TOTAL)

1 / 4

1

—

—

—

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

24 / 24

4

15

5

—

METALS

ANTIMONY

0 / 4

—

—

—

—

ARSENIC

0 / 4

—

—

—

—

BARIUM

0 / 4

—

—

—

—

BERYLLIUM

0 / 2

—

—

—

—

CADMIUM

0 / 4

—

—

—

—

CHROMIUM

4 / 4

—

1

3

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-103

-------
Table B-103

Soil Comparison to Benchmark Summary
5c Riverbank
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

COBALT

0 / 4

—

—

—

—

COPPER

2 / 4

2

—

—

—

LEAD

4 / 4

4

—

—

—

MERCURY

4 / 4

—

—

2

2

NICKEL

0 / 4

—

—

—

—

SILVER

1 / 4

1

—

—

—

THALLIUM

1 / 3

1

—

—

—

TIN

0 / 4

—

—

—

—

VANADIUM

4 / 4

3

1

—

—

ZINC

4 / 4

—

4

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-98 - B-107.xls B-103

-------
Table B-104

Soil Comparison to Benchmark Summary
6ab Floodplain
Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCBS

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

44 / 45

6

25

13

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-98 - B-107.xls B-104

-------
Table B-105

Soil Comparison to Benchmark Summary
6ab Riverbank
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCBS

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

3 / 3

—

3

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-98 - B-107.xls B-105

-------
Table B-106

Soil Comparison to Benchmark Summary

6cd Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

ACETOPHENONE

NBA

—

—

—

...

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 1

...

...

...

...

DIBENZOFURAN

NBA

—

—

—

...

DI-N-BUTYL PHTHALATE

0 / 1

—

—

—

...

1,4-DICHLOROBENZENE

0 / 2

—

—

—

...

2-METHYLNAPHTHALENE

0 / 2

—

—

—

...

4-METHYLPHENOL

NBA

—

—

—

...

4-NITROPHENOL

0 / 1

—

—

—

...

P-PHENYLENEDIAMINE

NBA

—

—

—

...

PAHs

ACENAPHTHYLENE

0 / 3

—

—

—

...

ANTHRACENE

0 / 3

—

—

—

...

BENZO(A)ANTHRACENE

0 / 6

—

—

—

...

BENZO(B)FLUORANTHENE

0 / 7

—

—

—

...

BENZO(K)FLUORANTHENE

0 / 7

—

—

—

...

BENZO(GHI)PERYLENE

0 / 6

—

—

—

...

BENZO(A)PYRENE

1 / 3

1

—

—

...

CHRYSENE

0 / 7

—

—

—

...

DIBENZO(A,H) ANTHRACENE

0 / 2

—

—

—

...

FLUORANTHENE

0 / 7

—

—

—

...

FLUORENE

0 / 1

...

...

...

...

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-106

-------
Table B-106

Soil Comparison to Benchmark Summary

6cd Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

INDENO( 1,2,3 -C,D)PYRENE

0 / 6

—

—

—

—

NAPHTHALENE

0 / 6

...

...

...

...

PHENANTHRENE

0 / 6

—

—

—

...

PYRENE

0 / 7

...

...

—

...

TOTAL PAH (USING 0)

NBA

—

—

—

...

TOTAL PAH (USING DL)

NBA

—

—

—

...

TOTAL PAH (USING HALF DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING 0)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING DL)

NBA

—

—

—

...

TOTAL PAH (HIGH) (USING HALF DL)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING 0)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING DL)

NBA

—

—

—

...

TOTAL PAH (LOW) (USING HALF DL)

NBA

—

—

—

...

DIOXINS/FURANS

TCDD (TOTAL)

5 / 6

3

2

...

...

TCDF (TOTAL)

3 / 7

3

...

...

...

TOTAL DIOXINS (USING 0)

NBA

—

—

—

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-106

-------
Table B-106

Soil Comparison to Benchmark Summary

6cd Floodplain
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCBS

AROCLOR-1254

NBA

...

...

...

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

41 / 43

15

15

11

...

METALS

ANTIMONY

0 / 4

...

...

...

...

ARSENIC

0 / 6

—

—

—

...

BARIUM

0 / 7

—

—

—

...

BERYLLIUM

0 / 7

...

...

...

...

CADMIUM

0 / 5

—

—

—

...

CHROMIUM

7 / 7

—

4

3

...

COBALT

0 / 7

...

...

...

...

COPPER

3 / 7

3

—

—

...

LEAD

7 / 7

7

...

...

...

MERCURY

7 / 7

—

—

4

3

NICKEL

0 / 7

...

...

...

...

SELENIUM

3 / 3

3

...

—

...

SILVER

3 / 6

3

...

...

...

THALLIUM

5 / 5

5

...

—

...

TIN

0 / 4

...

...

...

...

VANADIUM

7 / 7

1

6

—

...

ZINC

7 / 7

...

7

...

...

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-106

-------
Table B-106

Soil Comparison to Benchmark Summary

6cd Floodplain
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

INORGANICS

PERCENT SOLIDS

NBA

...

...

...

...

ORGANIC

TOTAL ORGANIC CARBON

NBA

...

...

...

...

Comparisons made using raw, not averaged data.

NBA = No Benchmark Available

Tables B-98 - B-107.xls B-106

-------
Table B-107

Soil Comparison to Benchmark Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX SEMIVOLATILES

BENZYL ALCOHOL

NBA

—

—

—

—

BIS(2-ETHYLHEXYL) PHTHALATE

0 / 6

—

—

—

—

2-METHYLNAPHTHALENE

0 / 1

—

—

—

—

PAHs

ACENAPHTHYLENE

0 / 2

—

—

—

—

ANTHRACENE

0 / 1

—

—

—

—

BENZO(A)ANTHRACENE

0 / 10

—

—

—

—

BENZO(B)FLUORANTHENE

0 / 10

—

—

—

—

BENZO(K)FLUORANTHENE

0 / 10

—

—

—

—

BENZO(GHI)PERYLENE

0 / 7

—

—

—

—

BENZO(A)PYRENE

0 / 8

—

—

—

—

CHRYSENE

0 / 14

—

—

—

—

DIBENZO(A,H) ANTHRACENE

0 / 1

—

—

—

—

FLUORANTHENE

0 / 16

—

—

—

—

INDENO(l,2,3-C,D)PYRENE

0 / 9

—

—

—

—

NAPHTHALENE

0 / 1

—

—

—

—

PHENANTHRENE

0 / 12

—

—

—

—

PYRENE

0 / 16

—

—

—

—

TOTAL PAH

NBA

—

—

—

—

TOTAL PAH (HIGH MW)

NBA

—

—

—

—

TOTAL PAH (LOW MW)

NBA

—

—

—

—

APP IX PESTICIDES

ALPHA-BHC

NBA

—

—

—

—

BETA-BHC

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-98 - B-107.xls B-107

-------
Table B-107

Soil Comparison to Benchmark Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

4,4'-DDE

NBA

—

—

—

—

HEPTACHLOR

NBA

—

—

—

—

DIOXINS/FURANS

TCDD (TOTAL)

4 / 17

4

—

—

—

TCDF (TOTAL)

0 / 20

—

—

—

—

TOTAL DIOXINS

NBA

—

—

—

—

TOTAL FURANS

NBA

—

—

—

—

HERBICIDES

2,4,5-T

NBA

—

—

—

—

2,4,5-TP (SILVEX)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

4 / 16

4

—

—

—

METALS

ANTIMONY

0 / 4

—

—

—

—

ARSENIC

0 / 20

—

—

—

—

BARIUM

0 / 20

—

—

—

—

BERYLLIUM

0 / 16

—

—

—

—

CADMIUM

0 / 1

—

—

—

—

CHROMIUM

20 / 20

—

20

—

—

COBALT

0 / 20

—

—

—

—

COPPER

0 / 19

—

—

—

—

LEAD

0 / 20

—

—

—

—

MERCURY

19 / 19

—

6

11

2

NBA = No benchmark available.

Tables B-98 - B-107.xls B-107

-------
Table B-107

Soil Comparison to Benchmark Summary
Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

NICKEL

0 / 20

—

—

—

—

SELENIUM

2 / 2

2

—

—

—

SILVER

0 / 1

—

—

—

—

THALLIUM

8 / 12

8

—

—

—

VANADIUM

20 / 20

17

3

—

—

ZINC

20 / 20

18

2

—

—

ORGANIC

TOTAL ORGANIC CARBON

NBA

—

—

—

—

INORGANICS

PERCENT SOLIDS

NBA

—

—

—

—

Comparisons made using raw, not averaged data.

NBA = No benchmark available.

Tables B-98 - B-107.xls B-107

-------
Table B-108

Fish Tissue Comparison to Benchmark Summary
5a Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALPHA-BHC

0 / 1

—

—

—

...

BETA-BHC

0 / 2

...

...

...

...

DELTA-BHC

0 / 1

...

...

...

...

GAMMA-CHLORDANE

0 / 1

—

—

—

...

CHLORP YRIF 0 S

0 / 2

—

—

—

...

0,P'-DDD

0 / 2

—

—

—

...

4,4'-DDD

0 / 2

—

—

—

...

4,4'-DDE

0 / 2

—

—

—

...

0 P'-DDT

2 / 2

—

2

—

...

4,4'-DDT

2 / 2

2

—

—

...

DIELDRIN

0 / 1

...

...

—

...

ENDOSULFANII

0 / 1

—

—

—

...

ENDRIN

0 / 1

—

—

—

...

HEPTACHLOR EPOXIDE

NBA

—

—

—

...

HEXACHLOROBENZENE

0 / 2

—

—

—

...

CIS-NONACHLOR

NBA

—

—

—

...

TRANS-NONACHLOR

NBA

—

—

—

...

OXYCHLORDANE

NBA

—

—

—

...

PENTACHLOROANISOLE

0 / 2

—

—

—

...

PENTACHLOROBENZENE

0 / 2

...

...

—

...

TOXAPHENE

0 / 1

...

...

...

...

NBA = No Benchmark Available

Tables B-108 - B-l 19.xls B-108

-------
Table B-108

Fish Tissue Comparison to Benchmark Summary
5a Small Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

1,2,3,4-TETRACHLOROBENZENE

0 / 2

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 2

...

...

...

...

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

...

...

...

...

TOTAL DIOXINS (USING DL)

NBA

—

—

—

...

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

...

TOTAL FURANS (USING 0)

NBA

—

—

—

...

TOTAL FURANS (USING DL)

NBA

—

—

—

...

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

...

PCBS

AROCLOR-1254

NBA

—

—

—

...

AROCLOR-1260

NBA

—

—

—

...

PCB, TOTAL

2 / 2

...

...

2

...

INORGANICS

PERCENT LIPIDS

NBA

...

...

...

...

NBA = No Benchmark Available

Tables B-108 - B-l 19.xls B-108

-------
Table B-109

Fish Tissue Comparison to Benchmark Summary
5a Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 3

—

—

—

—

ALPHA-BHC

0 / 36

—

—

—

—

BETA-BHC

0 / 19

—

—

—

—

DELTA-BHC

0 / 18

—

—

—

—

GAMMA BHC (LINDANE)

0 / 38

—

—

—

—

ALPHA-CHLORDANE

0 / 9

—

—

—

—

GAMMA-CHLORDANE

0 / 40

—

—

—

—

CHLORPYRIFOS

0 / 37

—

—

—

—

0,P'-DDD

0 / 46

—

—

—

—

4,4'-DDD

0 / 46

—

—

—

—

0,P'-DDE

0 / 32

—

—

—

—

4,4'-DDE

16 / 46

16

—

—

—

0,P'-DDT

45 / 46

6

34

5

—

4,4'-DDT

6 / 37

6

—

—

—

DIELDRIN

0 / 45

—

—

—

—

ENDOSULFANII

0 / 42

—

—

—

—

ENDRIN

0 / 24

—

—

—

—

HEPTACHLOR

0 / 19

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 46

—

—

—

—

MIREX

0 / 16

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-109

-------
Table B-109

Fish Tissue Comparison to Benchmark Summary
5a Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 46

—

—

—

—

PENTACHLOROBENZENE

0 / 46

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 46

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 45

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

46 / 46

2

6

38

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.	Tables B-108 - B-119.xls B-109

-------
Table B-110

Fish Tissue Comparison to Benchmark Summary
5a Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALPHA-BHC

0 / 2

—

—

—

—

BETA-BHC

0 / 2

—

—

—

—

GAMMA BHC (LINDANE)

0 / 2

—

—

—

—

ALPHA-CHLORDANE

0 / 2

—

—

—

—

GAMMA-CHLORDANE

0 / 2

—

—

—

—

CHLORPYRIFOS

0 / 2

—

—

—

—

0,P'-DDD

0 / 2

—

—

—

—

4,4'-DDD

0 / 2

—

—

—

—

4,4'-DDE

1 / 2

1

—

—

—

0,P'-DDT

1 / 2

—

1

—

—

4,4'-DDT

0 / 2

—

—

—

—

DIELDRIN

0 / 2

—

—

—

—

ENDOSULFANII

0 / 2

—

—

—

—

ENDRIN

0 / 1

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 2

—

—

—

—

MIREX

0 / 1

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

PENTACHLOROANISOLE

0 / 2

—

—

—

—

PENTACHLOROBENZENE

0 / 2

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 2

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 2

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-110

-------
Table B-110

Fish Tissue Comparison to Benchmark Summary
5a Large Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

2 / 2

—

1

1

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-110

-------
Table B-lll

Fish Tissue Comparison to Benchmark Summary
5bc Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALPHA-BHC

0 / 3

—

—

—

—

BETA-BHC

0 / 3

—

—

—

—

GAMMA BHC (LINDANE)

0 / 6

—

—

—

—

ALPHA-CHLORDANE

0 / 3

—

—

—

—

GAMMA-CHLORDANE

0 / 5

—

—

—

—

CHLORPYRIFOS

0 / 5

—

—

—

—

0,P'-DDD

0 / 6

—

—

—

—

4,4'-DDD

0 / 5

—

—

—

—

4,4'-DDE

1 / 5

1

—

—

—

0,P'-DDT

6 / 6

2

4

—

—

4,4'-DDT

0 / 6

—

—

—

—

DIELDRIN

0 / 5

—

—

—

—

ENDOSULFANII

0 / 5

—

—

—

—

HEPTACHLOR

0 / 1

—

—

—

—

HEXACHLOROBENZENE

0 / 6

—

—

—

—

MIREX

0 / 1

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

PENTACHLOROANISOLE

0 / 6

—

—

—

—

PENTACHLOROBENZENE

0 / 4

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 6

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 6

—

—

—

—

DIOXINS/FURANS

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-lll

-------
Table B-lll

Fish Tissue Comparison to Benchmark Summary
5bc Small Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

6 / 6

—

4

2

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-lll

-------
Table B-112

Fish Tissue Comparison to Benchmark Summary
5bc Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 7

—

—

—

—

ALPHA-BHC

0 / 56

—

—

—

—

BETA-BHC

0 / 63

—

—

—

—

DELTA-BHC

0 / 37

—

—

—

—

GAMMA BHC (LINDANE)

0 / 94

—

—

—

—

ALPHA-CHLORDANE

0 / 31

—

—

—

—

GAMMA-CHLORDANE

0 / 66

—

—

—

—

CHLORPYRIFOS

0 / 74

—

—

—

—

0,P'-DDD

0 / 94

—

—

—

—

4,4'-DDD

0 / 92

—

—

—

—

0,P'-DDE

0 / 22

—

—

—

—

4,4'-DDE

33 / 91

33

—

—

—

0,P'-DDT

93 / 94

14

77

2

—

4,4'-DDT

23 / 82

23

—

—

—

DIELDRIN

0 / 91

—

—

—

—

ENDOSULFANII

0 / 70

—

—

—

—

ENDRIN

0 / 34

—

—

—

—

HEPTACHLOR

0 / 32

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 94

—

—

—

—

MIREX

0 / 22

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-112

-------
Table B-112

Fish Tissue Comparison to Benchmark Summary
5bc Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 93

—

—

—

—

PENTACHLOROBENZENE

0 / 94

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 94

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 94

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

94 / 94

1

39

54

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-112

-------
Table B-113

Fish Tissue Comparison to Benchmark Summary
5bc Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 3

—

—

—

—

ALPHA-BHC

0 / 21

—

—

—

—

BETA-BHC

0 / 20

—

—

—

—

DELTA-BHC

0 / 8

—

—

—

—

GAMMA BHC (LINDANE)

0 / 32

—

—

—

—

ALPHA-CHLORDANE

0 / 4

—

—

—

—

GAMMA-CHLORDANE

0 / 31

—

—

—

—

CHLORPYRIFOS

0 / 23

—

—

—

—

0,P'-DDD

0 / 32

—

—

—

—

4,4'-DDD

0 / 32

—

—

—

—

0,P'-DDE

0 / 12

—

—

—

—

4,4'-DDE

18 / 32

18

—

—

—

0,P'-DDT

32 / 32

2

19

11

—

4,4'-DDT

6 / 16

6

—

—

—

DIELDRIN

0 / 30

—

—

—

—

ENDOSULFANII

0 / 17

—

—

—

—

ENDRIN

1 / 7

1

—

—

—

HEPTACHLOR

0 / 5

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 32

—

—

—

—

MIREX

0 / 6

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-113

-------
Table B-113

Fish Tissue Comparison to Benchmark Summary
5bc Large Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 31

—

—

—

—

PENTACHLOROBENZENE

0 / 32

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 32

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 31

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

32 / 32

—

2

28

2

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-113

-------
Table B-114

Fish Tissue Comparison to Benchmark Summary
6cd Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 1

—

—

—

—

ALPHA-BHC

0 / 4

—

—

—

—

BETA-BHC

0 / 4

—

—

—

—

DELTA-BHC

0 / 1

—

—

—

—

GAMMA BHC (LINDANE)

0 / 5

—

—

—

—

ALPHA-CHLORDANE

0 / 3

—

—

—

—

GAMMA-CHLORDANE

0 / 4

—

—

—

—

CHLORPYRIFOS

0 / 1

—

—

—

—

0,P'-DDD

0 / 5

—

—

—

—

4,4'-DDD

0 / 5

—

—

—

—

4,4'-DDE

0 / 5

—

—

—

—

0,P'-DDT

5 / 5

3

2

—

—

4,4'-DDT

0 / 4

—

—

—

—

DIELDRIN

0 / 5

—

—

—

—

ENDOSULFANII

0 / 4

—

—

—

—

ENDRIN

0 / 3

—

—

—

—

HEPTACHLOR

0 / 2

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 5

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

PENTACHLOROANISOLE

0 / 4

—

—

—

—

PENTACHLOROBENZENE

0 / 5

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 5

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-114

-------
Table B-114

Fish Tissue Comparison to Benchmark Summary
6cd Small Fish
Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

1,2,4,5-TETRACHLOROBENZENE

0 / 5

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

5 / 5

—

5

—

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-114

-------
Table B-115

Fish Tissue Comparison to Benchmark Summary
6cd Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 18

—

—

—

—

ALPHA-BHC

0 / 80

—

—

—

—

BETA-BHC

0 / 72

—

—

—

—

DELTA-BHC

0 / 59

—

—

—

—

GAMMA BHC (LINDANE)

0 / 105

—

—

—

—

ALPHA-CHLORDANE

0 / 53

—

—

—

—

GAMMA-CHLORDANE

0 / 76

—

—

—

—

CHLORPYRIFOS

0 / 81

—

—

—

—

0,P'-DDD

0 / 108

—

—

—

—

4,4'-DDD

0 / 108

—

—

—

—

0,P'-DDE

0 / 28

—

—

—

—

4,4'-DDE

47 / 108

46

1

—

—

0,P'-DDT

106 / 108

16

83

7

—

4,4'-DDT

3 / 63

3

—

—

—

DIELDRIN

2 / 99

2

—

—

—

ENDOSULFANII

0 / 78

—

—

—

—

ENDRIN

0 / 30

—

—

—

—

HEPTACHLOR

0 / 39

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 105

—

—

—

—

MIREX

0 / 7

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-115

-------
Table B-115

Fish Tissue Comparison to Benchmark Summary
6cd Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 103

—

—

—

—

PENTACHLOROBENZENE

0 / 106

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 106

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 108

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

108 / 108

—

52

56

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-115

-------
Table B-116

Fish Tissue Comparison to Benchmark Summary
6cd Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 6

—

—

—

—

ALPHA-BHC

0 / 22

—

—

—

—

BETA-BHC

0 / 21

—

—

—

—

DELTA-BHC

0 / 15

—

—

—

—

GAMMA BHC (LINDANE)

0 / 36

—

—

—

—

ALPHA-CHLORDANE

0 / 8

—

—

—

—

GAMMA-CHLORDANE

0 / 27

—

—

—

—

CHLORPYRIFOS

0 / 24

—

—

—

—

0,P'-DDD

0 / 36

—

—

—

—

4,4'-DDD

0 / 36

—

—

—

—

0,P'-DDE

0 / 22

—

—

—

—

4,4'-DDE

25 / 36

24

1

—

—

0,P'-DDT

36 / 36

1

21

14

—

4,4'-DDT

2 / 15

2

—

—

—

DIELDRIN

4 / 35

4

—

—

—

ENDOSULFANII

0 / 25

—

—

—

—

ENDRIN

0 / 8

—

—

—

—

HEPTACHLOR

0 / 7

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 36

—

—

—

—

MIREX

0 / 7

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-116

-------
Table B-116

Fish Tissue Comparison to Benchmark Summary
6cd Large Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 35

—

—

—

—

PENTACHLOROBENZENE

0 / 32

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 32

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 33

—

—

—

—

DIOXINS/FURANS

TOTAL DIOXINS (USING 0)

NBA

—

—

—

—

TOTAL DIOXINS (USING DL)

NBA

—

—

—

—

TOTAL DIOXINS (USING HALF DL)

NBA

—

—

—

—

TOTAL FURANS (USING 0)

NBA

—

—

—

—

TOTAL FURANS (USING DL)

NBA

—

—

—

—

TOTAL FURANS (USING HALF DL)

NBA

—

—

—

—

PCBS

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

39 / 39

—

4

34

1

METALS

LEAD

0 / 3

—

—

—

—

MERCURY

6 / 6

—

6

—

—

NICKEL

0 / 1

—

—

—

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-116

-------
Table B-117

Fish Tissue Comparison to Benchmark Summary
Background - Small Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 8

—

—

—

—

ALPHA-BHC

0 / 20

—

—

—

—

BETA-BHC

0 / 12

—

—

—

—

DELTA-BHC

0 / 5

—

—

—

—

GAMMA BHC (LINDANE)

0 / 18

—

—

—

—

ALPHA-CHLORDANE

0 / 19

—

—

—

—

GAMMA-CHLORDANE

0 / 19

—

—

—

—

CHLORPYRIFOS

0 / 12

—

—

—

—

0,P'-DDD

0 / 15

—

—

—

—

4,4'-DDD

0 / 20

—

—

—

—

0,P'-DDE

0 / 2

—

—

—

—

4,4'-DDE

0 / 20

—

—

—

—

0,P'-DDT

1 / 8

1

—

—

—

4,4'-DDT

0 / 19

—

—

—

—

DIELDRIN

0 / 16

—

—

—

—

ENDOSULFANII

0 / 18

—

—

—

—

ENDRIN

0 / 12

—

—

—

—

HEPTACHLOR

0 / 18

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 20

—

—

—

—

MIREX

0 / 10

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-117

-------
Table B-117

Fish Tissue Comparison to Benchmark Summary
Background - Small Fish
Housatonic River Site, OU2 - Pittsfield, MA





Number of Samples where

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 20

—

—

—

—

PENTACHLOROBENZENE

0 / 19

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 19

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 19

—

—

—

—

DIOXINS/FURANS

1,2,3,4,6,7,8-HPCDF

NBA

—

—

—

—

1,2,3,4,7,8,9-HPCDF

NBA

—

—

—

—

1,2,3,4,7,8-HXCDF

NBA

—

—

—

—

1,2,3,6,7,8-HXCDF

NBA

—

—

—

—

1,2,3,7,8-PECDF

NBA

—

—

—

—

2,3,4,6,7,8-HXCDF

NBA

—

—

—

—

2,3,4,7,8-PECDF

NBA

—

—

—

—

2,3,7,8-TCDD

NBA

—

—

—

—

2,3,7,8-TCDF

NBA

—

—

—

—

OCDD

NBA

—

—

—

—

OCDF

NBA

—

—

—

—

PCBS

AROCLOR-1242

NBA

—

—

—

—

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

1 / 20

1

—

—

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-117

-------
Table B-118

Fish Tissue Comparison to Benchmark Summary
Background - Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 61

—

—

—

—

ALPHA-BHC

0 / 106

—

—

—

—

BETA-BHC

0 / 90

—

—

—

—

DELTA-BHC

0 / 72

—

—

—

—

GAMMA BHC (LINDANE)

0 / 97

—

—

—

—

ALPHA-CHLORDANE

0 / 92

—

—

—

—

GAMMA-CHLORDANE

0 / 90

—

—

—

—

CHLORPYRIFOS

0 / 66

—

—

—

—

0,P'-DDD

0 / 109

—

—

—

—

4,4'-DDD

0 / 114

—

—

—

—

0,P'-DDE

0 / 56

—

—

—

—

4,4'-DDE

5 / 114

5

—

—

—

0,P'-DDT

0 / 39

—

—

—

—

4,4'-DDT

1 / 110

1

—

—

—

DIELDRIN

0 / 100

—

—

—

—

ENDOSULFANII

0 / 101

—

—

—

—

ENDRIN

0 / 58

—

—

—

—

HEPTACHLOR

0 / 96

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 113

—

—

—

—

MIREX

0 / 63

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-118

-------
Table B-118

Fish Tissue Comparison to Benchmark Summary
Background - Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 109

—

—

—

—

PENTACHLOROBENZENE

0 / 104

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 104

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 101

—

—

—

—

DIOXINS/FURANS

1,2,3,4,6,7,8-HPCDD

NBA

—

—

—

—

1,2,3,4,6,7,8-HPCDF

NBA

—

—

—

—

1,2,3,4,7,8,9-HPCDF

NBA

—

—

—

—

1,2,3,4,7,8-HXCDF

NBA

—

—

—

—

1,2,3,6,7,8-HXCDD

NBA

—

—

—

—

1,2,3,7,8,9-HXCDD

NBA

—

—

—

—

1,2,3,7,8-PECDF

NBA

—

—

—

—

2,3,4,6,7,8-HXCDF

NBA

—

—

—

—

2,3,4,7,8-PECDF

NBA

—

—

—

—

2,3,7,8-TCDD

NBA

—

—

—

—

2,3,7,8-TCDF

NBA

—

—

—

—

OCDD

NBA

—

—

—

—

OCDF

NBA

—

—

—

—

PCBS

AROCLOR-1242

NBA

—

—

—

—

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

PCB, TOTAL

1 / 114

1

—

—

—

METALS

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-118

-------
Table B-118

Fish Tissue Comparison to Benchmark Summary
Background - Medium Fish
Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

MERCURY

1 / 1

—

1

—

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-l 19.xls B-118

-------
Table B-119

Fish Tissue Comparison to Benchmark Summary
Background - Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

APP IX PESTICIDES

ALDRIN

0 / 10

—

—

—

—

ALPHA-BHC

0 / 10

—

—

—

—

BETA-BHC

0 / 12

—

—

—

—

DELTA-BHC

0 / 8

—

—

—

—

GAMMA BHC (LINDANE)

0 / 12

—

—

—

—

ALPHA-CHLORDANE

0 / 6

—

—

—

—

GAMMA-CHLORDANE

0 / 3

—

—

—

—

CHLORPYRIFOS

0 / 14

—

—

—

—

0,P'-DDD

0 / 13

—

—

—

—

4,4'-DDD

0 / 15

—

—

—

—

0,P'-DDE

0 / 7

—

—

—

—

4,4'-DDE

2 / 15

2

—

—

—

0,P'-DDT

0 / 6

—

—

—

—

4,4'-DDT

0 / 15

—

—

—

—

DIELDRIN

0 / 15

—

—

—

—

ENDOSULFANII

0 / 10

—

—

—

—

ENDRIN

0 / 5

—

—

—

—

HEPTACHLOR

0 / 9

—

—

—

—

HEPTACHLOR EPOXIDE

NBA

—

—

—

—

HEXACHLOROBENZENE

0 / 11

—

—

—

—

MIREX

0 / 5

—

—

—

—

CIS-NONACHLOR

NBA

—

—

—

—

TRANS-NONACHLOR

NBA

—

—

—

—

OXYCHLORDANE

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-119.xls B-119

-------
Table B-119

Fish Tissue Comparison to Benchmark Summary
Background - Large Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Frequency of Detected
Samples Exceeding
Benchmark

Number of Samples where

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PENTACHLOROANISOLE

0 / 9

—

—

—

—

PENTACHLOROBENZENE

0 / 9

—

—

—

—

1,2,3,4-TETRACHLOROBENZENE

0 / 13

—

—

—

—

1,2,4,5-TETRACHLOROBENZENE

0 / 10

—

—

—

—

DIOXINS/FURANS

1,2,3,4,6,7,8-HPCDD

NBA

—

—

—

—

1,2,3,4,6,7,8-HPCDF

NBA

—

—

—

—

1,2,3,4,7,8,9-HPCDF

NBA

—

—

—

—

1,2,3,4,7,8-HXCDF

NBA

—

—

—

—

1,2,3,6,7,8-HXCDD

NBA

—

—

—

—

1,2,3,6,7,8-HXCDF

NBA

—

—

—

—

1,2,3,7,8,9-HXCDF

NBA

—

—

—

—

1,2,3,7,8-PECDD

NBA

—

—

—

—

1,2,3,7,8-PECDF

NBA

—

—

—

—

2,3,4,6,7,8-HXCDF

NBA

—

—

—

—

2,3,4,7,8-PECDF

NBA

—

—

—

—

2,3,7,8-TCDD

NBA

—

—

—

—

2,3,7,8-TCDF

NBA

—

—

—

—

OCDD

NBA

—

—

—

—

OCDF

NBA

—

—

—

—

PCBS

AROCLOR-1242

NBA

—

—

—

—

AROCLOR-1248

NBA

—

—

—

—

AROCLOR-1254

NBA

—

—

—

—

AROCLOR-1260

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-119.xls B-119

-------
Table B-119

Fish Tissue Comparison to Benchmark Summary
Background - Large Fish
Housatonic River Site, OU2 - Pittsfield, MA







Number

of Samples where



Chemical

Frequency of Detected
Samples Exceeding
Benchmark

1<= HQ <10

10<= HQ <100

100<= HQ <1000

HQ >=1000

PCB, TOTAL

0 / 15

—

—

—

—

METALS

MERCURY

5 / 5

—

5

—

—

NICKEL

0 / 2

—

—

—

—

INORGANICS

PERCENT LIPIDS

NBA

—

—

—

—

NBA = No benchmark available.

Tables B-108 - B-119.xls B-119

-------
Table B-120

Analysis Options for Sediment Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Background

Reach 5a, MC

Reach 5a, SCOX

Reach 5a, VP

Number of
Detects

Number of
Samples

% Detects

Number of
Detects

Number of
Samples

% Detects

Analysis

Number of
Detects

Number of
Samples

% Detects

Analysis

Number of
Detects

Number of
Samples

% Detects

Analysis

SEMIVOLATILES









BIS(2-ETHYLHEXYL)PHTHALATE

9 | 20 | 45.0%

... | ... | ... | ...

... | ... | ... | ...

... | ... | ... | ...

PAHS









ACENAPHTHENE

1

20

5.0%

16

22

73%

2

...

...

...

...

...

...

...

...

ACENAPHTHYLENE



20

0.0%

20

22

91%

2

...

...

...

...

4

9

44.4%

2

ANTHRACENE

8

20

40.0%

22

22

100%

1

...

...

...

...

4

9

44.4%

1

BENZO(A)ANTHRACENE

11

20

55.0%

22

22

100%

1

1

2

50.0%

3

6

9

66.7%

1

BENZO(A)PYRENE

11

20

55.0%

22

22

100%

1

...

...

...

...

5

9

55.6%

1

BENZO(B)FLUORANTHENE

11

20

55.0%

22

22

100%

1

1

2

50.0%

3

5

9

55.6%

1

BENZO(GHI)PERYLENE

11

20

55.0%

22

22

100%

1

...

...

...

...

5

9

55.6%

1

BENZO(K)FLUORANTHENE

11

20

55.0%

22

22

100%

1

1

2

50.0%

3

5

9

55.6%

1

CHRYSENE

12

20

60.0%

22

22

100%

1

...

...

...

...

6

9

66.7%

1

DIBENZO(A,H) ANTHRACENE

8

20

40.0%

20

22

91%

1

1

2

50.0%

3

4

9

44.4%

1

FLUORANTHENE

14

20

70.0%

22

22

100%

1

...

...

...

...

7

9

77.8%

1

FLUORENE

6

20

30.0%

19

22

86%

1

...

...

...

...

...

...

...

...

INDENO(l,2,3-C,D)PYRENE

11

20

55.0%

22

22

100%

1

1

2

50.0%

3

5

9

55.6%

1

NAPHTHALENE

5

20

25.0%

21

22

95%



...

...

...

...

...

...

...

...

PHENANTHRENE

13

20

65.0%

22

22

100%

1

1

2

50.0%

3

6

9

66.7%

1

PYRENE

13

20

65.0%

22

22

100%

1

1

2

50.0%

3

6

9

66.7%

1

PESTICIDES









4,4'-DDD

3

21

14.3%

1

20

5%

2

...

...

...

...

2

12

16.7%

2

4,4'-DDE

2

20

10.0%

...

...

...

...

...

...

...

...

5

13

38.5%

2

METALS









ANTIMONY

13

20

65.0%

12

22

55%

1

...

...

...

...

4

9

44.4%

1

ARSENIC

15

20

75.0%

...

...

...

...

...

...

...

...

...

...

...

...

BARIUM

20

20

100.0%

22

22

100%

1

2

2

100.0%

3

9

9

100.0%

1

BERYLLIUM

19

20

95.0%

17

22

77%

1

2

2

100.0%

3

9

9

100.0%

1

CADMIUM

6

20

30.0%

...

...

...

...

...

...

...

...

4

9

44.4%

1

CHROMIUM

20

20

100.0%

22

22

100%

1

...

...

...

...

9

9

100.0%

1

COBALT

20

20

100.0%

1

1

100%

3

...

...

...

...

...

...

...

...

COPPER

20

20

100.0%

22

22

100%

1

...

...

...

...

9

9

100.0%

1

LEAD

20

20

100.0%

22

22

100%

1

2

2

100.0%

3

9

9

100.0%

1

MERCURY

9

20

45.0%

17

21

81%

1

...

...

...

...

6

9

66.7%

1

NICKEL

20

20

100.0%

20

22

91%

1

2

2

100.0%

3

9

9

100.0%

1

SELENIUM

1

20

5.0%

0

5

0%

2

...

...

...

...

1

9

11.1%

2

SILVER

5

20

25.0%

...

...

...

...

...

...

...

...

3

9

33.3%

2

THALLIUM

11

20

55.0%

9

22

41%

1

...

...

...

...

2

9

22.2%

3

TIN

6

20

30.0%

7

22

32%

1

1

2

50.0%

3

4

9

44.4%

1

VANADIUM

20

20

100.0%

22

22

100%

1

2

2

100.0%

3

9

9

100.0%

1

ZINC

20

20

100.0%

22

22

100%

1

...

...

...

...

9

9

100.0%

1

Analysis Options:

1	= Statistical Comparison

2	=Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xls B-120

-------
Table B-120, continued

Analysis Options for Sediment Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

Reach 5b, MC

Reach 5b,

scox



Reach 5b, VP



Number of

Number of



Number of

Number of





Number of

Number of





Number of

Number of





Chemical

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

SEMIVOLATILES









BIS(2-ETHYLHEXYL)PHTHALATE

9

20

45.0%

... | ... | ... | ...

... | ... | ... | ...

... | ... | ... | ...

PAHS









ACENAPHTHENE

1

20

5.0%

...

...

...

...

...

...

...

...

...

...

...

...

ACENAPHTHYLENE



20

0.0%

...

...

...

...

...

...

...

...

4

8

50.0%

2

ANTHRACENE

8

20

40.0%

4

6

66.7%

1

1

1

100%

3

4

8

50.0%

1

BENZO(A)ANTHRACENE

11

20

55.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

BENZO(A)PYRENE

11

20

55.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

BENZO(B)FLUORANTHENE

11

20

55.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

BENZO(GHI)PERYLENE

11

20

55.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

BENZO(K)FLUORANTHENE

11

20

55.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

CHRYSENE

12

20

60.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

DIBENZO(A,H) ANTHRACENE

8

20

40.0%

5

6

83.3%

1

1

1

100%

3

3

8

37.5%

1

FLUORANTHENE

14

20

70.0%

6

6

100.0%

1

1

1

100%

3

7

8

87.5%

1

FLUORENE

6

20

30.0%

4

6

66.7%

1

...

...

...

...

...

...

...

...

INDENO(l,2,3-C,D)PYRENE

11

20

55.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

NAPHTHALENE

5

20

25.0%

...

...

...

...

...

...

...

...

...

...

...

...

PHENANTHRENE

13

20

65.0%

5

6

83.3%

1

1

1

100%

3

7

8

87.5%

1

PYRENE

13

20

65.0%

6

6

100.0%

1

1

1

100%

3

8

9

88.9%

1

PESTICIDES









4,4'-DDD

3

21

14.3%

...

...

...

...

...

...

...

...

1

10

10.0%

2

4,4'-DDE

2

20

10.0%

...

...

...

...

...

...

...

...

1

8

12.5%

2

METALS









ANTIMONY

13

20

65.0%

3

6

50.0%

1

...

...

...

...

6

9

66.7%

1

ARSENIC

15

20

75.0%

...

...

...

...

...

...

...

...

7

9

77.8%

1

BARIUM

20

20

100.0%

6

6

100.0%

1

1

1

100%

3

9

9

100.0%

1

BERYLLIUM

19

20

95.0%

3

6

50.0%

1

1

1

100%

3

9

9

100.0%

1

CADMIUM

6

20

30.0%

...

...

...

...

...

...

...

...

5

8

62.5%

1

CHROMIUM

20

20

100.0%

6

6

100.0%

1

1

1

100%

3

9

9

100.0%

1

COBALT

20

20

100.0%

6

6

100.0%

1

...

...

...

...

9

9

100.0%

1

COPPER

20

20

100.0%

6

6

100.0%

1

1

1

100%

3

9

9

100.0%

1

LEAD

20

20

100.0%

6

6

100.0%

1

1

1

100%

3

9

9

100.0%

1

MERCURY

9

20

45.0%

5

5

100.0%

1

1

1

100%

3

9

9

100.0%

1

NICKEL

20

20

100.0%

...

...

...

...

1

1

100%

3

9

9

100.0%

1

SELENIUM

1

20

5.0%

...

...

...

...

...

...

...

...

1

8

12.5%



SILVER

5

20

25.0%

2

6

33.3%

2

1

1

100%

2

7

8

87.5%



THALLIUM

11

20

55.0%

4

6

66.7%

1

1

1

100%

3

5

9

55.6%

1

TIN

6

20

30.0%

4

6

66.7%

1

...

...

...

...

5

8

62.5%

1

VANADIUM

20

20

100.0%

6

6

100.0%

1

1

1

100%

3

9

9

100.0%

1

ZINC

20

20

100.0%

6

6

100.0%

1

1

1

100%

3

9

9

100.0%

1

Analysis Options:

1	= Statistical Comparison

2	=Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xls B-120

-------
Table B-120, continued

Analysis Options for Sediment Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

Reach 5c, MC



Reach 5c,

scox



Reach 5c, VP



Number of

Number of



Number of

Number of





Number of

Number of





Number of

Number of





Chemical

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

SEMIVOLATILES









BIS(2-ETHYLHEXYL)PHTHALATE

9

20

45.0%

8

13

62% | 1

... | ... | ... | ...

... | ... | ... | ...

PAHS









ACENAPHTHENE

1

20

5.0%

4

13

30.8%

2

...

...

...

...

...

...

...

...

ACENAPHTHYLENE



20

0.0%

6

13

46.2%

2

...

...

...

...

3

16

18.8%

2

ANTHRACENE

8

20

40.0%

7

13

53.8%

1

...

...

...

...

3

15

20.0%

2

BENZO(A)ANTHRACENE

11

20

55.0%

13

13

100.0%

1

...

...

...

...

12

16

75.0%

1

BENZO(A)PYRENE

11

20

55.0%

13

13

100.0%

1

...

...

...

...

8

15

53.3%

1

BENZO(B)FLUORANTHENE

11

20

55.0%

13

13

100.0%

1

...

...

...

...

11

16

68.8%

1

BENZO(GHI)PERYLENE

11

20

55.0%

13

13

100.0%

1

...

...

...

...

11

16

68.8%

1

BENZO(K)FLUORANTHENE

11

20

55.0%

13

13

100.0%

1

...

...

...

...

11

16

68.8%

1

CHRYSENE

12

20

60.0%

13

13

100.0%

1

...

...

...

...

12

17

70.6%

1

DIBENZO(A,H) ANTHRACENE

8

20

40.0%

10

13

76.9%

1

...

...

...

...

...

...

...

...

FLUORANTHENE

14

20

70.0%

13

13

100.0%

1

...

...

...

...

11

15

73.3%

1

FLUORENE

6

20

30.0%

5

13

38.5%

1

...

...

...

...

2

15

13.3%

2

INDENO(l,2,3-C,D)PYRENE

11

20

55.0%

13

13

100.0%

1

...

...

...

...

11

16

68.8%

1

NAPHTHALENE

5

20

25.0%

11

13

84.6%



...

...

...

...

9

17

52.9%

2

PHENANTHRENE

13

20

65.0%

11

13

84.6%

1

...

...

...

...

12

16

75.0%

1

PYRENE

13

20

65.0%

13

13

100.0%

1

...

...

...

...

13

16

81.3%

1

PESTICIDES









4,4'-DDD

3

21

14.3%

...

...

...

...

1

2

50%

2

1

15

6.7%

2

4,4'-DDE

2

20

10.0%

...

...

...

...

1

2

50%

2

...

...

...

...

METALS









ANTIMONY

13

20

65.0%

9

13

69.2%

1

...

...

...

...

8

15

53.3%

1

ARSENIC

15

20

75.0%

11

13

84.6%

1

2

2

100%

3

13

16

81.3%

1

BARIUM

20

20

100.0%

13

13

100.0%

1

2

2

100%

3

17

17

100.0%

1

BERYLLIUM

19

20

95.0%

12

13

92.3%

1

2

2

100%

3

17

17

100.0%

1

CADMIUM

6

20

30.0%

10

13

76.9%

1

...

...

...

...

14

16

87.5%

1

CHROMIUM

20

20

100.0%

13

13

100.0%

1

...

...

...

...

16

17

94.1%

1

COBALT

20

20

100.0%

13

13

100.0%

1

...

...

...

...

...

...

...

...

COPPER

20

20

100.0%

13

13

100.0%

1

2

2

100%

3

17

17

100.0%

1

LEAD

20

20

100.0%

13

13

100.0%

1

2

2

100%

3

17

17

100.0%

1

MERCURY

9

20

45.0%

10

13

76.9%

1

2

2

100%

3

15

17

88.2%

1

NICKEL

20

20

100.0%

13

13

100.0%

1

2

2

100%

3

17

17

100.0%

1

SELENIUM

1

20

5.0%

3

13

23.1%



...

...

...

...

8

16

50.0%

2

SILVER

5

20

25.0%

10

13

76.9%



...

...

...

...

10

16

62.5%

2

THALLIUM

11

20

55.0%

7

13

53.8%

1

2

2

100%

3

7

15

46.7%

1

TIN

6

20

30.0%

7

13

53.8%

1

2

2

100%

3

11

16

68.8%

1

VANADIUM

20

20

100.0%

13

13

100.0%

1

2

2

100%

3

17

17

100.0%

1

ZINC

20

20

100.0%

13

13

100.0%

1

2

2

100%

3

17

17

100.0%

1

Analysis Options:

1	= Statistical Comparison

2	=Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xls B-120

-------
Table B-120, continued

Analysis Options for Sediment Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

Reach 6ab, MC

Reach 6cd, POND



Number of

Number of



Number of

Number of





Number of

Number of





Chemical

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

SEMIVOLATILES







BIS(2-ETHYLHEXYL)PHTHALATE

9

20

45.0%

... | ... | ... | ...

... | ... | ... | ...

PAHS







ACENAPHTHENE

1

20

5.0%

...

...

...

...

...

...

...

...

ACENAPHTHYLENE



20

0.0%

...

...

...

...

...

...

...

...

ANTHRACENE

8

20

40.0%

...

...

...

...

1

1

100%

3

BENZO(A)ANTHRACENE

11

20

55.0%

2

2

100%

3

1

1

100%

3

BENZO(A)PYRENE

11

20

55.0%

2

2

100%

3

1

1

100%

3

BENZO(B)FLUORANTHENE

11

20

55.0%

2

2

100%

3

1

1

100%

3

BENZO(GHI)PERYLENE

11

20

55.0%

2

2

100%

3

1

1

100%

3

BENZO(K)FLUORANTHENE

11

20

55.0%

2

2

100%

3

1

1

100%

3

CHRYSENE

12

20

60.0%

2

2

100%

3

1

1

100%

3

DIBENZO(A,H)ANTHRACENE

8

20

40.0%

1

2

50%

3

1

1

100%

3

FLUORANTHENE

14

20

70.0%

2

2

100%

3

1

1

100%

3

FLUORENE

6

20

30.0%

...

...

...

...

...

...

...

...

INDENO(l,2,3-C,D)PYRENE

11

20

55.0%

2

2

100%

3

1

1

100%

3

NAPHTHALENE

5

20

25.0%

2

2

100%

2

...

...

...

...

PHENANTHRENE

13

20

65.0%

2

2

100%

3

1

1

100%

3

PYRENE

13

20

65.0%

2

2

100%

3

1

1

100%

3

PESTICIDES







4,4'-DDD

3

21

14.3%

...

...

...

...

...

...

...

...

4,4'-DDE

2

20

10.0%

...

...

...

...

...

...

...

...

METALS







ANTIMONY

13

20

65.0%

2

2

100%

3

1

1

100%

3

ARSENIC

15

20

75.0%

...

...

...

...

...

...

...

...

BARIUM

20

20

100.0%

2

2

100%

3

1

1

100%

3

BERYLLIUM

19

20

95.0%

2

2

100%

3

1

1

100%

3

CADMIUM

6

20

30.0%

2

2

100%

3

1

1

100%

3

CHROMIUM

20

20

100.0%

2

2

100%

3

1

1

100%

3

COBALT

20

20

100.0%

...

...

...

...

...

...

...

...

COPPER

20

20

100.0%

2

2

100%

3

1

1

100%

3

LEAD

20

20

100.0%

2

2

100%

3

1

1

100%

3

MERCURY

9

20

45.0%

2

2

100%

3

1

1

100%

3

NICKEL

20

20

100.0%

2

2

100%

3

1

1

100%

3

SELENIUM

1

20

5.0%

...

...

...

...

1

1

100%

2

SILVER

5

20

25.0%

2

2

100%

2

1

1

100%

2

THALLIUM

11

20

55.0%

2

2

100%

3

...

...

...

...

TIN

6

20

30.0%

2

2

100%

3

1

1

100%

3

VANADIUM

20

20

100.0%

2

2

100%

3

1

1

100%

3

ZINC

20

20

100.0%

2

2

100%

3

1

1

100%

3

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xls B-120

-------
Table B-121

Background Comparison Results - Sediment
Housatonic River Site, OU2, Pittsfield, MA

Chemical | Reach 5a, MC | Reach 5a, SCOX | Reach 5a, VP | Reach 5b, MC | Reach 5b, SCOX | Reach 5b, VP | Reach 5c, MC | Reach 5c, SCOX | Reach 5c, VP | Reach 6ab, MC | Reach 6cd, POND

Semivolatiles

BIS(2-ETHYLHEXYL)PHTHALATE | — | — | — | — | — | — | NS | — | — | — |

PAHs

ACENAPHTHENE

U











U









ACENAPTHYLENE

u



u





u

U



u





ANTHRACENE

NS



NS

NS

WP

NS

NS



u



WP

BENZO(A)ANTHRACENE

NS

WP

NS

NS

WP

NS

NS



NS

WP

WP

BENZO(A)PYRENE

NS



NS

NS

WP

NS

NS



NS

WP

WP

BENZO(B)FLUORANTHENE

NS

WP

NS

NS

WP

NS

NS



NS

WP

WP

BENZO(GHI)PERYLENE

NS



NS

NS

WP

NS

NS



NS

WP

WP

BENZO(K)FLUORANTHENE

NS

WP

NS

NS

WP

NS

NS



NS

WP

WP

CHRYSENE

NS



NS

NS

WP

NS

NS



NS

WP

WP

DIBENZO(A,H)ANTHRACENE

NS

WP

NS

NS

WP

NS

NS





WP

WP

FLUORANTHENE

D



NS

NS

WP

NS

NS



NS

WP

WP

FLUORENE

NS





NS





NS



U





INDENO(l,2,3-C,D)PYRENE

NS

WP

NS

NS

WP

NS

NS



NS

WP

WP

NAPHTHALENE

U











U



U

u



PHENANTHRENE

Da

WP

NS

NS

WP

NS

NS



NS

WP

WP

PYRENE

D

WP

NS

NS

WP

NS

NS



NS

WP

WP

Pesticides

4,4'-DDD

U



U





U



U

U





4,4'-DDE





U





U



U







Metals

ANTIMONY

NS



NS

NS



NS

D



D

WP

WP

ARSENIC











NS

NS

P

D





BARIUM

NS

WP

NS

NS

WP

NS

NS

WP

D

WP

WP

BERYLLIUM

NS

WP

NS

NS

WP

NS

NS

WP

D

WP

WP

CADMIUM





NS





D

D



D

p

p

CHROMIUM

NS



NS

NS

pb

D

D



D

p

p

COBALT

WP





NS



D

D









COPPER

NS



D

NS

WP

D

D

WP

D

p

p

LEAD

NS

WP

D

NS

WP

D

D

WP

D

WP

p

MERCURY

NS



D

NS

p

D

D

p

D

p

p

NICKEL

NS

p

D



p

D

D

pc

D

p

p

SELENIUM

U



U





U

U



U



u

SILVER





U

U

u

U

U



U

u

u

THALLIUM

NS



WP

NS

WP

NS

NS

p

NS

WP



TIN

NS

WP

NS

NS



D

D

WP

D

p

p

VANADIUM

NS

WP

NS

NS

WP

NS

NS

WP

NS

WP

WP

ZINC

NS



D

NS

p

D

D

WP

D

p

p

Results Options:

D = Significant using Dunnett's Test	U = Too many non-detects to determine using statistics

NS = Not significant using Dunnett's Test	WP = Within prediction interval

P = Outside of prediction interval	— No comparison made

a5a MC not signifintaly different using nonparametric comparisons, but significantly different using lognormal comparisons (data do not conform to lognormal model).

5b SCOX within background using a nonparametric prediction interval, but outside the lognormal prediction interval (data do not conform to lognormal model).
C5c SCOX within background using a lognormal prediction interval, but outside a nonparametric prediction interval (data do not conform to lognormal model).

Tables B-120 - B-127.xlsB-121

-------
Table B-122

Analysis Options for Surface Water Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

Reach 5a, MC

Reach 5b, MC



Number of

Number of



Number of

Number of





Number of

Number of





Chemical

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Volatiles





Vinyl Chloride

0

11

0%

4

10

40%

2

—

Semivolatiles





Bis(2-ethylhexyl)phthalate

1

34

3%

2

32

6%

2

—

PAHs





Fluoranthene

4

34

12%

—

—

—

—

5

29

17%

2

Pyrene

4

34

12%

5

32

16%

2

5

29

17%

2

Metals





Barium

33

34

97%

32

32

100%

1

29

29

100%

1

Cadmium

0

34

0%

2

32

6%

2

—

—

—

—

Copper

4

34

12%

6

32

19%

2

17

29

59%

2

Lead

4

34

12%

4

32

13%

2

4

29

14%

2

Silver

1

34

3%

—

—

—

—

2

29

7%

2

Zinc

15

34

44%

12

32

38%

1

11

29

38%

1

Metals, dissolved





Beryllium, dissolved

0

30

0%

1

30

3%

2

2

29

7%

2

Cadmium, dissolved

1

30

3%

1

30

3%

2

—

—

—

—

Cobalt, dissolved

1

30

3%

2

30

7%

2

—

—

—

—

Copper, dissolved

1

30

3%

3

30

10%

2

13

29

45%

2

Silver, dissolved

0

30

0%

3

30

10%

2

...

...

...

...

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-122

-------
Table B-122, continued

Analysis Options for Surface Water Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

Reach 5b, VP

Reach 5c, MC



Number of

Number of



Number of

Number of





Number of

Number of





Chemical

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Volatiles





Vinyl Chloride

0

11

0%

—

—

Semivolatiles





Bis(2-ethylhexyl)phthalate

1

34

3%

—

3

16

19%

2

PAHs





Fluoranthene

4

34

12%

—

—

—

—

—

—

—

—

Pyrene

4

34

12%

—

—

—

—

2

16

13%

2

Metals





Barium

33

34

97%

1

1

100%

3

15

16

94%

1

Cadmium

0

34

0%

—

—

—

—

—

—

—

—

Copper

4

34

12%

—

—

—

—

12

16

75%

2

Lead

4

34

12%

1

1

100%

2

—

—

—

—

Silver

1

34

3%

—

—

—

—

—

—

—

—

Zinc

15

34

44%

—

—

—

—

—

—

—

—

Metals, dissolved





Beryllium, dissolved

0

30

0%

—

—

—

—

1

14

7%

2

Cadmium, dissolved

1

30

3%

—

—

—

—

—

—

—

—

Cobalt, dissolved

1

30

3%

—

—

—

—

1

14

7%

2

Copper, dissolved

1

30

3%

—

—

—

—

—

—

—

—

Silver, dissolved

0

30

0%

...

...

...

...

...

...

...

...

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-122

-------
Table B-122, continued

Analysis Options for Surface Water Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

Reach 5c, VP

Reach 6cd Pond



Number of

Number of



Number of

Number of





Number of

Number of





Chemical

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Volatiles





Vinyl Chloride

0

11

0%

—

—

Semivolatiles





Bis(2-ethylhexyl)phthalate

1

34

3%

—

—

PAHs





Fluoranthene

4

34

12%

—

—

—

—

—

—

—

—

Pyrene

4

34

12%

—

—

—

—

1

13

8%

2

Metals





Barium

33

34

97%

3

4

75%

1

13

13

100%

1

Cadmium

0

34

0%

—

—

—

—

1

13

8%

2

Copper

4

34

12%

—

—

—

—

—

—

—

—

Lead

4

34

12%

—

—

—

—

—

—

—

—

Silver

1

34

3%

—

—

—

—

1

13

8%

2

Zinc

15

34

44%

—

—

—

—

3

13

23%

3

Metals, dissolved





Beryllium, dissolved

0

30

0%

—

—

—

—

—

—

—

—

Cadmium, dissolved

1

30

3%

—

—

—

—

1

14

7%

2

Cobalt, dissolved

1

30

3%

—

—

—

—

—

—

—

—

Copper, dissolved

1

30

3%

—

—

—

—

—

—

—

—

Silver, dissolved

0

30

0%

...

...

...

...

...

...

...

...

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-122

-------
Table B-123

Analysis Options for Surface Water Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Reach 5a, MC

Reach 5b, MC

Reach 5b, VP

Reach 5c, MC

Reach 5c, VP

Reach 6cd Pond

Volatiles

Vinyl Chloride

U

—

—

—

—

—

Semivolatiles

Bis(2-ethylhexyl)phthalate

U

—

—

U

—

—

PAHs

Fluoranthene

—

U

—

—

—

—

Pyrene

u

u

—

u

—

U

Metals

Barium

NS

NS

WP

NS

NS

NS

Cadmium

u

—

—

—

—

U

Copper

u

u

—

u

—

—

Lead

u

u

u

—

—

—

Silver

—

u

—

—

—

u

Zinc

NS

D*

—

—

—

WP

Metals, dissolved

Beryllium, dissolved

u

u

—

u

—

—

Cadmium, dissolved

u

—

—

—

—

u

Cobalt, dissolved

u

—

—

u

—

—

Copper, dissolved

u

u

—

—

—

—

Silver, dissolved

u

...

...

...

...

...

Results Options:

D = Significant using Dunnett's Test	U = Too many non-detects to determine using statistics

NS = Not significant using Dunnett's Test WP = Within prediction interval
P = Outside of prediction interval	— No comparison made

*5b MC signifintaly different using nonparametric comparisons, but not significantly different using lognormal comparisons (data do
not conform to lognormal model).

Tables B-120 - B-127.xlsB-123

-------
Table B-124

Analysis Options for Soil Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

5a FP

5a RB



Number of

Number of



Number of

Number of





Number of

Number of





CAPTION

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Semivolatiles





BUTYLBENZYLPHTHALATE

0

20

0%

2

23

9%

2

2

10

20%

2

DIBENZOFURAN

0

20

0%

9

23

39%

2

8

10

80%

2

4-METHYLPHENOL

0

20

0%

...

...

...

...

...

...

...

...

N-NITROSO-DI-N-BUTYLAMINE

0

20

0%

...

...

...

...

...

...

...

...

P-PHENYLENEDIAMINE

0

11

0%

...

...

...

...

...

...

...

...

PAHs





BENZO(A)ANTHRACENE

10

20

50%

...

...

...

...

10

10

100%

1

BENZO(A)PYRENE

8

20

40%

22

24

92%

1

9

10

90%

1

BENZO(K)FLUORANTHENE

10

20

50%

...

...

...

...

9

10

90%

1

CHRYSENE

14

20

70%

...

...

...

...

10

10

100%

1

PYRENE

16

20

80%

24

24

100%

1

10

10

100%

1

PYRIDINE

0

20

0%

...

...

...

...

1

10

10%

2

Pesticides/Herbicides





4,4'-DDT

0

16

0%

2

18

11%

2

...

...

...

...

ENDOSULFAN SULFATE

0

20

0%

...

...

...

...

...

...

...

...

ENDRIN ALDEHYDE

0

20

0%

...

...

...

...

1

9

11%

2

2,4,5-T

2

6

33%

1

1

100%

2

...

...

...

...

Metals





ARSENIC

20

20

100%

...

...

...

...

9

10

90%

1

CADMIUM

1

20

5%

...

...

...

...

...

...

...

...

CHROMIUM

20

20

100%

24

24

100%

1

10

10

100%

1

COPPER

19

20

95%

24

24

100%

1

10

10

100%

1

CYANIDE

0

20

0%

...

...

...

...

...

...

...

...

LEAD

20

20

100%

24

24

100%

1

10

10

100%

1

MERCURY

19

20

95%

23

24

96%

1

10

10

100%

1

NICKEL

20

20

100%

24

24

100%

1

...

...

...

...

SELENIUM

2

20

10%

2

23

9%

2

1

10

10%

2

SILVER

1

20

5%

...

...

...

...

...

...

...

...

THALLIUM

12

20

60%

17

24

71%

1

4

10

40%

1

VANADIUM

20

20

100%

24

24

100%

1

10

10

100%

1

ZINC

20

20

100%

24

24

100%

1

10

10

100%

1

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-124

-------
Table B-124, continued

Analysis Options for Soil Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

5b FP

5b RB



Number of

Number of



Number of

Number of





Number of

Number of





CAPTION

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Semivolatiles





BUTYLBENZYLPHTHALATE

0

20

0%

1

16

6%

2

1

4

25%

2

DIBENZOFURAN

0

20

0%

9

16

56%

2

3

4

75%

2

4-METHYLPHENOL

0

20

0%

3

15

20%

2

...

...

...

...

N-NITROSO-DI-N-BUTYLAMINE

0

20

0%

...

...

...

...

...

...

...

...

P-PHENYLENEDIAMINE

0

11

0%

...

...

...

...

...

...

...

...

PAHs





BENZO(A)ANTHRACENE

10

20

50%

...

...

...

...

...

...

...

...

BENZO(A)PYRENE

8

20

40%

16

16

100%

1

...

...

...

...

BENZO(K)FLUORANTHENE

10

20

50%

...

...

...

...

...

...

...

...

CHRYSENE

14

20

70%

...

...

...

...

...

...

...

...

PYRENE

16

20

80%

16

16

100%

1

4

4

100%

1

PYRIDINE

0

20

0%

...

...

...

...

...

...

...

...

Pesticides/Herbicides





4,4'-DDT

0

16

0%

...

...

...

...

...

...

...

...

ENDOSULFAN SULFATE

0

20

0%

1

16

6%

2

...

...

...

...

ENDRIN ALDEHYDE

0

20

0%

...

...

...

...

...

...

...

...

2,4,5-T

2

6

33%

...

...

...

...

...

...

...

...

Metals





ARSENIC

20

20

100%

...

...

...

...

...

...

...

...

CADMIUM

1

20

5%

...

...

...

...

...

...

...

...

CHROMIUM

20

20

100%

16

16

100%

1

4

4

100%

1

COPPER

19

20

95%

16

16

100%

1

4

4

100%

1

CYANIDE

0

20

0%

2

16

13%

2

...

...

...

...

LEAD

20

20

100%

16

16

100%

1

4

4

100%

1

MERCURY

19

20

95%

16

16

100%

1

4

4

100%

1

NICKEL

20

20

100%

...

...

...

...

...

...

...

...

SELENIUM

2

20

10%

2

16

13%

2

...

...

...

...

SILVER

1

20

5%

11

16

69%

2

4

4

100%

2

THALLIUM

12

20

60%

9

16

56%

1

2

4

50%

3

VANADIUM

20

20

100%

16

16

100%

1

4

4

100%

1

ZINC

20

20

100%

16

16

100%

1

4

4

100%

1

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-124

-------
Table B-124, continued

Analysis Options for Soil Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

5c FP

5c RB



Number of

Number of



Number of

Number of





Number of

Number of





CAPTION

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Semivolatiles





BUTYLBENZYLPHTHALATE

0

20

0%

1

19

5%

2

...

...

...

...

DIBENZOFURAN

0

20

0%

3

19

16%

2

...

...

...

...

4-METHYLPHENOL

0

20

0%

4

19

21%

2

1

4

25%

2

N-NITROSO-DI-N-BUTYLAMINE

0

20

0%

1

19

5%

2

...

...

...

...

P-PHENYLENEDIAMINE

0

11

0%

...

...

...

...

...

...

...

...

PAHs





BENZO(A)ANTHRACENE

10

20

50%

...

...

...

...

...

...

...

...

BENZO(A)PYRENE

8

20

40%

17

19

89%

1

...

...

...

...

BENZO(K)FLUORANTHENE

10

20

50%

...

...

...

...

...

...

...

...

CHRYSENE

14

20

70%

...

...

...

...

...

...

...

...

PYRENE

16

20

80%

18

19

95%

1

...

...

...

...

PYRIDINE

0

20

0%

...

...

...

...

...

...

...

...

Pesticides/Herbicides





4,4'-DDT

0

16

0%

...

...

...

...

...

...

...

...

ENDOSULFAN SULFATE

0

20

0%

...

...

...

...

...

...

...

...

ENDRIN ALDEHYDE

0

20

0%

...

...

...

...

...

...

...

...

2,4,5-T

2

6

33%

...

...

...

...

...

...

...

...

Metals





ARSENIC

20

20

100%

19

19

100%

1

...

...

...

...

CADMIUM

1

20

5%

17

19

89%

2

...

...

...

...

CHROMIUM

20

20

100%

19

19

100%

1

4

4

100%

1

COPPER

19

20

95%

19

19

100%

1

4

4

100%

1

CYANIDE

0

20

0%

...

...

...

...

...

...

...

...

LEAD

20

20

100%

19

19

100%

1

4

4

100%

1

MERCURY

19

20

95%

18

19

95%

1

4

4

100%

1

NICKEL

20

20

100%

19

19

100%

1

...

...

...

...

SELENIUM

2

20

10%

5

19

26%

2

...

...

...

...

SILVER

1

20

5%

14

19

74%

2

4

4

100%

2

THALLIUM

12

20

60%

9

19

47%

1

3

4

75%

1

VANADIUM

20

20

100%

19

19

100%

1

4

4

100%

1

ZINC

20

20

100%

19

19

100%

1

4

4

100%

1

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-124

-------
Table B-124, continued

Analysis Options for Soil Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

6cd FP



Number of

Number of



Number of

Number of





CAPTION

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Semivolatiles



BUTYLBENZYLPHTHALATE

0

20

0%

...

...

...

...

DIBENZOFURAN

0

20

0%

1

7

14%

2

4-METHYLPHENOL

0

20

0%

1

7

14%

2

N-NITROSO-DI-N-BUTYLAMINE

0

20

0%

...

...

...

...

P-PHENYLENEDIAMINE

0

11

0%

1

6

17%

2

PAHs



BENZO(A)ANTHRACENE

10

20

50%

...

...

...

...

BENZO(A)PYRENE

8

20

40%

3

7

43%

1

BENZO(K)FLUORANTHENE

10

20

50%

...

...

...

...

CHRYSENE

14

20

70%

...

...

...

...

PYRENE

16

20

80%

...

...

...

...

PYRIDINE

0

20

0%

...

...

...

...

Pesticides/Herbicides



4,4'-DDT

0

16

0%

...

...

...

...

ENDOSULFAN SULFATE

0

20

0%

...

...

...

...

ENDRIN ALDEHYDE

0

20

0%

...

...

...

...

2,4,5-T

2

6

33%

...

...

...

...

Metals



ARSENIC

20

20

100%

...

...

...

...

CADMIUM

1

20

5%

...

...

...

...

CHROMIUM

20

20

100%

7

7

100%

1

COPPER

19

20

95%

7

7

100%

1

CYANIDE

0

20

0%

...

...

...

...

LEAD

20

20

100%

7

7

100%

1

MERCURY

19

20

95%

7

7

100%

1

NICKEL

20

20

100%

...

...

...

...

SELENIUM

2

20

10%

3

7

43%

2

SILVER

1

20

5%

6

7

86%

2

THALLIUM

12

20

60%

5

7

71%

1

VANADIUM

20

20

100%

7

7

100%

1

ZINC

20

20

100%

7

7

100%

1

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-124

-------
Table B-125

Background Comparison Results - Soil
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

5a FP

5a RB

5b FP

5b RB

5c FP

5c RB

6cd FP

Semivolatiles

BUTYLBENZYLPHTHALATE

U

U

U

U

U

...

—

DIBENZOFURAN

U

U

U

U

U

...

U

4-METHYLPHENOL

...

...

u

...

u

U

U

N-NITROSO-DI-N-BUTYLAMINE

...

...

...

...

u

...

...

P-PHENYLENEDIAMINE

...

...

...

...

...

...

u

PAHs

BENZO(A)ANTHRACENE

...

T

...

...

...

...

...

BENZO(A)PYRENE

NS

D

NS

...

NS

...

D

BENZO(K)FLUORANTHENE

...

D

...

...

...

...

...

CHRYSENE

...

T

...

...

...

...

...

PYRENE

D

D

D

D

D

...

...

PYRIDINE

...

U

...

...

...

...

...

Pesticides/Herbicides

4,4'-DDT

U

...

...

...

...

...

...

ENDOSULFAN SULFATE

...

...

U

...

...

...

...

ENDRIN ALDEHYDE

...

u

...

...

...

...

...

2,4,5-T

u

...

...

...

...

...

...

Metals

ARSENIC

...

NS

...

...

NS

...

...

CADMIUM

...

...

...

...

U

...

...

CHROMIUM

D

D

D

D

D

D

D

COPPER

D

D

D

D

D

D

D

CYANIDE

...

...

U

...

...

...

...

LEAD

D

D

D

D

D

D

D

MERCURY

NS

D

D

D

D

D

D

NICKEL

D

...

...

...

D*

...

...

SELENIUM

U

U

U

...

U

...

U

SILVER

...

...

U

U

U

U

U

THALLIUM

NS

NS

NS

WP

NS

NS

NS

VANADIUM

NS

NS

NS

NS

NS

NS

D

ZINC

D

D

D

D

D

D

D

Results Options:

D = Significant using Dunnett's Test	U = Too many non-detects to determine using statistics

NS = Not significant using Dunnett's Test or T-test WP = Within prediction interval
P = Outside of prediction interval	— No comparison made

T = Significant using T-Test

Non-parametric result. Not significantly different if log-normal distribution is assumed. Shapiro-Wilks p-value = 0.002.

Tables B-120-B-127.xlsB-1

-------
Table B-126

Analysis Options for Fish Chemicals for Comparison to Background
Housatonic River Site, OU2 - Pittsfield, MA



Background

5a

5bc

6cd



Number of

Number of



Number of

Number of





Number of

Number of





Number of

Number of





Chemical

Detects

Samples

% Detects

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Detects

Samples

% Detects

Analysis

Small Fish

Pesticides







4,4'-DDE

20

20

100%

...

...

...

...

5

6

83%

1

...

...

...

...

0,P'-DDT

8

20

40%

2

2

100%

3

6

6

100%

1

5

5

100%

1

4,4'-DDT

19

20

95%

2

2

100%

3

...

...

...

...

...

...

...

...

HEPTACHLOR EPOXIDE

15

20

75%

1

2

50%

3

...

...

...

...

4

5

80%

1

CIS-NONACHLOR

20

20

100%

2

2

100%

3

5

6

83%

1

5

5

100%

1

TRANS-NONACHLOR

20

20

100%

2

2

100%

3

6

6

100%

1

5

5

100%

1

OXYCHLORDANE

20

20

100%

1

2

50%

3

5

6

83%

1

...

...

...

...

Medium Fish

Pesticides







4,4'-DDE

114

114

100%

46

46

100%

1

91

94

97%

1

108

108

100%

1

0,P'-DDT

39

114

34%

46

46

100%

1

94

94

100%

1

108

108

100%

1

4,4'-DDT

110

114

96%

37

46

80%

1

82

94

87%

1

63

108

58%

1

DIELDRIN

100

114

88%

...

...

...

...

...

...

...

...

99

107

93%

1

HEPTACHLOR EPOXIDE

53

114

46%

31

46

67%

1

15

94

16%



55

108

51%

1

CIS-NONACHLOR

106

114

93%

46

46

100%

1

87

94

93%

1

107

108

99%

1

TRANS-NONACHLOR

112

114

98%

46

46

100%

1

94

94

100%

1

108

108

100%

1

OXYCHLORDANE

101

114

89%

20

46

43%

1

67

94

71%

1

40

108

37%

1

Large Fish

Pesticides







4,4'-DDE

15

15

100%

2

2

100%

3

32

32

100%

1

36

36

100%

1

0,P'-DDT

6

15

40%

2

2

100%

3

32

32

100%

1

36

36

100%

1

4,4'-DDT

15

15

100%

...

...

...

...

16

32

50%

1

15

36

42%

1

DIELDRIN

15

15

100%

...

...

...

...

...

...

...

...

35

36

97%

1

ENDRIN

5

15

33%

...

...

...

...

7

32

22%



...

...

...

...

BIEPTACHLOR EPOXIDE

3

15

20%

1

2

50%

2

11

32

34%



20

36

56%

2

OXYCHLORDANE

9

15

60%

1

2

50%

3

26

32

81%

1

17

36

47%

1

CIS-NONACHLOR

15

15

100%

2

2

100%

3

31

32

97%

1

34

36

94%

1

TRANS-NONACHLOR

15

15

100%

2

2

100%

3

30

32

94%

1

35

36

97%

1

Metals







MERCURY

5

5

100%

___ | ___ | ___ | ___

___ | ___ | ___ | ___

6

6

100%

i

Analysis Options:

1	= Statistical Comparison

2	= Non-stastical Decision

3	= Prediction Interval around Background Data

— = No background comprions because of low frequency of detection or not exceeding benchmarks.

Tables B-120 - B-127.xlsB-126

-------
Table B-127

Background Comparison Results - Fish
Housatonic River Site, OU2 - Pittsfield, MA



Small Fish

Medium Fish

Large Fish

Chemicals

5a

5bc

6cd

5a

5bc

6cd

5a

5bc

6cd

Pesticides

4,4'-DDE

—

NS

—

D

D

D

WP

D

D

0,P'-DDT

P

D

D

D

D

D

P

D

D

4,4'-DDT

P

—

—

D

D

D

—

D

D

DIELDRIN

—

—

—

—

—

T

—

—

T

ENDRIN

—

—

—

—

—

—

—

P

—

HEPTACHLOR EPOXIDE

WP

—

T

D

WP

D

u

U

U

CIS-NONACHLOR

P

D

D

D

D

D

p

D

D

TRANS-NONACHLOR

WP

NS

NS

D

D

D

p

D

D

OXYCHLORDANE

p*

T

—

D

D

D

WP

D

D

Metals

MERCURY

...

...

...

...

...

...

...

...

NS

Results Options:

D = Significant using Dunnett's Test	U = Too many non-detects to determine using statistics

NS = Not significant using Dunnett's Test or T-test WP = Within prediction interval
P = Outside of prediction interval	— No comparison made

T = Significant using T-Test

Within background using log-normal assumption for prediction interval, outside of nonparametric interval. Shapiro-Wilks test
for log-normality yields p-value of 0.046.

Tables B-120 - B-127.xlsB-127

-------
Table B-128

Tier I Evaluation Results - Sediment
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Geomorph



5a

5b

5c

6a b

6cd

Chemical

MC&AB

scox

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

Pond

Semivolatiles

Acetophenone

ND

ND

NBA

ND

ND

ND

ND

ND

NBA

ND

NA

NA

ND

B is (2 -ethy lhexy l)phthalate

NE

NE

NE

NE

NE

NE

X/NBA

ND

NE

NE

NA

NA

ND

Butylbenzyl phthalate

ND

ND

ND

NE

ND

ND

ND

ND

ND

ND

NA

NA

ND

Dibenzofuran

X/NBA

ND

ND

NE

NE

ND

X/NBA

ND

NE

ND

NA

NA

ND

Di-n-butylphthalate

ND

ND

NE

NE

NE

ND

ND

ND

ND

ND

NA

NA

ND

1,2-Dichlorobenzene

ND

ND

ND

ND

ND

ND

X/NBA

ND

ND

ND

NA

NA

ND

1,3-Dichlorobenzene

ND

ND

ND

ND

ND

ND

NE

ND

ND

NE

NA

NA

ND

1,4-Dichlorobenzene

NE

ND

NE

NE

NE

NE

X/NBA

ND

NE

X/NBA

NA

NA

NE

Diethyl phthalate

ND

ND

ND

NE

ND

NE

ND

ND

ND

ND

NA

NA

ND

Methapyrilene

ND

ND

ND

ND

ND

ND

NBA

ND

ND

ND

NA

NA

ND

2-Methy lnaphthalene

NBA

ND

NBA

NBA

NBA

NBA

NBA

ND

NBA

NBA

NA

NA

NBA

2-Methylphenol

ND

ND

ND

ND

ND

ND

X/NBA

ND

ND

ND

NA

NA

ND

4-Methylphenol

NBA

ND

NBA

ND

NBA

NBA

NBA

ND

NBA

ND

NA

NA

NBA

PAHs

Acenaphthene

X/NBA

ND

NE

NE

NE

ND

X/NBA

ND

ND

ND

NA

NA

NE

Ac enaphthy lene

X/NBA

ND

X/NBA

NE

NE

X/NBA

X/NBA

ND

X/NBA

ND

NA

NA

ND

Anthracene

XX

ND

XL

XL

XL

XL

XX

ND

XL

NE

NA

NA

XL

B enzo(a)anthr ac ene

XX

XL

XL

XL

XL

XL

XX

ND

XX

XL

NA

NA

XL

Benzo(a)pyrene

XX

NE

XL

XL

XL

XL

XX

ND

XX

XL

NA

NA

XL

B enzo(b)fluoranthene

NBA

NBA

NBA

NBA

NBA

NBA

NBA

ND

NBA

NBA

NA

NA

NBA

B enzo(k)fluoranthene

XL

NE

XL

XL

XL

XL

XL

ND

XL

XL

NA

NA

XL

B enzo(g, h, I)pery lene

XX

XL

XX

XX

XL

XX

XX

ND

XX

XL

NA

NA

XX

Chrysene

XX

NE

XX

XL

XL

XL

XX

NE

XX

XL

NA

NA

XL

Dibenzo(a,h)anthracene

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

ND

X/NBA

X/NBA

NA

NA

X/NBA

Fluoranthene

XX

NE

XL

XL

XL

XL

XX

NE

XL

XL

NA

NA

XL

Fluorene

XX

ND

NE

XL

NE

NE

XX

ND

XL

ND

NA

NA

NE

Indeno(l ,2,3-cd)pyrene

XX

XL

XX

XX

XL

XX

XX

ND

XX

XL

NA

NA

XX

Naphthalene

XX

ND

NE

NE

NE

NE

XX

ND

XL

NA

NA

NA

NE

Phenanthrene

XX

XL

XL

XL

XL

XL

XX

NE

XL

XL

NA

NA

XL

Pyrene

XX

XL

XX

XL

XL

XL

XX

NE

XL

XL

NA

NA

XL

Total PAH (using 0)

XX

XL

XL

XL

XL

XL

XX

NE

XX

XL

NA

NA

XL

Total PAH (using DL)

XX

XX

XL

XL

XL

XL

XX

XX

XX

XL

NA

NA

XL

Total PAH (high) (using 0)

XX

XX

XX

XX

XX

XX

XX

NE

XX

XX

NA

NA

XX

Total PAH (high) (using DL)

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

NA

NA

XX

Total PAH (low) (using 0)

XX

XL

XX

XX

XX

XX

XX

XL

XX

XL

NA

NA

XL

Total PAH (low) (using DL)

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

NA

NA

XX

Pentachlorobenzene

NE

ND

ND

NE

ND

ND

ND

ND

ND

ND

NA

NA

ND

Phenol

ND

ND

ND

ND

ND

X/NBA

X/NBA

ND

ND

ND

NA

NA

ND

1,2,4,5 - T etrachl or ob enzene

NBA

ND

ND

ND

ND

ND

ND

ND

ND

ND

NA

NA

ND

1,2,4-Trichlorobenzene

NE

ND

NE

ND

NE

NE

NE

ND

NE

ND

NA

NA

NE

Pesticides



























alpha-BHC

ND

ND

XL

ND

ND

ND

ND

ND

ND

ND

NA

NA

ND

beta-BHC

ND

ND

XL

ND

ND

ND

ND

ND

ND

ND

NA

NA

ND

Tables B-128 - B-135.xlsB-128

-------
Table B-128

Tier I Evaluation Results - Sediment
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Geomorph



5a

5b

5c

6a b

6cd

Chemical

MC&AB

scox

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

Pond

4,4'-DDD

XL

ND

XX

ND

ND

XX

ND

XX

ND

ND

NA

NA

ND

4,4'-DDE

ND

ND

XX

ND

ND

ND

ND

XX

ND

ND

NA

NA

ND

4,4'-DDT

ND

ND

ND

ND

NA

XX

ND

ND

XL

ND

NA

NA

NA

Endrin aldehyde

NBA

NBA

ND

ND

ND

ND

ND

ND

ND

ND

NA

NA

ND

Heptachlor

ND

ND

X/NBA

ND

ND

ND

ND

ND

ND

ND

NA

NA

ND

Dioxins/Furans

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NA

NA

NBA

PCBs

PCBs, Total

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

Metals

Antimony

BG

ND

BG

BG

ND

BG

NBA

ND

NBA

BG

NA

NA

BG

Arsenic

NE

NE

NE

NE

NE

NE

BG

XL

XL

NE

NA

NA

NE

Barium

BG

BG

BG

BG

BG

BG

BG

BG

NBA

NBA

NA

NA

BG

Beryllium

BG

BG

BG

BG

BG

BG

BG

NBA

NBA

BG

NA

NA

BG

Cadmium

ND

ND

BG

ND

NE

XL

XX

ND

XX

XX

NA

NA

XL

Chromium

BG

NE

BG

BG

XL

XX

XX

NE

XX

XX

NA

NA

XX

Cobalt

NE

NE

NE

NE

NE

NE

NE

NE

NE

NE

NA

NA

NE

Copper

BG

NE

XL

BG

BG

XL

XX

BG

XX

XX

NA

NA

XX

Lead

BG

BG

XX

BG

BG

XX

XX

BG

XX

BG

NA

NA

XX

Mercury

BG

NE

XL

BG

XL

XX

XX

XL

XX

XX

NA

NA

XX

Nickel

BG

XL

XL

NE

XL

XL

XL

XL

XX

XL

NA

NA

XL

Selenium

ND

ND

NBA

ND

ND

NBA

NBA

ND

NBA

ND

NA

NA

NBA

Silver

NE

ND

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

ND

X/NBA

X/NBA

NA

NA

X/NBA

Thallium

BG

ND

BG

BG

BG

BG

BG

NBA

BG

BG

NA

NA

ND

Tin

BG

BG

BG

BG

ND

NBA

NBA

BG

NBA

NBA

NA

NA

NBA

Vanadium

BG

BG

BG

BG

BG

BG

BG

BG

BG

BG

NA

NA

BG

Zinc

BG

NE

XL

BG

XL

XL

XX

BG

XX

XL

NA

NA

XL

Conventionals (only those with benchmarks)

Ammonia as N

NE

NA

NA

NA

NA

NA

NE

NA

NA

NA

NA

NA

NA

Cyanide

ND

ND

ND

ND

ND

ND

X/NBA

ND

ND

ND

NA

NA

ND

BG = Not different from background.

FOD = Frequency of detection <5%

NA = Not analyzed

NBA = No benchmark available

ND = Not detected

NE = Not exceeding benchmark

XL = Chemical exceeds both low and high-end benchmark.

XX = Chemcial exceeds both low and high-end benchmark.

XXX = Chemical exceeds low, and both high-end benchmarks based on assumed TOC content.

X/NBA = Chemical exceeds low benchmark, but no high-end benchmark available.	

Chemical screens-in based on > 5% FOD and at least one ¦ >1 based on the highest available benchmark.

Chemical screens-in based on > 5% FOD and at least one HQ >1 based on the low-end benchmark, but does not exceed high-end benchmark.

Chemical screens-in based on >5% FOD and NBA.

Tables B-128 - B-135.xlsB-128

-------
Table B-129

Tier I Evaluation Results - Surface Water
Housatonic River Site, OU2 - Pittsfleld, MA



Modeling Reach/Geomorph



5a

5b

5c

6cd

Chemical

MC& AB

scox

VP

MC& AB

VP

MC& AB

SCOX

VP

Pond

Volatiles

Vinyl Chloride

NBA

NA

NA

ND

NA

ND

NA

NA

ND

Semivolatiles

Bis(2-ethylhexyl)phthalate

x

NA

NA

FOD

ND

x

NA

ND

NE

PAHs

Fluor anthene

NE

NA

NA

X

ND

NE

NA

ND

NE

Pyrene

X

NA

NA

X

ND

X

NA

ND

X

Dioxins/F urans

NBA

NA

NBA

NBA

NBA

NBA

NA

NBA

NBA

PCBs

PCBs. Total

X

NA

X

X

X

X

NA

X

X

PCBs. Dissolved

X

NA

NA

X

NA

ND

NA

NA

FOD

Metals

Barium

BG

NA

NA

BG

BG

BG

NA

BG

BG

Cadmium

X

NA

NA

FOD

ND

ND

NA

ND

X

Copper

X

NA

NA

X

NE

X

NA

NE

NE

Lead

X

NA

NA

X

X

ND

NA

NE

ND

Silver

FOD

NA

NA

X

ND

ND

NA

ND

X

Zinc

BG

NA

NA

X

ND

NE

NA

NE

BG

Metals, dissolved

Beryllium, dissolved

NBA

NA

NA

NBA

NA

NBA

NA

NA

ND

Cadmium, dissolved

FOD

NA

NA

FOD

NA

ND

NA

NA

X

Cobalt, dissolved

NBA

NA

NA

FOD

NA

NBA

NA

NA

ND

Copper, dissolved

NE

NA

NA

X

NA

NE

NA

NA

NE

Silver, dissolved

NBA

NA

NA

ND

NA

ND

NA

NA

ND

Conventionals (only those with benchmarks)

Ammonia as N

X

X

NA

X

NA

X

X

NA

X

Sulfide

X

NA

NA

X

X

X

NA

NA

X

BG = Site concentrations not statistically different from background.

FOD = Frequency of detection <5%

NA = Not analyzed

NBA = No benchmark available

ND = Not detected

NE = Not exceeding benchmark

Chemical screens-in based on > 5% FOD and at least one HQ >1.

Chemical screens-in based on >5% FOD and NBA.

Tables B-128 - B-135.xlsB-129

-------
Table B-130

Tier I Evaluation Results - Soil
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Geomorph



5a

5b

5c

6ab

6cd

Chemical

FP

RB

FP

RB

FP

RB

FP

RB

FP

Semivolatiles

fs

















Acetophenone

ND

ND

ND

ND

ND

ND

NA

NA

NBA

Bis(2-ethylhexyl)phthalate

NE

NE

NE

NE

NE

NE

NA

NA

NE

Butylbenzylphthalate

NBA

NBA

NBA

NBA

NBA

ND

NA

NA

ND

Dibenzofuran

NBA

NBA

NBA

NBA

NBA

ND

NA

NA

NBA

Di-n-butyl phthalate

FOD

NE

NE

NE

NE

ND

NA

NA

NE

1,4-Dichlorobenzene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Diethylphthalate

NE

ND

ND

ND

ND

ND

NA

NA

ND

Dimethylphthalate

ND

NE

ND

ND

ND

ND

NA

NA

ND

4-Methylphenol

ND

ND

NBA

ND

NBA

NBA

NA

NA

NBA

4-Nitrophenol

ND

ND

ND

ND

ND

ND

NA

NA

NE

n-Nitroso-di-n-butylamine

ND

ND

ND

ND

NBA

ND

NA

NA

ND

PAHs

Acenaphthene

NE

NE

NE

NE

NE

ND

NA

NA

ND

Acenaphthylene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Anthracene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Benzo(a)anthracene

NE

X

NE

NE

NE

NE

NA

NA

NE

Benzo(b)fluoranthene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Benzo(k)fluoranthene

NE

X

NE

NE

NE

NE

NA

NA

NE

Benzo(ghi)perylene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Benzo(a)pyrene

X

X

X

NE

X

NE

NA

NA

X

Chrysene

NE

X

NE

NE

NE

NE

NA

NA

NE

Dibenzo(a,h)anthracene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Fluoranthene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Fluorene

NE

NE

NE

NE

NE

ND

NA

NA

NE

Indeno( 1,2,3-c,d)pyrene

NE

NE

NE

NE

NE

NE

NA

NA

NE

2-Methylnaphthalene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Naphthalene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Phenanthrene

NE

NE

NE

NE

NE

NE

NA

NA

NE

Pyrene

X

X

X

X

X

NE

NA

NA

NE

Total PAH

NBA

NBA

NBA

NBA

NBA

NBA

NA

NA

NBA

Total PAH (High MW)

NBA

NBA

NBA

NBA

NBA

NBA

NA

NA

NBA

Total PAH (Low MW)

NBA

NBA

NBA

NBA

NBA

NBA

NA

NA

NBA

Pentachlorobenzene

NE

ND

ND

ND

ND

ND

NA

NA

ND

Phenol

ND

ND

NE

ND

ND

ND

NA

NA

ND

p-Phenylenediamine

ND

ND

ND

ND

ND

ND

NA

NA

NBA

Pyridine

ND

NBA

ND

ND

ND

ND

NA

NA

ND

1,2,4-Trichlorobenzene

NE

NE

NE

NE

NE

NE

NA

ND

ND

Pesticides

Tables B-128 - R-135.xlsB-130

-------
Table B-130

Tier I Evaluation Results - Soil
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Geomorph



5a

5b

5c

6ab

6cd

Chemical

FP

RB

FP

RB

FP

RB

FP

RB

FP

4,4-DDT

X

ND

ND

ND

ND

ND

NA

NA

ND

Endosulfan sulfate

ND

ND

NBA

ND

ND

ND

NA

NA

ND

Endrin aldehyde

ND

NBA

ND

ND

ND

ND

NA

NA

ND

2,4,5-T

NBA

ND

ND

NA

ND

NA

NA

NA

NA

Dioxins/Furans

TCDD (Total)

X

x

x

x

x

x

NA

NA

X

TCDF (Total)

X

X

X

X

X

X

NA

NA

X

PCBs

Total PCBs

X

X

X

X

X

X

X

X

X

Metals

Antimony

NE

NE

NE

ND

NE

NE

NA

NA

NE

Arsenic

NE

BG

NE

NE

BG

NE

NA

NA

NE

Barium

NE

NE

NE

NE

NE

NE

NA

NA

NE

Beryllium

NE

NE

NE

NE

NE

NE

NA

NA

NE

Cadmium

NE

ND

NE

NE

X

NE

NA

NA

NE

Chromium

X

X

X

X

X

X

NA

NA

X

Cobalt

NE

NE

NE

NE

NE

NE

NA

NA

NE

Copper

X

X

X

X

X

X

NA

NA

X

Lead

X

X

X

X

X

X

NA

NA

X

Mercury

BG

X

X

X

X

X

NA

NA

X

Nickel

X

NE

NE

NE

X

NE

NA

NA

NE

Selenium

X

X

X

ND

X

ND

NA

NA

X

Silver

NE

NE

X

X

X

X

NA

NA

X

Thallium

BG

BG

BG

BG

BG

BG

NA

NA

BG

Tin

NE

NE

NE

NE

NE

NE

NA

NA

NE

V anadium

BG

BG

BG

BG

BG

BG

NA

NA

X

Zinc

X

X

X

X

X

X

NA

NA

X

Inorganics

Cyanide

FOD

ND

X

ND

ND

ND

NA

NA

ND

BG = Not different from background.

FOD = Frequency of detection <5%

NA = Not analyzed

NBA = No benchmark available

ND = Not detected

NE = Not exceeding benchmark

Chemical screens-in based on > 5% FOD and at least one HQ >1.
Chemical screens-in based on >5% FOD and NBA.

Tables B-128 - R-135.xlsB-130

-------
Table B-131

Tier I Evaluation Results - Fish
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Fish Size



5a

5bc

6cd

Chemical

Small

Medium

Large

Small

Medium

Large

Small

Medium

Large

Semivolatiles

Hexachlorobenzene

NE

NE

NE

NE

NE

NE

NE

NE

NE

Pentachloroanisole

NE

NE

NE

NE

NE

NE

NE

NE

NE

Pentachlorobenzene

NE

NE

NE

NE

NE

NE

NE

NE

NE

1,2,3,4-T etrachl or ob enzene

NE

NE

NE

NE

NE

NE

NE

NE

NE

1,2,4,5 - T etrachl or ob enzene

NE

NE

NE

NE

NE

NE

NE

NE

NE

Pesticides

Aldrin

ND

NE

ND

ND

NE

NE

NE

NE

NE

alpha-BHC

NE

NE

NE

NE

NE

NE

NE

NE

NE

beta-BHC

NE

NE

NE

NE

NE

NE

NE

NE

NE

delta-BHC

NE

NE

ND

ND

NE

NE

NE

NE

NE

gamma-BHC (Lindane)

ND

NE

NE

NE

NE

NE

NE

NE

NE

alpha-Chlordane

ND

NE

NE

NE

NE

NE

NE

NE

NE

gamma-Chlordane

NE

NE

NE

NE

NE

NE

NE

NE

NE

Chlorpyrifos

NE

NE

NE

NE

NE

NE

NE

NE

NE

o,p'-DDD

NE

NE

NE

NE

NE

NE

NE

NE

NE

4,4'-DDD

NE

NE

NE

NE

NE

NE

NE

NE

NE

o,p'-DDE

ND

NE

ND

ND

NE

NE

ND

NE

NE

4,4'-DDE

NE

X

X

X

X

X

NE

X

X

o,p'-DDT

X

X

X

X

X

X

X

X

X

4,4'-DDT

X

X

NE

NE

X

X

NE

X

X

Dieldrin

NE

NE

NE

NE

NE

NE

NE

X

X

Endosulfan II

NE

NE

NE

NE

NE

NE

NE

NE

NE

Endrin

NE

NE

NE

ND

NE

X

NE

NE

NE

Heptachlor

ND

NE

ND

NE

NE

NE

NE

NE

NE

Heptachlor Epoxide

NBA

NBA

NBA

ND

NBA

NBA

NBA

NBA

NBA

Mirex

ND

NE

NE

NE

NE

NE

ND

NE

NE

cis-Nonachlor

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

trans-Nonachlor

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

Oxychlordane

NBA

NBA

NBA

NBA

NBA

NBA

ND

NBA

NBA

Toxaphene

NE

ND

ND

ND

ND

ND

ND

FOD

FOD

Dioxins/Furans

Total Dioxins

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

Total Furans

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

PCBs

Total PCBs

X

X

X

X

X

X

X

X

X

Metals

Lead

NA

NA

NA

NA

NA

NA

NA

NA

NE

Mercury

NA

NA

NA

NA

NA

NA

NA

NA

BG

Nickel

NA

NA

NA

NA

NA

NA

NA

NA

NE

NA = Not analyzed

NBA = No benchmark available

ND = Not detected

NE = Not exceeding benchmark	

Chemical screens-in based on > 5% FOD and at least one HQ >1.

Chemical screens-in based on >5% FOD and NBA.

Tables B-128 - B-135.xls B-131

-------
Table B-132

Tier II Evaluation Results - Sediment
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Geomorph



5a

5b

5c

6a b

6cd

Chemical

MC&AB

scox

VP

MC&AB

scox

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

Pond

Semivolatiles

Acetophenone

...

...

NBA

...

...

...

...

...

NBA

...

...

...

...

B is (2 -ethy lhexy l)phthalate

...

...

...

...

...

...

X/NBA

...

...

...

...

...

...

Butylbenzyl phthalate

...

...

...

...

...

...

...

...

...

...

...

...

...

Dibenzofuran

X/NBA

...

...

...

...

...

X/NBA

...

...

...

...

...

...

Di-n-butylphthalate

—

...

...

...

...

...

...

...

...

...

...

...

...

1,2-Dichlorobenzene

—

...

...

...

...

...

X/NBA

...

...

...

...

...

...

1,3-Dichlorobenzene

...

...

...

...

...

...

...

...

...

...

...

...

...

1,4-Dichlorobenzene

...

...

...

...

...

...

X/NBA

...

...

X/NBA

...

...

...

Diethyl phthalate

...

...

...

...

...

...

...

...

...

...

...

...

...

Methapyrilene

...

...

...

...

...

...

NBA

...

...

...

...

...

...

2-Methy lnaphthalene

NBA

...

NBA

NBA

NBA

NBA

NBA

...

NBA

NBA

...

...

NBA

2-Methylphenol

...

...

...

...

...

...

X/NBA

...

...

...

...

...

...

4-Methylphenol

NBA

...

NBA

...

NBA

NBA

NBA

...

NBA

...

...

...

NBA

PAHs

Acenaphthene

X/NBA

...

...

...

...

...

X/NBA

...

...

...

...

...

...

Ac enaphthy lene

X/NBA

...

XL

...

...

X/NBA

X/NBA

...

X/NBA

...

...

...

...

Anthracene

XX

...

XL

XL

XL

XL

XX

...

XL

...

...

...

XL

B enzo(a)anthr ac ene

XX

XL

XL

XL

XL

XL

XX

...

XX

XL

...

...

XL

Benzo(a)pyrene

XX

...

XL

XL

XL

XL

XX

...

XX

XL

...

...

XL

B enzo(b)fluoranthene

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

NBA

NBA

...

...

NBA

B enzo(k)fluoranthene

XL

...

XL

XL

XL

XL

XL

...

XL

XL

...

...

XL

B enzo(g, h, I)pery lene

XX

XL

XX

XX

XL

XX

XX

...

XX

XL

...

...

XX

Chrysene

XX

...

XX

XL

XL

XL

XX

...

XX

XL

...

...

XL

Dibenzo(a,h)anthracene

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

...

X/NBA

X/NBA

...

...

X/NBA

Fluoranthene

XX

...

XL

XL

XL

XL

XX

...

XL

XL

...

...

XL

Fluorene

XX

...

...

XL

...

...

XX

...

XL

...

...

...

...

Indeno(l ,2,3-cd)pyrene

XX

XL

XX

XX

XL

XX

XX

...

XX

XL

...

...

XX

Naphthalene

XX

...

...

...

...

...

XX

...

XL

...

...

...

...

Phenanthrene

XX

XL

XL

XL

XL

XL

XX

...

XL

XL

...

...

XL

Pyrene

XX

XL

XX

XL

XL

XL

XX

...

XL

XL

...

...

XL

Total PAH (using 0)

XX

XL

XL

XL

XL

XL

XX

...

XX

XL

...

...

XL

Total PAH (using DL)

XX

XX

XL

XL

XL

XL

XX

XX

XX

XL

...

...

XL

Total PAH (using half DL)

XX

XL

XL

XL

XL

XL

XX

XL

XX

XL

...

...

XL

Total PAH (high) (using 0)

XX

XX

XX

XX

XX

XX

XX

...

XX

XX

...

...

XX

Total PAH (high) (using DL)

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

...

...

XX

Total PAH (high) (using half DL)

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

...

...

XX

Total PAH (low) (using 0)

XX

XL

XX

XX

XX

XX

XX

XL

XX

XL

...

...

XL

Total PAH (low) (using DL)

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

...

...

XX

Total PAH (low) (using half DL)

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

...

...

XX

Pentachlorobenzene

...

...

...

...

...

...

...

...

...

...

...

...

...

Phenol

...

...

...

...

...

X/NBA

X/NBA

...

...

...

...

...

...

1,2,4,5 - T etrachl or ob enzene

NBA

...

...

...

...

...

...

...

...

...

...

...

...

1,2,4-Trichlorobenzene

...

...

...

...

...

...

...

...

...

...

...

...

...

Tables B-128 - B-135.xlsB-132

-------
Table B-132

Tier II Evaluation Results - Sediment
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Geomorph



5a

5b

5c

6a b

6cd

Chemical

MC&AB

scox

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

Pond

Pesticides



alpha-BHC

...

...

XL

...

...

...

...

...

...

...

...

...

...

beta-BHC

...

...

XL

...

...

...

...

...

...

...

...

...

...

4,4'-DDD

XL

—

XX

—

—

XX

—

XX

—

—

—

—

—

4,4'-DDE

—

—

XX

—

—

—

—

XX

—

—

—

—

—

4,4'-DDT

—

—

—

—

—

XX

—

—

XL

—

—

—

—

Endrin aldehyde

NBA

NBA

—

—

—

—

—

—

—

—

—

—

—

Heptachlor

...

...

X/NBA

...

...

...

...

...

...

...

...

...

...

Dioxins/Furans

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

...

...

NBA

PCBs

PCBs, Total

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

XX

Metals

Antimony

...

...

...

...

...

...

NBA

...

NBA

...

...

...

...

Arsenic

—

—

—

—

—

—

—

XL

XL

—

—

—

—

Barium

—

—

—

—

—

—

—

—

NBA

NBA

—

—

—

Beryllium

—

—

—

—

—

—

—

NBA

NBA

—

—

—

—

Cadmium

...

...

...

...

...

XL

XX

...

XX

XX

...

...

XL

Chromium

...

...

...

...

XL

XX

XX

...

XX

XX

...

...

XX

Cobalt

—

—

—

—

—

—

—

—

—

—

—

—

—

Copper

—

—

XL

—

—

XL

XX

—

XX

XX

—

—

XX

Lead

—

—

XX

—

—

XX

XX

—

XX

—

—

—

XX

Mercury

—

—

XL

—

XL

XX

XX

XL

XX

XX

—

—

XX

Nickel

...

XL

XL

...

XL

XL

XL

XL

XX

XL

...

...

XL

Selenium

...

...

NBA

...

...

NBA

NBA

...

NBA

...

...

...

NBA

Silver

—

—

X/NBA

X/NBA

X/NBA

X/NBA

X/NBA

—

X/NBA

X/NBA

—

—

X/NBA

Thallium

...

...

NBA

...

...

...

...

NBA

...

...

...

...

...

Tin

...

...

...

...

...

NBA

NBA

...

NBA

NBA

...

...

NBA

Vanadium

—

—

—

—

—

—

—

—

—

—

—

—

—

Zinc

...

...

XL

...

XL

XL

XX

...

XX

XL

...

...

XL

— = Not considered in Tier II Evaluation
clear = Unlikely COPC
Potential COPC
Likely COPC

Tables B-128 - B-135.xlsB-132

-------
Table B-133

Tier II Evaluation Results - Surface Water
Housatonic River Site, OU2 - Pittsfield, MA



Modeling Reach/Geomorph



5a

5b

5c

6cd

Chemical

MC& AB

scox

VP

MC&AB

VP

MC&AB

SCOX

VP

Pond

Volatiles

Vinyl Chloride

NBA

...

...

...

...

...

...

...

...

Semivolatiles

Bis(2-ethylhexyl )phthalate

x

...

...

...

...

X

...

...

...

PAHs

Fluoranthene

...

—

—

X

—

...

—

...



Pyrene

X

—

...

X

...

X

...

...

X

Total PAHs

NBA

—

...

NBA

NBA

NBA

...

...

NBA

Dioxins/Furans

NBA

...

NBA

NBA

NBA

NBA

...

NBA

NBA

PCBs

PCBs. Total

X

...

X

X

X

X

...

X

X

PCBs. Dissolved

X

...

...

X

...

...

...

...

...

Metals

Barium

...

...



...



...

—

...

...

Cadmium

X

...

...

...

...

...

...

...

X

Copper

X

...

...

X

...

X

...

...

...

Lead

X

...

...

X

X

...

...

...

...

Silver

...

...



X

...

...

...

...

X

Zinc

...

...

...

X

...

...

...

...

...

Metals, dissolved

Beryllium, dissolved

NBA

...

...

NBA

...

NBA

...

...

...

Cadmium, dissolved

...

...

...

...

—

...

—

...

X

Cobalt, dissolved

NBA

...

...

...

...

NBA

...

...

...

Copper, dissolved

...

...



X

...

...

...

...

...

Silver, dissolved

NBA

...

...

...

...

...

...

...

...

Conventionals (only those with benchmarks)

Ammonia as N

X

X

...

X

...

X

X

...

X

Sulfide

X

...

...

X

X

X

...

...

X

— = Not considered in Tier II Evaluation
clear = Unlikely COPC
PotentialCOP^^^^^^^^^^

Likely COPC

Tables B-128 - B-135.xlsB-133

-------
Table B-134

Tier II Evaluation Results - Soil
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Modelin

g Reach/Geomorph

5a

5b

5c

6ab

6cd

FP

RB

FP

RB

FP

RB

FP

RB

FP

Semivolatiles



















Acetophenone

—

—

—

—

—

—

—

—

NBA

Bis(2-ethylhexyl)phthalate

—

—

—

—

—

—

—

—

—

Butylbenzylphthalate

NBA

NBA

NBA

NBA

NBA

—

—

—

—

Dibenzofuran

NBA

NBA

NBA

NBA

NBA

—

—

—

NBA

Di-n-butyl phthalate

—

—

—

—

—

—

—

—

—

1,4-Dichlorobenzene

—

—

—

—

—

—

—

—

—

Diethylphthalate

—

—

—

—

—

—

—

—

—

Dimethylphthalate

—

—

—

—

—

—

—

—

—

4-Methylphenol

—

—

NBA

—

NBA

NBA

—

—

NBA

4-Nitrophenol

—

—

—

—

—

—

—

—

—

n-Nitroso-di-n-butylamine

—

—

—

—

NBA

—

—

—

—

PAHs

Acenaphthene

—

—

—

—

—

—

—

—

—

Acenaphthylene

—

—

—

—

—

—

—

—

—

Anthracene

—

—

—

—

—

—

—

—

—

Benzo(a)anthracene

—

X

—

—

—

—

—

—

—

Benzo(b)fluoranthene

—

—

—

—

—

—

—

—

—

Benzo(k)fluoranthene

—

X

—

—

—

—

—

—

—

Benzo(ghi)perylene

—

—

—

—

—

—

—

—

—

Benzo(a)pyrene

X

X

X

—

X

—

—

—

X

Chrysene

—

X

—

—

—

—

—

—

—

Dibenzo(a,h)anthracene

—

—

—

—

—

—

—

—

—

Fluoranthene

—

—

—

—

—

—

—

—

—

Fluorene

—

—

—

—

—

—

—

—

—

Indeno(l ,2,3-c,d)pyrene

—

—

—

—

—

—

—

—

—

2-Methylnaphthalene

—

—

—

—

—

—

—

—

—

Naphthalene

—

—

—

—

—

—

—

—

—

Phenanthrene

—

—

—

—

—

—

—

—

—

Pyrene

X

X

X

X

X

—

—

—

—

Total PAH

NBA

NBA

NBA

NBA

NBA

NBA

—

—

NBA |

Tables B-128 - B-135.xlsB-134

-------
Table B-134

Tier II Evaluation Results - Soil
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Modelin

g Reach/Geomorph

5a

5b

5c

6ab

6cd

FP

RB

FP

RB

FP

RB

FP

RB

FP

Total PAH (High MW)

NBA

NBA

NBA

NBA

NBA

NBA

—

—

NBA 1

Total PAH (Low MW)

NBA

NBA

NBA

NBA

NBA

NBA

—

—

NBA

Pentachlorobenzene

—

—

—

—

—

—

—

—

—

Phenol

—

—

—

—

—

—

—

—

—

p-Phenylenediamine

—

—

—

—

—

—

—

—

NBA

Pyridine

—

NBA

—

—

—

—

—

—

—

1,2,4-Trichlorobenzene

—

—

—

—

—

—

—

—

—

Pesticides

4,4'-DDT

X

—

—

—

—

—

—

—

—

Endosulfan sulfate

—

—

NBA

—

—

—

—

—

—

Endrin aldehyde

—

NBA

—

—

—

—

—

—

—

2,4,5-T

NBA

—

—

—

—

—

—

—

—

Dioxins/Furans

TCDD (Total)

X

X

X

X

X

X

—

—

x 1

TCDF (Total)

X

X

X

X

X

X

—

—

X

PCBs

Total PCBs

X

X

X

X

X

X

X

X

X I

Metals

Antimony

—

—

—

—

—

—

—

—

—

Arsenic

—

—

—

—

—

—

—

—

—

Barium

—

—

—

—

—

—

—

—

—

Beryllium

—

—

—

—

—

—

—

—

—

Cadmium

—

—

—

—

X

—

—

—

—

Chromium

X

X

X

X

X

X

—

—

x

Cobalt

—

—

—

—

—

—

—

—

—

Copper

X

X

X

X

X

X

—

—

X

Lead

X

X

X

X

X

X

—

—

x

Mercury

—

X

X

X

X

X

—

—

x

Nickel

X

—

—

—

X

—

—

—

—

Selenium

X

X

X

—

X

—

—

—

X

Silver

—

—

X

X

X

X

—

—

X I

Tables B-128 - B-135.xlsB-134

-------
Table B-134

Tier II Evaluation Results - Soil
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Modelin

g Reach/Geomorph

5a

5b

5c

6ab

6cd

FP

RB

FP

RB

FP

RB

FP

RB

FP

Thallium

—

—

—

—

—

—

—

—

—

Tin

—

—

—

—

—

—

—

—

—

Vanadium

—

—

—

—

—

—

—

—

X

Zinc

X

X

X

X

X

X

—

—

x I

Inorganics

Cyanide

—

—

X

—

—

—

—

—

—

— = Not considered in Tier II Evaluation
clear = Unlikely COPC
Potential COPC
Likely COPC

Tables B-128 - B-135.xlsB-134

-------
Table B-135

Tier II Evaluation Results - Fish
Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Modeling Reach/Fish Size

5 a

5bc

6cd

Small | Medium | Large

Small | Medium | Large

Small | Medium | Large

Semivolatiles

Hexachlorobenzene

—

—

—

—

—

—

—

—

—

Pentachloroanisole

___

___

___

___

___

___

___

___

___

Pentachlorobenzene

___

___

___

___

___

___

___

___

___

1,2,3,4-Tetrachlorobenzene

___

___

___

___

___

___

___

___

___

1,2,4,5-Tetrachlorobenzene

—

—

—

—

—

—

—

—

—

Pesticides

Aldrin

—

—

—

—

—

—

—

—

—

alpha-BHC

___

___

___

___

___

___

___

___

___

beta-BHC

___

___

___

___

___

___

___

___

___

delta-BHC

___

___

___

___

___

___

___

___

___

gamma-BHC (Lindane)

___

___

___

___

___

___

___

___

___

alpha-Chlordane

___

___

___

___

___

___

___

___

___

gamma-Chlordane

___

___

___

___

___

___

___

___

___

Chlorpyrifos

___

___

___

___

___

___

___

___

___

o,p'-DDD

___

___

___

___

___

___

___

___

___

4,4'-DDD

___

___

___

___

___

___

___

___

___

o,p'-DDE

___

...

...

...

...

...

...

...

...

4,4'-DDE

___

X

X

X

X

X

___

X

X

o,p'-DDT

X

X

X

X

X

X

X

X

X

4,4'-DDT

X

X

___

___

X

X

___

X

X

Dieldrin

___

___

___

___

___

___

___

X

X

Endosulfan II

___

___

___

___

___

___

___

___

___

Endrin

___

___

___

___

___

X

___

___

___

Heptachlor

___

...

...

...

...

...

...

...

...

Heptachlor Epoxide

NBA

NBA

NBA

—

NBA

NBA

NBA

NBA

NBA

Mirex

___

___

___

___

___

___

___

___

___

cis-Nonachlor

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

trans-Nonachlor

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

Oxychlordane

NBA

NBA

NBA

NBA

NBA

NBA

___

NBA

NBA

Toxaphene

—

—

—

—

—

—

—

—

—

Dioxins/Furans

Total Dioxins

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

Total Furans

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

NBA

PCBs

Total PCBs | X | X | X | X | X | X | X | X | X

Metals

Lead

—

—

—

—

—

—

—

—

—

Mercury

___

___

___

___

___

___

___

___

___

Nickel

...

...

...

...

...

...

...

...

...

— = Not considered in Tier II Evaluation
clear = Unlikely COPC
Potential COPC
Likely COPC

Tables B-128 - B-135.xls B-135

-------
Table B-136

Chemicals Evaluated in Tier III Sediment Evaluation without Statistical Background Comparisons* or that Are Significantly Higher than Background

Housatonic River Site, OU2 - Pittsfield, MA

Chemical

Half-Mile

Landfill and Below to Facility

Frequency of Detection

# of Detected
Samples Exceeding
Benchmark

# of SQLs Exceeding
Benchmark



# of Detected
Samples Exceeding
Benchmark

# of SQLs Exceeding
Benchmark

Low | High

Low | High

Frequency of Detection

Low | High

Low | High

Semivolatiles



Acetophenone

3 / 57

NBA

NBA

NBA

NBA

ND

...

...

NBA

NBA

Dibenzofuran

18 / 57

2

NBA

18

NBA

12 / 74

0

NBA

34

NBA

1,4-Dichlorobenzene

32 / 57

17

NBA

26

NBA

6 / 74

0

NBA

70

NBA

2-Methylnaphthalene

23 / 57

NBA

NBA

NBA

NBA

13 / 74

NBA

NBA

NBA

NBA

2-Methylphenol (o-cresol)

2 / 57

FOD

FOD

56

NBA

ND

...

...

69

NBA

4-Methylphenol

5 / 57

NBA

NBA

NBA

NBA

3 / 66

FOD

FOD

NBA

NBA

PAHs



Acenaphthene

27 / 57

22

NBA

31

NBA

11 / 74

4

NBA

65

NBA

Acenaphthylene

24 / 57

11

NBA

34

NBA

4 / 74

0

NBA

73

NBA

Phenol

3 / 57

3

NBA

55

NBA

ND

...

...

69

NBA

Pesticides



alpha-BHC

ND

...

...

44

27

2 / 74

FOD

FOD

1

0

beta-BHC

2 / 54

FOD

FOD

44

14

7 / 66

3

0

8

0

4,4'-DDD

ND

...

...

48

38

1 / 74

FOD

FOD

20

0

4,4'-DDE

ND

...

...

55

38

2 / 74

FOD

FOD

75

0

4,4'-DDT

ND

...

...

41

25

ND

...

...

45

0

Endrin aldehyde

ND

...

...

NBA

NBA

ND

...

...

NBA

NBA

Metals



Antimony

14 / 54

NBA

NBA

NBA

NBA

21 / 74

NBA

NBA

NBA

NBA

Barium

52 / 54

NBA

NBA

NBA

NBA

75 / 75

NBA

NBA

...

...

Beryllium

28 / 54

NBA

NBA

NBA

NBA

34 / 74

NBA

NBA

NBA

NBA

Cadmium

21 / 50

4

1

0

0

15 / 74

0

0

0

0

Chromium

54 / 54

2

0

...

...

74 / 75

0

0

0

0

Copper

53 / 54

13

6

0

0

56 / 75

0

0

0

0

Lead

54 / 54

15

7

...

...

74 / 75

1

0

0

0

Mercury

20 / 54

4

2

0

0

18 / 75

0

0

0

0

Selenium

16 / 54

NBA

NBA

NBA

NBA

13 / 69

NBA

NBA

NBA

NBA

Silver

4 / 54

3

NBA

0

NBA

2 / 74

FOD

FOD

0

NBA

Tin

28 / 54

NBA

NBA

NBA

NBA

6 / 74

NBA

NBA

NBA

NBA

Zinc

ND

...

...

0

0

75 / 75

1

1

...

...

*Background comparisons not made because of insufficient number of detects in background.

FOD = Frequency of detection less than 5%, therefore, benchmark comparisons not made.
NBA = No benchmark available.

ND = Not detected.

Tables B-136 - B-138.xls B-136

-------
Table B-137

Chemicals Evaluated in Tier III Surface Water Evaluation without Statistical Background Comparisons* or that Are Significantly Higher than Background

Housatonic River Site, OU2 - Pittsfield, MA



Half-Mile Surface Water

Landfill and Below to Facility





# of Detected

# of SQLs



# of Detected

# of SQLs





Samples Exceeding

Exceeding



Samples Exceeding

Exceeding

Chemical

Frequency of Detection

Benchmark

Benchmark

Frequency of Detection

Benchmark

Benchmark

Volatiles



Vinyl Chloride

ND

—

NBA

ND

—

NBA

Semivolatiles



Benzo(a)anthracene

2 / 17

0

18

4 / 30

2

29

Benzo(a)pyrene

3 / 17

1

18

4 / 30

3

29

Fluoranthene

3 / 17

2

18

4 / 30

3

30

PAHs



Pyrene

3 / 17

3

18

5 / 30

4

29

Metals



Copper, dissolved

1 / 17

0

1

2 / 31

0

0

Copper

3 / 17

0

0

7 / 30

1

2

Lead

ND

—

12

3 / 30

3

17

Silver

ND

—

18

ND

—

30

Zinc

5 / 17

0

0

11 / 30

0

0

*Background comparisons not made because of insufficient number of detects in background.

FOD = Frequency of detection less than 5%, therefore, benchmark comparisons not made.
NBA = No benchmark available.

ND = Not detected.

Tables B-136 - B-138.xlsB-137

-------
Table B-138

Chemicals Evaluated in Tier III Soil Evaluation without Statistical Background Comparisons* or that Are Significantly Higher than Background

Housatonic River Site, OU2 - Pittsfield, MA



Half-Mile Soil

Other Facility Areas





# of Detected

# of SQLs



# of Detected

# of SQLs





Samples Exceeding

Exceeding



Samples Exceeding

Exceeding

Chemical

Frequency of Detection

Benchmark

Benchmark

Frequency of Detection

Benchmark

Benchmark

Semivolatiles



Dibenzofuran

44 / 119

NBA

NBA

70 / 153

NBA

NBA

4-Methylphenol

39 / 119

NBA

NBA

19 / 125

NBA

NBA

PAHs



Benzo(a)pyrene

111 / 119

30

0

124 / 160

86

38

Pyrene

113 / 119

38

0

132 / 163

92

33

Metals



Chromium

119 / 119

119

—

81 / 81

81

—

Lead

118 / 119

65

0

o

00

o

00

16

—

Mercury

114 / 119

114

5

31 / 56

32

25

Selenium

38 / 119

38

81

25 / 65

25

65

Vanadium

ND

—

—

ND

—

—

Zinc

ND

...

...

ND

...

...

*Background comparisons not made because of insufficient number of detects in background.

FOD = Frequency of detection less than 5%, therefore, benchmark comparisons not made.
NBA = No benchmark available.

ND = Not detected.

Tables B-136 - B-138.xlsB-138

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Unlikely COPCs

2-Methylphenol

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 1/82 samples had detected concentrations of 2-methylphenol. Of the 81 samples with non-
detects, the SQL exceeded the low benchmark in all 81 samples. (A high benchmark was not available.)

¦ Detection limit issues were of concern in each subarea except for 6ab SCOX and vernal pools, where semivolatiles were
not analyzed.

¦ Not detected in background; therefore, comparison not made.

¦ 2-Methylphenol was detected in only 2/57 samples in the Half-Mile Reach; however, all of the SQLs exceed the
benchmark.

¦ 2-Methylphenol was not detected in sediment adjacent to the landfill down to the Half-Mile Reach; however, all of the
SQLs (69 samples) exceeded the benchmark.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

Alpha-BHC

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 1/81 samples had detected concentrations of alpha-BHC. The detected concentration
exceeded the low benchmark only with a HQ =1.3. Of the 80 samples with non-detects, the SQL exceeded the low and
high benchmark in 75 and 45 samples, respectively.

¦ Detection limit issues were of concern in each subarea except for 6ab SCOX and vernal pools, where pesticides were
not analyzed.

¦ Insufficient number of detects in background to make comparisons. Detected in 1 /18 samples in background at 1.4E-
02 mg/kg and once in the PSA at 7.6E-03 mg/kg.

¦ Alpha-BHC was not detected in the Half-Mile Reach; however, out of 55 samples, the SQLs exceeded the low
benchmark in 44 samples and the high benchmark in 27 samples.

¦ Alpha-BHC was detected in sediment adjacent to the landfill down to the Half-Mile Reach in 2 / 74 samples. In the 72
nondetected samples, only 1 SQL exceeded the low benchmark and none exceeded the high.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

Beta-BHC

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 1/81 samples had detected concentrations of beta-BHC. The detected concentration
exceeded the low benchmark only with a HQ = 3.2. Of the 80 samples with non-detects, the SQL exceeded the low and
high benchmark in 76 and 34 samples, respectively.

¦ Detection limit issues were of concern in each subarea except for 6ab SCOX and vernal pools, where pesticides were
not analyzed.

¦ Insufficient number of detects in background to make comparisons. Detected in 2 /18 samples in background (range =
2.6E-02 to 3.5E-02 mg/kg) and once in the PSA at 1.6E-02 mg/kg.

¦ Beta-BHC was detected in the Half-Mile Reach in 2 / 54 samples. In the 52 non-detected samples, the SQLs exceeded
the low benchmark in 44 samples and the high benchmark in 14 samples.

¦ Beta-BHC was detected in sediment adjacent to the landfill down to the Half-Mile Reach in 7 / 66 samples. Of the 7
detected samples, 3 exceeded the low benchmark and none exceeded the high. In the 59 non-detected samples, 8
exceeded the low benchmark, and none exceeded the high.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Phenol

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 2/81 samples had detected concentrations of phenol. Of the 79 samples with non-detects,
the SQL exceeded the low benchmark in all 79 samples. (A high benchmark was not available).

¦ Detection limit issues were of concern in each subarea except for 6ab SCOX and vernal pools, where semivolatiles were
not analyzed.

¦ Not detected in background; therefore, comparison not made.

¦ Phenol was detected in the Half-Mile Reach in 3 / 57 samples. All three detected samples exceeded the benchmark. In
all of the non-detected samples, the SQLs exceeded the benchmark.

¦ Phenol was not detected in sediment adjacent to the landfill down to the Half-Mile Reach; however, the SQLs in all 69
of the samples exceeded the benchmark.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

Potential COPCs

Bis(2-ethylhexyl)phthalate

¦ Categories considered in the Tier III evaluation: 5c main channel; either not detected or below benchmarks in rest of
PSA sediment.

¦ Has low benchmark only.

¦ Detected in 8/13 samples, exceeds low benchmark in 1/8 detected samples with a HQ of 3.2.

¦ Statistically not different from background.

¦ Eliminate as a sediment COPC due to generally low frequency of detection, low potential for toxicity (number of times
benchmark exceeded and on HQ) and similarity with background.

Dibenzofuran

Yes

¦ Categories considered in the Tier III evaluation: 5a main channel and aggrading bars; either not detected or below
benchmarks in rest of PSA sediment, except for 5c main channel and aggrading bars, which was eliminated in Tier II.

¦ Has low benchmark only.

¦ Detected in 15/22 samples, exceeds low benchmark in 2 / 15 detected samples with the maximum detected
concentration yielding a HQ = 4.8.

¦ Insufficient number of detects in background to make comparisons. Detected in 1 / 20 samples in background at 3.4E-
02 mg/kg. The concentration range in 5a main channel is 3.2E-02 to 2.0E+00 mg/kg.

¦ Dibenzofuran was detected in 18/57 samples in the Half-Mile Reach, with 2 exceeding the benchmark. In the non-
detected samples, 18 SQLs exceeded the benchmark.

¦ Dibenzofuran was detected in the sediment adjacent to the landfill down to the Half-Mile Reach in 12 / 74 samples.
None of the detected samples exceeded the benchmark, but 34 of the SQLs in the non-detected samples exceeded the
benchmark.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Acenaphthylene

Yes

¦	Categories considered in the Tier III evaluation: 5a main channel, 5a vernal pool, 5b vernal pools, and 5c main channel.

¦	Has low benchmark only.

¦	In 5a main channel, detected in 20 / 22 samples, exceeds benchmark in 5 / 20 with the maximum detected concentration
yielding a HQ =18.

¦	In 5a vernal pools, detected in 4 / 8 samples, exceeds benchmark in 1 / 4 with a HQ =1.5. Eliminate from subarea.

¦	In 5b vernal pools, detected in 4 / 8 samples, exceeds benchmark in 1 / 4 with a HQ =1.1. Eliminate from subarea.

¦	In 5c main channel, detected in 6 /13 samples, exceeds benchmark in 1 / 6 with a HQ = 33.

¦	Not detected in background; therefore, comparison not made. SIM analyses not performed. Almost all other PAHs are
within background concentrations.

¦	Acenaphthylene was detected in the Half-Mile Reach in 24 / 57 samples, with 11 exceeding the benchmark. In the non-
detected samples, all of the SQLs exceeded the benchmark.

¦	Acenaphthylene was detected in the sediment adjacent to the landfill down to the Half-Mile Reach in 4 / 74 samples.
None of the detected samples exceeded the benchmark, but all of the SQLs in the non-detected samples exceeded the
benchmark.

Anthracene

Yes

¦	Categories considered in the Tier III evaluation: 5a main channel; 5a vernal pools; 5b main channel, SCOX, and vernal
pools; 5c main channel, and 6cd pond.

¦	Has both low and high benchmark.

¦	In 5a main channel, detected in all 22 samples, exceeds low benchmark in 17 / 22 samples and high benchmark in 4 / 22
samples with the maximum detected concentration yielding HQs = 192 and 13 using low and high benchmarks,
respectively.

¦	In 5a vernal pool detected in 4 / 8, exceeds low benchmark only in 2 / 4 with the maximum detected concentration
yielding a HQ of 1.9. Eliminate from this subarea.

¦	In 5b main channel detected in 4 / 5, exceeds low benchmark only in all four detected samples with the maximum
detected concentration yielding a HQ of 2.4. Eliminate from this subarea.

¦	In 5b SCOX, only one sample available, exceeds low benchmark only with a HQ of 1.4. Eliminate from this subarea.

¦	5b vernal pools detected in 4 / 8, exceeds low benchmark only in 2 / 4 with the maximum detected concentration
yielding a HQ of 1.3. Eliminate from this subarea.

¦	In 5c main channel, detected in 7 / 13 samples, exceeds low benchmark in 5 / 7 samples and high benchmark in 1 / 7
samples with the maximum detected concentration yielding HQs = 245 and 17 using low and high benchmarks,
respectively.

¦	In 6cd pond, only one sample available, exceeds low benchmark only with a HQ of 1.4. Eliminate from this subarea.

¦	Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Benzo(a)pyrene

Yes

¦ Categories considered in the Tier III evaluation: 5a main channel and vernal pools; 5b main channel, SCOX, and vernal
pools; 5c main channel and vernal pools; 6ab main channel; and 6cd pond.

¦ Has both low and high benchmark.

¦ In 5a main channel, detected in all 22 samples, exceeds low benchmark in 20 / 22 samples and high benchmark in 5 / 22
samples with the maximum detected concentration yielding HQs = 957 and 6.3 using low and high benchmarks,
respectively.

¦ In 5a vernal pools detected in 5 / 8, exceeds low benchmark only in 4 / 5, HQ using maximum detected concentration =
9.3.

¦ In 5b main channel detected in 5 / 6, exceeds low benchmark only in 4 / 5, HQ using maximum detected concentration
= 3.1.

¦ In 5b SCOX, only one sample, exceeds low benchmark only with a HQ of 4.3.

¦ In 5b vernal pools detected in 7 / 8, exceeds low benchmark only in 5 / 7, HQ using maximum detected concentration =
4.3.

¦ In 5c main channel, detected in all 13 samples, exceeds low benchmark in 11 /13 samples and high benchmark in 1 /13
samples with the maximum detected concentration yielding HQs = 100 and 13 using low and high benchmarks,
respectively.

¦ In 5c vernal pools, detected in 8 / 14 samples, exceeds low benchmark in 5 / 8 samples and high benchmark in 1 / 8
samples with the maximum detected concentration yielding HQs = 29 and 3 using low and high benchmarks,
respectively.

¦ In 6ab main channel detected in 2 / 2, exceeds low benchmark only in 2 / 2, HQ using maximum detected concentration
= 1.9. Eliminate from this subarea.

¦ In 6cd pond, only one sample, exceeds low benchmark only with a HQ of 2.4. Eliminate from this subarea.

¦ Note that all of the areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

B enzo (k)fluoranthene

Yes

¦	Categories considered in the Tier III evaluation: All areas except for 5a SCOX, 5c SCOX, 6ab SCOX, and 6ab vernal
pools.

¦	Has both low and high benchmark.

¦	In 5a main channel, detected in 22 / 22, exceeds low benchmark only in 20 / 22 with the maximum detected
concentration yielding a HQ = 40.

¦	In 5a vernal pools, detected in 5 / 8, exceeds low benchmark only in 2 / 5 with the maximum detected concentration
yielding a HQ = 5.

¦	In 5b main channel, detected in 5 / 6, exceeds low benchmark only in 4 / 5 with the maximum detected concentration
yielding a HQ = 2. Eliminate from this subarea.

¦	In 5b SCOX, only one sample, exceeds low benchmark only with a HQ = 2.8. Eliminate from this subarea.

¦	In 5b vernal pools, detected in 7 / 8, exceeds low benchmark only in 4 / 7 with the maximum detected concentration
yielding a HQ = 2.8. Eliminate from this subarea.

¦	In 5c main channel, detected in 13 / 13, exceeds low benchmark only in 10 / 13 with the maximum detected
concentration yielding a HQ = 50.

¦	In 5c vernal pools, detected in 11 /15, exceeds low benchmark only in 3 /11 with the maximum detected concentration
yielding a HQ = 24.

¦	In 6ab main channel, detected in 2/2, exceeds low benchmark only in 1 / 2 with a HQ =1.1. Eliminate from subarea.

¦	In 6cd pond, only one sample. Exceeds low benchmark only with a HQ = 1.7. Eliminate from subarea.

¦	Note that all of the areas are within background concentrations..

Fluorene

Yes

¦	Categories considered in the Tier III evaluation: 5a main channel, 5b main channel, 5c main channel, and 5c vernal
pools.

¦	Has both low and high benchmark.

¦	In 5a main channel, detected in 19 / 22, exceeds low benchmark in 14 / 19 samples and high benchmark in 4 / 19
samples with the maximum detected concentration yielding HQs = 52 and 7.5 using low and high benchmarks,
respectively.

¦	In 5b main channel detected in 4 / 6, exceeds low benchmark only in 1 / 4 with a HQ of 2. Eliminate from this subarea.

¦	In 5c main channel, detected in 5 /13, exceeds low and high benchmark in 1 / 5 samples with HQs = 129 and 19 using
the low and high benchmarks, respectively.

¦	In 5c vernal pools detected in 2/15, exceeds low benchmark only in 1 / 2 with a HQ of 1. Eliminate from this subarea.

¦	Note that all of the areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Naphthalene

Yes

¦ Of concern are 5a main channel, 5c main channel, and 5c vernal pools.

¦ Has both low and high benchmark.

¦ In 5a main channel, detected in 21 / 22 samples, exceeds low benchmark in 7 / 21 samples and high benchmark in 1 / 21
samples with the maximum detected concentration yielding HQs = 27 and 8.4 using low and high benchmarks,
respectively.

¦ In 5c main channel, detected in 11 /13 samples, exceeds low benchmark in 4 /11 samples and high benchmark in 1 /11
samples with the maximum detected concentration yielding HQs = 34 and 11 using low and high benchmarks,
respectively.

¦ In 5c vernal pool, detected in 9 /17 samples, exceeds low benchmark only in 1 / 9 with a HQ = 1.5. Eliminate from this
subarea.

¦ Note that all of the areas are within background concentrations.

4,4'-DDT

Not
quantitative;
SQL issues

¦ Categories considered in the Tier III evaluation: 5b vernal pools.

¦ Has both low and high benchmark.

¦ Detected in 1 / 8, exceeds low benchmark with a HQ of 673 and the high benchmark with a HQ of 45.

¦ Over the entire PSA, only 2/61 sediment samples had detected concentrations of 4,4'-DDT. Of the 59 samples with
non-detects, the SQL exceeded the low and high benchmark in 57 and 41 samples, respectively.

¦ Detection limit issues were of concern in each subarea except for 5a SCOX, 5b main channel and SCOX, 5c SCOX, 6ab
SCOX and vernal pools, and 6cd pond.

¦ Not detected in background; therefore, comparison not made.

¦ 4,4'-DDT was not detected in the Half-Mile Reach; however, the SQLs exceeded the low and high benchmarks in 41
and 25 of the 45 samples, respectively.

¦ 4,4,'-DDT also was not detected adjacent to the landfill down to the Half-Mile Reach; however, the SQLs in 45 / 75
samples exceed the low benchmark.

¦ Recommend leaving as a COPC in every subarea it was detected (i.e., 5b VP and 5c VP) and discussing uncertainty
associated with the high SQLs in the remaining subareas.

Arsenic

¦ Categories considered in the Tier III evaluation: 5c SCOX and vernal pools.

¦ Has both low and high benchmark.

¦ In 5c SCOX, detected in 2 / 2, exceeds low benchmark only in 1 / 2 with a HQ =1.1. Higher than the background
prediction interval; however, based on HQ, low potential for toxicity. Eliminate from subarea.

¦ In 5c vernal pools, detected in 13 / 15 samples, exceeds low benchmark only in 3 / 13 samples with the maximum
detected concentration yielding a HQ =1.5. Statistically higher than background; however, based on HQ, low potential
for toxicity. Eliminate from subarea.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Nickel

No

¦	Categories considered in the Tier III evaluation: all subareas except for 5a main channel, 5b main channel, 6ab SCOX
and vernal pools.

¦	Has both low and high benchmark.

¦	In 5a SCOX, detected in 2 / 2 samples, exceeds low benchmark only in 1 /2 samples with a HQ = 1.0. Eliminate from
subarea.

¦	In 5a vernal pools, detected in 8 / 8 samples, exceeds low benchmark only in 3 / 9 samples with the maximum detected
concentration yielding a HQ =1.1. Eliminate from subarea.

¦	In 5b SCOX, only one sample, exceeds low benchmark only with a HQ =1.1. Eliminate from subarea.

¦	In 5b vernal pools, detected in 9 / 9 samples, exceeds low benchmark only in 4 / 9 samples with the maximum detected
concentration yielding a HQ = 1.2. Eliminate from subarea.

¦	In 5c main channel, 13 / 13 samples, exceeds low benchmark only in 7 / 13 with the maximum detected concentration
yielding a HQ = 1.2. Eliminate from subarea.

¦	In 5c SCOX, detected in 2 / 2 samples, exceeds low benchmark only in 1 / 2 with a HQ =1.1. Eliminate from subarea.

¦	In 5c vernal pools, detected in 16 /16 samples, exceeds low benchmark in 10 /17 and high benchmark in 1 /17 with the
maximum detected concentration yielding HQs = 2.2 and 1.0. Eliminate from subarea.

¦	In 6ab main channel, detected in 2 / 2 samples, exceeds low benchmark only in 1 / 2 with a HQ = 1.2. Eliminate from
subarea.

¦	In 6cd pond, only one sample, exceeds low benchmark only with a HQ =1.2. Eliminate from subarea.

¦	Is statistically higher than background or higher than the background prediction interval in all subareas presented above;
however, based on HQs, low potential for toxicity. Eliminate as a sediment COPC.

Thallium

No

¦	Categories considered in the Tier III evaluation: 5c SCOX. Either within background or not detected in other subareas.

¦	No benchmark available.

¦	In 5c SCOX, detected in 2 / 2 samples. Concentrations higher than the background prediction interval.

¦	Eliminate from subarea and as a sediment COPC since not a COPC anywhere else in sediment due to either non-detects
or being below background.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Zinc

¦ Categories considered in the Tier III evaluation: 5a vernal pool, 5b SCOX, 5b vernal pool, 5c main channel and
aggrading bars, 5c vernal pools, 6ab main channel, 6cd pond.

¦ Has both low and high benchmark.

¦ In 5a vernal pools, detected in 8 / 8 samples, exceeds low benchmark only in 6 / 9 with the maximum detected
concentration yielding a HQ = 2.4. Statistically higher than background; however, based on HQ, low potential for
toxicity. Eliminate from subarea.

¦ In 5b SCOX, only one sample available, exceeds low benchmark only with a HQ = 1.8. Concentration higher than the
background prediction interval; however, based on HQ, low potential for toxicity. Eliminate from subarea.

¦ In 5b vernal pool, detected in 9 / 9 samples, exceeds low benchmark only in 8 / 9 with the maximum detected
concentration yielding a HQ = 2.3. Statistically higher than background; however, based on HQ, low potential for
toxicity. Eliminate from subarea.

¦ In 5c main channel, detected in 13 /13 samples, exceeds low benchmark in 8 /13 and high benchmark in 3 /13 with the
maximum detected concentration yielding HQs = 7.8 and 2.1 using the low and high benchmark, respectively.
Statistically higher than background.

¦ In 5c vernal pools, detected in 16 /16 samples, exceeds low benchmark in 13 /17 and high benchmark in 3 /17 with the
maximum detected concentration yielding HQs = 4.3 and 1.1 using the low and high benchmark, respectively.
Statistically higher than background.

¦ In 6ab main channel, detected in 2 / 2, exceeds low benchmark only in 2 / 2 with the maximum detected concentration
yielding a HQ = 3.1. Concentrations higher than the background prediction interval; however, based on HQ, low
potential for toxicity. Eliminate from subarea.

¦ In 6cd pond, only one sample available, exceeds low benchmark only with a HQ = 3.7. Concentration higher than the
background prediction interval.

¦ Consistently evaluated in Tier III in the main channel from reach 5c downstream; however, zinc is not detected in the
Half-Mile Reach and the SQLs for zinc do not exceed benchmarks. Since zinc is not likely site-related, eliminate as
sediment COPC.

Likely COPCs

Acetophenone

¦ Categories considered in the Tier III evaluation: 5a vernal pools based on frequency of detection.

¦ No benchmark available.

¦ Detected in 4 / 8 samples.

¦ Insufficient number of detects in background to make comparisons. Detected in 3 / 20 samples in background (range =
6.0E-02 to 7.0E-01 mg/kg). The range of detected concentrations in 5a vernal pools is 3.3E-02 and 7.4E-02 mg/kg.

¦ Not detected in the main channel between the confluence and Woods Pond.

¦ Detected in less than 10% of the samples from the Half-Mile Reach and in sediment adjacent to the landfill to the
facility (3 / 57 and 7 / 74, respectively).

¦ Eliminate as sediment COPC since may not be site related and detected concentrations appear to be within background.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

1,4-Dichlorobenzene

No

¦	Categories considered in the Tier III evaluation: 5c main channel and 6ab main channel.

¦	Low benchmark only.

¦	In 5c main channel, detected in 11 / 13, exceeds benchmark in 5 / 11 detected samples with the maximum detected
concentration yielding a HQ = 2.4. Eliminate from subarea.

¦	In 6ab main channel, detected in 2 / 2 samples, exceeds benchmark in 1 / 2 detected samples with a HQ = 1.3.
Eliminate from subarea.

¦	Insufficient number of detects in background to make comparisons. Detected in 1 / 20 samples in background at 6.2E-
02 mg/kg. The maximum detected concentrations in 5c main channel and 6ab main channel are 8.3E-01 and 4.5E-01
mg/kg, respectively. However, based on HQs, low potential for toxicity, eliminate as a sediment COPC.

2-Methylnaphthalene

No

¦	Categories considered in the Tier III evaluation: 5a main channel; 5b main channel, SCOX, and vernal pool; 5c main
channel; 6ab main channel; and 6cd pond based on frequency of detect.

¦	No benchmark available.

¦	In 5a main channel, detected in 18 / 22 samples. Maximum detected concentration = 6.7E-02 mg/kg.

¦	In 5b main channel, detected in 3 / 6 samples. Maximum detected concentration = 6.8E-02 mg/kg.

¦	In 5b SCOX, only one sample available sample. Concentration = 6.7E-02 mg/kg.

¦	In 5b vernal pool, detected in 2 / 8 samples. Maximum detected concentration = 6.8E-02 mg/kg.

¦	In 5c main channel, detected in 11 /13 samples. Maximum detected concentration = 2.2E+00 mg/kg.

¦	In 6ab main channel, detected in 1 / 2 samples. Detected concentration = 6.3E-02 mg/kg.

¦	In 6cd pond, only one sample available sample. Maximum detected concentration = 8.6E-02 mg/kg.

¦	Insufficient number of detects in background to make comparisons. Detected in 3 / 20 samples in background (range =
2.5E-02 to 6.0E-02 mg/kg).

¦	2-Methylnaphthalene was detected in the Half-Mile Reach in 23 / 57 samples and adjacent to the landfill down to the
Half-Mile Reach in 13 / 74 samples.

¦	Since most detected concentrations fall within the range of background concentrations, and detected both upstream of
and adjacent to the facility, suggest 2-methylnaphthalene is not site-related and recommend eliminating as a sediment
COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

4-Methylphenol

No

¦	Categories considered in the Tier III evaluation: 5a vernal pools, 5b SCOX, 5b vernal pools, 5c main channel, and 6cd
pond based on frequency of detection.

¦	No benchmark available.

¦	In 5a vernal pools, detected in 4 / 8 samples. Maximum detected concentration = 1.0E+00 mg/kg.

¦	In 5b SCOX, only one sample available sample. Concentration = 3.7E-02 mg/kg.

¦	In 5b vernal pools, detected in 3 / 8 samples. Maximum detected concentration = 5.1E+00 mg/kg.

¦	In 5c main channel, detected in 3 /13 samples. Maximum detected concentration = 4.0E-01 mg/kg.

¦	In 6cd pond, only one sample available sample. Maximum detected concentration = 8.8E-01 mg/kg.

¦	Insufficient number of detects in background to make comparisons. Detected in 1 / 20 samples in background at 1.6
mg/kg.

¦	4-Methylphenol was detected in the Half-Mile Reach in 5 / 57 samples and adjacent to the landfill down to the Half-
Mile Reach in 3 / 66 samples.

¦	Since most detected concentrations fall below the detected background concentration, and detected both upstream of
and adjacent to the facility, suggest 4-methylphenol is not site-related and recommend eliminating as a sediment COPC.

Acenaphthene

Yes

¦	Categories considered in the Tier III evaluation: 5a main channel and 5c main channel

¦	Low benchmark only.

¦	In 5a main channel, detected in 16 / 22, exceeds low benchmark in 8 / 16 with maximum detected concentration
yielding a HQ =19.

¦	In 5c main channel, detected in 4 / 13, exceeds low benchmark in 1 / 4 with the maximum detected concentration
yielding a HQ = 44.

¦	Insufficient number of detects in background data to perform background comparisons. Detected in 1 / 20 background
samples at 6.1E+00 mg/kg. The maximum detected concentrations in 5a main channel and 5c main channel are
1.7E+00 and 3.9E+00 mg/kg, respectively. Almost all other PAHs are within background concentrations.

¦	Acenaphthene was detected in the Half-Mile Reach in 27 / 57 samples, with 22 exceeding the benchmark. In the non-
detected samples, 31 of the SQLs exceeded the benchmark.

¦	Acenaphthene was detected in the sediment adjacent to the landfill down to the Half-Mile Reach in 11 / 74 samples.
Four of the detected samples exceeded the benchmark, and the SQLs in 65 of the non-detected samples exceeded the
benchmark.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Benzo(a)anthracene

Yes

¦	Categories considered in the Tier III evaluation: all subareas except for 5c SCOX, 6ab SCOX, and 6ab vernal pools.

¦	Has both low and high benchmark.

¦	In 5a main channel, detected in 22 / 22 samples exceeds low benchmark in 21 / 22, with the maximum detected
concentration yielding HQ = 139. Exceeds the high benchmark in 6 / 22, with the maximum detected concentration
yielding a HQ =14.

¦	In 5a SCOX, detected in 1 / 2, exceeds low benchmark only with a HQ of 1.0. Eliminate from this subarea.

¦	In 5a vernal pools, detected in 6 / 8, exceeds low benchmark only in 4 / 6 with the maximum detected concentration
yielding a HQ of 9.3.

¦	In 5b main channel, detected in 5 / 6, exceeds low benchmark only in 4 / 5 with the maximum detected concentration
yielding a HQ of 4.2.

¦	In 5b SCOX, only one sample. Exceeds low benchmark only with a HQ = 4.4.

¦	In 5b vernal pools, detected in 7 / 8, exceeds low benchmark only in 6 / 7 with the maximum detected concentration
yielding a HQ of 4.7.

¦	In 5c main channel, detected in 13 / 13, exceeds low benchmark in 11 / 13 with the maximum detected concentration
yielding a HQ = 185. Exceeds high benchmark in 1 /13 with a HQ =19.

¦	In 5c vernal pools, detected in 12 / 15, exceeds low benchmark in 10 / 12 with the maximum detected concentration
yielding a HQ = 73. Exceeds high benchmark in 1 /12 with a HQ = 7.5.

¦	In 6ab main channel, detected in 2 / 2, exceeds low benchmark only 2/2 with the maximum detected concentration
yielding a HQ of 1.6. Eliminate from this subarea.

¦	In 6cd pond only one sample, exceeds low benchmark only with a HQ = 2.8. Eliminate from this subarea.

¦	Note that all of these areas are within background concentrations.

B enzo (b)fluoranthene

Yes

¦	Categories considered in the Tier III evaluation: all areas except for 5c SCOX, 6ab SCOX, and 6ab vernal pools based
on frequency of detection (>50%).

¦	Benchmarks not available.

¦	In 5a main channel, detected in 22 / 22 samples.

¦	In 5a SCOX, detected in 1 / 2 samples.

¦	In 5a vernal pools, detected in 5 / 8 samples.

¦	In 5b main channel, detected in 5 / 6 samples.

¦	In 5b SCOX, only one sample available sample.

¦	In 5b vernal pools, detected in 7 / 8 samples.

¦	In 5c main channel, detected in 13 /13 samples.

¦	In 5c vernal pools, detected in 11 /15 samples.

¦	In 6ab main channel, detected in 2 / 2 samples.

¦	In 6cd pond, only one sample available sample.

¦	Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Benzo(g,h,I)perylene

Yes

¦ Categories considered in the Tier III evaluation: all areas except for 5c SCOX, 6ab SCOX, and 6ab vernal pools.

¦ Has both low and high benchmark.

¦ In 5a main channel, detected in 22 / 22 samples, exceeds low benchmark in 22 / 22 detects and high benchmark in 13 /
22 detects with the maximum detected concentration yielding HQs = 292 and 14 using low and high benchmarks,
respectively.

¦ In 5a SCOX, detected in 1 / 2, and exceeds low benchmark only with a HQ = 5.4

¦ In 5a vernal pool, detected in 5 / 8, exceeds low benchmark in 5 / 5 detects and high benchmark in 2 / 5 detects with the
maximum detected concentration yielding HQs = 50 and 2.3 using low and high benchmarks, respectively.

¦ In 5b main channel, detected in 5 / 6 samples, exceeds low benchmark in 5 / 5 samples and high benchmark in 1 / 5
detects with the maximum detected concentration yielding HQs = 32 and 1.5 using low and high benchmarks,
respectively.

¦ In 5b SCOX, only one sample. Exceeds low benchmark only with HQ = 18.5.

¦ In 5b vernal pools, detected in 7 / 8 samples, exceeds low benchmark in 7 / 7 detects and high benchmark in 2 / 7
detects with the maximum detected concentration yielding HQs = 47 and 2.2 using the low and high benchmarks,
respectively.

¦ In 5c main channel, detected in 13 /13 samples, exceeds low benchmark in 13 /13 detects and high benchmark in 4 /13
detects with the maximum detected concentration yielding HQs = 377 and 18 using low and high benchmarks,
respectively.

¦ In 5c vernal pools, detected in 11 /15 samples, exceeds low benchmark in 11 /11 detects and high benchmark in 2 /11
detects with the maximum detected concentration yielding HQs = 92 and 4.3 using low and high benchmarks,
respectively.

¦ In 6ab main channel, detected 2/2, exceeds low benchmark only in 2 / 2 with the maximum detected concentration
yielding a HQ of 19.3.

¦ In 6cd pond, only one sample available, exceeds low and high benchmarks with HQs = 22 and 1.0 using low and high
benchmarks, respectively.

¦ Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Chrysene

Yes

¦ Categories considered in the Tier III evaluation: all areas except for 5a SCOX, 5c SCOX, 6ab SCOX, and 6ab vernal
pools.

¦ Has both low and high benchmark.

¦ In 5a main channel, detected in 22 / 22 samples, exceeds low benchmark in 21 / 22 detects and high benchmark in 6 / 22
detects with the maximum detected concentration yielding HQs = 78 and 10 using low and high benchmarks,
respectively.

¦ In 5a vernal pools, detected in 6 / 8 samples, exceeds low benchmark in 4 / 6 detects and high benchmark in 1 / 6 detects
with the maximum detected concentration yielding HQs = 7.8 and 1.0 using low and high benchmarks, respectively.

¦ In 5b main channel, detected in 5 / 6, exceeds low benchmark in 4 / 5 with the maximum detected concentration
yielding a HQ of 3.0. Eliminate from subarea.

¦ In 5b SCOX only one sample. Exceeds low benchmark only with HQ = 4.1.

¦ In 5b vernal pools, detected in 7 / 8, exceeds low benchmark only in 6 / 7 with the maximum detected concentration
yielding a HQ of 4.1.

¦ In 5c main channel, detected in 13 /13 samples, exceeds low benchmark in 11 /13 detects and high benchmark in 1 /13
detects with the maximum detected concentration yielding HQs = 84 and 11 using low and high benchmarks,
respectively.

¦ In 5c vernal pools, detected in 12 / 16, exceeds low benchmark in 10 /12 detects and high benchmark in 1 /12 detects
with the maximum detected concentration yielding HQs = 49 and 6.4 using low and high benchmarks, respectively.

¦ In 6ab main channel, detected in 2 / 2, exceeds low benchmark only in 2 / 2 with the maximum detected concentration
yielding a HQ of 1.9. Eliminate from this subarea

¦ In 6cd pond, only one sample. Exceeds low benchmark only with HQ = 2.3. Eliminate from this subarea.

¦ Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Dibenzo(a,h)anthracene

Yes

¦	Categories considered in the Tier III evaluation: all areas except for 5c SCOX, 6ab SCOX, and 6ab vernal pools.

¦	Has low benchmark only.

¦	In 5a main channel, detected in 20 / 22 samples, exceeds low benchmark in 20 / 20 detected with the maximum detected
concentration yielding a HQ of 70.

¦	In 5a SCOX, detected inl / 2 samples, exceeds low benchmark with HQ = 27.

¦	In 5a vernal pools, detected in 4 / 8, exceeds low benchmark in 4 / 4 with the maximum detected concentration yielding
aHQ = 6.1.

¦	In 5b main channel, detected in 5 / 6, exceeds low benchmark in 3 / 5 with the maximum detected concentration
yielding a HQ = 2.2. Eliminate from this subarea.

¦	In 5b SCOX only one sample. Exceed low benchmark with HQ = 2.5. Eliminate from this subarea.

¦	In 5b vernal pools, detected in 3 / 8, exceeds low benchmark in 3 / 3 with the maximum detected concentration yielding
a HQ = 6.4.

¦	In 5c main channel, detected in 10 / 13, exceeds low benchmark in 9 / 10 with the maximum detected concentration
yielding a HQ = 70.

¦	In 5c vernal pool, detected in 2 /13, exceeds low benchmark in 2 / 2 with the maximum detected concentration yielding
aHQ = 23.

¦	In 6ab main channel, detected in 1 / 2 samples, exceeds low benchmark with a HQ =1.1. Eliminate from this subarea.

¦	In 6cd pond, only one sample. Exceeds low benchmark with a HQ = 2.4. Eliminate from this subarea.

¦	Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Fluoranthene

Yes

¦	Categories considered in the Tier III evaluation: all areas except for 5a SCOX, 5c SCOX, 6ab SCOX, and 6ab vernal
pools.

¦	Has both low and high benchmark.

¦	In 5a main channel, detected in 22 / 22 samples, exceeds low benchmark in 19 / 22 detects and high benchmark in 6 / 22
detects with the maximum detected concentration yielding HQs = 47 and 9.0 using low and high benchmarks,
respectively.

¦	In 5a vernal pool, detected in 7 / 8, exceeds low benchmark only in 2 / 7 with the maximum detected concentration
yielding a HQ = 3.1. Eliminate from this subarea.

¦	In 5b main channel, detected in 6 / 6, exceeds low benchmark only in 3 / 6 with the maximum detected concentration
yielding a HQ = 2. Eliminate from this subarea.

¦	In 5b SCOX, only one sample. Exceeds low benchmark only with a HQ = 2.6. Eliminate from this subarea.

¦	In 5b vernal pools, detected in 7 / 8, exceeds low benchmark only in 3 / 7 with the maximum detected concentration
yielding a HQ = 2.0. Eliminate from this subarea.

¦	In 5c main channel, detected in 13 / 13, exceeds low benchmark in 7 / 13 with the maximum detected concentration
yielding a HQ = 95. The high benchmark is exceeded once, with the HQ greater than 10.

¦	In 5c vernal pools, detected in 11 /14, exceeds benchmark in 2 /11 with the maximum detected concentration yielding
a HQ = 3.5.

¦	6ab main channel, detected in 2 / 2, exceeds low benchmark only in 1 / 2 with an HQ =1.1. Eliminate from subarea.

¦	6cd pond, only one sample. Exceeds low benchmark only with an HQ =1.3. Eliminate from subarea.

¦	Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Indeno( 1,2,3 -cd)pyrene

Yes

¦	Categories considered in the Tier III evaluation: all areas except for 5c SCOX, 6ab SCOX, and 6ab vernal pools.

¦	Has both low and high benchmark.

¦	In 5a main channel, detected in 22 / 22 samples, exceeds low benchmark in 22 / 22 detects and high benchmarks in 14 /
22 detects with the maximum detected concentration yielding HQs = 259 and 13 using low and high benchmarks,
respectively.

¦	In 5a SCOX, detected in 1 / 2 with a HQ for the low benchmark of 3.8.

¦	In 5a vernal pools, detected in 5 / 8 samples, exceeds low benchmark in 5 / 5 detects and high benchmark in 2 / 5 detects
with the maximum detected concentration yielding HQs = 34 and 2.4 using low and high benchmarks, respectively.

¦	In 5b main channel, detected in 5 / 6, exceeds low benchmark in 5 / 5 and high benchmark in 1 / 5 with the maximum
detected concentration yielding HQs = 19.4 and 1.4 using low and high benchmarks, respectively.

¦	In 5b SCOX, only one sample. Exceeds low benchmark with an HQ = 13.

¦	In 5b vernal pools, detected in 1 /1 sample, exceeds the low benchmark only with a HQ = 29.

¦	In 5c main channel, detected in 13 /13 samples, exceeds the low benchmark in 13 /13 detects and the high benchmark
in 5 /13 detects with the maximum detected concentration yielding HQs = 294 and 21 using low and high benchmarks,
respectively.

¦	In 5c vernal pools, detected in 11 / 15, exceeds low benchmark, in 11 / 11 with the maximum detected concentration
yielding a HQ = 88.2. 2/11 exceed the high benchmark, with the maximum detected concentration yielding a HQ =
6.25.

¦	In 6ab main channel, detected in 2 / 2 samples, exceeds the low benchmark only in 2 / 2, with the maximum detected
concentration yielding a HQ = 5.9.

¦	In 6cd pond, only one sample available, exceeds both the low and high benchmark with HQs =16 and 1.1 using low and
high benchmarks, respectively.

¦	Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Phenanthrene

Yes

¦	Categories considered in the Tier III evaluation: all areas except for 5c SCOX, 6ab SCOX, and 6ab vernal pools.

¦	Has both low and high benchmark.

¦	In 5a main channel, detected in 22 / 22, exceeds low benchmark in 20 / 22 with the maximum detected concentration
yielding a HQ = 142. Exceeds high benchmark in 7 / 22 with the maximum detected concentration yielding a HQ = 25.

¦	In 5a SCOX, detected in 1 / 2, exceeds low benchmark only with a HQ of 1.2. Eliminate from subarea.

¦	In 5a vernal pools, detected in 6 / 8, exceeds low benchmark only in 4 / 6 with the maximum detected concentration
yielding a HQ = 2.8. Eliminate from subarea.

¦	In 5b main channel, detected in 5 / 6, exceeds low benchmark only in 4 / 5 with the maximum detected concentration
yielding a HQ = 4.1.

¦	In 5b SCOX, only one sample, exceeds low benchmark with a HQ = 3.1.

¦	In 5b vernal pools, detected in 7 / 8, exceeds the low benchmark only in 4 / 7, with the maximum detected concentration
yielding a HQ = 2.2. Eliminate from subarea.

¦	In 5c main channel, detected in 11 / 13, exceeds the low benchmark in 7 / 11 and the high benchmark in 1 / 11. The
maximum detected concentration yields HQs of 265 and 46 for low and high benchmarks, respectively.

¦	In 5c vernal pools, detected in 12 / 15, exceeds the low benchmark only in 3 / 12, with the maximum detected
concentration yielding a HQ = 4.9.

¦	In 6ab main channel, detected in 2 / 2, exceeds the low benchmark only in 1 / 2, with the maximum detected
concentration yielding a HQ = 1.6. Eliminate from subarea.

¦	In 6cd pond, only one sample, exceeds low benchmark only with a HQ =1.8. Eliminate from subarea.

¦	Note that all of these areas are within background concentrations.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Pyrene

Yes

¦ Categories considered in the Tier III evaluation: all areas except for 5c SCOX, 6ab SCOX, and 6ab vernal pools.

¦ Has both low and high benchmark.

¦ In 5a main channel, detected in 22 / 22, exceeds low benchmark in 21 / 22 samples with the maximum detected
concentration yielding a HQ = 113. Exceeds high benchmark in 8 / 22 with the maximum detected concentration
yielding a HQ =15.

¦ In 5a SCOX, detected in 1 / 2, exceeds low benchmark only with a HQ of 2.5. Eliminate from subarea.

¦ In 5a vernal pools, detected in 6 / 8, exceeds low benchmark in 4 / 6 with the maximum detected concentration yielding
a HQ = 8.7. Exceeds high benchmark in 1 / 6 with the maximum detected concentration yielding a HQ =1.1. Eliminate
from subarea.

¦ In 5b main channel, detected in 5 / 6, exceeds low benchmark in 4 / 6 with the maximum detected concentration
yielding a HQ = 3.8.

¦ In 5b SCOX, only one sample, exceeds low benchmark with a HQ = 5.1.

¦ In 5b vernal pools, detected in 8 / 9, exceeds the low benchmark only in 6 / 8, with the maximum detected concentration
yielding a HQ = 6.2.

¦ In 5c main channel, detected in 13 / 13, exceeds the low benchmark in 11 / 13 and the high benchmark in 2 / 13. The
maximum detected concentration yields HQs of 185 and 24 for low and high benchmarks, respectively.

¦ In 5c vernal pools, detected in 13 / 15, exceeds the low benchmark only in 11 / 13, with the maximum detected
concentration yielding a HQ = 7.7.

¦ In 6ab main channel, detected in 2 / 2, exceeds the low benchmark only in 2 / 2, with the maximum detected
concentration yielding a HQ = 2.8. Eliminate from subarea.

¦ In 6cd pond, only one sample, exceeds low benchmark only with a HQ of 3. Eliminate from subarea.

¦ Note that all of these areas are within background concentrations.

4,4'-DDD

Not
quantitative;
SQL issue

¦ Categories considered in the Tier III evaluation: 5c SCOX.

¦ Has both low and high benchmark.

¦ Detected in 1 / 2 samples, concentration exceeds both low and high benchmark with HQs = 16 and 2.9 using the low
and high benchmarks, respectively.

¦ Insufficient number of detects in background to make comparisons. Detected in background in 3 / 21 samples (range =
3.0E-02 to 4.0E-01 mg/kg). Detected only once in 5c SCOX at 8.0E-02 mg/kg.

¦ Over the entire PSA, only 5/82 samples had detected concentrations of 4,4'-DDD. Of the 77 samples with non-detects,
the SQL exceeded the low and high benchmark in 75 and 63 samples, respectively. Detection limit issues were of
concern in each subarea except for 6ab SCOX and vernal pools.

¦ 4,4'-DDD was not detected in the Half-Mile Reach; however, 48 and 38 of the 55 SQLs from the non-detected samples
exceeded the low and high benchmarks, respectively.

¦ 4,4'-DDD was detected in the sediment adjacent to the landfill down to the Half-Mile Reach in 1 / 74 samples. The
detected concentration exceeds the low benchmark and the SQLs in 20 of the non-detected samples exceeded the low
benchmark.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

4,4'-DDE

Not

¦ Categories considered in the Tier III evaluation: 5a vernal pools and 5c SCOX.

quantitative;

¦ Has both low and high benchmark.

SQL issue

¦ In 5a vernal pools, detected in 5 / 11 samples, exceeded low and high benchmarks in all 5 detected samples, with the
maximum detected concentration yielding HQs = 633 and 17 using the low and high benchmarks, respectively.

¦ In 5c SCOX, detected in 1 / 2 samples, exceeded both low and high benchmark with HQs = 54 and 5.4 using the low
and high benchmarks, respectively.

¦ Insufficient number of detects in background to make comparisons. Detected in background in 2 / 20 samples (both at
same concentration, 5.4E-02 mg/kg). The maximum detected concentrations in 5a vernal pools and 5c SCOX are
2.0E+00 and 1.7E-01 mg/kg, respectively.

¦ Over the entire PSA, only 6/82 samples had detected concentrations of 4,4'-DDE. Of the 76 samples with non-detects,
the SQL exceeded the low and high benchmark in 76 and 62 samples, respectively. Detection limit issues were of
concern in each subarea except for 5ab SCOX and vernal pools.

¦ 4,4'-DDE was not detected in the Half-Mile Reach; however, 55 and 38 of the 55 SQLs from the non-detected samples
exceeded the low and high benchmarks, respectively.

¦ 4,4'-DDE was detected in the sediment adjacent to the landfill down to the Half-Mile Reach in 2 / 74 samples. The
detected concentrations exceed the low benchmark and all of the SQLs in the non-detected samples exceeded the low
benchmark.

¦ Recommend leaving as a COPC in every subarea it was detected (i.e., 5a VP and 5c SCOX) and discussion uncertainty
associated with the high SQLs in the remaining subareas.

Endrin aldehyde

¦ Categories considered in the Tier III evaluation: 5a SCOX based on frequency of detection.

¦ No benchmark available.

¦ In 5a SCOX, detected in 1 / 2 samples.

¦ Detected in PSA sediment only in 5a main channel and 5a SCOX.

¦ Not detected in background; therefore, comparison not made.

¦ Not detected in samples from the Half-Mile Reach or in sediment adjacent to the landfill to the facility.

¦ Assuming not site-related; therefore, eliminating.

Dioxins/Furans

Yes

¦ Detected in most samples everywhere analyzed.

¦ Similar mechanism of action of some PCB congeners

PCBs

Yes

¦ Detected and exceeding benchmarks in most samples analyzed.

Antimony

Yes

¦ Categories considered in the Tier III evaluation: 5c main channel and vernal pools.

¦ No benchmark available.

¦ In 5c main channel, detected in 9 /13 samples.

¦ In 5c vernal pools, detected in 8 /14 samples.

¦ Statistically higher than background.

¦ Antimony was detected in 14 / 54 samples in the Half-Mile Reach and in 21 / 74 samples in sediment adjacent to the
landfill down to the Half-Mile Reach.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Barium

Yes

¦	Categories considered in the Tier III evaluation: 5c vernal pools and 6ab main channel.

¦	No benchmark available.

¦	In 5c vernal pools, detected in 16 /16 samples.

¦	In 6ab main channel, detected in 2 / 2 samples.

¦	Statistically higher than background or higher than the background prediction interval.

¦	Barium was detected in 52 / 54 samples in the Half-Mile Reach and in 75 / 75 samples in sediment adjacent to the
landfill down to the Half-Mile Reach.

Beryllium

Yes

¦	Categories considered in the Tier III evaluation: 5c SCOX and vernal pools based

¦	No benchmark available.

¦	In 5c SCOX, detected in 2 / 2 samples.

¦	In 5c vernal pools, detected in 16 /16 samples.

¦	Statistically higher than background or higher than the background prediction interval.

¦	Beryllium was detected in 28 / 54 samples in the Half-Mile Reach and in 34 / 74 samples in sediment adjacent to the
landfill down to the Half-Mile Reach.

Cadmium

Yes

¦	Categories considered in the Tier III evaluation: 5b vernal pool, 5c main channel, 5c vernal pool, 6ab main channel, and
6cd pond.

¦	Has both low and high benchmarks.

¦	In 5b vernal pools, detected in 5 / 8 samples, exceed low benchmark only in 3 / 5 detected samples with maximum
detected concentration yielding HQ = 2.3. Eliminate from subarea.

¦	In 5c main channel, detected in 10 / 13 samples, exceeds low benchmark in 8 / 10 detects and 5/10 detects with the
maximum detected concentration yielding HQs = 17 and 3.3 using low and high benchmarks, respectively.

¦	In 5c vernal pools, detected in 13 /15 samples, exceeds low benchmark in 10 /14 detects and high benchmark in 3 /14
detects with HQs = 7.9 and 1.6 using low and high benchmarks, respectively.

¦	In 6ab main channel, detected in 2 / 2 samples, exceeds low and high benchmark in 1 / 2 with HQs = 6.1 and 1.2 using
low and high benchmarks, respectively.

¦	In 6cd pond, only one sample available, exceeds low benchmark only with HQ = 2.8. Eliminate from subarea.

¦	Statistically higher than background or concentrations higher than prediction interval for all of the subareas of concern.

¦	Cadmium was detected in 21 / 50 samples in the Half-Mile Reach, with 1 detected concentration exceeding the
benchmark. None of the SQLs in the non-detected samples exceeded the benchmark.

¦	Cadmium was detected in sediment adjacent to the landfill down to the Half-Mile Reach in 15 / 74 samples. None of
the detected concentrations or SQLs in the non-detected samples exceeded the benchmark.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Chromium

Yes

¦ Categories considered in the Tier III evaluation: 5b SCOX, 5b vernal pools, 5c main channel, 5c vernal pools, 6ab main
channel, and 6cd pond.

¦ Has both low and high benchmarks.

¦ In 5b SCOX, only one sample available, exceeds low benchmark only with HQ = 1.7. Eliminate from subarea.

¦ In 5b vernal pools, detected in 9 / 9 samples, exceeds low benchmark in 8 / 9 samples and high benchmark in 1 / 9
samples with the maximum detected concentration yielding HQs = 3.0 and 1.2. Eliminate from subarea.

¦ In 5c main channel, detected in 13 /13 samples, exceeds the low benchmark in 8 /13 and the high benchmark in 7 /13
with the maximum detected concentration yielding HQs =11 and 4.1 using low and high benchmarks, respectively.

¦ In 5c vernal pools, detected in 16 /16 samples, exceeds the low benchmark in 12 /16 and the high benchmark in 8 /16
with the maximum detected concentration yielding HQs = 5.9 and 2.3 using low and high benchmarks, respectively.

¦ In 6ab main channel, detected in 2 / 2 samples, exceeds the low benchmark in 2 / 2 and high benchmark 1/2 with the
maximum detected concentration yielding HQs = 3.6 and 1.4 using low and high benchmarks, respectively.

¦ In 6cd pond, only one sample available samples, exceeds the low and high benchmark with HQs = 4.4 and 1.7 using low
and high benchmarks, respectively.

¦ Statistically higher than background or concentrations higher than prediction interval for all of the subareas of concern.

¦ Chromium was detected in 54 / 54 samples in the Half-Mile Reach, with 2 detected concentrations exceeding the
benchmark.

¦ Chromium was detected in sediment adjacent to the landfill down to the Half-Mile Reach in 74 / 75 samples. None of
the detected concentrations or SQL in the non-detected sample exceeded the benchmark.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Copper

Yes

¦ Categories considered in the Tier III evaluation: 5a vernal pools, 5b vernal pools, 5c main channel, 5c vernal pools, 6ab
main channel, and 6cd pond.

¦ Has both low and high benchmarks.

¦ In 5a vernal pools, detected in 8 / 8 samples, exceeds the low benchmark only in 5 / 9 detects with the maximum
detected concentration yielding a HQ = 3.7.

¦ In 5b vernal pools, detected in 9 / 9 samples, exceeds the low benchmark only in 8 / 9 detects with the maximum
detected concentration yielding a HQ = 3.8.

¦ In 5c main channel, detected in 13 /13 samples, exceeds the low benchmark in 10 /13 detects and the high benchmark
in 7 /13 detects with the maximum detected concentration yielding HQs = 8.5 and 1.8 using low and high benchmarks,
respectively.

¦ In 5c vernal pools, detected in 16 /16 samples, exceeds the low benchmark in 13 /17 detects and high benchmark in 7 /
17 detects with the maximum detected concentration yielding HQs = 7.9 and 1.7 using low and high benchmarks,
respectively.

¦ In 6ab main channel, detected in 2 / 2 samples, exceeds the low benchmark in 1 / 2 detects and high benchmark in 1 / 2
detects with the maximum detected concentration yielding HQs = 5.6 and 1.2 using low and high benchmarks,
respectively.

¦ In 6cd pond, only one sample available, exceeds the low and high benchmark with HQs = 5.9 and 1.3 using low and
high benchmarks, respectively.

¦ Statistically higher than background or concentrations higher than prediction interval for all of the subareas of concern.

¦ Copper was detected in 53 / 54 samples in the Half-Mile Reach, with 13 detected concentrations exceeding the low
benchmark and 5 detected concentrations exceeding the high benchmark. The SQL in the non-detected sample did not
exceed the benchmark.

¦ Copper was detected in sediment adjacent to the landfill down to the Half-Mile Reach in 56 / 75 samples. None of the
detected concentrations or SQLs in the non-detected samples exceeded the benchmark.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Lead

Yes

¦ Categories considered in the Tier III evaluation: 5a vernal pools, 5b vernal pools, 5c main channel and vernal pools and
6cd pond.

¦ Has both low and high benchmarks.

¦ In 5a vernal pools, detected in 8 / 8 samples, exceeds the low benchmark in 8 / 9 detects and high benchmark in 1 / 9
detects with the maximum detected concentration yielding HQs = 6.6 and 1.8 using low and high benchmarks,
respectively.

¦ In 5b vernal pools, detected in 9 / 9 samples, exceeds the low benchmark in 8 / 9 detects and high benchmark in 2 / 9
detects with the maximum detected concentration yielding HQs = 5.2 and 1.5 using low and high benchmarks,
respectively.

¦ In 5c main channel, detected in 13 /13 samples, exceeds the low benchmark in 9 /13 detects and high benchmark in 7 /
13 detects with the maximum detected concentration yielding HQs =8.6 and 2.4 using low and high benchmarks,
respectively.

¦ In 5c vernal pools, detected in 16 /16 samples, exceeds the low benchmark in 14 /17 detects and high benchmark in 7 /
17 detects with the maximum detected concentration yielding HQs = 8.4 and 2.3 using low and high benchmarks,
respectively.

¦ In 6cd pond, only one sample available, exceeds the low and high benchmark with HQs = 6.9 and 1.9 using low and
high benchmarks, respectively.

¦ Statistically higher than background or concentrations higher than prediction interval for all of the subareas of concern.

¦ Lead was detected in 54 / 54 samples in the Half-Mile Reach, with 15 detected concentration exceeding the low
benchmark and 7 detected concentrations exceeding the high benchmark.

¦ Lead was detected in sediment adjacent to the landfill down to the Half-Mile Reach in 74 / 75 samples, with 1 detected
concentration exceeding the low benchmark. None of the detected concentrations exceeded the high benchmark and the
SQL in the non-detected sample did not exceed either benchmark.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Mercury

Yes

¦ Categories considered in the Tier III evaluation: 5a vernal pools; 5b SCOX and vernal pools; 5c main channel, SCOX,
and vernal pools; 6ab main channel, and 6cd pond.

¦ Has both low and high benchmarks.

¦ In 5a vernal pools, detected in 6 / 8 samples, exceeds the low benchmark only in 6 / 6 detects with the maximum
detected concentration yielding a HQ = 5.1.

¦ In 5b SCOX, only one sample available samples, exceeds the low benchmark only with a HQ = 3.4.

¦ In 5b vernal pools, detected in 9 / 9 samples, exceeds the low benchmark in 9 / 9 detects and high benchmark in 1 / 9
detects with the maximum detected concentration yielding HQs = 6.1 and 1.0 using low and high benchmarks,
respectively.

¦ In 5c main channel, detected in 10 /13 samples, exceeds the low benchmark in 10 /10 detects and high benchmark in 6
/ 10 detects with the maximum detected concentration yielding HQs = 14 and 2.4 using low and high benchmarks,
respectively.

¦ In 5c SCOX, detected in 2 / 2 samples, exceeds the low benchmark only in 1 / 2 detects with the maximum detected
concentration yielding a HQ = 2.5.

¦ In 5c vernal pools, detected in 15 /16 samples, exceeds the low benchmark in 13 /15 detects and high benchmark in 8 /
15 detects with the maximum detected concentration yielding HQs = 9.4 and 1.6 using low and high benchmarks,
respectively.

¦ In 6ab main channel, detected in 2 / 2 samples, exceeds the low benchmark in 2 / 2 detects and high benchmark in 1 / 2
detects with the maximum detected concentration yielding HQs = 6.1 and 1.0 using low and high benchmarks,
respectively.

¦ In 6cd pond, only one sample available, exceeds the low and high benchmark with HQs = 9.4 and 1.6 using low and
high benchmarks, respectively.

¦ Statistically higher than background or concentrations higher than prediction interval for all of the subareas of concern.

¦ Mercury was detected in 20 / 54 samples in the Half-Mile Reach, with 4 detected concentrations exceeding the low
benchmark and 2 detected concentrations exceeding the high benchmark.. None of the SQLs in the non-detected
samples exceeded the benchmark.

¦ Mercury was detected in sediment adjacent to the landfill down to the Half-Mile Reach in 18 / 75 samples. None of the
detected concentrations or SQLs in the non-detected samples exceeded the benchmark.

¦ Is bioaccumulative.

Selenium

Yes

¦ Categories considered in the Tier III evaluation: 5c vernal pools and 6cd pond based on frequency of detection.

¦ No benchmark available.

¦ In 5c vernal pools, detected in 8 /15 samples.

¦ In 6cd pond, only one sample available sample.

¦ Insufficient number of detects in background to make comparisons. Detected in 1 / 20 samples in background at 9.4E-
01 mg/kg. Detected in 5c vernal pools at a range of 7.4E-01 to 3.9E+00 and 6cd pond at 6.9E-01 mg/kg.

¦ Selenium was detected in the Half-Mile Reach in 16 / 57 samples and in sediment adjacent to the landfill down to the
Half-Mile Reach in 13 / 69 samples.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Silver

Yes

¦	Categories considered in the Tier III evaluation: 5a vernal pools; 5b main channel, SCOX, and vernal pools; 5c main
channel and vernal pools; 6ab main channel; and 6cd pond.

¦	Has low benchmark only.

¦	In 5a vernal pools, detected in 3 / 8 samples, exceeds the low benchmark in 3 / 3 detects with the maximum detected
concentration yielding a HQ = 3.8.

¦	In 5b main channel, detected in 2 / 6 samples, exceeds the low benchmark in 1 / 2 detects with the maximum detected
concentration yielding a HQ = 1.7. Eliminate from subarea.

¦	In 5b SCOX, only one sample available samples, exceeds the low benchmark with a HQ = 3.6.

¦	In 5b vernal pools, detected in 7 / 8 samples, exceeds the low benchmark in 6 / 7 detects with the maximum detected
concentration yielding a HQ = 8.0.

¦	In 5c main channel, detected in 10 / 13 samples, exceeds the low benchmark in 9 / 10 detects with the maximum
detected concentration yielding a HQ =18.

¦	In 5c vernal pools, detected in 10 / 15 samples, exceeds the low benchmark in 10 / 10 detects with the maximum
detected concentration yielding a HQ = 20.

¦	In 6ab main channel, detected in 2 / 2 samples, exceeds the low benchmark in 2 / 2 detects with the maximum detected
concentration yielding a HQ =13.

¦	In 6cd pond, only one sample available, exceeds the low benchmark with a HQ = 9.2.

¦	Insufficient number of detects in background to make comparisons. Detected in background in 5 / 20 samples (range =
4.8E-01 to 2.7E+00). Maximum detected concentrations in 5a vernal pools and 5b main channel and SCOX (1.9E+00,
8.7E-01, and 1.8E+00, respectively) are less than the maximum detected background concentration. Maximum detected
concentrations in 5b vernal pools, 5c main channel, 5c vernal pools, 6ab main channel, and 6cd pond (4.0E+00,
9.1E+00, 1.0E+01, 2.6E+01, and 4.6E+00, respectively) exceed the maximum detected background concentration.

¦	Silver was detected in 4 / 54 samples in the Half-Mile Reach and exceeds the benchmark in 3 / 4 detected samples.
None of the SQLs from the non-detected samples exceeded the benchmark.

¦	Silver was detected in the sediment adjacent to the landfill down to the Half-Mile Reach in 2 / 74 samples. None of the
detected concentrations or SQLs in the non-detected samples exceed the low benchmark.

¦	HQs seem to increase as progress downstream.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-139

Tier III Evaluation Results - Sediments

Chemical

COPC?

Discussion

Tin

Yes

¦	Categories considered in the Tier III evaluation: 5b vernal pools, 5c main channel, 5c vernal pools, 6ab main channel,
and 6cd pond based on frequency of detection.

¦	No benchmark available.

¦	In 5b vernal pools, detected in 5 / 8 samples.

¦	In 5c main channel, detected in 7 /13 samples.

¦	In 5c vernal pools, detected in 11 /15 samples.

¦	In 6ab main channel, detected in 2 / 2 samples.

¦	In 6cd pond, only one sample available sample.

¦	Statistically higher than background or concentrations higher than prediction interval for all of the subareas of concern.

¦	Tin was detected in 28 / 54 samples in the Half-Mile Reach, and in sediment adjacent to the landfill down to the Half-
Mile Reach in 75 / 75 samples

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-140

Tier III Evaluation Results - Surface Water

Chemical

COPC?

Discussion

Originally Eliminated During Tier I Evaluation

Benzo(a)anthracene

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 3/112 samples had detected concentrations of benzo(a)anthracene. The SQLs for the non-detected
samples exceeded the benchmark in all samples by at least a factor of 100. The non-detected samples with SQLs that exceed
the benchmarks did not have elevated detection limits (e.g., due to matrix interference). They were not analyzed using the
more sensitive SIM method; therefore, detections below the benchmark were not achievable.

¦ Not detected in background.

¦ Benzo(a)anthracene was detected in the Half-Mile Reach in 2 /17 samples, with none of the detected concentrations exceeding
the benchmark. In the non-detected samples, all of the SQLs exceeded the benchmark.

¦ Benzo(a)anthracene was detected in the surface water adjacent to the landfill down to the Half-Mile Reach in 4 / 30 samples,
with 2 exceeding the benchmark. In the non-detected samples, all of the SQLs exceeded the benchmark.

¦ Within background concentrations in sediment.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

Benzo(a)pyrene

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 3/112 samples had detected concentrations of benzo(a)pyrene. The SQLs for the non-detected
samples exceeded the benchmark in all samples by at least a factor of 100. The non-detected samples with SQLs that exceed
the benchmarks did not have elevated detection limits (e.g., due to matrix interference). They were not analyzed using the
more sensitive SIM method; therefore, detections below the benchmark were not achievable.

¦ Not detected in background.

¦ Benzo(a)pyrene was detected in the Half-Mile Reach in 3 / 17 samples, with 1 detected concentration exceeding the
benchmark. In the non-detected samples, all of the SQLs exceeded the benchmark.

¦ Benzo(a)pyrene was detected in the surface water adjacent to the landfill down to the Half-Mile Reach in 4 / 30 samples, with
3 concentrations exceeding the benchmark. In the non-detected samples, all of the SQLs exceeded the benchmark.

¦ Within background concentrations in sediment.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

Unlikely COPCs

Fluoranthene

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 12/112 samples had detected concentrations of fluoranthene. Of the 100 samples with non-detects,
the SQL exceeded the benchmark in all 100 samples by at least a factor of 100. The non-detected samples with SQLs that
exceed the benchmarks did not have elevated detection limits (e.g., due to matrix interference). They were not analyzed using
the more sensitive SIM method; therefore, detections below the benchmark were not achievable.

¦ Insufficient number of detects in background to make comparisons. Detected in 4 / 40 samples in background (range = 1.8E-
02 to 2.6E-02 |ig/L). PSA detection range = 1.5E-02 to 8.0E-01 |ig/L.

¦ Fluoranthene was detected in the Half-Mile Reach in 3 / 17 samples, with 2 exceeding the benchmark. In the non-detected
samples, all of the SQLs exceeded the benchmark.

¦ Fluoranthene was detected in the surface water adjacent to the landfill down to the Half-Mile Reach in 4 / 30 samples, with 3
exceeding the benchmark. In the non-detected samples, all of the SQLs exceeded the benchmark.

¦ Within background concentrations in sediment.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-140

Tier III Evaluation Results - Surface Water

Chemical

COPC?

Discussion

Pyrene

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 13/112 samples had detected concentrations of pyrene. Of the 99 samples with non-detects, the
SQL exceeded the benchmark in all 99 samples by at least a factor of 100. The non-detected samples with SQLs that exceed
the benchmarks did not have elevated detection limits (e.g., due to matrix interference). They were not analyzed using the
more sensitive SIM method; therefore, detections below the benchmark were not achievable.

¦ Insufficient number of detects in background to make comparisons. Detected in 4 / 40 samples in background (range = 2.1E-
02 to 2.9E-02 |ig/L). PSA detection range = 2.0E-02 to 8.0E-01 |ig/L.

¦ Pyrene was detected in the Half-Mile Reach in 3 / 17 samples, with all of the detected and SQL values exceeding the
benchmark.

¦ Pyrene was detected in the surface water adjacent to the landfill down to the Half-Mile Reach in 5 / 30 samples, with 4
exceeding the benchmark. In the non-detected samples, all of the SQLs exceeded the benchmark.

¦ Within background concentrations in sediment.

¦ Recommend eliminating from quantitative evaluation and discussing the uncertainty associated with the high SQLs.

Silver, total

Not
quantitative;
SQL issues

¦ Over the entire PSA, only 4 /102 samples had detected concentrations of silver, total (5a main channel, 5b main channel, 6cd
pond). Of the 98 samples with non-detects, the SQL exceeded the benchmark in all samples.

¦ Insufficient number of detects in background to make comparisons. Detected in 1 / 36 samples in background at 2.3E+00
|ig/L. PSA detection range = 1.3E+00 to 2.8E+00 |ig/L.

¦ Silver, total was not detected in the Half-Mile Reach or adjacent to the landfill down to the Half-Mile Reach. All of the SQLs
exceeded the benchmark.

¦ Recommend elimination from quantitative evaluation and discussing the uncertainty with the high SQLs.

Potential COPCs

Vinyl Chloride

¦ Categories considered in the Tier III evaluation: 5a main channel.

¦ No benchmark available.

¦ In 5a main channel, detected in 4 /10. No other detects in PSA.

¦ Not detected in background.

¦ No other volatile contamination in surface water.

¦ Not detected in 7 surface water samples from the Half-Mile Reach or in the 9 samples from adjacent to the landfill to the Half-
Mile Reach.

¦ Not a sediment COPC.

¦ Eliminate as surface water COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-140

Tier III Evaluation Results - Surface Water

Chemical

COPC?

Discussion

Copper, Dissolved

¦ Categories considered in the Tier III evaluation: 5b main channel.

¦ In 5b main channel, detected in 13 / 29, with 6/13 exceeding the benchmark based on the minimum detected hardness. Using
the minimum detected hardness derived benchmark value, the highest HQ = 3.5. When the benchmark is corrected for sample-
specific hardness, the maximum detected concentration yields an HQ =1.3. Eliminate from subarea.

¦ Insufficient number of detects in background to make comparisons. Detected in 1 / 32 background samples at 1.0E+00 |ig/L.
Detected in 5b main channel at a range = 1.3E+00 to 1.74E+01 |ig/L.

¦ Sediment COPC; however, given the hardness of the Housatonic River, there appears to be a low potential for toxicity in
surface water.

Zinc, total

¦ Categories considered in the Tier III evaluation: 5b main channel

¦ In 5b main channel, detected 11/29, with 2/11 exceeding the benchmark with the maximum detected concentration yielding a
HQ of 2.0.

¦ Zinc, total in the main channel of the other PSA subareas is present at less than the benchmark, or at concentrations equal to or
less than background. If the distribution is assumed to be lognormal, the comparison shows no difference between the site and
background.

¦ Not as sediment COPC.

Likely COPCs

Dioxins/Furans

Yes

¦ Detected in most samples everywhere analyzed.

¦ Similar mechanism of action of some PCB congeners

PCBs

Yes

¦ Detected and exceeding benchmarks in most samples analyzed.

Copper, total

¦ Categories considered in the Tier III evaluation: Reach 5b and 5c, main channel

¦ In Reach 5b main channel, detected in 17 / 29 samples, 9/17 exceed the benchmark with the maximum detected concentration
yielding a HQ = 1.7. Eliminate from this subarea.

¦ In Reach 5c main channel, detected in 12 /16 samples, 2/12 exceed the benchmark with the maximum detected concentration
yielding a HQ =1.1. Eliminate from this subarea.

¦ Insufficient number of detects in background to make comparisons. Detected in 4 / 36 background samples (range = 1.0E+00
to 1.9E+01 |ig/L). Maximum detected concentrations in Reach 5b and 5c main channel were 8.6E+00 and 5.7E+00|ig/L.
respectively.

¦ Sediment COPC; however, given the hardness of the Housatonic River, there appears to be a low potential for toxicity in
surface water and the concentrations appear to be similar to background.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-140

Tier III Evaluation Results - Surface Water

Chemical

COPC?

Discussion

Lead, total

No

¦	Potential issue for Reach 5b vernal pool.

¦	Only one sample in data set.

¦	Hazard quotient of 3.4 when corrected for sample-specific hardness.

¦	Insufficient number of detects in background to make comparisons. Detected in 4 / 36 background samples (range = 1.1E+00
to 5.3E+01 |ig/L). Detected in 5b vernal pools at 6.9E+00 |ig/L.

¦	Lead was not detected in the Half-Mile Reach; however 12 / 17 of the SQLs for the non-detected samples exceeded the
benchmark.

¦	Lead was detected in the surface water adjacent to the landfill down to the Half-Mile Reach in 3 / 30 samples, with all 3
detects exceeding the benchmark. In the non-detected samples, 17 of the SQLs exceeded the benchmark.

¦	Total suspended solids were not analyzed in this sample; therefore, cannot determine if the higher lead concentration due to
turbidity.

¦	Sediment COPC (in 5a vernal pool, 5b vernal pool, 5c main channel and vernal pool, and 6cd pond); however, given the
hardness of the Housatonic River, there appears to be a low potential for toxicity in surface water and the concentrations
appear to be similar to background.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-141

Tier III Evaluation Results - Soil

Chemical

COPC?

Discussion

Potential COPCs

Butylbenzylphthalate

No

¦	Categories considered in the Tier III evaluation: 5b riverbank soils.

¦	No benchmark available.

¦	In 5b riverbank, detected in 1 / 4 samples.

¦	Not detected in background.

¦	Not a COPC in sediment; detected in only 1/81 sediment samples.

¦	Eliminate as a soil COPC.

4-Methylphenol

No

¦	Categories considered in the Tier III evaluation: 5c riverbank soils.

¦	No benchmark available.

¦	In 5c riverbank, detected in 1 / 3 samples.

¦	Not detected in background.

¦	4-Methylphenol was detected in the floodplain in 39/119 samples in the Half-Mile Reach and in 19 / 125 soil samples at
other areas at the GE facility.

¦	Not a sediment COPC.

¦	Eliminate as a soil COPC.

B enzo (k)fluoranthene

No

¦	Categories considered in the Tier III evaluation: 5a riverbank soils.

¦	Detected in 9 /10 samples, exceeds benchmark in 1 / 9 samples with an HQ = 1.2.

¦	Statistically higher than background however, based on HQ, low potential for toxicity. Eliminate from subarea.

¦	Eliminate as soil COPC.

Nickel

No

¦	Categories considered in the Tier III evaluation: 5c floodplain.

¦	Detected in 18 /18 samples, exceeds benchmark in 2 /18 with the maximum detected concentration yielding a HQ = 1.0.

¦	Statistically higher than background however, based on HQ, low potential for toxicity. Eliminate from subarea.

¦	Not a sediment COPC.

¦	Eliminate as soil COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-141

Tier III Evaluation Results - Soil

Chemical

COPC?

Discussion

Selenium

Yes

¦	Categories considered in the Tier III evaluation: 5c and 6cd floodplain.

¦	In 5c floodplain, detected in 5 / 18, exceeds benchmark in all 5 samples with the maximum detected concentration yielding a
HQ = 11.4.

¦	In 6cd floodplain, detected in 3 / 7, exceeds benchmark in all 3 samples with the maximum detected concentration yielding a
HQ = 6.2.

¦	Insufficient number of detects in background to make comparisons. Detected in 2 / 19 samples (concentrations of 8.8E-01
and 1.1E+00 mg/kg). Maximum detected concentration in 5c and 6cd floodplains of 2.4E+00 and 1.3E+00 mg/kg,
respectively.

¦	Selenium was detected in the floodplain of 38 / 119 samples in the Half-Mile Reach. All of the detected concentrations and
SQLs from the non-detected samples exceed the benchmark.

¦	Selenium was detected in 25 / 65 soil samples at other areas at the GE facility. All of the detected concentrations and SQLs
from the non-detected samples exceed the benchmark.

¦	Sediment COPC.

Likely COPCs

Dibenzofuran

Yes

¦	Categories considered in the Tier III evaluation: 5a floodplain and riverbank and 5b floodplain and riverbank based on
frequency of detection only.

¦	No benchmark available.

¦	In 5a floodplain, detected in 9 / 23 samples.

¦	In 5a riverbank, detected in 8 /10 samples.

¦	In 5b floodplain, detected in 9 /16 samples.

¦	In 5b riverbank, detected in 3 / 4 samples.

¦	Not detected in background.

¦	Dibenzofuran was detected in the floodplain in 44 /119 samples in the Half-Mile Reach and in 70 / 153 soil samples at other
areas at the GE facility.

¦	Sediment COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-141

Tier III Evaluation Results - Soil

Chemical

COPC?

Discussion

Benzo(a)pyrene

Yes

¦	Categories considered in the Tier III evaluation: 5a floodplain and riverbank, 5b floodplain, 5c floodplain, and 6cd
floodplain.

¦	In 5a floodplain, detected in 22 / 24, exceeds benchmark in 7 / 22 with the maximum detected concentration yielding a HQ =
2.3. Not statistically different from background. Eliminate from subarea.

¦	In 5a riverbank, detected in 9 /10, exceeds benchmark in 8 / 9 with the maximum detected concentration yielding HQ = 15.7.
Statistically higher than background.

¦	In 5b floodplain, detected in 16 /16, exceeds benchmark in 2 / 16 with the maximum detected concentration yielding HQ =
1.1. Not statistically different from background. Eliminate from subarea.

¦	In 5c floodplain, detected in 16 / 18, exceeds benchmark in 2 111 (including duplicates) with the maximum detected
concentration yielding HQ = 23. Not statistically different from background. Eliminate from subarea.

¦	In 6cd floodplain, detected in 3 / 7, exceeds benchmark in 1/ 3 with a HQ = 1.6. Statistically higher than background.
Eliminate from subarea.

¦	Benzo(a)pyrene was detected in the floodplain of 111 / 119 samples in the Half-Mile Reach, with 30 / 111 detected
concentrations exceeding the benchmark. None of the SQLs from the non-detected samples exceed the benchmark.

¦	Benzo(a)pyrene was detected in 124 / 160 soil samples at other areas at the GE facility, with 86 / 124 detected concentrations
exceeding the benchmark. In the non-detected samples, 38 of the SQLs exceed the benchmark.

¦	Sediment COPC.

Pyrene

Yes

¦	Categories considered in the Tier III evaluation: 5a floodplain and riverbank; 5b riverbank; and 5c floodplain.

¦	In 5a floodplain, detected in 24 / 24, exceeds benchmark in 8 / 24 with the maximum detected concentration yielding a HQ =
3.3. Eliminate from subarea.

¦	In 5a riverbank, detected in 10 / 10, exceeds benchmark in 7 /10 with the maximum detected concentration yielding a HQ =
11.5.

¦	In 5b riverbank, detected in 4 / 4, exceeds benchmark in 1 / 4 with a HQ =1.1. Eliminate from subarea.

¦	In 5c floodplain, detected in 17/18, exceeds benchmark in 3 /18 with the maximum detected concentration yielding a HQ =
2.5. Eliminate from subarea.

¦	Statistically higher than background in all four subareas noted above.

¦	Pyrene was detected in the floodplain of 113/119 samples in the Half-Mile Reach, with 38 / 113 detected concentrations
exceeding the benchmark. None of the SQLs from the non-detected samples exceed the benchmark.

¦	Pyrene was detected in 132 / 163 soil samples at other areas at the GE facility, with 92 / 132 detected concentrations
exceeding the benchmark. In the non-detected samples, 33 of the SQLs exceed the benchmark.

¦	Sediment COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-141

Tier III Evaluation Results - Soil

Chemical

COPC?

Discussion

2,4,5-T

No

¦	Categories considered in the Tier III evaluation: 5a floodplain soils.

¦	No benchmark available.

¦	Analyzed and detected in one sample.

¦	Insufficient number of samples to perform background comparisons. Detected in 2 / 6 samples at concentrations of 6.0E-03
and 1.1E-02 mg/kg. Detected at 2.4E-01 mg/kg in 5a floodplains.

¦	Detected only once in sediment. Not a sediment COPC.

¦	Eliminate as COPC.

Dioxin/Furans

Yes

¦	Of concern everywhere analyzed (i.e., all categories except for 6ab floodplain and riverbank).

¦	Detected in >50% of the samples and TCDD (total) exceeds the benchmark greater than 50% of the time.

PCBs

Yes

¦ Of concern everywhere.

Chromium

Yes

¦	Detected in every sample analyzed (analyzed in all categories except for 6ab floodplain and riverbank).

¦	Exceeds benchmark in every sample with all HQs >10.

¦	Statistically higher than background in every subarea.

¦	Chromium was detected in the floodplain of 119 / 119 samples in the Half-Mile Reach, and in 81 / 81 soil samples at other
areas at the GE facility. All of the concentrations exceed the benchmark.

¦	Sediment COPC.

Copper

No

¦	Categories considered in the Tier III evaluation: 5a riverbank, 5b riverbank, 5c floodplain and riverbank, and 6cd floodplain.

¦	In 5a riverbank, detected in 10 /10, exceeds benchmark in 2 /10, with the maximum detected concentration yielding a HQ =
1.6. Eliminate from subarea.

¦	In 5b riverbank, detected in 4 / 4, exceeds benchmark in 1 / 4 with a HQ =1.1. Eliminate from subarea.

¦	In 5c floodplain, detected in 18 /18, exceeds benchmark in 10 /19 with the maximum detected concentration yielding a HQ =
2.9. Eliminate from subarea.

¦	In 5c riverbank, detected in 3 / 3, exceeds benchmark in 2 / 4 with the maximum detected concentration yielding a HQ = 1.7.
Eliminate from subarea.

¦	In 6cd floodplain, detected in 7 / 7, exceeds benchmark in 3 / 7 with the maximum detected concentration yielding a HQ =
2.1. Eliminate from subarea.

¦	Statistically higher than background in all subareas presented above; however, based on HQs, low potential for toxicity,
eliminate as a soil COPC.

¦	Sediment COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-141

Tier III Evaluation Results - Soil

Chemical

COPC?

Discussion

Lead

Yes

¦	Of concern everywhere analyzed (i.e., all categories except for 6ab floodplain and riverbank).

¦	In 5a floodplain, detected in 24 / 24, exceeds benchmark in 11 / 24 with the maximum detected concentration yielding a HQ =

3.1.	Eliminate from subarea.

¦	In 5a riverbank, detected in 10 /10, exceeds benchmark in 8 /10, with the maximum detected concentration yielding a HQ =

3.2.	Eliminate from subarea.

¦	In 5b floodplain, detected in 16 /16, exceeds benchmark in 14 /16 with the maximum detected concentration yielding a HQ =
4.2. Eliminate from subarea.

¦	In 5b riverbank, detected in 4 / 4, exceeds benchmark in 4 / 4 with the maximum detected concentration yielding a HQ = 2.4.
Eliminate from subarea.

¦	In 5c floodplain, detected in 18 /18, exceeds benchmark in 14 /19 with the maximum detected concentration yielding a HQ =
6.0.

¦	In 5c riverbank, detected in 3 / 3, exceeds benchmark in 4 / 4 with the maximum detected concentration yielding a HQ = 2.8.
Eliminate from subarea.

¦	In 6cd floodplain, detected in 7 / 7, exceeds benchmark in 7/ 7 with the maximum detected concentration yielding a HQ = 4.4.

¦	Statistically higher than background in all subareas; however, based on HQs, low potential for toxicity, eliminate from 5a
floodplain, 5a riverbank, 5b floodplain, 5b riverbank, and 5c riverbank.

¦	Lead was detected in the floodplain of 118 / 119 samples in the Half-Mile Reach, with 65/118 detected concentrations
exceeding the benchmark. None of the SQLs from the non-detected samples exceed the benchmark.

¦	Lead was detected in 80 / 80 soil samples at other areas at the GE facility, with 16 / 80 detected concentrations exceeding the
benchmark.

¦	Sediment COPC.

Mercury

Yes

¦	Categories considered in the Tier III evaluation: 5a riverbank, 5b floodplain and riverbank, 5c floodplain and riverbank, and
6cd floodplain

¦	Detected in 57/58 samples within these areas total.

¦	HQs range from 118 to 3,333.

¦	Statistically higher than background in all of these areas.

¦	Mercury was detected in the floodplain of 114 / 119 samples in the Half-Mile Reach, with all of the detected concentrations
and SQLs from the non-detected samples exceeding the benchmark.

¦	Mercury was detected in 31 / 56 soil samples at other areas at the GE facility, with 32 detected concentrations exceeding the
benchmark. In the non-detected samples, all 25 of the SQLs exceed the benchmark.

¦	Sediment COPC.

¦	Bioaccumulative.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-141

Tier III Evaluation Results - Soil

Chemical

COPC?

Discussion

Silver

No

¦	Categories considered in the Tier III evaluation: 5b riverbank, 5c floodplain and riverbank, and 6cd floodplain

¦	In 5b riverbank, detected in 4 / 4, exceeds benchmark in 1 / 4 with a HQ =1.3. Eliminate from subarea.

¦	In 5c floodplain, detected in 13/18, exceeds benchmark in 7 /14 with the maximum detected concentration yielding a HQ =
2.9. Eliminate from subarea.

¦	In 5c riverbank, detected in 3 / 3, exceeds benchmark in 1 / 4 with a HQ = 3.2. Eliminate from subarea.

¦	In 6cd floodplain, detected in 6 / 7, exceeds benchmark in 3 / 6 with the maximum detected concentration yielding a HQ =
1.8. Eliminate from subarea.

¦	Insufficient number of detects in background to make comparisons. Detected in 1 / 20 samples at 4.2E-01 mg/kg. Detected
in 5b riverbank, 5c floodplain and riverbank, and 6cd floodplain at 2.6E+00, 5.7E+00, 6.3E+00, and 3.5E+00 mg/kg,
respectively.

¦	Although site concentrations are higher than the one detected background concentration, based on HQ, low potential for
toxicity. Eliminate as soil COPC.

Vanadium

No

¦	Categories considered in the Tier III evaluation: 6cd floodplain.

¦	In 6cd floodplain, detected in 7 / 7, exceeds benchmark in 7 / 7 with the HQ ranging from 9 to 14.

¦	Statistically higher than background.

¦	Vanadium is not detected in the soil or sediment in the Half-Mile Reach. In addition, vanadium is not detected in the soil
from Newell to Lyman Street or in any of the specific facility operable units. The sample quantitation limits in these areas to
not exceed benchmarks. Does not appear to be site-related.

¦	Not a COPC in sediment.

¦	Eliminate as a soil COPC.

Zinc

No

¦	Of concern everywhere analyzed (i.e., all categories except for 6ab floodplain and riverbank).

¦	Detected in 81 / 82 samples.

¦	Exceeds benchmark in every sample with the maximum detected concentration in each subarea yielding a HQ between 19 and
45.

¦	Statistically higher than background in each subarea.

¦	Zinc is not detected in the soil or sediment in the Half-Mile Reach. In addition, zinc is not detected in the soil from Newell to
Lyman Street or in any of the specific facility operable units. The sample quantitation limits in these areas to not exceed
benchmarks. Does not appear to be site-related.

¦	Not a sediment COPC.

¦	Eliminate as a soil COPC.





O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-142

Tier III Evaluation Results - Fish

Chemical

COPC?

Discussion

Potential COPCs

Dieldrin

No

¦	Categories considered in the Tier III evaluation: 6cd large fish.

¦	In 6cd large fish, detected in 35 / 36 samples, exceeds benchmark in 4 / 35 detects with the maximum detected concentration
yielding a HQ = 2.3.

¦	Statistically higher than background.

¦	Not detected in sediment, surface water, or soil.

¦	Eliminate as COPC.

Likely COPCs

4,4'-DDE

Yes

¦	Categories considered in the Tier III evaluation: 5a medium and large fish; 5bc small, medium, and large fish; and 6cd
medium and large fish.

¦	In 5a medium fish, detected in 46 / 46 samples, exceeds benchmark in 16 / 46 samples with the maximum detected
concentration yielding a HQ = 2. Statistically higher than background; however, based on HQ, low potential for toxicity.
Eliminate from subarea/category.

¦	In 5a large fish, detected in 2/2, exceeds benchmark in one samples with a HQ = 1.1. Not statistically different from
background. Eliminate from subarea/category.

¦	In 5bc small fish, detected in 5 / 6, exceeds benchmark in 1 / 5 with a HQ =1.1. Statistically not different from background.
Eliminate from subarea/category.

¦	In 5bc medium fish, detected in 91 / 94, exceeds benchmark in 33 / 91, with the maximum detected concentration yielding a
HQ = 2.6. Statistically higher than background; however, based on HQ, low potential for toxicity, eliminate from
subarea/category.

¦	In 5bc large fish, detected 32 / 32, exceeds benchmark in 18 / 32, with the maximum detected concentration yielding a HQ =
6.6. Statistically higher than background.

¦	In 6cd medium fish, detected 108 / 108, exceeds benchmark in 47 / 108, with the maximum detected concentration yielding a
HQ = 10. Statistically higher than background.

¦	In 6cd large fish, detected in 36 / 36, exceeds benchmarks in 25 / 36, with the maximum detected concentration yielding a HQ
= 12. Statistically higher than background.

¦	Sediment COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-142

Tier III Evaluation Results - Fish

Chemical

COPC?

Discussion

0,p'-DDT

Yes

¦	Categories considered in the Tier III evaluation: all categories.

¦	In 5a small fish, detected in 2 / 2 samples, exceeds benchmark in both samples with the maximum detected concentration
yielding a HQ = 33. Concentrations higher than the background prediction interval.

¦	In 5a medium fish, detected in 46 / 46 samples, exceeds benchmark in 45/46 samples with the maximum detected
concentration yielding a HQ =132. Statistically higher than background.

¦	In 5a large fish, detected in 2 / 2, exceeds benchmark in one samples with a HQ = 54. Concentration higher than the
background prediction interval.

¦	In 5bc small fish, detected in 6 / 6, exceeds benchmark in all samples, with the maximum detected concentration yielding a
HQ = 28. Statistically higher than background.

¦	In 5bc medium fish, detected in 94 / 94, exceeds benchmark in 93 / 94, with the maximum detected concentration yielding a
HQ = 413. Statistically higher than background.

¦	In 5bc large fish, detected 32 / 32, exceeds benchmark in 32 / 32, with the maximum detected concentration yielding 215.
Statistically higher than background.

¦	In 6cd small fish, detected in 5 / 5, exceeds benchmark in 5 / 5, with the maximum detected concentration yielding 11.
Statistically higher than background.

¦	In 6cd medium fish, detected 108 / 108, exceeds benchmark in 106 /108, with the maximum detected concentration yielding a
HQ = 215. Statistically higher than background.

¦	In 6cd large fish, detected in 36 / 36, exceeds benchmarks in 36 / 36, with the maximum detected concentration yielding a HQ
= 257. Statistically higher than background.

¦	Not analyzed in sediment.

4,4'-DDT

Yes

¦	Categories considered in the Tier III evaluation: 5a small and medium fish; 5bc medium and large fish; and 6cd large fish.

¦	In 5a small fish, detected in 2 / 2 samples, exceeds benchmark in both samples with the maximum detected concentration
yielding a HQ = 2. Concentration higher than the background prediction interval; however, based on HQ, low potential for
toxicity. Eliminate from subarea.

¦	In 5a medium fish, detected in 37 / 46 samples, exceeds benchmark in 6 / 37 samples with the maximum detected
concentration yielding a HQ = 8. Statistically higher than background.

¦	In 5bc medium fish, detected in 82 / 94, exceeds benchmark in 23 / 91, with the maximum detected concentration yielding a
HQ = 3.8. Statistically higher than background.

¦	In 5bc large fish, detected 16 / 32, exceeds benchmark in 6 / 16, with the maximum detected concentration yielding 9.6.
Statistically higher than background.

¦	In 6cd medium fish, detected 63 / 108, exceeds benchmark in 3 / 63, with the maximum detected concentration yielding a HQ
= 1.9. Statistically higher than background; however, based on HQ, low potential for toxicity. Eliminate from
subarea/category.

¦	In 6cd large fish, detected in 15 / 36, exceeds benchmarks in 2 / 15, with the maximum detected concentration yielding a HQ
= 9.1. Statistically higher than background.

¦	Sediment COPC.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-142

Tier III Evaluation Results - Fish

Chemical

COPC?

Discussion

Heptachlor epoxide

Yes

¦	Categories considered in the Tier III evaluation: 5a small, medium, and large fish; 5bc large fish; and 6cd small, medium and
large fish based on frequency of detection only.

¦	No benchmark available.

¦	In 5a small fish, detected in 1 / 2 samples. Concentrations within the background prediction interval.

¦	In 5a medium fish, detected in 31 / 46 samples. Statistically higher than background.

¦	In 5a large fish, detected in 1 / 2 samples at 1.63E-03 mg/kg. Insufficient number of samples to perform background
comparison. Detected in 3 /15 samples in background at a range of 9.28E-04 to 2.31E-03 mg/kg.

¦	In 5bc large fish, detected in 11 / 32 samples with the range of detected concentrations of 2.05E-03 to 8.09E-03 mg/kg.
Insufficient number of background samples to perform comparison. Detected in 3 / 15 samples in background at a range of
9.28E-04 to 2.31E-03 mg/kg.

¦	In 6cd small fish, detected in 4 / 5 samples. Statistically higher than background.

¦	In 6cd medium fish, detected in 55 / 108 samples. Statistically higher than background.

¦	In 6cd large fish, detected in 20 / 36 samples with the range of detected concentrations of 1.41E-03 to 4.21E-02 mg/kg.
Insufficient number of background samples to perform comparison. Detected in 3 / 15 samples in background at a range of
9.28E-04 to 2.31E-03 mg/kg.

¦	Not detected in sediment, surface water, or soil.

Cis-Nonachlor

Yes

¦	Categories considered in the Tier III evaluation: each subarea/category based on frequency of detection only.

¦	No benchmark available.

¦	In 5a small fish, detected in 2 / 2 samples. Concentrations higher than the background prediction interval.

¦	In 5a medium fish, detected in 46 / 46 samples. Statistically higher than background.

¦	In 5a large fish, detected in 2 / 2 samples. Concentrations higher than the background prediction interval.

¦	In 5bc small fish, detected in 5 / 6 samples. Statistically higher than background.

¦	In 5bc medium fish, detected in 87 / 94 samples. Statistically higher than background.

¦	In 5bc large fish, detected in 31 / 32 samples. Statistically higher than background.

¦	In 6cd small fish, detected in 5 / 5 samples. Statistically higher than background.

¦	In 6cd medium fish, detected in 107 /108 samples. Statistically higher than background.

¦	In 6cd large fish, detected in 34 / 36 samples. Statistically higher than background.

¦	Not detected in sediment, surface water, or soil.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-142

Tier III Evaluation Results - Fish

Chemical

COPC?

Discussion

Trans-Nonachlor

Yes

¦	Categories considered in the Tier III evaluation: each subarea/category based on frequency of detection only.

¦	No benchmark available.

¦	In 5a small fish, detected in 2 / 2 samples. Concentrations within the background prediction interval.

¦	In 5a medium fish, detected in 46 / 46 samples. Statistically higher than background.

¦	In 5a large fish, detected in 2 / 2 samples. Concentrations higher than the background prediction interval

¦	In 5bc small fish, detected in 6 / 6 samples. Not statistically different from background.

¦	In 5bc medium fish, detected in 94 / 94 samples. Statistically higher than background.

¦	In 5bc large fish, detected in 30 / 32 samples. Statistically higher than background.

¦	In 6cd small fish, detected in 5 / 5 samples. Not statistically different from background.

¦	In 6cd medium fish, detected in 108 /108 samples. Statistically higher than background.

¦	In 6cd large fish, detected in 35 / 36 samples. Statistically higher than background.

¦	Not detected in sediment, surface water, or soil.

Oxychlordane

Yes

¦	Categories considered in the Tier III evaluation: every subarea/category except for 6cd small fish based on frequency of
detection only.

¦	No benchmark available.

¦	In 5a small fish, detected in 1 / 2 samples. Concentrations higher than the background prediction interval.

¦	In 5a medium fish, detected in 20 / 46 samples. Statistically higher than background.

¦	In 5a large fish, detected in 1 / 2 samples. Not statistically different from background.

¦	In 5bc small fish, detected in 5 / 6 samples. Statistically higher than background.

¦	In 5bc medium fish, detected in 67 / 94 samples. Statistically higher than background.

¦	In 5bc large fish, detected in 26 / 32 samples. Statistically higher than background.

¦	In 6cd medium fish, detected in 40 / 108 samples. Statistically higher than background.

¦	In 6cd large fish, detected in 17 / 36 samples. Statistically higher than background.

¦	Not detected in sediment, surface water, or soil.

Dioxins/Furans

Yes

¦ Of concern based on frequency of detection only, no benchmark derived.

PCBs

Yes

¦ Detected in >50% of samples and exceeds benchmark in >50% of detected samples.

O:\20123001.096\ERA\App B\Tables B-57 B-139 to B-142.doc

-------
Table B-143

Nondetected Chemicals with SQLs Greater than Benchmarks

Sediment

Surface Water

Soil

Aldrin

4,4'-DDD

Aldrin

Gamma-BHC

4,4'-DDT

Dieldrin

Chlordane

Chlordane

Endosulfan I

Dieldrin

Dieldrin

Endosulfan I

Endosulfan

Endosulfan II

Endrin

Endrin

Endrin

Gamma-BHC

Heptachlor epoxide

Gamma-BHC

Heptachlor

Methoxychlor

Heptachlor

Heptachlor Epoxide

Benzyl alcohol

Heptachlor epoxide

Methoxychlor

Hexachl orob enzene

Methoxychlor

1,2-Dichlorobenzene



Toxaphene

2,4-Dichl orophenol



4-Bromophenyl phenyl ether

2,4-Dimethylphenol



4-Nitrophenol

2,4-Dinitrotoluene



Aniline

2-Chlorophenol



Anthracene

Bis(2-chloroethyl)ether



Benzyl alcohol

Bis(2-chloroisopropyl)ether



Dibenzofuran

Hexachl orobutadi ene



Dinoseb





Hexachl orob enzene





Hexachl orobutadi ene





Hexachlorocyclopentadiene





Pentachl orob enzene





Pentachl orophenol





Phenol



O:\20123001.096\ERA\App B\Tables B-143 - B-148.doc

-------
Table B-144

Detected Chemicals with SQLs Greater than Benchmarks Recommended
for Uncertainty Analysis Discussion Instead of Quantitative Evaluation

Sediment

Surface Water

2-Methylphenol

Benzo(a)anthracene

Phenol

Benzo(a)pyrene

alpha-BHC

Fluoranthene

beta-BHC

Pyrene

4,4'-DDD

Silver, total

4,4'-DDE



4,4'-DDT



O:\20123001.096\ERA\App B\Tables B-143 - B-148.doc

-------
Table B-145

Final COPCs - Sediment

Chemical

5a

5b

5c

6ab

6cd

MC&
AB

SCOX

VP

MC&
AB

SCOX

VP

MC&AB

SCOX

VP

MC&AB

SCOX

VP

Pond

Semivolatiles

Dibenzofuran

X

—

—

—

—

—

—

—

—

—

—

—

—

PAHs

Acenaphthene

X

—

—

—

—

—

X

—

—

—

—

—

—

Acenaphthylene

X

—

—

—

—

—

X

—

—

—

—

—

—

Anthracene

X

—

—

—

—

—

X

—

—

—

—

—

—

Benzo(a)anthracene

X

—

X

X

X

X

X

—

X

—

—

—

—

Benzo(b)fluoranthene

X

X

X

X

X

X

X

—

X

X

—

—

X

B enzo (k)fluoranthene

X

—

X

—

—

—

X

—

X

—

—

—

—

Benzo(g,h,I)perylene

X

X

X

X

X

X

X

—

X

X

—

—

X

Benzo(a)pyrene

X

—

X

X

X

X

X

—

X

—

—

—

—

Chrysene

X

—

X

—

X

X

X

—

X

—

—

—

—

Dibenzo(aJi)antliracene

X

X

X

—

—

X

X

—

X

—

—

—

—

Fluoranthene

X

—

—

—

—

—

X

—

X

—

—

—

—

Fluorene

X

—

—

—

—

—

X

—

—

—

—

—

—

Indeno (1,2,3 -CD)py rene

X

X

X

X

X

X

X

—

X

X

—

—

X

Naphthalene

X

—

—

—

—

—

X

—

—

—

—

—

—

Phenanthrene

X

—

—

X

X

—

X

—

X

—

—

—

—

Pyrene

X

—

—

X

X

X

X

—

X

—

—

—

—

Dioxins/Furans

X

X

X

X

X

X

X

X

X

X

—

—

X

PCBs

X

X

X

X

X

X

X

X

X

X

X

X

X

Metals

Antimony

—

—

—

—

—

—

X

—

X

—

—

—

—

Barium

—

—

—

—

—

—

—

—

X

X

—

—

—

Beryllium

—

—

—

—

—

—

—

X

X

—

—

—

—

Cadmium

—

—

—

—

—

—

X

—

X

X

—

—

—

Chromium

—

—

—

—

—

—

X

—

X

X

—

—

X

Copper

—

—

X

—

—

X

X

—

X

X

—

—

X

O:\20123001.096\ERA\App B\Tables B-143 - B-148.doc

-------
Table B-145

Final COPCs - Sediment

Chemical

5a

5b

5c

6ab

6cd

MC&
AB

scox

VP

MC&
AB

SCOX

VP

MC&AB

scox

VP

MC&AB

scox

VP

Pond

Lead

—

—

X

—

—

X

X

—

X

—

—

—

X

Mercury

—

—

X

—

X

X

X

X

X

X

—

—

X

Selenium

—

—

—

—

—

—

—

—

X

—

—

—

X

Silver

—

—

X

—

X

X

X

—

X

X

—

—

X

Tin

—

—

—

—

—

X

X

—

X

X

—

—

X

O:\20123001.096\ERA\App B\Tables B-143 - B-148.doc

-------
Table B-146
Final COPCs - Surface Water

Chemical

Modeling Reach/Geomorph

5a

5

)

5c

6cd

MC& AB

VP

MC& AB

VP

MC & AB

VP

Pond

Dioxins/Furans

X

X

X

X

X

X

X

PCBs

X

X

X

X

X

X

X

O:\20123001.096\ERA\App B\Tables B-143 - B-148.doc

-------
Table B-147
Final COPCs - Soil

Chemical

Modeling Reach/Geomorph

5a

5

)

5c

6cd

Floodplain

Riverbank

Floodplain

Riverbank

Floodplain

Riverbank

Pond

Semivolatiles















Dibenzofuran

X

X

X

X

—

—

—

PAHs















Benzo(a)pyrene

—

X

—

—

—

—

—

Pyrene

—

X

—

—

—

—

—

Dioxins/Furans

X

X

X

X

X

X

X

PCBs

X

X

X

X

X

X

X

Metals















Chromium

X

X

X

X

X

X

X

Lead

—

—

—

—

X

—

X

Mercury

—

X

X

X

X

X

X

Selenium

—

—

—

—

X

—

X

O:\20123001.096\ERA\App B\Tables B-143 - B-148.doc

-------
Table B-148
Final COPCs - Risk Assessment Fish

Chemical

Modeling Reach/Fish Size

5a

5bc

6cd

Small

Medium

Large

Small

Medium

Large

Small

Medium

Large

Pesticides

4,4'-DDE

—

—

—

—

—

X

—

X

X

0,p'-DDT

X

X

X

X

X

X

X

X

X

4,4'-DDT

—

X

—

—

X

X

—

—

X

Heptachlor epoxide

X

X

X

—

—

X

X

X

X

Cis-Nonachlor

X

X

X

X

X

X

X

X

X

Trans-nonachlor

X

X

X

X

X

X

X

X

X

Oxychlordane

X

X

X

X

X

X

—

X

X

Dioxins/Furans

X

X

X

X

X

X

X

X

X

PCBs

X

X

X

X

X

X

X

X

X

O:\20123001.096\ERA\App B\Tables B-143 - B-148.doc

-------




/d



Click Here to go to
Figures Folder



/





MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PREERA FOR PDF.DOC

7/11/2003

-------
ATTACHMENT B.2
FISH TISSUE RE-ANALYSIS AND INTERPRETATION OVERVIEW

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

ATTACHMENT B.2
FISH TISSUE RE-ANALYSIS AND INTERPRETATION OVERVIEW

After a detailed review of the pesticide analysis results for fish tissue, it was decided by the risk
assessment team that a more sensitive evaluation of pesticide concentrations in fish tissues was
warranted to confirm or deny any potential interference from PCB concentrations. It was agreed
that this analysis was necessary before making a final determination of whether or not these
contaminants should be incorporated in the baseline risk assessment process for the PSA. To
achieve this objective, 10 fish tissues extracts that were keep frozen after the initial sample
analysis, were submitted to GERG for re-analysis using a more sensitive analytical method
(GC/MS). Chemicals included in this re-analysis were limited to 11 key pesticides that were
identified by the human health and ecological risk assessment teams as contaminants of potential
concern for one or both of the baseline risk assessments being conducted for the Lower
Housatonic River.

The results of the GC/MS re-analysis indicated that pesticide concentrations and detection limits
associated with the initial GC/ECD analysis may have overestimated actual pesticide
concentrations in fish tissue. The following list of chemicals and percent reduction (in average
concentrations) reflect how much the GC/ECD method may have overestimated actual
concentrations:

4,4'-DDD: 84 % reduction

4,4'-DDE: 82% reduction

4,4'-DDT: 92% reduction

0,P'-DDD: 99% reduction

0,P'-DDE: 93% reduction

0,P'-DDT: 99% reduction

Cis-Nonachlor: 99% reduction

Dieldrin: 94% reduction

Heptachlor epoxide: not detected by GC/MS

Oxychlordane: 82% reduction

Trans-Nonachlor: 11% reduction

While the results of this re-analysis does not, by itself, suggest the elimination of any pesticide as
contaminants of potential concern in fish tissue (with the possible exception of heptachlor
epoxide), it does suggest that the incorporation and analysis of these contaminants as
contaminants of potential concern in the ecological risk assessment should be done only after
careful evaluation of the relative impact each contaminant is likely to have on each assessment
endpoint assessed using fish tissue data. Therefore, the analysis and use of this data throughout
the baseline ERA was done on an endpoint by endpoint basis, and justification for their exclusion
was provided whenever warranted.

MK01 |O:\20123001,096\ERA_PB\APP B\APPB_PB_PREERA.DOC

-------
APPENDIX C
SUPPORTING TECHNICAL INFORMATION

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

-------
TABLE OF CONTENTS

Section	Page

C. SUPPORTING TECHNICAL INFORMATION	C-l

C. 1 APPROACH TO REVIEW OF HISTORICAL DATA SETS	C. 1-1

C. 1.1 Protocol for Review of Historical Data Sets for Usability in the

Housatonic River Project	C. 1-3

C.1.2 Summary Scores for Data Sets Identified and Evaluated	C.l-15

C. 1.3 Narrative Summaries for Evaluated Data Sets	C. 1-21

C.2 APPROACH FOR TREATING NON-DETECTS IN DATA ANALYSIS

I OR THE HOUSATONIC RIVER ERA	C.2-1

C.2.1 Background	C.2-1

C.2.2 Available Approaches	C.2-2

C.2.3 Discussion of Approaches	C.2-3

C.2.4 Approach for the Housatonic River ERA	C.2-4

C.2.5 Duplicate Samples with Non-Detects	C.2-7

C.2.6 Case Study	C.2-7

C.2.7 References	C.2-9

C.3 APPROACH TO SPATIAL WEIGHTING OF tPCB

CONCENTRATIONS IN FLOODPLAIN SOILS	C.3-1

C.4 APPROACH FOR USING MONTE CARLO AND PROBABILITY

BOUNDS ANALYSES	C.4-1

C.4.1 Introduction	C.4-1

C.4.2 Monte Carlo Simulation	C.4-1

C.4.3 Interval Probability or Probability Bounds	C.4-6

C.4.4 References	C.4-8

C.5 APPROACH FOR CALCULATING EXPOSURE POINT

CONCENTRATIONS	( .5-1

C.5.1 Introduction	C.5-1

C.5.2 UCL Estimation Methods	C.5-3

C.5.3 Spatial Weighting	C.5-20

C.5.4 Outliers	C.5-30

C.5.5 References	C.5-31

C.6 APPLICATIONS OF A WEIGHT-OF-EVIDENCE APPROACH	C.6-1

C.6.1	Introduction	C.6-1

C.6.2	Available Approaches	C.6-2

C.6.3	Approach	C.6-6

C.6.4	References	C.6-6

C.7 ANALYSIS OF PCB CONGENER COMPOSITION	C.7-1

C.7.1 Introduction	C.7-1

C.7.2 Methods	C.7-3

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	111	7/10/2003

-------
TABLE OF CONTENTS
(Continued)

Section	Page

C.7.3 Results	C.7-15

C.7.4 Conclusions	C.7-24

C.7.5 References	C.7-25

C.8 SUMMARY OF DATA COLLECTION ACTIVITIES	C.8-1

C.8.1 Introduction	C.8-1

C.8.2 Systematic Sampling	C.8-3

C.8.3 Discrete Sampling	C.8-7

C.8.4 Water Quality Sampling and Modeling Studies	C.8-44

C.8.5 References	C.8-54

C.9 SUMMARY OF ANALYTICAL METHODS	C.9-1

C.9.1 Introduction	C.9-1

C.9.2 Analytical Overview	C.9-1

C.9.3 Analysis of Water for PCBs	C.9-2

C.9.4 Analysis of Soil and Sediment for PCBs	C.9-4

C.9.5 Analysis of Biological Tissues for PCBs	C.9-8

C.9.6 Other Chemistry Analyses	C.9-11

C.9.7 Supplemental Analyses	C.9-11

C.9.8 References	C.9-16

C. 10 APPROACH FOR CALCULATING TOXIC EQUIVALENCE (TEQ)	C.10-1

C.10.1 Introduction	C.10-1

C. 10.2 Congener Non-Detects	C.10-1

C.10.3 Congener Co-Elution	C.10-3

C.10.4 References	C.10-9

C. 11 SUMMARY OF ANALYTICAL VARIABILITY	C. 11-1

C. 11.1 Introduction	C. 11-1

C.11.2 Quality Control Sample Types	C. 11-1

C.11.3 Results	C. 11-2

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	IV	7/10/2003

-------
LIST OF ACRONYMS

AD CP

Acoustic Doppler Current Profiler

Ah

aryl hydrocarbon

BB&L

Blasland, Bouck, and Lee

BW

body weight

CDD

chlorinated dibenzo-p-dioxin

CERC

Columbia Environmental Research Center

CLT

Central Limit Theorem

COC

contaminant of concern

COPC

contaminant of potential concern

DL

detection limit

DOC

dissolved organic carbon

DOM

dissolved organic matter

DQI

data quality indicator

DQO

data quality objective

EDTA

ethylene-diaminetetra-acetate

EE/CA

Engineering Evaluation and Cost Analysis

EPA

U.S. Environmental Protection Agency

EPC

exposure point concentration

EPRI

Electric Power Research Institute

ERA

ecological risk assessment

FIR

food intake rate

FMR

free metabolic rate

GC/ECD

gas chromatograph with electron capture detector

GE

General Electric Company

GERG

Geochemical and Environmental Research Group

HQ

hazard quotient

HRGC/HRMS

high-resolution gas chromatography /high-resolution mass spectrometry

HRMS

high-resolution mass spectrometry

IDW

inverse distance weighting

KS

Kolmogorov-Smirnov

LCL

lower confidence limit

LOAEL

lowest observed adverse effect level

LRMS

low-resolution mass spectrometry

MATC

maximum acceptable threshold concentration

MDEP

Massachusetts Department of Environmental Protection

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	V	7/10/2003

-------
LIST OF ACRONYMS
(Continued)

MDFW

Massachusetts Division of Fisheries and Wildlife

MSU

Michigan State University

MVUE

minimum-variance unbiased estimators

^g/kg

micrograms per kilogram

Hg/L

micrograms per liter

ng/L

nanograms per liter

NEA

Northeast Analytical, Inc.

OC

organic carbon

ORNL

Oak Ridge National Laboratory

PAH

polycyclic aromatic hydrocarbon

PAL

Project Action Limit

PARCC

precision, accuracy, representativeness, completeness, and comparability

PCB

polychlorinated biphenyl

PCH

polychlorinated hydrocarbon

P/E

performance evaluation

PL-PCB

planar polychlorinated biphenyl

PPb

parts per billion

ppm

parts per million

PQL

Practical Quantitation Limit

PRP

potentially responsible party

PSA

Primary Study Area

QA/QC

quality assurance/quality control

QAPP

Quality Assurance Project Plan

QC

quality control

RI

Remedial Investigation

RME

reasonable maximum exposure

RPD

relative percent difference

SAP/DCAQAP

Sampling and Analysis Plan and Data Collection and Analysis Quality Assurance



Plan

SI

Supplemental Investigation

SIM

selected ion monitoring

SIWP

Supplemental Investigation Work Plan

SOP

standard operating procedure

SOW

Scope of Work

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC VI 7/10/2003

-------
LIST OF ACRONYMS
(Continued)

SPE

solid-phase extraction

STM

short-term measure

SVOC

semivolatile organic compound

TCDD

2,3,7,8-tetrachlorodibenzo-p-dioxin

TEF

toxic equivalency factor

TEQ

toxic equivalence

TIE

toxicity identification evaluation

TOC

total organic carbon

TOM

total organic matter

tPCB

total PCB

TSS

total suspended solids

UCL

upper confidence limit

USACE

U.S. Army Corps of Engineers

USFWS

U.S. Fish and Wildlife Service

USGS

U.S. Geological Survey

VDC

vertical definition core

WHO

World Health Organization

WOE

weight-of-evidence

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	Vll	7/10/2003

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

APPENDIX C
SUPPORTING TECHNICAL INFORMATION

INTRODUCTION

Appendix C to the GE/Housatonic River Site Ecological Risk Assessment (ERA) comprises a
series of 11 technical documents or "white papers" that discuss the approach to a variety of
issues applicable to the ERA in a generic sense. These supporting technical documents have
been prepared at various times during the approximately 3 years of focused investigations
performed under the Supplemental Investigation Work Plan (SIWP) to augment the information
documented in the SIWP or in other SOPs, or to describe an approach or activity that was
common to more than one effort. Collectively, they represent the consensus of the project team
and were used as appropriate to guide the conduct of the work.

In general, technical documents dealing with issues that required resolution early in the risk
assessment process (e.g., Appendix C.l, Approach to Review of Historical Data Sets, and
Appendix C.6, Applications of a Weight-of-Evidence Approach in the Housatonic River ERA)
were prepared earlier in this 3-year period, while documents dealing with issues that evolved
from the completion of the sampling and analysis programs (e.g., Appendix C.8, Data Collection
Activities, and Appendix C.l 1, Summary of Analytical Variability) were prepared more recently.
In some cases, the issue dealt with in an appendix is focused specifically on the ecological risk
assessment process (e.g., Appendix C.6, which deals with a technical issue unique to ecological
risk assessment). In other appendices, the technical issue is applicable to risk assessment
generally and applies equally to the Human Health Risk Assessment being conducted for the
GE/Housatonic River site (e.g., Appendix C.5, which concerns calculation of exposure point
concentrations and their associated confidence intervals). In yet other cases, the appendix
documents procedures that apply across the entire spectrum of studies being conducted by EPA
at the GE/Housatonic River site (e.g., Appendix C.9, which provides an overview of analytical
methods).

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C-l

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

The 11 individual technical documents included in this appendix are as follows:

¦	Appendix C. 1: Approach to Review of Historical Data Sets

¦	Appendix C.2: Approach for Treating Non-Detects in Data Analysis

¦	Appendix C.3: Approach to Spatial Weighting of tPCB Concentrations in Floodplain
Soils

¦	Appendix C.4: Approach for Using Monte Carlo and Probability Bounds Analyses

¦	Appendix C.5: Approach for Calculating Exposure Point Concentrations

¦	Appendix C.6: Applications of a Weight-of-Evidence Approach

¦	Appendix C.7: Analysis of PCB Congener Composition

¦	Appendix C.8: Summary of Data Collection Activities

¦	Appendix C.9: Summary of Analytical Methods

¦	Appendix C. 10: Approach for Calculating Toxic Equivalence (TEQ)

¦	Appendix C. 11: Summary of Analytical Variability

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C-2

-------
APPENDIX C.1
APPROACH TO REVIEW OF HISTORICAL DATA SETS

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

-------
1

APPENDIX C.1
APPROACH TO REVIEW OF HISTORICAL DATA SETS

In addition to the large volume of data collected specifically to support this Ecological Risk
Assessment (ERA) and other studies of the Housatonic River conducted for the U.S.
Environmental Protection Agency (EPA), a large number of studies of the river have been
conducted historically over the last approximately 25 years. Much of this work has been
conducted by and for General Electric Company (GE) and was provided to EPA as part of both
formal and informal data exchange agreements between the parties. Additional studies and data
sets, generally smaller and less comprehensive, were not included in these data exchanges and/or
were developed independently of EPA and GE. In total, these additional studies and data sets
comprise nearly 100 additional studies.

At the initiation of the ERA, it was decided that the assessment would draw upon all identified
and available sources of information and data to the extent that those data could be demonstrated
to be of known quality, or could at least be evaluated to the extent necessary to identify potential
limitations on their use. To achieve this objective, the Housatonic River Project Quality
Assurance Office developed a protocol for data set evaluation based on existing EPA guidance
documents. The protocol describes a suite of criteria and a ranking system for each criterion that
can be used to develop an overall score for the quality and "usability" of the data set. This
procedure was applied to each of the identified data sets, and a score ranging from A (indicating
a very robust data set) to D (indicating a data set with significant use limitations) was assigned.
In addition, a brief narrative was prepared documenting the reasons for the score(s) assigned and
providing additional background information on the data set. All this information was provided
to the risk assessors for their use in determining if and how certain data would be considered in
the ERA.

This appendix describes the procedure followed to evaluate these various data sets and provides
the results of that evaluation for all data sets known to pertain to the ERA for the Housatonic
River. Appendix C.l.l is the evaluation protocol, including a summary matrix of the evaluation
criteria with a brief summary of the guidance used to assign scores to each data set. Appendix

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C.l-1

-------
1	C. 1.2 is a spreadsheet that lists the data sets identified and evaluated and provides a few pertinent

2	facts about each, including the summary score assigned for overall usability. Appendix C.1.3

3	includes the narrative summaries for each of the evaluated data sets.

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C.l-2

-------
APPENDIX C.1.1

PROTOCOL FOR REVIEW OF HISTORICAL DATA SETS FOR
USABILITY IN THE HOUSATONIC RIVER PROJECT

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

C.1.1 Protocol for Review of Historical Data Sets for Usability in the Housatonic
River Project

C.1.1.1 Background

A number of historical data sets exist for the Housatonic River (see Appendix C. 1.2). These data
sets needed to be evaluated to determine if and how the data might be used in the ecological risk
assessment and other components of the Housatonic River Project. The evaluation process must
be rigorous and transparent. This protocol describes a procedure comprising six criteria that
were used to determine the usability of data sets and provides guidance on the application of
these criteria to the review of historical data sets.

C. 1.1.2 Recommended Procedure

The process for evaluating data sets against the six criteria is summarized in Table C.l-1,
"Proposed Decision Criteria Matrix." The six criteria to be used in evaluating historical data sets
are as follows:

Criterion 1:	Overall quality and level of detail in reports

Criterion 2:	Formal documentation of procedures

Criterion 3:	Analytical methods used and detection limits achieved

Criterion 4:	Data review, validation, and quality assurance

Criterion 5:	Assessment of data quality indicators

Criterion 6:	Data history and overall apparent data quality

These evaluation criteria are similar to those described in Guidance for Data Usability in Risk
Assessment (EPA 1992), but have been modified to better fit the needs of the Housatonic River
Project. EPA Criterion III ("Data Sources") was found to be not applicable because it deals with
determining whether a single study is sufficiently comprehensive to have considered all or most
COPCs. As this issue has already been adequately investigated using the current data, it is not a
factor in evaluating the usability of historical data sets. Criterion 6, which does not appear in the
EPA guidance, was created to allow consideration of the age of a data set and to allow a
somewhat more subjective evaluation of the apparent overall quality of the study from which it
was developed. Each of the six criteria is defined in terms of four levels of usability:

¦ Level A: Acceptable, unrestricted use

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	C 1 3	7/10/2003

-------
1	¦ Level B: Acceptable, some use restrictions may apply

2	¦ Level C: Conditionally acceptable for limited uses

3	¦ Level D: Conditionally acceptable, use with caution

4

5	The remainder of this protocol provides detailed guidance for evaluating each data set and

6	assigning a score for each criterion. In addition to a separate score for each of the criteria, each

7	data set was assigned an overall score that was equivalent to the lowest score applied to any

8	single criterion, e.g., a data set that was ranked Level A for four of the criteria and Level B for

9	two was considered Level B overall. It is important to note that the results of this procedure do

10	not determine whether a data set may be used for a particular purpose but rather are intended to

11	alert investigators to potential limitations in the data. The decision to use or not use a given data

12	point or data set remains the responsibility of the individual investigator and must be made in the

13	context of the particular study.

14	Criterion 1: Overall quality and level of detail in reports

15	Overview: This criterion applies to the technical report and/or narrative that accompanies a data

16	set. This information is needed to evaluate the study design and procedures, allowing a

17	determination of the likely overall quality of the data. It also allows the data evaluator to

18	determine if the procedures were followed properly or if there were any deviations from the

19	work plan. In general, the more of this type of information that is provided to support a data set,

20	the greater the degree of confidence in the data. Isolated data sets, i.e., those that are not

21	supported by sufficient background information, cannot be evaluated fully for usability and

22	therefore, can be considered usable only if the investigator considers carefully the potential

23	issues surrounding their use and employs caution in any decision to use such data.

24	Like all criteria discussed in this protocol, four different conditions are described that result in a

25	data set being scored from Level A (the highest score, indicating a data set can be used without

26	restriction) to Level D (conditionally usable with caution). Data evaluators should score data

27	sets following these descriptions. It is recognized that this process is somewhat subjective, and

28	evaluators are expected to use professional judgment in awarding a score.

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C.l-4

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Level A: Acceptable, unrestricted use

For this criterion, a Level A data set must be accompanied by a narrative report that provides
complete details of the study design and includes at least some discussion of the underlying
reasons for selecting the stated sampling locations and methods. The sampling locations must be
provided accurately and precisely, and the procedure(s) used to locate the stations should also be
provided. The analytical methods followed should be fully described, including supporting
information such as detection limits, qualifiers, and procedures for handling non-detects.

Level B: Acceptable, some use restrictions may apply

A Level B data set is one accompanied by a narrative report that generally provides an adequate
description of the study and its methods, but does not meet the stringent requirements for Level
A. Examples of deficiencies that might cause a data set to be downgraded to Level B include
failure to specify how sampling stations were located, or failure to specify how non-detects were
treated. In such cases, the data are considered to be generally usable, but some consideration
should be given to the potential for reaching erroneous conclusions if, to continue with the two
previous examples, the sampling locations were only approximately located or if non-detects
were reported as blanks or zero values. This evaluation must be performed in the context of the
actual use of the data by each investigator.

Level C: Conditionally acceptable for limited uses

The intention of the Level C score for this criterion is to identify data sets that are accompanied
by reports that are largely insufficient for proper evaluation, but that may contain data on
parameters for which such limitations are less important, or for uses, such as trend analysis, that
may not require data that can be rigorously reviewed. It is also intended to apply to data sets that
may have certain critical historical data that are necessary for a particular study and cannot be
obtained from another source. In such cases, the investigators must proceed carefully and
understand the limitations that will likely be imposed on their conclusions.

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C.l-5

-------
1 Level D: Conditionally acceptable, use with caution

2 Level D data sets, in general, will exist independently of a written narrative report and therefore,

3 will not be dependable with regard to design and methodology. Such data sets should not be

4 used unless there is no reasonable alternative data source.

5 Criterion 2: Formal documentation of procedures

6 Overview: This criterion applies to what is thought of as "formal" quality assurance

7 documentation that is currently required for all studies done under contract to EPA and is also

8 typically prepared for studies that have a reasonable probability of being closely scrutinized,

9 particularly as part of legal proceedings. This documentation consists of four general types of

10 records: Work Plans and/or Quality Assurance Project Plans (QAPP), chain-of-custody, standard

11 operating procedures (SOPs) or protocols, and field/analytical records.

12 Work Plans, which may be separate from or combined with a QAPP, describe the procedures to

13 be employed in a study to ensure that the work is conducted properly and completely. They are

14 expected to be complete and prepared in sufficient detail so that different properly trained

15 professionals could conduct the work scope in precisely the same manner.

16 Chain-of-custody, at a minimum, allows the reviewer to ensure that a data point is clearly linked

17 to a particular geographic location and date/time. So-called "full-scale" chain-of-custody is the

18 documentation that also ensures a particular sample has been handled properly and not tampered

19 with. In general, full-scale chain-of-custody is necessary for enforcement or cost recovery.

20 SOPs or protocols are written detailed procedures that describe clearly how the components of a

21 study (typically, field and laboratory procedures) are to be conducted. In general, the term SOP

22 applies to "standardized" procedures that are usually part of a company's routine way of

23 conducting business and are applicable to all projects; protocols are specialized or non-routine

24 procedures that may be prepared for a specific project or task. The same level of detail applies to

25 either, and the two terms are intended to be equivalent for the purposes of this data evaluation.

26 SOPs/protocols may be incorporated into Work Plans or QAPPs or may be stand-alone

27 documents.

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C.l-6

-------
1	Field and analytical records are less standardized, but are intended to provide a permanent record

2	of what was actually done as part of the study. Such records may be critical to resolving issues

3	in data interpretation and are necessary if a data set is to achieve a Level A rating.

4	Level A: Acceptable, unrestricted use

5	To achieve a Level A rating, a data set must be accompanied by the full suite of documentation

6	just described, including full-scale chain-of custody.

7	Level B: Acceptable, some use restrictions may apply

8	Level B for this criterion is intended to describe data sets that in general have the documentation

9	just described, but for which the documentation may be insufficient, inadequate, or poorly

10	prepared in some areas that are deemed to be non-critical. For example, a data set that appears to

11	have SOPs in place for the majority of the field procedures but is lacking SOPs for some

12	procedures may be graded Level B. Similarly, a data set that was sent to a recognized analytical

13	laboratory and analyzed using standard procedures may be graded Level B even if the actual

14	SOP from the laboratory cannot be obtained. This rating would also apply to data sets for which

15	the necessary documentation is not currently available but can be easily accessed or provided by

16	a third party if necessary.

17	Level C: Conditionally acceptable for limited uses

18	Level C for this criterion is primarily intended to apply to data sets that are lacking much of the

19	necessary documentation but are believed to be of high quality because of the evaluator's

20	knowledge regarding the source, i.e., the company or principal investigator. It is also intended to

21	apply to data derived from recognized laboratories that may no longer be in business or may be

22	difficult to correspond with for other reasons. In these cases, it is assumed that the study was

23	conducted in a manner consistent with a documented Level A or B study, but the documentation

24	was never prepared or is otherwise unavailable.

25	Level D: Conditionally acceptable, use with caution

26	Data sets for which none or very little of the required documentation is available and about

27	which there is insufficient information to qualify for Level C will carry the warning "use with

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	^ -> q 7/10/2003

-------
1	caution"—the choice of whether such data are acceptable for use in a study or for a particular

2	purpose will remain the responsibility of the individual investigator.

3	Criterion 3: Analytical methods used and detection limits achieved

4	Overview: This criterion concerns both the actual analytical methods used to develop the data

5	and the application of those methods to achieve sufficiently low detection limits. In general, it is

6	preferable that the methods used in a study are routine and federally documented. In practice,

7	this means either approved EPA methods or ASTM methods, with the EPA methods generally

8	being preferred. Although other types of analytical methods may be usable for a particular study

9	or study component if properly documented, there is an element of uncertainty introduced.

10	Detection limits actually achieved must be sufficiently low in comparison with concentrations

11	that are known or likely to be of concern for the particular project, by which is meant the "end

12	use" project (in this case the Housatonic River Project), not necessarily the project for which the

13	data were originally developed. The general expectation is that the Practical Quantitation Limit

14	(PQL) should be below the Project Action Limit (PAL), which dictates that the Method

15	Detection Limit (MDL) should generally be less than 20% of the PAL. MDLs and PQLs near

16	the PAL introduce additional uncertainty and may also compromise the identification of

17	particular analytes.

18	Level A: Acceptable, unrestricted use

19	To achieve a Level A rating, all analytes of interest in the data set must have been quantified

20	using standard EPA-approved analytical methods current as of the date the study was conducted,

21	or well-documented and accepted ASTM methods. MDLs achieved must be as specified in the

22	method descriptions. For truly unrestricted use, the PQLs should be at or below concentrations

23	known or expected, based on other information such as EPA guidance or criteria, to be of

24	concern (PALs).

25	Level B: Acceptable, some use restrictions may apply

26	The Level B rating for this criterion is intended to apply to those data sets that were developed

27	using non-standard methods, but which have been sufficiently documented to satisfy the

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	C 1 8	7/10/2003

-------
1	evaluator that the data are equivalent in quality to data developed using EPA or ASTM methods.

2	Level B data sets for this criterion would also include data that were analyzed by EPA or ASTM

3	methods that have since been revised to improve detection limits or analyte identification but

4	which were current at the time of the study. Implicit in this criterion is the assessment that the

5	modification to the procedures does not in some way invalidate the previous version of the

6	method.

7	Level C: Conditionally acceptable for limited uses

8	Level C data sets for this criterion would include data that were developed using non-standard

9	methods that have not been well-documented but which are believed to be of sufficient quality to

10	be used with consideration of their potential limitations. In general, this level is intended to

11	apply to data sets that might have been developed using experimental or developmental methods

12	by highly qualified firms, laboratories, or individuals.

13	Level D: Conditionally acceptable, use with caution

14	Data sets developed using unknown analytical methods, developed using non-standard or poorly

15	documented methods about which nothing more is known, or data sets developed using methods

16	that are otherwise considered to derive from questionable methods will be judged to be Level D.

17	Criterion 4: Data review, validation, and quality assurance

18	Overview: This criterion deals with the range and variety of QA and QC methods available to

19	ensure that the data are of known quality. These include such methods and procedures as blank

20	samples, spikes, and duplicates. Further, it also concerns the review conducted on the data

21	following receipt from the analytical laboratory; such review typically falls into two categories:

22	various data completeness reviews, and formal validation. The latter usually requires that the

23	appropriate QC procedures were built into the sample collection and analysis process.

24	Level A: Acceptable, unrestricted use

25	The Level A rating for this criterion is reserved for data sets that have undergone a formal

26	validation process. Although it is preferable that all data in the data set were part of batches that

27	were formally validated, the data set may be judged to be Level A if the level of validation was

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	^ -> q	7/10/2003

-------
1	reduced to a subset of the data for well-documented reasons consistent with known quality of

2	laboratory performance. For example, the WESTON tissue data being developed by GERG

3	would be considered a Level A data set in spite of the fact that currently only approximately 15%

4	of the data receive formal validation. This reduction was warranted by consistently high

5	performance at GERG, which allowed the level of validation to be reduced as a cost-saving

6	measure.

7	Level B: Acceptable, some use restrictions may apply

8	Level B data sets are those which have been subjected to a rigorous data review that has been

9	fully described and documented but which have not received formal data validation. Such a

10	review would typically include examination of completeness and should be accompanied by data

11	for blanks and duplicates. Another example of a Level B data set for this criterion might be a

12	data set that is accompanied by satisfactory data from performance evaluation (P/E) samples but

13	has not had formal data validation. It is assumed that a Level B study would have been

14	conducted with established written QA/QC procedures and that a review is conducted to ensure

15	compliance with these procedures.

16	Level C: Conditionally acceptable for limited uses

17	A rating of Level C will be applied to data sets that have received limited documented review or

18	for which QA/QC procedures were not properly specified, but which are believed to be of

19	reasonable quality due to other known factors.

20	Level D: Conditionally acceptable, use with caution

21	Data sets that have received no documented review or for which the level of review is not known

22	will be considered Level D.

23	Criterion 5: Assessment of data quality indicators

24	Overview: Data quality indicators (DQIs) are a means of defining data quality in terms of data

25	quality objectives. This criterion is concerned with the following five DQIs: precision, accuracy,

26	representativeness, completeness, and comparability (PARCC) and the additional DQI of

27	sensitivity, which is related to Criterion 3. As part of the evaluation of a data set for Criterion 5,

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	^ ia	7/10/2003

-------
1 each of these DQIs must be evaluated against the goals established in the planning phase of the

2 study. A detailed description of the individual DQIs and their application is beyond the scope of

3 this protocol but both are readily available from EPA and other sources.

4 Level A: Acceptable, unrestricted use

5 To achieve a rating of Level A, data sets must have been developed as part of a study that had

6 pre-defined DQIs for all or most of the six parameters. Further, each of the DQIs should have

7 been substantially achieved by the study. Alternatively, if a study failed to achieve one or more

8 of its established DQIs but then provided a discussion of the implications of that failure and

9 concluded that the DQOs were still achieved, that study could also receive a Level A rating at the

10 discretion of the evaluator.

11 Level B: Acceptable, some use restrictions may apply

12 For this criterion, Level B is intended to apply to data sets that were developed without formal

13 DQIs being established as part of the planning process, but which did evaluate (or allow the

14 evaluator to obtain) the DQIs achieved after the fact. In effect, this rating indicates that DQIs

15 were achieved that were consistent with those for Level A data sets and that would likely have

16 been established had the planning process included them.

17 Level C: Conditionally acceptable for limited uses

18 Level C data sets include those data sets that also did not have DQIs established in the planning

19 phase of the study and, further, appear to have not satisfied what might be considered reasonable

20 standards for one or more of the non-critical DQI parameters (i.e., completeness, comparability).

21 For example, 90% is a typical completeness goal. A data set that established a completeness

22 goal of 90% and achieved it would (for this one parameter) be considered Level A. A data set

23 that achieved 90% completeness in the absence of a specified goal would be Level B. A data set

24 that achieved 70% completeness would be Level C. Data from such a data set may be used if, at

25 the discretion of the investigator, the failure to achieve a reasonable completeness did not unduly

26 limit or bias the data for a particular analyte.

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C.l-11

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

Level D: Conditionally acceptable, use with caution

Data sets are considered to be Level D for this criterion if it is not possible to evaluate the typical
DQIs or if the study failed to achieve a reasonable result for one or more of the critical DQIs.

Criterion 6: Data history and overall apparent data quality

Overview: This criterion is somewhat more subjective than the preceding ones and is intended to
allow the evaluator to exercise a greater degree of professional judgment regarding a data set.
Because of changes in methodology, both field and analytical, and the inability at times to obtain
answers to specific questions for older data sets, their use can be questionable. In addition, it is
recognized that conditions in the study area are changeable with time and data developed some
years previous may not represent present conditions. This criterion also recognizes that trained
evaluators may use many indicators, including personal knowledge of individuals and
organizations, that are not easily captured in an objective rating scheme.

Level A: Acceptable, unrestricted use

Level A will apply only to data sets developed in whole or in substantial part recently, typically
defined as within the last 5 years, and for which the evaluator has no reason to question their
validity. In addition, to qualify for Level A, the study that produced the data must have used
methods that are consistent with current practice and there should be some objective indication
that the proposed methods were actually followed conscientiously by the individuals conducting
the work. In effect, this rating indicates that the study is fully equivalent to the work currently
being conducted by WESTON and its subcontractors.

Level B: Acceptable, some use restrictions may apply

Level B for this criterion is essentially equivalent to Level A, but the study and data are older
than 5 years or the stringent standards of Level A with regard to methods and practices are either
not satisfied or cannot be determined. To qualify for Level B, however, the study must still have
produced data that are equivalent to what would have been produced using current
methodologies. Nonetheless, investigators should examine such data sets carefully to ensure that
the particular data and data uses would not be invalidated by the age of the data.

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC	C 1 12	7/10/2003

-------
1	Level C: Conditionally acceptable for limited uses

2	Level C applies if, in the professional opinion of the evaluator, portions of the data appear to be

3	of questionable quality based primarily on the methods used and/or the apparent adherence to

4	those methods during the performance of the work. Other data from the study may be usable,

5	but investigators should exercise caution and should use such data only if necessary.

6	Level D: Conditionally acceptable, use with caution

7	Data sets will be considered Level D if, in the professional opinion of the evaluator, the data are

8	of questionable quality due to methodology or any other reason. This assessment may be made

9	in spite of acceptable performance on any or all of the more objective criteria discussed above.

10	C.1.1.3 References

11	EPA (U.S. Environmental Protection Agency). 1992. Guidance for Data Usability in Risk

12	Assessment (Part A) Final. Office of Emergency and Remedial Response, Washington, DC.

13	PB92-963356.

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

C.l-13

-------
Table C.1-1

Proposed Decision Criteria Matrix for Evaluating Usability of Historical Data in Ecological Risk Assessment

Level A - Acceptable,
unrestricted use

Level B - Acceptable, some use
restrictions may apply

Level C - Conditionally
acceptable for limited uses

Level D - Conditionally
acceptable, use with caution

Criterion 1: Overall quality of
and level of detail in report(s)

Accompanying report provides
complete description of study
design and sample location(s) with
justification and rationale.

Report is generally complete and
well written but lacks sufficient
detail in a few areas. Sampling
locations specified, but not located
with GPS or equivalent.

Accompanying report is
incomplete but does provide
sufficient information for one or
more parameters of interest.
Sampling locations may not be
well specified.

No information available on
background and conduct of study.
Significant questions regarding
sampling locations.

Criterion 2: Formal
documentation of procedures

Work Plan, Quality Assurance
Plan, chain-of-custody records,
SOPs, and similar field and
laboratory documentation exists
and is available for review.

Documentation exists for most
areas but is insufficient or lacking
in a few areas considered non-
critical.

Documentation generally not
available but sufficient
information is known or available
via other sources to establish
validity of field and analytical
procedures.

Documentation non-existent, not
available for review, or status
unknown.

Criterion 3: Analytical methods
used and detection limits
achieved

Analytical procedures follow
documented standard methods
such as EPA or ASTM.

Analytical procedures non-
standard but sufficiently
documented to establish validity
of and ensure confidence in data.

Analytical procedures non-
standard and not well documented,
but data are believed to be valid
due to other information provided.

Insufficient information provided
or available via other sources to
establish validity of data.

Criterion 4: Data review,
validation, and quality
assurance

Study incorporated all or most of
the full range of QA/QC
procedures, e.g., blanks, spikes,
dups, data review, and data
validation.

Study generally employed and
documented established QA/QC
procedures but did not conduct
data validation.

Non-standard or incomplete
QA/QC procedures were
followed.

No QA/QC procedures employed
or documented.

Criterion 5: Assessment of data
quality indicators

Study had established data quality
indicators and data substantially
meet all acceptability criteria for
completeness, comparability,
representativeness, precision, and
accuracy.

Data quality indicators not
established, but data appear to
meet minimum standards for
DQIs.

Data quality indicators not
established; data appear to not
satisfy minimum standards for one
or more noncritical DQIs.

Data fail to meet minimum
standards for one or more critical
DQIs, or not possible to evaluate
DQIs.

Criterion 6: Data History and
Overall Apparent Data Quality

Data are recent (i.e., within past 5
years), reported in standard units,
and are reasonable and internally
consistent. Methods followed
meet current standards for
scientific investigation and were
followed consistently.

Data appear to be of acceptable
quality but derive from a study
conducted prior to 1995. Methods
may not meet current standards
but are judged to have produced
data equivalent to current
methodologies.

Portions of the data appear to be
of questionable quality due to age,
changes in methods, and/or failure
to follow current standards for
scientific investigation.

The overall data quality is
questionable due to outmoded
methodologies, poor performance,
and/or apparent lack of
consistency with current
standards.

MK01\O:\20123001.096\ERA_PB\ERA_APC1_TblC1-1_PB.xls C. 1-14 7/10/2003

-------
APPENDIX C.1.2

SUMMARY SCORES FOR DATA SETS IDENTIFIED AND EVALUATED

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

-------
Table C.1-2

Summary Scores for Data Sets Identified and Evaluated

Serial

RefNo.

Title

Comments

Estimate of Samples for Data Entry

Score

Reviewer*

1

02-0215

Academy ofNatural Sciences of Philadelphia. 1991. PCB Concentration in
Fishes from the Housatonic River, Connecticut, 1984 to 1990.

Upgradable to be based upon further review. In GE deliverable
(fish). Age distribution data. Composite PCB results and
averages. Report references RFW-02-214

194 PCB fish results for 1990. Separate table
for replicates.

C

SC

2

02-0249

Academy ofNatural Sciences of Philadelphia. 1993. PCB Concentrations in
Fishes from the Housatonic River, Connecticut, 1984-1992.

Most data contained in previous reports. Most current report
should be used as PCB quantitation may have changed. Report
contains results for two different methods for insect and crayfish
PCB quantification.

145 PCB fish results for 1992. Separate table
for replicates.

B

SC

3

02-0050

Academy ofNatural Sciences of Philadelphia. 1995. PCB Concentrations in
Fishes and Benthic Insects from the Housatonic River, Connecticut, 1984 to
1994.

In GE deliverable (fish)

168 fish results for 1994. Report also contains
fish age distribution. Separate table for
replicate analysis

B

KMS

4

02-0014

Beck, G. 1982. PCBs in Housatonic River Fish - Statistical Analysis.

PCB (fish)

476

C

KMS

5

02-0100

Blasland, BouckandLee. 1996. Evaluation of Housatonic River Sediment
and Floodplain Soil Data on Hazardous Constituents to Assess the Need for
Further Sampling.

AppIX



B

SC

6

02-0036

Blasland, Bouck and Lee, Inc. 1994. Housatonic River Floodplain
Properties, Results of Supplemental Site Characterization.

PCB

37

C

SC

7

04-0003

Blasland, Bouck and Lee, Inc. 1994. MCP Supplemental Phase II Scope of
Work and Proposal for RCRA Facility Investigation of Housatonic River and
Silver Lake.





B

PH

8

04-0006

Blasland, Bouck and Lee, Inc. 1994. Report on Silver Lake Short-Term
Measure Evaluation and Related Activities.

PCB

17

B

DJG

9

04-0004

Blasland, BouckandLee, Inc. 1996. Supplemental Phase II/RCRA Facility
Investigation Report for Housatonic River and Silver Lake, Volume I of II.
Report prepared for General Electric Company, Pittsfield, MA by BBL,
6723 Towpath Road, Syracuse, NY 13214.

No data tables with individual results contained in report.



B

SC

10

04-0004

Blasland, BouckandLee, Inc. 1996. Supplemental Phase II/RCRA Facility
Investigation Report for Housatonic River and Silver Lake, Volume II of II.
Report prepared for General Electric Company, Pittsfield, MA by BBL,
6723 Towpath Road, Syracuse, NY 13214.

GRSz, SpGr, Bulk Density, Cs-137, TO-10



B

SC

11

02-0237

Blasland and Bouck Engineers. 1984. Housatonic River Study, 45-Day
Interim Report, Remedial Alternatives Evaluation.

No data presented in report.



N/A

SC

12

02-0238

Blasland & Bouck Engineers, P.C. November 1984. Housatonic River
Study 90 Day Interim Report, Remedial Alternatives Evaluation, Flow and
Velocity Control, River Channelization, In-situ Impoundment.

No data presented in report.



N/A

SC

13

02-0234

Blasland and Bouck Engineers. 1985. Housatonic River Study, 135-Day
Interim Report, Remedial Alternatives Evaluation.

No data presented in report.



N/A

SC

14

02-0062

Blasland and Bouck Engineers. 1986. Housatonic River Study, 135-Day
Interim Report, Assessment of Remedial Alternatives (Addendum).

No data presented in addendum.



N/A

SC

15

02-0235

Blasland and Bouck Engineers. 1989. Housatonic River Velocity and
Sedimentation Control Pilot Study.

8 PCB results, no specific locations



D

SC

16

02-0071

Blasland and Bouck Engineers. 1991. MCP Interim Phase II Report/Current
Assessment Summary for Housatonic River.

App IX/PCB 137Frog/Fish/Turtle/Inv

780

C

SC

MK01\20123001.096\ERA_PB\ERA_APC1_TblC1-2_PB.xls	C. 1-15	7/10/2003

-------
Table C.1-2

Summary Scores for Data Sets Identified and Evaluated

Serial

RefNo.

Title

Comments

Estimate of Samples for Data Entry

Score

Reviewer*

17

02-0038/02-
0039

Blasland and Bouck Engineers. 1992. Addendum to MCP Interim Phase II
Report/Current Assessment Summary for Housatonic River, Volumes 1 and
2.

App IX/PCB



B

PH

18

02-0109

Blasland and Bouck Engineers. 1992. Plan for Evaluation of Need for Short-
Term Measures in Floodplain of Housatonic River.

Eval with 02-0097



A

SC

19

02-0103

Blasland and Bouck Engineers. 1992. Identification of Extent of, and Land
Use and Property Ownership within, Potentially Affected Area of
Housatonic River Floodplain.

App IX/PCB



C

SC

20

02-0097

Blasland and Bouck Engineers. 1992. Summary of Housatonic River
Floodplain Property Sampling and Analysis.

PCB/TOC Eval with 02-0109



A

SC

21

02-0109

Blasland and Bouck Engineers. 1992. Evaluation of Need for Short-Term
Measures in the Housatonic River Floodplain.





A

SC

22

02-0037

Blasland and Bouck Engineers. 1993. Report on January 1993 Housatonic
River Floodplain Property Sampling and Analysis.





B

DS

23

02-0240

Blasland and Bouck Engineers. 1993. Evaluation of Potential Short-T erm
Measures for Properties within the Housatonic River Floodplain.





N/A

PH

24

02-0241

Blasland and Bouck Engineers. 1993. Short-Term Measure Proposals for the
"60-day" Residential Properties within the Housatonic River Floodplain.





B

PH

25

02-0242

Blasland and Bouck Engineers. 1993. Short-Term Measure Proposals for the
"90-day" and "120-day" Properties within the Housatonic River Floodplain.





B

PH

26

04-0005

Blasland and Bouck Engineers. 1993. Silver Lake Data Summary.

Rpt data from previous rpt - small set inorg/PCB



C/D

SC

27

02-0102

Chadwick and Associates, Inc. 1993. Fisheries Investigation of the
Housatonic River, Massachusetts. Report prepared for General Electric Co.
by Chadwick and Associates, 5575 S. Sycamore Street, Littleton, CO.



0

C

DJG

28

02-0047

Chadwick and Associates, Inc. 1994. Aquatic Ecology Assessment of the
Housatonic River, Massachusetts, 1993. Report prepared for General
Electric Co., Pittsfield, MA, by Chadwick and Associates, Inc., 5575 S.
Sycamore Street, Littleton, CO.



0

C

DJG

29

02-0243

ChemRisk. 1993. Risk Assessment to Evaluate the Need for Short-Term
Measures in the Floodplain of the Housatonic River.





B

PH

30

02-0239

ChemRisk. 1993. Preliminary Health and Environmental Assessment
Proposal for the Housatonic River, Silver Lake, and their Floodplains.



0

N/A

SC

31

02-0104

ChemRisk; JSA Environmental; Chadwick Ecological Consultants;
ENVIRON. 1997. Work Plan for the Ecological Risk Assessment of the
Housatonic River Site, Volume I - Work Plan Text, Tables, and Figures,
Volume II - Appendices A (Sampling Maps) and B (Protocols). Report
prepared for General Electric Company, Pittsfield, MA, by ChemRisk, a
Division of McLaren/Hart, 1685 Congress Street, Portland, ME.





B

PH

MK01\20123001.096\ERA_PB\ERA_APC1_TblC1-2_PB.xls	C.l-16	7/10/2003

-------
Table C.1-2

Summary Scores for Data Sets Identified and Evaluated

Serial

RefNo.

Title

Comments

Estimate of Samples for Data Entry

Score

Reviewer*

32

02-0048

ChemRisk; S.G. Martin and Associates. 1994. Evaluation of the Terrestrial
Ecosystem of the Housatonic River Valley. Report prepared for General
Electric Company, Pittsfield, MA, by ChemRisk, a Division of
McLaren/Hart, 1685 Congress Street, Portland, ME, in collaboration with
S.G. Martin and Associates, 7121 N. County Road 9, Wellington, CO.

Report contains historical PCB concentrations plotted on maps.



C

SC

33

02-0251

ENVIRON Corporation. 1996. Preliminary Habitat and Biota Impact
Assessment of Sediment Remediation Technologies, GE Housatonic River
Study Area.

PCB/Gr Sz/TOC



B

PH

34

02-0016

Frink, C.R., Sawhney, B.L.,Kulp, K.P., Fredette, C.G. 1982.
Polychlorinated Biphenyls in Housatonic River Sediments in Massachusetts
and Connecticut; Determination, Distribution, and Transport. Connecticut
Agricultural Experiment Station Bulletin 800.

Sed 348 and water 54



B

SC

35

02-0025

Gay, F.G. and Frimpter, M.H. 1985. Distribution of Polychlorinated
Biphenyls in the Housatonic River and Adjacent Aquifer, Massachusetts.
USGS Water Supply Paper 2266.

Samples not located by GPS or lat/long

50

C

DJG

36

99-0276

GZA GeoEnvironmental, Inc. 1991. Sediment Sampling and Analysis Data
Report, Rising Paper Company, Great Barrington, Massachusetts.





C

DJG

37

02-0098

Harrington Engineering and Construction, Inc. 1996. Report on the
Preliminary Investigation of Corrective Measures for Housatonic River and
Silver Lake Sediment. Report prepared for General Electric Company,
Pittsfield, MA, by Harrington Engineering and Construction.



0

N/A

DJG

38

02-0203

Kapish, J. 1981. Summary of Fish PCB Data for 1977.

PCB (fish)



D

SC

39

02-0246

Kearsley, J. 1998. Inventory and Vegetation Classification of Floodplain
Forest Communities in Massachusetts. Final report submitted to the U.S.
Environmental Protection Agency in fulfillment of Wetland Protection-State
Development Grant #CD 001976-01-1 by the Massachusetts Natural
Heritage & Endangered Species Program, Massachusetts Division of
Fisheries and Wildlife, Route 135, Westborough, MA.





C

DS

40

02-0201

Kulp, K.P. 1991. Concentration and transport of Polychlorinated Biphenyls
in the Housatonic River Between Great Barrington, Massachusetts, and
Kent, Connecticut, 1984-1988.

PCB (water)

32

C

KMS

41

02-0248

Lawler, Matusky and Skelly Engineers. 1988. Housatonic River PCB
Sediment Management Study - Chapter 6, Program for Monitoring the
Natural Recovery of the River.





C

DJG

42

02-0090

Lawler, Matusky and Skelly Engineers. 1991. Ambient Trend Monitoring
and PCB Fate and Transport Model.

PCB (air)



B

SC

43

02-0099

Lawler, Matusky and Skelly Engineers, Inc. 1994. Housatonic River
Connecticut Cooperative Agreement - Task IV.B, PCB Fate and Transport
Model: Additional Monitoring and Model Verification.

In GE deliv/PCB (soil/water)

119/99/61/218/158

B

SC

44

02-0247

Lawler, Matusky and Skelly Engineers. 1985. Housatonic River PCB
Management and Study (Chapters 1 through 5).





D

DS

45

99-0403

Neumuth, E.J. 1991. Ninety-first Christmas Bird Count, Central Berkshire,
Massachusetts. American Birds 45(4): 621.

No analytical data



N/A

SC

MK01\20123001.096\ERA_PB\ERA_APC1_TblC1-2_PB.xls	C. 1-17	7/10/2003

-------
Table C.1-2

Summary Scores for Data Sets Identified and Evaluated

Serial

RefNo.

Title

Comments

Estimate of Samples for Data Entry

Score

Reviewer*

46

99-0404

Neumuth, E.J. 1992. Ninety-second Christmas Bird Count, Central
Berkshire, Massachusetts. American Birds 46(4): 613.

No analytical data



N/A

SC

47

99-0405

Neumuth, E.J. 1993. Ninety-third Christmas Bird Count, Central Berkshire,
Massachusetts. American Birds 47(4): 580.

No analytical data



N/A

SC

48

99-0406

Neumuth, E.J. 1994. Ninety-fourth Christmas Bird Count, Central
Berkshire, Massachusetts. FieldNotes 48(4): 456.

No analytical data



N/A

SC

49

99-0407

Neumuth, E.J. 1995. Ninety-fifty Christmas Bird Count, Central Berkshire,
Massachusetts. FieldNotes 49(4): 426-427.

No analytical data



N/A

SC

50

02-0030

Stewart Laboratories, Inc. 1982. Housatonic River Study, 1980 and 1982
Investigations.

PCB

721 fish, 892 sediment

C

SC

51

00-0309

Woodlot Alternatives. 1998. Preliminary Report - Wetland Characterization
and Function-Value Assessment Housatonic River from Newell Street to
Woods Pond. Report prepared for Tech Law, Inc., 160 Washington Street,
Boston, MA, by Woodlot Alternatives, Inc., 122 Main Street, Topsham,

ME.

Wetland characterization - no analytical.



N/A

DJG

52

01-0135

Zorex Environmental Engineers, Inc. 1992. Ambient Air Monitoring for
PCBs, August 20, 1991 through August 14, 1992. Report prepared for
General Electric, Pittsfield, MA, by Zorex Environmental Engineers.

PCB air)

19

B

DJG

53

01-0137

Zorex Environmental Engineers, Inc. 1993. Ambient Air Monitoring for
PCBs, May 4, 1993 to August 17, 1993. Report prepared for General
Electric, Pittsfield, MA, by Zorex Environmental Engineers.

PCB air)



B

SC

54

02-0254

Zorex Environmental Engineers, Inc.; Berkshire Environmental Consultants,
Inc. 1996. Ambient Air Monitoring for PCB, May 10, 1995 through
August 24, 1995. Report to General Electric Company, Pittsfield, MA by
Zorex Environmental Engineers, 247 South Street, Pittsfield, MA, and
Berkshire Environmental Consultants, 152 North Street, Pittsfield, MA.

Document in 04-0004



A

SC

55

02-0255

Coles, J.F. 1999. Length-age Relations and PCB Content of Mature White
Suckers from the Connecticut and Housatonic River Basins. Northeastern
Naturalist 6(3):263-275.





C

DJG

56

02-0121

Smith, S.B. and J.F. Coles. 1997. Endocrine Biomarkers. Organochlorine
Pesticides, and Congener Specific Polychlorinated Biphenyls (PCBs) in
Largemouth Bass (Micropterus salmoides) from Woods Pond, Housatonic
River, Massachusetts, September 1994 and May 1995. U.S. Geological
Survey Administrative Report. Prepared in cooperation with the U.S.
Environmental Protection Agency. Reston, Virginia.

No units for results. Pest/Congener

37

C

SC

57

02-0209

Breault, R.F. and S.L. Harris. 1997. Geographical Distribution and
Potential for Adverse Biological Effects of Selected Trace Elements and
Organic Compounds for Streambed Sediment in the Connecticut,
Housatonic, and Thames River Basins, 1992-94. U.S. Geological Survey
Water-Resources Investigation Report 97-4169, 24 pp.

Score may increase with additional information. No lat/long or
individual sample results. References for data, RFW 02-0210, 02

0212 and 02-0025



C

SC

58

02-0210

Coles, J.F. 1996. Organochlorine Compounds and Trace Elements in Fish
Tissue and Ancillary Data for the Connecticut, Housatonic, and Thames
River Basin Study Unit, 1992-94. U.S. Geological Survey Open-File
Report 96-358, 26 pp.

Portion of review scored lower than B. Mtls/Pest (fish). 32 loc w/
lat/long, 42 metal results, 32 PCB (fish), apprx 250 wt, lgth,
sex data.



B

SC

MK01\20123001.096\ERA_PB\ERA_APC1_TblC1-2_PB.xls	C. 1-18	7/10/2003

-------
Table C.1-2

Summary Scores for Data Sets Identified and Evaluated

Serial

RefNo.

Title

Comments

Estimate of Samples for Data Entry

Score

Reviewer*

59

02-0213

Garabedian, S.O., etal. 1998. Water Quality in the Connecticut,
Housatonic, and Thames River Basins, Connecticut, Massachusetts,
New Hampshire, New York, and Vermont, 1992-95. USGS Circular
1155.

Portion of review scored lower than B. Report references RFW 02
0209 and 02-0210.



B

SC

60

02-0212

Harris, Sandra L. 1997. Inorganic and Organic Constituents and Grain-Siz*
Distribution in Streambed Sediment and Ancillary Data for the Connecticut,
Housatonic, and Thames River Basins Study Unit, 1992-1994. USGS
Open-File Report 96-397.

Portion of review scored lower than B. Mtls/Pest/BNA/Inorg
(sed). 43 Sites w/ lat/long

43 locations w/ multiple analytes.

B

sc

61

02-0149

Blasland, Bouck & Lee. 1996. Report on Lower Housatonic River Sediment
PCB Sampling, December 1996.

PCB (sed)

79

B

KMS

62

02-0202

GEI Consultants. 1999. Task 220: Exploratory Site Investigation,
Replacement of Stevenson Dam Bridge CT Rt. 34) Monroe/Oxford,
Connecticut, September 3, 1999.

PCB (soil)

21

A

KMS

63

02-0206

CTDEP, Bureau of Water Management. 1992. Interdepartmental Message,
USGS 1992 Housatonic River Sediment PCB Data, December 21, 1992.

PCB (sed)

7

D

KMS

64

02-0204

State of Connecticut Department of Health Services. 1979. Housatonic River
PCB Fish Log Book.

PCB (fish)

392

D

KMS

65

02-0205

Sawhney, Frink and Glowa. 1981. PCBs in Housatonic River: Determination
and Distribution, Journal of Environmental Quality.



0

C

KMS

66

02-0207

Connecticut Post. 1993. Higher PCBs in Housatonic Feared, May 23,
1993.



0

D

KMS

67

02-0208

Environmental Research Institute, University of Connecticut. 2000.
Candlewood PCB Data, March 21, 2000.

PCB (no loc)

14

B

KMS

68

02-0218

CTDEP, Letter to Mr. Richard Thibedeau, Massachusetts Department of
Environmental Management from Michael J. Harder, April 6, 1994.

PCB (fish)

42

C

KMS

69

02-0221

Letter to D. Wetstone from RW. Berliner, Yale University, re: Review of
Health Services Study of PCB-contaminatedFish from the Housatonic River,
June 21, 1981.





D

KMS

70

02-0222

State of Connecticut Department of Health Services, Bureau of Health
Promotion & Disease Prevention. April 14, 1981. PCBs in the Housatonic
River Fish, Fact Sheet.





D

KMS

71

02-0220

Memorandum to M. Harder from C.G. Fredette, re: Summary of 1992
CTDEP Housatonic PCB Monitoring re: Rising Dam, Great Barrington,

MA, May 18, 1993.

PCB (sed)

10

C

KMS

72

02-0211

Coles, James F. 1998. Organochlorine Compounds in Fish Tissue from the
Connecticut, Housatonic and Thames River Basins Study Unit, 1992-1994,
Water-Resources Investigations Report 98-4075.

Pest (fish). Report references RFW 02-0050, 02-0025, 02-0212
and 02-0210.

3

B

KMS

73

02-0214

Academy of Natural Sciences of Philadelphia, Division of Environmental
Research. 1990. PCB Concentrations in Fishes from the Housatonic River,
Connecticut in 1984, 1986, and 1988, Report No. 89-30F, January 11,
1990.

In GE deliv (fish). A small number of sed data located in report.

1296

B

KMS

74

02-0216

Academy of Natural Sciences of Philadelphia, Patrick Center for
Environmental Research. 1999. PCB Concentrations in Fishes and Benthic
Insects from the Housatonic River, Connecticut in 1984 to 1998, Report
No. 99-1 OF, November 15, 1999.

No individual sample results. Summaries and averages. Report
references RFW 02-0050 and Academy report w/ fish data from

1984 to 1996.



N/A

SC

75

02-0217

Housatonic River Flood PCBs. 1984.

Report contains only result summaries



D

KMS

MK01\20123001.096\ERA_PB\ERA_APC1_TblC1-2_PB.xls	C. 1-19	7/10/2003

-------
Table C.1-2

Summary Scores for Data Sets Identified and Evaluated

Serial

RefNo.

Title

Comments

Estimate of Samples for Data Entry

Score

Reviewer*

76

00-0489

BBL. 2000. Background Soil Data Assessment for the GE-
Pittsfield/Housatonic River Site, December 15, 2000.

Soil samples previously reported

96

C

DJG

77

99-0397

USGS Database Query.

Document is a table with result. No supporting documentation
regarding work plans, procedures, orDQIs.

195

D

SC

78

99-0396

PCB - Fish Data.

Data from CT DEP and Dept. of Health files. Most of the
information is pre-1984 fish data.

See scorecard analysis section

D

SC

79

02-0111

Academy of Natural Sciences of Philadelphia, Patrick Center for
Environmental Research. 1997. PCB Concentrations in Fishes and Benthic
Insects from the Housatonic River, Connecticut in 1984 to 1996, Report
No. 97-8F, July 3, 1997.

No individual sample results. Summaries and averages.

Data in GE deliverable

B

SC

80

02-0200

Blasland and Bouck Engineers. 1988. First Quarterly Status Report:
Housatonic River Velocity & Sedimentation Control Pilot Study. August

1988.

Six cores collected in Woods Pond and backwater areas for PCB
analysis. Four of the cores were also analyzed for isotope dating.

Data in GE deliverable

C

SC

*List of Reviewers

SC

= Scott Campbell

KMS

= Kelly Spittler

PH

= Pam Hoskins

DJG

= Douglas J. Godfrey

DS

= Dyana Sagges

MK01\20123001.096\ERA_PB\ERA_APC1_TblC1-2_PB.xls	C.l-20	7/10/2003

-------
APPENDIX C.1.3
NARRATIVE SUMMARIES FOR EVALUATED DATA SETS

MK01 |O:\20123001.096\ERA_PB\ERA_APC_PB.DOC

-------
Title: PCB Concentrations in Fishes From the Housatonic River, Connecticut in 1984 to 1990.
Authors: Division of Environmental Research, Academy of Natural Sciences of Philadelphia,
August 1991.

Reference Number: RFW-02-0215
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is generally complete and well written but lacks sufficient detail in a few
areas. GPS points for sampling locations are not provided. Written location descriptions are
provided for the four fish sampling locations.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Report contains sections on sample locations, methods, result interpretation,
precision and replicate analysis, and a quality control statement. Tracking of samples and COC
procedures are described in detail, although the documents are not provided in the report.

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: Sampling and handling procedures followed ANSP Procedure No. P-14-04 (Rev.
01). The report states that background samples were not required due to previous studies
conducted in 1988. Differences between methods over the course of studies are documented.
The differences are not considered to introduce biases in concentrations, which would affect
between-year comparisons. The total Aroclors in fish were calculated, based in part on the
method of Carnahan and Wagner (unpublished data). Further review of the PCB quantitation
(page 13) should be reviewed by an experienced laboratory quality control specialist or data
validator. The study only quantified Aroclors 1254 and 1260 by calculating the concentration of
congeners from peak numbers 52 to 82 and 178 to 203, respectively.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: Approximately every tenth sample was a sample replicate injection. EPA certified
samples were no longer available for use for 1988 and 1990 samples. No other information is
provided regarding the unavailability of PE samples. The data reviewer only observed results for
Aroclors 1254 and 1260. No calculations were preformed to determine other potential Aroclors
in the samples.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: No data quality indicators noted in this report.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: The data appear to be of questionable quality based on methods.

General comments: The PCB total calculation method should be thoroughly examined to
determine if the quantitation techniques is a valid approach. Report contains analytical data for
1984, 1986, 1988, and 1990. 1984 through 1988 data first published in RFW 02-0214.
Approximately 194 PCB results for fish tissue from 1990. Report contains replicate results for
fish collected in 1984, 1986, 1987, and 1988

Overall Score: Level C - Conditionally acceptable for limited uses. Score may change
depending upon additional review.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC C 1 21 07/10/03

-------
Title: PCB Concentrations in Fishes and Benthic Insects from the Housatonic River,
Connecticut in 1984 to 1992.

Authors: Academy of Natural Sciences of Philadelphia, prepared for General Electric Company
31 August 1993

Reference Number: RFW-02-0249
Reviewer: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: A complete description of the study design, scope of work and rationale were
provided. Thorough background was addressed. Sampling protocol and preservation were
discussed. Specific locations and depths were recorded.

Criterion 2: Formal documentation of procedures. Level: A

Comments: Samples were collected, processed and preserved utilizing procedures referenced
within this report. Report also includes sections on crayfish and aquatic insect sampling and
processing. There was no evidence of a formalized QAPP and no indication that field QA/QC
samples were taken, but existence of quality control procedures were indicated. Locations were
discussed in-depth and background sampling was presented. Extensive statistical analyses were
performed. Chain of Custody procedures explained but actual forms not available in this report.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: A detailed description of the analytical procedure was presented in the study text.
Did not reference approved EPA method, but mirrored EPA protocol.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: There were no specific statements concerning data review, evaluation or validation.
Based on the source of the sample results and extensive content within text, the quality of the
analytical results is not in question.

Criterion 5: Assessment of data quality indicators. Level: A

Comments: Various QA/QC samples were documented, including spikes, standards, replicates
and off-site splits. According to the text, the median coefficient of variation of TPCB for 1998
replicate analysis was 0.12 and the median standard deviation of LNTPCB was 0.11. There were
certain differences between the 1990 and 1992 analytical methods. The 1990 method employed
an external standard, whereas the 1992 method uses the internal standard method for
quantitation. The 1990 method used a non-PCB surrogate, whereas the 1992 method measured
recovery using three PCB surrogates added to the extraction solvent immediately before
beginning extractions. The 1992 method used a slower temperature ramp to increase resolution
of the chromatogram. According to the authors of the report, the different methods may produce
a potential bias toward elevated PCB estimates in pre-1992 studies. However, the authors
indicate that TPCB estimates should be comparable across all their studies.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j 07/10/03

-------
Criterion 6: Data history and overall apparent data quality.	Level: B, caution

Comments: Data History was discussed within the body of the report. Overall intentions of
sampling effort were specifically addressed. Due to the age of the older, some results may not be
of the same quality; however, the study most likely produced data that are equivalent to what
would have been produced using current methodologies. Caution should be used when
comparing data from different years, as the method of TPCB calculation may differ.

General comments: The sample data were well presented, extensive background information
was provided and the intent of scope was discussed. The paper was well written and organized.
Report contains summaries, averages, and age distribution data for 1984, 1986, 1988, 1990 and
1992. There are approximately 145 PCB fish results for fish collected in 1992. There are
separate tables for replicate analysis.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-23

-------
Title: PCB Concentrations in Fishes and Benthic Insects from the Housatonic River,

Connecticut in 1984 to 1994.

Authors: Academy of Natural Sciences (1995)

Reference Number: RFW-02-0050

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Criterion 2: Formal documentation of procedures. Level: A

Comments: Samples were collected, processed and preserved utilizing procedures referenced
within this report. There was no evidence of a formalized QAPP and no indication that field
QA/QC samples were taken, but existence of quality control procedures were indicated.

Locations were discussed in-depth and background sampling was presented. Extensive statistical
analyses were performed.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: A detailed description of the analytical procedure was presented in the study text. A
copy of the Laboratory Standard Operating Procedure was provided as an appendix. Did not
reference approved EPA method, but mirrored EPA protocol.

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: No indication of a Quality Assurance Plan. There were no specific statements
concerning data review, evaluation or validation. Based on the source of the sample results and
extensive content within text, the quality of the analytical results is not in question.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: DQI indicators were not established within this report. Various QA/QC samples
were documented, including spikes, standards, replicates and off-site splits.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data are over five years old (1994). Data History was discussed within the body of
the report. Overall intentions of sampling effort were specifically addressed.

General comments: This study examined the extent of PCB contamination in fish and benthic
insect samples. Age determination was performed along with the other analytical parameters.
The sample data were well presented, extensive background information was provided and the
intent of scope was discussed. The paper was well written and organized. Report references
RFW 02-(214, 215,0249). Approximately 168 individual PCB results have been quantified for
1994 data. A separate table is in the report for replicate analysis and fish age distribution.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-24

-------
Title: PCBs in Housatonic River Fish, Statistical Analysis

Authors: Beck (1982)

Reference Number: RFW-02-0014

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: A description of the statistical study design, scope of work and rationale were
provided. Brief background was addressed. Sampling protocol and preservation were not
discussed. Specific locations and depths were not recorded.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Formal documentation was not presented with this statistical analysis, associated
studies, sampling protocol and/or data sources were not outlined. There was no evidence of a
formalized QAPP and no indication that field QA/QC samples were taken.

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: A detailed description of the analytical procedure was not presented in the study
text. No reference of approved EPA method.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: No indication of a Quality Assurance Plan. There were no statements concerning
data review, evaluation or validation.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: DQI indicators were not established within this report. Additional references were
not provided. QA samples were not documented. Based on the information provided, the
reviewer could not fully evaluate the assessment of DQIs

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Data are over five years old (1980/81). Data History was discussed within the body
of the report. Overall intentions of sampling effort were specifically addressed.

General comments: This report is comprised of a statistical analysis of PCB concentrations
versus fish attribute (i.e. species, location, length, and % lipids). The statistical analysis process
lacked detail. The overall trends were discussed and in several cases, significantly disputed.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-25

-------
Title: Evaluation of Housatonic River Sediment and Floodplain Soil Data on Hazardous

Constituent to Assess Needfor Further Sampling.

Authors: BBL - September 1996

Reference Number: RFW-02-0100

Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is a statistical analysis of existing (up to 1996) non-PCB data. BBL used the
data to determine that no additional floodplain or sediment non-PCB samples are warranted
downstream. Evaluator is unable to comment on the quality of the statistics; however, there
appears to be a high level of detail. Details of sample collection and analysis are presented in the
Phase II/RFI Report.

Criterion 2: Formal documentation of procedures. Level: B

Criterion 4: Data review, validation, and quality assurance. Level: B, caution

Comments: The level of data validation for all of the data is unknown. Based on previous BBL
reports regarding to the Housatonic River, some type of data validation has generally been
performed. The more recent data may be of higher quality and useability than the pre-1995 data.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: Study does not have established DQI indicators. The report is based on several
previous reports in which data was collected for various reasons.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality but a number of the data derives from
studies conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

General comments: The data user should be cautious of some of the older data. Footnotes
associated with each table should be available for the data user to determine the quality of
individual results. A number of the results are estimated, and other potential problems include
duplicate analysis not within control limits, possible method blank contamination, and
concentrations exceeding calibration range.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j 2g 07/10/03

-------
Title: Housatonic River Floodplain Properties - Results of Supplemental Site Characterization
Sampling.

Authors: Blasland, Bouck & Lee, Inc. February, 1994
Reference Number: RFW-02-0036
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is generally complete and well written but lacks sufficient detail in a few
areas.

Criterion 2: Formal documentation of procedures. Level: B

Comments: According to BBL, all sample procedures and analytical-based analysis were
performed in accordance with protocols presented in the MDEP-approved "MCP Sampling and
Analysis Plan". Sample collection procedures and shipping protocols are described in detail.
Copies of COCs are included in this report. Documentation exists for most areas but is
insufficient or lacking in a few areas considered non-critical.

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: Samples were analyzed for PCBs using EPA SW-846 Method 8080. Due to false
positive results from immunoassay field screening kits, additional samples and archived samples
were submitted to a separate laboratory and analyzed for PCBs using EPA SW-846 Method
8081. Holding time for the archived samples had expired prior to laboratory analysis; however,
the results are believed, by BBL and MDEP, to be useful for PCB screening due to the non-
volatile nature of PCBs. Analytical procedures non-standard but sufficiently documented to
establish validity of and ensure confidence in data.

Criterion 4: Data review, validation, and quality assurance. Level: B, caution

Comments: Sample 17-2-6G appears to be double spiked for surrogate. The surrogate
recoveries are outside QC limits. According to the laboratory case narratives, the first eight
samples submitted in October 1993 underwent routine laboratory level I QC. The second and
third groups of five and nine samples respectively, submitted in October and November
underwent routine laboratory level III QC. Some sample result detection limits are elevated due
to interference encountered during analysis. Lining through some sample IDs and writing the
correct number above the typed mistake has changed sample IDs; however, the person making
the correction did not initial and date the correction.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: No data quality indicators noted in this report.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Portions of the data appear to be of questionable quality due to age, changes in
methods, and/or failure to follow current standards for scientific investigation.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j 07/10/03

-------
General comments:

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC	q j	07/10/03

-------
Title: MCP Supplemental Phase II Scope of Work and Proposal for RCRA Facility Investigation

of Housatonic River and Silver Lake

Authors: Blasland, Bouck, and Lee, June 1994

Reference Number: RFW-04-0003

Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is an Addendum to the MCP Interim Phase IIReport, dated March 1992,
and attempts to address MDEP comments and fill data gaps from the Interim report. The work
plan includes discussion of existing (up to 1994) PCB and non-PCB data. The report also
includes results of biota and additional sediment and bank soil sampling. Details of sample
collection and analysis are said to be incompliance with the Sampling and Analysis Plan and
Data Collection and Analysis Quality Assurance Plan (SAP/DC AO AP - Blasland, Bouck, and
Lee, May 1994).

Criterion 2: Formal documentation of procedures. Level: B

Comments: This work plan draws heavily on the results of past sampling efforts to determine
new sample locations/rationale. The data used to formulate the work plan sampling strategy
were collected from 1982 to 1994 during several different sampling programs. No formal
documentation procedures for those sampling efforts are provided in the work plan; however,
this reviewer is aware that appropriate planning documents were prepared past sampling efforts
by Blasland and Bouck. In addition, MDEP comments on the initial Interim Phase IIReport
(completed in 1992) indicates overall DEP satisfaction that correct protocols in the literature, and
as stated in the planning documents, were adhered to in the Blasland and Bouck 1992 effort.
Standard procedures for the new sampling effort are not included in the Work Plan, but are
incorporated by reference to the SAP/DCAQAP (BB&L, 1994).

Criterion 3: Analytical methods used and detection limits achieved. Level: B, caution
Comments: It will be assumed that all historic data incorporated into this report follow
documented standard methods such as EPA methodologies. The assumption is based upon
previous BBL reports, the planning documents listed in the references, and the laboratories that
analyzed samples. Laboratories that analyzed the floodplain and sediment samples include IT
Analytical, and CompuChem Labs, Research Triangle Park. Laboratories performing biota
analyses include Hazleton Laboratories (frog legs). However, the analytical methods (and their
associated detection limits) to be used in this effort are not described in the work plan. It is
assumed these details are included in the SAP/DCAQAP, but this is not certain.

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: The level of data validation for all of the new data to be collected data is unknown,
based on the work plan alone. Based on previous BBL reports regarding to the Housatonic
River, some type of data validation has generally been performed. Since this work plan is dated
June 1994, it may be of greater quality and useability than the pre-1994 data summarized in the
work plan.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j ^ 07/10/03

-------
Criterion 5: Assessment of data quality indicators.	Level: C	

Comments: Plan does not have established DQI indicators. As far as the reviewer can tell, no
DQIs were formally established for the sampling addressed in this plan. These DQIs may be in
the SAPDCAQAP, but this would need to be verified, before assessing the quality of the data
collected as a result of the effort described in this work plan.

Criterion 6: Data history and overall apparent data quality.	Level: B	

Comments: Historic data appear to be of acceptable quality but the data derive from studies
conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

General comments: The data user should be cautious of some of the older data. Footnotes
associated with each table should be available for the data user to determine the quality of
individual results. A number of the results are estimated, and other potential problems include
insufficient QC sample frequency, possible method blank contamination, and concentrations
exceeding calibration range.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-30

-------
Title: Report on Silver Lake Short-Term Measure Evaluation and Related Activities.

Author: Blasland, Bouck and Lee, Inc. (1994)

Reference Number: 04-0006
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: The report provides a description of study design with justification and rationale.
Sample location information given, but not specific.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation indicates samples were collected following the MCP Sampling and
Analysis Plan. Laboratory identified in report, methods referenced in Table 1.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: Specific documentation of analytical procedures is not present in report; however,
Table 1 provides methods used by laboratory (SW-846).

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: Evaluation is listed as having been performed. Review consisted of holding times,
field/prep blank evaluation, precision and accuracy evaluated to specified guidance limits.
Detailed information on the data review is not present.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: Data quality indicators not reported. A brief mention in the report states "data
presented above appear to be acceptable". No further information on data quality exists.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality, but derive from a study conducted prior to
1995. Methods may not meet current standards but are judged to have produced data equivalent
to current methodologies.

General comments: Analytical data presented in the report are of acceptable quality.
Overall Score: Level B - Acceptable, Some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-

-------
Title: Blasland, Bouck and Lee, Inc: Supplemental Phase II/RCRA Facility Investigation Report
for Housatonic River and Silver Lake, Volume II of II.

Authors: Report prepared for General Electric Company, Pittsfield, MA by BBL, 6723 Towpath
Road, Syracuse, NY 13214.

Reference Number: RFW-04-0004
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: The primary focus of this report is the presentation and evaluation of data generated
pursuant to the MCP Supplemental Phase II Scope of Work and Proposal for RCRA Facility
Investigation of Housatonic River and Silver Lake. This report covers activities performed and
data collected received through the end of 1995. Details of sample collection and analysis are
said to be in compliance with the Sampling and Analysis Plan and Data Collection and Analysis
Quality Assurance Plan (SAP/DCAQAP - Blasland, Bouck, and Lee, May 1994). Laboratory
analytical data sheets for Phase II/RFI analyses performed between June 1994 and December
1995 were to be compiled and submitted under separate cover. Overall this report appears to be
of good quality and detail. Data appears in text or in summary tables, no data tables are present
in this report. Note: Volume II contains the report Ambient Air Monitoring Report for PCB May
10, 1995 through August 24, 1995 by Zorex Environmental Engineers, which received an overall
score of A by the review process.

Criterion 2: Formal documentation of procedures. Level: B

Comments: This report draws somewhat on the results of past sampling efforts to determine
new sample locations/rationale. No formal documentation procedures for sampling efforts are
provided in the report; however, this reviewer is aware that appropriate planning documents were
prepared past sampling efforts by Blasland and Bouck. In addition, MDEP comments on the
initial Interim Phase IIReport (completed in 1992) indicates overall DEP satisfaction that
correct protocols in the literature, and as stated in the planning documents, were adhered to in the
Blasland and Bouck 1992 effort. Standard procedures for the new sampling effort are not
included in the report, but are incorporated by reference to the SAP/DCAQAP (BB&L, 1994).

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: It will be assumed that all historic data incorporated into this report follow
documented standard methods such as EPA methodologies. The assumption is based upon
previous BBL reports, the planning documents listed in the references, and the laboratories that
analyzed samples. Laboratories that analyzed the floodplain and sediment samples include IT
Analytical, and CompuChem Labs, Research Triangle Park. Laboratories performing biota
analyses include Hazleton Laboratories (frog legs). However, the analytical methods (and their
associated detection limits) to be used in this effort are not described in the work plan. It is
assumed these details are included in the SAP/DCAQAP, but this is not certain.

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: The level of data validation for all of the new data to collected data is unknown,
based on the report alone. Based on previous BBL reports regarding to the Housatonic River,
some type of data validation has generally been performed. The laboratory data sheets are to be
submitted under separate cover and at a later date than this report.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j ^2 07/10/03

-------
Criterion 5: Assessment of data quality indicators.	Level: C

Comments: Plan does not have established DQI indicators. As far as the reviewer can tell, no
DQIs were formally established for the sampling addressed in this plan. These DQIs may be in
the SAPDCAQAP, but this would need to be verified, before assessing the quality of the data
collected as a result of the effort described in this work plan.

Criterion 6: Data history and overall apparent data quality.	Level: B

Comments: Historic data appear to be of acceptable quality but the data derive from studies
conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

General comments:

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-

-------
Title: Housatonic River Study 45 Day Interim Report - Remedial Alternatives Evaluation

Sediment Disposal Sites.

Authors: Blasland, Bouck & Lee October 1984

Reference Number: RFW 02-0237

Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Comments: The scope of this document is limited to the evaluation of potential local disposal
sites, identification of special engineering and environmental considerations; and identification
of applicable Federal, State and local regulatory requirements. None of the following criterion
has been sufficiently satisfied to warrant extended comments.

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: N/A

Comments: No data review, validation, or quality assurance discussed in this document.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: N/A

Comments: No data analyses are discussed in this document.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-34

-------
Title: Housatonic River Study 90-Day Interim Report, Remedial Alternatives Evaluation, Flow

and Velocity Control, River Channelization, In-situ Impoundment.

Authors: Blasland & Bouck Engineers November 1984

Reference Number: RFW 02-0238

Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Comments: The scope of this document is to evaluate remedial alternatives for PCB reduction
and remediation in the Housatonic River. No analytical data are presented in this report and as
such, none of the following criteria has been sufficiently satisfied to warrant extended comments.

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: N/A

Comments: No data review, validation, or quality assurance discussed in this document.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: N/A

Comments: No data analyses are discussed in this document.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-35

-------
Title: Housatonic River Study 135 Day Interim Report - Assessment of Remedial Alternatives.
Authors: Blasland, Bouck & Lee May 1985
Reference Number: RFW 02-0234
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Comments: The document contains information on dredging techniques/assessments, river
channelization, in-situ impoundment, flow and sedimentation control and information regarding
biodegradation of PCBs. None of the following criterion has been sufficiently satisfied to
warrant extended comments.

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: N/A

Comments: No data review, validation, or quality assurance discussed in this document.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: N/A

Comments: No data analyses are discussed in this document.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-36

-------
Title: Housatonic River Study 135 Day Interim Report - Assessment of Remedial Alternatives
(Addendum).

Authors: Blasland, Bouck & Lee September 1986
Reference Number: RFW 02-0062
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Comments: This document is an addendum to a previous report and was written to satisfy EPA
and MDEP comments on the 1985 report with the same title. The document contains
information on dredging techniques/assessments and information regarding biodegradation of
PCBs. None of the following criterion has been sufficiently satisfied to warrant extended
comments.

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: N/A

Comments: No data review, validation, or quality assurance discussed in this document.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: N/A

Comments: No data analyses are discussed in this document.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-

-------
Title: Housatonic River Velocity and Sedimentation Control Pilot Study.

Authors: Blasland, Bouck & Lee March 1989
Reference Number: RFW 02-0235
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: This document is a Pilot Study designed to demonstrate that physical modifications
performed on the river system will increase sedimentation in upstream reaches, as such,
relatively little (8 PCB and TSS results) analytical data are presented in this report. None of the
following criteria has been sufficiently satisfied to warrant extended comments.

Criterion 2: Formal documentation of procedures. Level: D

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: D
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No data review, validation, or quality assurance discussed in this document.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: No data analyses are discussed in this document.

General comments: Report contains 8 PCB and 8 TSS results.
Overall Score: D-Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-38

-------
Title: MCP Interim Phase II Report/Current Assessment Summary For Housatonic River.
Authors: BBL - September 1991
Reference Number: RFW-02-0071
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: Report provides complete description of study design with justification and
rationale.

Criterion 2: Formal documentation of procedures. Level: C

Comments: The data used to support this study was collected from the early 1980s and early
1990s. Field documentation is acceptable. The documentation is generally not available but
sufficient information is know or available via other sources to establish validity of analytical
procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: Estimates of PCB mass in sediments should not be viewed as highly accurate. The
estimates made by Stewart and Frink studies in the early 1980s are subject to substantial errors
from various sources. Analytical methods for PCB extraction of different types of matrixes not
observed. Some of the water data contain interferences during extraction or other laboratory
issues that may not be discussed in the text.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: This report references an Appendix H for QA/QC results for the analytical data
associated with the MCP investigation. Appendix H could not be located.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: Study does not have established DQI indicators.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality but all of the data derives from studies
conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

General comments: Analytical methods, level of QA/QC and validation are unknown. Review
of Appendix H may help to increase the report to a more favorable and trustworthy source of
data. Report references RFW 02-0215, RFW 02-0014, RFW 02-0235, RFW 02-0203, RFW 02-
0201, RFW 02-0247, RFW 02-0248, RFW 02-0090, AND RFW 02-0030. Samples collected in
1990 for PCB analysis by BBL are in the GEDB. Reviewer did not locate App. IX results for
HCSE or HCRP samples. No sample IDs are provided for water samples collected in 1989,
1990, and 1991 for Tables 5-3A, 5-4, 5-5, 5-6A, 5-6B, 5-7. No CS-137 results in GE DB for
18K(H)_ or 17E_. Table 9-1 is fish data collected from the 1982 Stewart report. Table 9-2
contains fish data from RFW 02-0214, RFW 02-0203, and others not reviewed. A large number
of maps are available for sample location. Volume III of III not reviewed.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> 07/10/03

-------
Title: Addendum toMCP Interim Phase II Report/Current Assessment Summary for Housatonic
River.

Authors: Blasland & Bouck Engineers, P.C. - August 1992
Reference Number: RFW-02-0038 (vol.1) and RFW-02-0039 (vol. 2)

Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is an Addendum to the MCP Interim Phase IIReport, dated March 1992,
and attempts to address MDEP comments and fill data gaps from the Interim report. The report
includes discussion of revised statistical analysis of existing (up to 1992) PCB and non-PCB
data. The report also includes results of biota and additional sediment and bank soil sampling.
BBL used the data to determine that no additional floodplain or sediment non-PCB samples are
warranted downstream. Details of sample collection and analysis are said to have been presented
in the Interim Phase IIReports for the Housatonic River (March 1992) and Unkamet Brook Area
(April 1992), as well as in the Phase II Scope of Work, and Sampling and Analysis Plan
(Blasland and Bouck, June 1990 and September 1990, respectively.

Criterion 2: Formal documentation of procedures. Level: B

Comments: The data used to support this study was collected from 1990 to 1996 during several
different sampling programs. No formal documentation procedures are provided in this report;
however, a reference list, indicating that appropriate planning documents were prepared for the
effort is provided. In addition, MDEP comments on the initial Interim Phase IIReport (included
as Appendix A of this Addendum) indicates overall DEP satisfaction that correct protocols in the
literature, and as stated in the planning documents, were adhered to.

Criterion 3: Analytical methods used and detection limits achieved. Level: A, caution
Comments: It will be assumed that all data incorporated into this report follow documented
standard methods such as EPA methodologies. The assumption is based upon previous BB&L
reports, the planning documents listed in the references, and the laboratories that analyzed
samples. Laboratories that analyzed the floodplain and sediment samples include IT Analytical,
and CompuChem Labs, Research Triangle Park. Laboratories performing biota analyses include
Hazleton Laboratories (frog legs). However, in the MDEP comments on the initial Interim
Phase II Report, it was noted that insufficient numbers of QC samples, such as duplicate and
method blanks were run for certain data sets, and that some of the data interpretation was
questionable. This Addendum responds to the DEP comments, and Appendix G (Analytical
data) does contain duplicate, method blank, and spike/surrogate recovery information, which
appear to be sufficient, but actual validation of the data would be needed to know for sure.

Criterion 4: Data review, validation, and quality assurance. Level: B, caution

Comments: The level of data validation for all data is unknown. Based on previous BB&L
reports regarding to the Housatonic River, some type of data validation has generally been
performed. Data is from 1992 and previous may be of lesser quality and useability than the post-
1995 data.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> /ia 07/10/03

-------
Criterion 5: Assessment of data quality indicators. Level: C

Comments: Study does not have established DQI indicators. The report is based on several
previous reports in which data was collected for various reasons, with some additional sampling
(biota) added from this Addendum. As far as the reviewer can tell, no DQIs were formally
established for the additional biota sampling addressed in this Addendum.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Historic data appear to be of acceptable quality but some of the data derive from
studies conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

General comments: The data user should be cautious of some of the older data. Footnotes
associated with each table should be available for the data user to determine the quality of
individual results. A number of the results are estimated, and other potential problems include
insufficient QC sample frequency, possible method blank contamination, and concentrations
exceeding calibration range.

This report references RFW 02-0030 and RFW 02-0130. Sample results in Table 2-9 are located
in the GEDB. Inorganic or App. IX results for BM-92-l(4), FPL-1-*, and HCSE-* were not
located in GE DB. Data presented in Table 5-2 is located in GE DB. Frog data in Table 6-1
was not located in GE DB. Property sample appear to be loaded into GE DB. All volumes and
appendix have been located for RFW 02-0038/0039.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-41

-------
Title: Evaluation of Needfor Short Term Measures In the Housatonic River Floodplain.

Note: The analytical results and quality control information is located in the report titled
Summary of Housatonic River Floodplain Property and Sampling and Analysis. For the
purposes of this review, the two aforementioned reports have been evaluated together.

Authors: BBL - November and Octoberl992.

Reference Number: RFW 02-0109 and RFW 02-0097
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: The two accompanying reports provides complete description of study design with
justification and rationale.

Criterion 2: Formal documentation of procedures. Level: A

Comments: Samples were collected using protocols outlined in the MCP Sampling and
Analysis Plan. Field procedures are described in good detail. Samples were packaged and
shipped with Chain-of-Custody form by overnight courier. Any deviations or laboratory
problems are included in laboratory narrative reports.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: All samples were analyzed for PCBs by US EPA SW-846 Method 8080.
Approximately half of the samples were analyzed for TOC using US EPA Method 9060. The
TOC results should be considered nonpurgeable organic carbon since the sample preparation
may result in loss of volatiles.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: The evaluation of QA/QC components of the PCB data involved review of sample
handling, analysis of QC procedures for consistency with the MCP Sampling and Analysis Plan,
and the analysis of QA data (blanks, duplicate, and MS/MSD analyses).

Criterion 5: Assessment of data quality indicators. Level: A

Comments: Each report was reviewed for laboratory accuracy, laboratory precision, and
evidence of blank contamination, COC completeness, and several other data quality indicators.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality but some of the data derives from studies
conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

General comments: The PCB method SW-846 Method 8080 was updated to SW-846 Method
8082 in 1996; however, the analytical results are judged to have been produced data equivalent
to current methodologies.

Overall Score: Level A - Acceptable, unrestricted use - see above General comments section.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j ^ 07/10/03

-------
Title: Identification of Extent of and Land Use and Property Ownership Within, Potentially

Affected Area of Housatonic River Floodplain.

Authors: Blasland & Bouck Engineers, August, 1992

Reference Number: RFW-02-0103

Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: MDEP noted inconsistencies between the PCB concentrations identified in the
narrative text and the figures. Both the MDEP and an independent data validation company
judge the report to be lacking in overall quality of and level of detail. Report is incomplete but
does provide sufficient information for one or more parameters of interest.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation indicates sample collection for Appendix IX sediments between
Silver Lake Outfall and Elm Street Bridge were collected following the protocols in the MCP
Sampling and Analysis Plan. In a comment letter the MDEP states "The attribution of some of
the elevated PCB levels in samples to filtering and spiking problems must be better
documented". According to independent data validators for MDEP, several laboratory
deficiencies were not fully described by GE's consultant and inaccurate conclusions may be
based on unacceptable data and statistical methods.

Criterion 3: Analytical methods used and detection limits achieved. Level: C
Comments: Sediments were analyzed for PCBs by Method SW-846 8080 and surficial
sediments were analyzed for TOCs by Method 9060. See comments below.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: The MDEP determined that the 1990-1991 MCP work was not performed consistent
with the QA/QC Plan. Deficiencies noted include: sample holding times were exceeded in
several instances; COC procedures were not described; the required number of sample spikes,
equipment blanks, blind field duplicates were not analyzed; and the number of sediment samples
for hazardous constituents were below that which was required.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: DQIs not established; data appear to not satisfy minimum standards for one or more
non-critical DQIs.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Portions of the data appear to be of questionable quality due to age, changes in
methods, and/or failure to follow current standards for scientific investigation.

General comments: GE's consultant's conclusion of the data appears incorrect based on
comments from MDEP and Lawrence Experiment Station.

Most of the data in this report appears to be in the existing GE database, with the exception of
some appendix IX results.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> 07/10/03

-------
Title: Evaluation of Needfor Short Term Measures In the Housatonic River Floodplain.

Authors: BBL - November and Octoberl992.

Reference Number: RFW 02-0109 and RFW 02-0097
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: The two accompanying reports provides complete description of study design with
justification and rationale.

Criterion 2: Formal documentation of procedures. Level: A

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Criterion 4: Data review, validation, and quality assurance. Level: A

Criterion 5: Assessment of data quality indicators. Level: A

Comments: Each report was reviewed for laboratory accuracy, laboratory precision, and
evidence of blank contamination, COC completeness, and several other data quality indicators.

Criterion 6: Data history and overall apparent data quality. Level: B

Overall Score: Level A - Acceptable, unrestricted use - see above General comments section.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> a a 07/10/03

-------
Title: Report on January 1993 Housatonic River Floodplain Property Sampling and Analysis.
Authors: Blasland & Bouck Engineers , P.C.

Reference Number:RFW-02-0037
Reviewer: D. Sagges

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: A complete description of the area, study design and scope of work are provided.
This investigation was approved by MDEP in May 8,1992 and modified by a subsequent
agreement between GE and MDEP. Sampling procedures, logbook information and sample
handling were all discussed. In addition, sampling locations were well documented with use area
maps and site photographs.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Samples were collected utilizing procedures established in the MCP Sampling and
Analysis Plan. Field procedures were described in good detail. The complete laboratory reports
were provided, which included the COCs, QA/QC samples and some sample attribute
information. Laboratory SOPs were not provided.

Criterion 3: Analytical methods used and detection limits achieved. Level: A
Comments: Samples were analyzed for PCBs by Method SW-846 8080.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: In section 2.3 of this report," based on a review of sample holding times, COCs,
results of blanks, duplicate analyses, MS/MSD analyses the Housatonic River floodplain PCB
data presented appear to be acceptable with respect to USEPA Region 1 guidelines." There were
no statements concerning data validation.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: DQI indicators were not established within this report. Additional references were
not provided. QA samples were documented. Based on the information provided, the reviewer
could not fully evaluate the assessment of DQIs, but the data does appear to meet minimum
standards.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data is over five years old (1993). Site history and data history were not discussed
within the body of the report. Overall intentions of sampling effort were specifically addressed
and approved methodologies were performed.

General comments: Overall the report was complete, but lacked support documentation. Based
on the data sources, the integrity of the data is not suspect.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-45

-------
Title: Evaluation of Potential Short-Term Measure Proposals for Properties Within the

Housatonic River Floodplain

Authors: Blasland, Bouck, and Lee, April 1993

Reference Number: RFW 02-0240

Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Comments: This is a proposal, rather than a data report, and as such, no data is presented or
discussed. The level of detail presented for each of the possible Short-Term measure (STM)
alternatives is minimal, because it was not known at the time this document was written exactly
which residential properties would undergo STMs. The document acknowledges that data gaps
exist, and that the properties would be chosen once the data gaps are filled.

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal procedures are discussed in this document, because that type of
discussion is not appropriate for this evaluation, which is conceptual in nature.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No analytical methods are discussed in this document.

Criterion 4: Data review, validation, and quality assurance. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: N/A

Comments: No data analyses are discussed in this document.

General comments: This document is an informal evaluation of STMs, without regard to site-
specific data. The evaluation contains minimal detail, as is appropriate for the type of evaluation
it is.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-46

-------
Title: Short-Term Measure Proposals for the "60-Day " Residential Properties within the

Housatonic River Floodplain

Authors: Blasland, Bouck, and Lee, September 1993

Reference Number: RFW 02-0241

Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: This is a proposal, rather than a data report; however for each property evaluated a
synopsis of existing data for the property is presented. The floodplain data summarized for each
property is said to have been collected in August 1992 and January 1993. A planning document,
entitled MCI* Sampling and Analysis Plan (Blasland and Bouck, September 1990) is included in
the references portion of the document. However, it is unclear within the document who
collected the 1992 and 1993 historic samples upon which the STM proposal is based. It appears
to have been done by Blasland and Bouck. If that is true, then at least some provision for overall
data quality of the historic data has been provided for in the MCP SAP. However, since no data
packages are included for the historic data, Data Quality Indicators (DQIs - such as precision,
accuracy, and completeness), and attainment of Data Quality Objectives (DQOs - i.e., did
laboratory quantitation limits meet required criteria?) cannot be fully assessed. In addition, the
level of detail presented with regard to historic data is limited, but adequate for the purposes of
the document (a proposal for Short-Term Measures at several floodplain properties). Historic
data are presented in data boxes on site figures, and quoted in ranges (max./min.) within the
document text.

Criterion 2: Formal documentation of procedures. Level: B, caution

Comments: This work plan draws heavily on the results of past sampling efforts to determine
new sample locations/rationale. The data used to formulate the new sampling strategy were
collected from 1922 to 1993 during several different sampling programs. No formal
documentation procedures for those sampling efforts are provided in the work plan; however,
this reviewer is aware that appropriate planning documents were prepared past sampling efforts
by Blasland and Bouck. The Proposals identify data gaps, and proposed sampling schemes for
filling the gaps. Based on the reference list, it is presumed that details of proposed sample
collection and analysis are in compliance with the Sampling and Analysis Plan and Data
Collection and Analysis Quality Assurance Plan (SAP/DCAQAP - Blasland and Bouck, June
1993). In addition, Attachment F to the document presents a comprehensive SOP for analysis of
PCBs using EnSYS Corporation's PCB RISc™ Immunoassay field test kits.

Criterion 3: Analytical methods used and detection limits achieved. Level: B, caution
Comments: It will be assumed that all historic data incorporated into this report follow
documented standard methods such as EPA methodologies. The assumption is based upon
previous BBL reports, the planning documents listed in the references, and the laboratories that
analyzed samples. However, the analytical methods (and their associated detection limits) used
in past sampling efforts are not described in the document. It is assumed these details are
included in the 1992 SAP/DCAQAP, but this is not certain. For proposed sampling efforts,
Attachment F does contain an SOP for PCB analysis via immunoassay. Within the method, it is
implied that the method detection limit can meet the 10 ppm goal set by MDEP for total PCBs in
soil.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> a*-j 07/10/03

-------
Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: The level of data validation for all of the new data to be collected data is unknown,
based on the document alone. Based on previous BBL reports regarding to the Housatonic
River, some type of data validation has generally been performed. Since this work plan is dated
September 1993, it may be of lesser quality and useability than post-1995 data.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: Document does not have established DQI indicators. As far as the reviewer can tell,
no DQIs were formally established for the historic or proposed sampling addressed in this plan.
These DQIs may be in the 1992 and/or 1993 SAPDCAQAPs, but this would need to be verified,
before assessing the quality of the data collected as a result of the effort described in this work
plan. In addition, since no data packages are included for the historic data, Data Quality
Indicators (DQIs - such as precision, accuracy, and completeness), and attainment of Data
Quality Objectives (DQOs - i.e., did laboratory quantitation limits meet required criteria?) cannot
be fully assessed. However, based on knowledge of other BB&L documents, this reviewer
believes the historic data, and data generated from proposed activities would meet minimum DQI
standards.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Historic data appear to be of acceptable quality but the data derive from studies
conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies. Data gleaned from proposed sampling
would likely be pre-1995 as well.

General comments: The data user should be cautious of some of the older data. Potential
problems could include insufficient QC sample frequency, inability to properly evaluate data
against DQOs and DQIs, use of inappropriate or outdated analytical methods, and reported QC
results outside of acceptable ranges.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply. The
source of the historic data in the document should be confirmed, and analytical data packages
obtained before actually incorporating data into risk assessments, etc.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-48

-------
Title: Short-Term Measure Proposals for the "90-Day " and "120-Day " Residential Properties

Within the Housatonic River Floodplain.

Authors: Blasland, Bouck, and Lee (BB&L), October 1993

Reference Number: RFW 02-0242

Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: This is a proposal, rather than a data report; however for each property evaluated a
synopsis of existing data for the property is presented. The document is set up in exactly the
same way as Short-Term Measure Proposals for the "60-Day " Residential Properties Within the
Housatonic River Floodplain (BB&L, September 1993). The floodplain data summarized for
each property is said to have been collected in August 1992 and January 1993, with the
exception of Parcel 29-1, whose floodplain data was collected in 1988 and 1989. A planning
document, entitled MCP Sampling and Analysis Plan (Blasland and Bouck, September 1990) is
included in the references portion of the document. However, it is unclear within the document
who collected the 1992 and 1993 historic samples upon which the STM proposal is based. It
appears to have been done by Blasland and Bouck. If that is true, then at least some provision
for overall data quality of the historic data has been provided for in the MCP SAP. However,
since no data packages are included for the historic data, Data Quality Indicators (DQIs - such as
precision, accuracy, and completeness), and attainment of Data Quality Objectives (DQOs - i.e.,
did laboratory quantitation limits meet required criteria?) cannot be fully assessed. In addition,
the level of detail presented with regard to historic data is limited, but adequate for the purposes
of the document (a proposal for Short-Term Measures at several floodplain properties). Historic
data are presented in data boxes on site figures, and quoted in ranges (max./min.) within the
document text.

Criterion 2: Formal documentation of procedures. Level: B, caution

Criterion 3: Analytical methods used and detection limits achieved. Level: B, caution
Comments: It will be assumed that all historic data incorporated into this report follow
documented standard methods such as EPA methodologies. The assumption is based upon
previous BBL reports, the planning documents listed in the references, and the laboratories that
analyzed samples. However, the analytical methods (and their associated detection limits) used
in past sampling efforts are not described in the document. It is assumed these details are
included in the 1992 SAP/DCAQAP, but this is not certain.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> aq 07/10/03

-------
Criterion 4: Data review, validation, and quality assurance. Level: B

Criterion 5: Assessment of data quality indicators. Level: B

Criterion 6: Data history and overall apparent data quality. Level: B

General comments: The data user should be cautious of some of the older data, especially the
pre-1990 data from Parcel 29-1. Potential problems could include insufficient QC sample
frequency, inability to properly evaluate data against DQOs and DQIs, use of inappropriate or
outdated analytical methods, and reported QC results outside of acceptable ranges.

This document came with an enclosure, which provides preliminary cost estimates for
implementation of the STMs for the "60-Day" properties discussed in the BB &L report dated
September 1993. The cost estimates are conceptual in nature, and do not go into a great level of
detail. However, they do provide an excellent source of information regarding required tasks and
work sequencing for the STMs chosen for each of the subject properties.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-50

-------
Title: Silver Lake Data Summary.

Authors: Blasland, Bouck & Lee, Inc. November 1993
Reference Number: RFW-04-0005
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: Quality and detail of report is lacking. Most of the sampling results presented in the
report were collected from previous investigations for the MCP Phase II report, the Stewart
Laboratories baseline study in 1980 and 1982, and outfall discharge sampling conducted by GE
during unauthorized releases. The quality of these previous documents should be considered
when using results from this report.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation generally not available but sufficient information is know or
available via other sources to establish validity of field and analytical procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: C/l)

Comments: Analytical procedures not well documented, but data are believed to be valid due to
other information (refer to references in report) provided. Some of the samples were analyzed at
the GE Laboratory and no supporting documentation is provided. The validity of these results
cannot be validated.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: QA/QC procedures and data validation may differ depending upon the report in
which the data was collected in conjunction with.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: No data quality indicators noted in this report.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Portions of the data appear to be of questionable quality due to age, changes in
methods, and/or failure to follow current standards for scientific investigation.

General comments: Some of the data located on Figure 1 was collected and analyzed under
separate reports, which may have received different comments and an overall score than this
report. Data analyzed by GE's Plant Laboratory should be used with caution; however, the
analytical results determined by this lab detected generally high concentrations of PCBs in Silver
Lake sediments.

No sample results found during a data search in GE floodplain and river sample results database.
Report contains approximately 100 PCB sediment results, 35 PCB surface water results, and 6
appendix IX samples. The only sample location information is on the map with sample results.

Overall Score: Level C/D - Conditionally acceptable for limited uses (some of the data are
found in previous reports). The PCB data analyzed by GE Labs is conditionally acceptable (use
with caution) due to lack of documentation - unless documentation exists under separate cover.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-51

-------
Title: Fisheries Investigation of the Housatonic River, Massachusetts.

Author: Chadwick& Associates, Inc. (March 1993).

Reference Number: 02-0102
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: The report provides a basic description of study design with justification and
rationale. Sample location information given, but not specific. Study is of fish counts, and a
qualitative discussion of the river habitats. Limited chemical parameters used in describing
water quality.

Criterion 2: Formal documentation of procedures. Level: D

Comments: No procedures documented. Fish were weighed and measured and released. Fish
were caught by electrofishing. No information was supplied on the analysis of water quality
parameters.

Criterion 3: Analytical methods used and detection limits achieved. Level: D
Comments: Specific documentation of analytical procedures is not present in report.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No review or validation of data was mentioned.

Criterion 5: Assessment of data quality indicators.

Level: D

Comments: Data quality indicators not reported.

Criterion 6: Data history and overall apparent data quality.

Level: C

Comments: Portions of the data appear to be of questionable quality due to age, and lack of

clear DQI and review.

General comments: Analytical data presented in this fish population study are: temperature,
dissolved oxygen, pH, total ammonia, nitrate, and orthophosphate. No other chemical analysis
was performed.

Overall Score: Level C/D - Conditionally acceptable for limited uses, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-52

-------
Title: Aquatic Ecology Assessment of the Housatonic River, Massachusetts 1993.

Author: Chadwick& Associates, Inc. (March 1994).

Reference Number: RFW 02-0047
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: The report provides a basic description of study design with justification and
rationale. Sample location information given, but may not be sufficiently specific to identify
individual sample locations with confidence. Study is of fish and benthic invertebrate counts,
and a qualitative discussion of the river habitats. Limited chemical parameters used in
describing water quality.

Criterion 2: Formal documentation of procedures. Level: D

Comments: No procedures documented. Fish were weighed and measured and released. Fish
were caught by electrofishing. Benthic invertebrates were counted. No information was
supplied on the analysis of water quality parameters.

Criterion 3: Analytical methods used and detection limits achieved. Level: D
Comments: Specific documentation of analytical procedures is not present in report.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No review or validation of data was mentioned.

Criterion 5: Assessment of data quality indicators.

Level: D

Comments: Data quality indicators not reported.

Criterion 6: Data history and overall apparent data quality.

Level: C

Comments: Portions of the data appear to be of questionable quality

due to age, and lack of

clear DQI and review.

General comments: Analytical data presented in this aquatic ecology assessment study are:
temperature, dissolved oxygen, pH, total ammonia, nitrate, and orthophosphate. No other
chemical analysis was performed.

Overall Score: Level C/D - Conditionally acceptable for limited uses, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-53

-------
Title: Revised Risk Assessment To Evaluate the Need For Short-term measures in the Flood
Plain of the Housatonic River.

Authors: ChemRisk, A Division of McLaren/Hart, April 12, 1993
Reference Number: RFW 02-0243
Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Thee risk assessment appears to be very well thought out, and that a large body of
scientific literature has been used in developing and supporting the conclusions of the risk
assessment. Many of the studies cited directly contradict similar studies or risk assessment
protocols espoused by EPA and DEP, and this fact is stated throughout the assessment. Data
from previous studies (from 1990 to 1993) were incorporated into the risk calculations. Detailed
data packages were not provided, as these were likely included in previous reports.

Criterion 2: Formal documentation of procedures. Level: B, caution

Comments: Although no protocols for the collection and analysis of the data being used in the
assessment are provided in the report, Section 1 of the report states "... procedures and
methodologies for collecting the samples are discussed in detail in B&B - 1991." This citation
refers to the document entitled MCP Interim Phase II Report/Current Assessment Summary for
the Housatonic River, by Blasland and Bouck (1991). The reviewer recalls the referenced
document does not itself present a lot of protocols, but it does incorporate comprehensive
planning documents, which do include sample collection protocols.

The historic data used to support this risk assessment, and to aid in the selection of assessment
and measurement endpoints, were collected from 1990 to 1993 during several different sampling
programs. No formal documentation of the data quality from these sampling programs is
provided. A planning document, entitled MCP Sampling and Analysis Plan (Blasland and
Bouck, September 1990) is known to have existed prior to initiation of the earliest sampling
event (December 1990). However, it is unclear within the document who collected the 1992 and
1993 historic samples upon which the risk assessment is based. It appears to have been done by
Blasland and Bouck. If that is true, then at least some provision for overall data quality of the
historic data has been provided for in the MCP SAP.

The observations above notwithstanding, the data are old, and should be used with caution.

Criterion 3: Analytical methods used and detection limits achieved. Level: A, caution
Comments: It is assumed that historic data incorporated into the risk assessment were collected
as part of programs that followed documented standard methods such as EPA methodologies.
The assumption is based upon other BB&L reports. However, due to the age of the risk
assessment, which pre-dates the formal Sampling and Analysis Plan and Data Collection and
Analysis Quality Assurance Plan (SAP/DC AQAP), written in 1994, the data may be of lesser
quality than "post-SAP/DCAQAP" data. Based on other document reviews this reviewer knows
that at least some of the BB&L reports contain duplicate, method blank, and spike/surrogate
recovery information, which appears to be sufficient, but actual validation of the data would be
needed to know for sure.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> ca 07/10/03

-------
Historic data are presented in data boxes on site figures, and tabulated within Section 2 of the
text. Within the tables, reference is made to a PCB quantitation limit of 0.05 ppm, but it is
unclear whether this applies to each sample, or was strictly the project quantitation limit being
strived for.

Criterion 4: Data review, validation, and quality assurance. Level: B, caution

Comments: The level of data validation for all of the historic data cited is unknown. Based on
other BB&L reports regarding the Housatonic River, some type of data validation has generally
been performed. Data is from 1993 and previous, and may be of lesser quality and useability
than post-1995 data.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: The risk assessment does not refer to any established DQI indicators. No formal
data validation procedures are described. As far as the reviewer can tell, no DQIs were formally
established for the sampling addressed in this assessment. However, since no data packages are
included for the historic data, Data Quality Indicators (DQIs - such as precision, accuracy, and
completeness), and attainment of Data Quality Objectives (DQOs - i.e., did laboratory
quantitation limits meet required criteria?) cannot be fully assessed. In addition, the level of
detail presented with regard to historic data is limited, but adequate for the purposes of the
document (a risk assessment, based on data already presented to regulatory agencies under
separate cover).

Criterion 6: Data history and overall apparent data quality. Level: B, caution

General comments: The data user should be cautious of some of the older data. No validation
flags are provided, and it is unclear from the tabulated data provided whether DQOs were
consistently achieved, or whether DQIs were within acceptable ranges.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-55

-------
Title: Preliminary Health and Environmental Assessment Proposal for the Housatonic River,

Silver Lake, and their Floodplains.



Authors: ChemRisk, April 28, 1993



Reference Number: RFW 02-0239



Reviewed by: Scott Campbell



Criterion 1: Overall quality of and level of detail in report(s).

Level: N/A

Comments:

Criterion 2: Formal documentation of procedures.

Level: N/A

Comments:

Criterion 3: Analytical methods used and detection limits achieved.

Level: N/A

Comments:

Criterion 4: Data review, validation, and quality assurance.

Level: N/A

Comments:

Criterion 5: Assessment of data quality indicators.

Level: N/A

Comments:

Criterion 6: Data history and overall apparent data quality.

Level: N/A

Comments:

General comments: No analytical data are presented in this report.
Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-56

-------
Title: Work Plan for the Ecological Risk Assessment of the Housatonic River Site.

Authors: ChemRisk, A Division of McLaren/Hart, May 24, 1997
Reference Number: RFW-02-0104 (vols.l&2)

Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: The study design in this work plan appears to be very well thought out, and that a
large body of scientific literature has been used in formulating the study design. Data from
previous studies (from 1991 to 1996) were incorporated into the study strategy, as well as the
requirements for the Consent Order and regulatory agency comments on a previous draft of the
work plan. Rationale is provided for each type of proposed sampling and sampling locations. In
addition, protocols for evaluating the data collected and incorporation into an Ecological Risk
Assessment is presented.

Criterion 2: Formal documentation of procedures. Level: A

Comments: The historic data used to support this study, and to aid in the selection of
assessment and measurement endpoints were collected from 1982 to 1996 during several
different sampling programs. No formal documentation of the data quality from these sampling
programs is provided; however, a reference list is provided, which indicates that for at least some
of the studies, formal planning documents (work plans and quality assurance project plans) were
prepared. Some of the studies are quite old, and should be used with caution. However, the
document under review is a formal work plan for the collection, analysis, evaluation, and use in
risk characterization of additional data, including sediment, floodplain soil, water column, avian,
fish, reptile, insect, and mammal samples. In addition, for the proposed sample collection and
evaluation in this work plan, formal procedures are provided in Volume 2. Most of the
procedures have the same format, which includes terminology, equipment list, mobilization and
training, field procedures, analytical procedures, data evaluation procedures, schedule, and
references. Therefore, this reviewer believes that as long as the work plan is adhered to, a viable
ecological risk assessment would result.

Criterion 3: Analytical methods used and detection limits achieved. Level: A, caution
Comments: It will be assumed that all historic data incorporated into the introductory sections of
this work plan follow documented standard methods such as EPA methodologies. The
assumption is based upon previous BB&L reports, the planning documents listed in the
references, and the laboratories that analyzed samples. Laboratories that analyzed the floodplain
and sediment samples include IT Analytical, and CompuChem Labs, Research Triangle Park.
Laboratories performing biota analyses include Hazleton Laboratories (frog legs). Based on
other document reviews this reviewer knows that at least some of the BB&L reports contain
duplicate, method blank, and spike/surrogate recovery information, which appears to be
sufficient, but actual validation of the data would be needed to know for sure.

Criterion 4: Data review, validation, and quality assurance. Level: B, caution

Comments: The level of data validation for all of the historic data cited in the introductory
sections of the work plan is unknown. Based on previous BB&L reports regarding to the
Housatonic River, some type of data validation has generally been performed. Data is from 1992
and previous may be of lesser quality and useability than the post-1995 data.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> rn 07/10/03

-------
Criterion 5: Assessment of data quality indicators. Level: B

Comments: Study design does not have established DQI indicators. No formal data validation
procedures are described. However, this may be the result of the fact that many of the data will
be more "exotic" than the soil, sediment and water samples normally collected as part of an
environmental study. In many of the protocols in Volume 2, it is stated that data will be
reviewed for accuracy by a qualified scientist other than the one entering the data. The design is
based on several previous reports in which data was collected for various reasons. As far as the
reviewer can tell, no DQIs were formally established for the additional sampling addressed in
this work plan.

Criterion 6: Data history and overall apparent data quality. Level: B

General comments: The data user should be cautious of some of the older data. Footnotes
associated with each table should be available for the data user to determine the quality of
individual results. A number of the results are estimated, and other potential problems include
insufficient QC sample frequency, possible method blank contamination, and concentrations
exceeding calibration range.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-58

-------
Title: Evaluation of the Terrestrial Ecosystem of the Housatonic River Valley
Authors: ChemRisk, A Division of Mclaren/Hart - July 26, 1994
Reference Number: RFW-02-0048
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: This study was designed to evaluate the effects of PCBs on terrestrial wildlife by
comparing the health and wildlife populations within the floodplain downstream of New Lenox
Road bridge, to the health of reference populations that share the characteristics and stressors of
the target population, with the exception of exposure to PCBs. Note: No data tables with
analytical results are contained in this report. A relatively small amount of data is presented on
maps.

Criterion 2: Formal documentation of procedures. Level: C

Comments: This report draws heavily on the results of past sampling efforts to determine study
locations/rationale. The data used to formulate the work plan sampling strategy were collected
from 1981 to 1994 during several different sampling programs. No formal documentation
procedures for those sampling efforts are provided in the report; however, this reviewer is aware
that appropriate planning documents were prepared past sampling efforts by Blasland and
Bouck. However, the quality of data and documentation procedures for older data is less certain.

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: It will be not assumed that all historic data incorporated into this report follow
documented standard methods such as EPA methodologies. The age of some of the data and
lack of documentation regarding the quality of the data limit the reviewer's ability to effectively
score criterion 3. ChemRisk used GE data from previous studies conducted in 1981-1982, 1990-
1991, and 1994.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: The level of data validation for older data is unknown. Based on previous BBL
reports regarding to the Housatonic River, some type of data validation has generally been
performed for more recent (post-1990) data.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: Plan does not have established DQI indicators. As far as the reviewer can tell, no
DQIs were formally established for the sampling addressed in this plan.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Portions of the data appear to be of questionable quality based primarily on lack of
documentation. Other data may be useable but risk assessors should exercise caution and should
only use such with caution.

General comments: The data user should be cautious of some of the older data.
Overall Score: Level C - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ -> rQ 07/10/03

-------
Title: Preliminary Habitat and Biota Assessment of Sediment Remediation Technologies - GE

Housatonic River Study Area (Appendix A to the Report on Preliminary Investigation Corrective

Measures for Housatonic River and Silver Lake Sediment by HEC)

Authors: ENVIRON Corporation

Reference Number: RFW- 02-0251

Reviewed by: Pam Hoskins

Criterion 1: Overall quality of and level of detail in report(s).	Level: B	

Comments: Report is an Appendix to the Report on Preliminary Investigation Corrective
Measures for Housatonic River and Silver Lake Sediment by HEC 1996. The Report provides an
introduction, which gives a delineation of the study area, based on past studies by BBL (1991,
1992, 1996). Studies of aquatic and terrestrial communities from Chadwick and Associates
(1994) and ChemRisk (1994) were used in establishing baseline information about these
communities for use in the impact evaluations. Both are good, up-to-date sources, although they
draw on pre-1995 analytical data. Many other historic literature sources were used to derive
toxicity values, information on various species, etc., but this reviewer is not an expert in these
areas, and cannot comment on the quality of these sources.

Sample collection and analysis methods for tasks performed after 1994 are assumed to be in
compliance with the Sampling and Analysis Plan and Data Collection and Analysis Quality
Assurance Plan (SAP/DCAQAP - Blasland, Bouck, and Lee, May 1994).

Criterion 2: Formal documentation of procedures.	Level: B	

Comments: This document draws exclusively on the results of past sampling efforts to
determine possible effects of certain remedial actions on the habitat and biota in select reaches of
the river. The data used to delineate the study area were collected from 1982 to 1994 during
several different sampling programs. The study area includes The Housatonic River between the
GE facility and Woods Pond Dam, Woods Pond, and bordering vegetated wetlands and other
habitats adjacent to, and within the 10-year floodplain of the Housatonic River and Woods Pond.
No formal documentation procedures for those sampling efforts are provided in the report;
however, this reviewer is aware that appropriate planning documents were prepared for past
sampling efforts by Blasland and Bouck. In addition, MDEP comments on the initial Interim
Phase IIReport (completed by BBL in 1992) indicates overall DEP satisfaction that correct
protocols in the literature, and as stated in the planning documents, were adhered to in the
Blasland and Bouck 1992 effort.

Standard procedures for the newer sampling efforts also are not included in the report, but are
assumed to be incorporated via the SAP/DCAQAP (BB&L, 1994).

Wetlands/aquatic habitats were evaluated by ENVIRON biologists in support of this report. In
addition to a two-day site reconnaissance, the evaluation included a review of historic reports
dating as far back as 1988 (USDA soil survey of Berkshire County). While some of this
information may be considered old, much of it is physical or geographical in nature, and may
still be considered "useable". Physiograhic information for as far back as 1975 (Lawler,

Matusky, and Skelly Engineers, Inc.) was used; but again, since this information is not analytical
in nature, it is considered "useable".

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC	^ ->	07/10/03

-------
The most recent updates of the EPA Risk Assessment Guidance (USEPA 1994 - ecological) and
the MDEP Guidance on Ecological Risk Assessment were considered in the evaluations
documented in this report, as well as EPA Guidance on Conducting Remedial
Investigations/Feasibility Studies (USEPA 1988). This helped ensure the evaluations are sound
and incompliance with accepted procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: B, caution
Comments: It is assumed that all historic data incorporated into this report follow documented
standard methods such as EPA methodologies. The assumption is based upon previous BBL
reports. However, the analytical methods (and their associated detection limits) used in these
past sampling and analysis efforts are not described in the report. It is assumed these details are
included in the SAP/DCAQAP, or in other work plans and reports for data collected prior to
1994, but this is not certain.

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: The level of data validation for the data collected data is unknown, based on the
report alone. Based on previous BBL reports regarding to the Housatonic River, some type of
data validation has generally been performed. Data from 1994 or later, is likely to be of greater
quality and useability than the pre-1994 data summarized in the report, because of its more
recent dating, and also because of the existence of the SAPDCAQAP at that time.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: Report does not reference established DQ indicators. As far as the reviewer can
tell, no DQIs were formally established for the sampling addressed in this report. These DQIs
may be in the SAPDCAQAP, but this would need to be verified, before assessing the quality of
the data collected as a result of the effort described in this work plan.

The majority of the tabulated data in this report are representations of habitat, biotic species,
avian species, plant species, rare, threatened, or endangered species, etc. Very little new
analytical data are presented, particularly for sediment and surface water. Data quality indicators
for these tabulated data are not provided.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Historic data appear to be of acceptable quality but much of the data derive from
studies conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

General comments: The data user should be cautious of some of the older data, whether
analytical or more qualitative in nature.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-61

-------
Title: PolychlorinatedBiphenyls in Housatonic River Sediments in Massachusetts and
Connecticut: Determination, Distribution, and Transport. A cooperative study by The
Connecticut Agricultural Experiment Station, the Connecticut Department of Environmental
Protection, and the U.S. Geological Survey.

Authors: C.R. Fink, B.L. Sawhney, K.P. Kulp, and C.GFredette, December, 1982
Reference Number: RFW-02-0016
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: Report provides complete description of study design with justification and
rationale.

Criterion 2: Formal documentation of procedures. Level: B

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: Split sediment samples were analyzed by USGS and CAES laboratories. The report
does not mention a Method number for PCB analysis. According to the report, the method of
PCB analysis of suspended sediment is described in detail by Gorelits and Brown (1972,

Methods for analysis of organic substances in water. U.S.G.S. Survey Techniques of Water
Resources Investigations, Book 5, Chap. A, 40 p.). This method uses hexane extraction and gas
chromatography for PCB concentration. Table 3. On page 7 the report presents the analytical
results for split samples. Caution should be used with this data due to the large RPDs between
the separate laboratory results for some of the data.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: The report did not contain a section on the level of review, validation, or quality
assurance associated with the data. The USGS and CAES laboratories analyzed a comparison of
20 split samples. A two-way analysis of variance showed the differences between sites were at
the 0.01 level, but differences between labs was not statistically significant. Only final data are
presented in the report and no results for QA/QC samples could be located.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: Unable to evaluate, precision and accuracy of analytical method is unknown.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality but some of the data potentially derive from
studies conducted prior to 1995. Methods may not meet current standards but are judged to have
produced data equivalent to current methodologies.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j ^ 07/10/03

-------
General comments: This report is a public release document developed by the USGS. The
format of the report differs from scientific reports associated with the Housatonic River Project.
Although some criterions are evaluated with a Level C rating, the source of the data (USGS) will
be considered for the overall level of confidence in the data.

Latitude and longitude have been determined for each sample location.

Sediment samples: 348
Water: 54

Overall Score: Level B: Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-63

-------
Title: Distribution of Polychlor incited Biphenyls in the Housatonic River and Adjacent Aquifer,
Massachusetts.

Author: Gay, F.G., Frimpter, M.H. (1984) U. S. Geological Survey-Water-supply Paper 2266.
Reference Number: RFW 02-0025
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s).	Level: B	

Comments: Report provides a description of study design, with justification and rationale.
Sample locations specified, but not located with GPS or equivalent.

Criterion 2: Formal documentation of procedures.	Level: C	

Comments: Field procedures are described in good detail. Sample handling, and analysis not
documented.

Criterion 3: Analytical methods used and detection limits achieved. Level: C
Comments: Specific documentation of analytical procedures is not provided.

Criterion 4: Data review, validation, and quality assurance.	Level: C

Comments: No information provided on the review and validation of the data.

Criterion 5: Assessment of data quality indicators.

Level: D

Comments: Data quality indicators not reported or discussed.

Criterion 6: Data history and overall apparent data quality.

Level: C

Comments: Data appear to be of acceptable quality. Methods are judged to have produced data

equivalent to current methodologies.



General comments:

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-64

-------
Title: Sediment Sampling and Analysis Data Report, Rising Paper Company, Great Barrington,
Massachusetts.

Note: Prepared for Rising Paper Company Division of Fox River Paper Company.

Author: GZA GeoEnvironmental, Inc. (1991)

Reference Number: 99-0276
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Details of sample collection and analysis are presented in the report. New data
(1991) presented in appendices. Sampling locations identified.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Procedures are identified, specific QA/QC requirements not specified.

Criterion 3: Analytical methods used and detection limits achieved. Level: B
Comments: A description of the analytical methods used for the 1991 data is present.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: Detailed information on the data review is not present. Quality assurance data for
metals and PCB analysis appear acceptable.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Laboratory reports contain DQIs, however no criteria are indicated in this report.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality but derive from a study conducted prior to
1995. Methods may not meet current standards but are judged to have produced data equivalent
to current methodologies.

General comments: Analytical data presented in the report are of acceptable quality. 1991 data
not previously reported.

Overall Score: Level C - Conditionally acceptable for limited use.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-65

-------
Title: Report on the Preliminary Investigation of Corrective Measures for Housatonic River and
Silver Lake Sediment.

Note: Excerpts from report. Prepared for General Electric Co.
Author: Harrington Engineering and Construction, Inc. (1996).
Reference Number: RFW 02-0098
Reviewed by: Doug Godfrey



Criterion 1: Overall quality of and level of detail in report(s).

Level: N/A

Comments: The report discusses the disposal of dredged sediments. No chemical data is
present.

Criterion 2: Formal documentation of procedures.

Level: N/A

Comments: No procedures documented.

Criterion 3: Analytical methods used and detection limits achieved.

Level: N/A

Comments: Specific documentation of analytical procedures is not present in report.

Criterion 4: Data review, validation, and quality assurance.

Level: N/A

Comments: No review or validation of data was mentioned.

Criterion 5: Assessment of data quality indicators.

Level: N/A

Comments: Data quality indicators not reported.

Criterion 6: Data history and overall apparent data quality.

Level: N/A

Comments: No data exists in the report as provided.

General comments: Report incorporates data obtained from previous reports (Blasland &
Bouck, P.C. and Jacobs Engineering Group.

Overall Score: Level N/A- Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-66

-------
Title: PCB Data From Our Files For 1977 Fish.

Authors: State of CT, Dept. of Health Services 10/8/91
Reference Number: RFW-02-0203
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s).	Level: D	

Comments: No information is available on background and conduct of study.

Criterion 2: Formal documentation of procedures.	Level: D	

Comments: Documentation non-existent, not available for review, or status unknown.

Criterion 3: Analytical methods used and detection limits achieved. Level: D	

Comments: Insufficient information provided or available via other sourced to establish validity
of data.

Criterion 4: Data review, validation, and quality assurance.	Level: D

Comments: No QA/QA procedures employed or documented.

Criterion 5: Assessment of data quality indicators.	Level: D

Comments: Not possible to evaluate DQIs.

Criterion 6: Data history and overall apparent data quality.	Level: D	

Comments: The overall data quality is questionable due to outmoded methodologies, poor
performance and/or apparent lack of consistency with current standards.

General comments: This report is a memo/letter with results attached. The text of the report
indicates, "These results are questionable and should not be used for any summary report".

Overall Score: Level D - Conditionally acceptable for limited uses

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-67

-------
Title: Inventory and Vegetation Classification of Floodplain Forest Communities in
Massachusetts. Final report submitted to the USEPA in fulfillment of Wetland Protection-State
Development Grant #CD 001976-01-1 by the Massachusetts Heritage & Endangered Species
Program, Massachusetts Division of Fisheries and Wildlife, Route 135, Westborough , MA

Authors: Jennifer Kearsley, Wetlands Plant Ecologist
Reference Number: RFW 02-0246
Reviewer: D. Sagges

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: A complete description of the area, study design and scope of work are provided.
This investigation was performed to the USEPA in fulfillment of Wetland Protection-State
Development Grant #CD-001976-01-1. The study was designed to show the variability in
floodplain forest vegetation environments. Sampling procedures, plot sampling and sample
handling were all discussed. In addition, sampling locations were well documented with
Floodplain Forest Landscape Analysis Summaries, USGS topographic maps, Massachusetts
Natural Heritage & Endangered Species Program Natural Community Site Description sheets
and EO Specs.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Samples were collected utilizing procedures in a given study area of specific size
and identified topographic or vegetation unit. Samples were collected using counts of vegetation
data from plots and soil characteristic data from soil pits. No QAPP was generated, no field
QA/QC samples were taken, no COCs and no sample attribute information.

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: Approved EPA analytical methods were not used; however, the data source would
indicate implementation of appropriate methods. Data were analyzed using numerical
classification (TWINSPAN) and ordination techniques (DCA) along with other statistical
methods.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan. There were no statements concerning
data review, evaluation or validation.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: No data quality indicators were noted in this report.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Data is recent (1997). Site history and data history were not discussed within the
body of the report. Overall intentions of sampling effort were specifically addressed and
approved methodologies were performed.

General comments: Overall the report was complete, but lacked support documentation. Based
on the data sources, the integrity of the data is not suspect.

Overall Score: Level C - Conditionally acceptable for limited uses, report is useful for plant
populations and soil typing.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC C 1 68 07/10/03

-------
Title: Concentration and Transport of PCBs in Housatonic River Between Great Barrington,

MA and Kent, CT, 1984-1988.

Authors: Kulp (USGS)

Reference Number: RFW-02-0201

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: Report is lacking detail. Does reference previous investigations performed by DEP
and USGS conducted 1979-80. Streamflow and sedimentation are discussed in some detail;
however, scope of work and sampling protocols are not significantly addressed. Sampling
locations were provided.

Criterion 2: Formal documentation of procedures. Level: D

Comments: Documentation was not provided, which includes lack of reference to a Workplan,
QAPP, SAP or required project paperwork, i.e. COCs, sample attributes and laboratory SOPs. In
addition, appropriate references were also not available.

Criterion 3: Analytical methods used and detection limits achieved. Level: D

Comments: No specific reference to approved methods. Brief statement that linked back to other
investigations.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan; QC samples/frequency were not
addressed. There was no statement concerning data review, evaluation and/or data validation.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not established within the study document. Additional
references were not provided. QA samples were not documented or presented in the text.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Data is over five years old (1984-1988). Data History was discussed within the
body of the report. Overall intentions of sampling effort were discussed, but there is no
confirmation of inherent data quality or laboratory procedures.

General comments: Overall the report was very brief. It was organized; however lacked
substance. The scope of the project, intended quality procedures and support documentation
were not provided.

Overall Score: Level C - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-69

-------
Title: Housatonic River PCB Sediment Management Study - Chapter 6, Program for Monitoring
the Natural Recovery of the River.

Author: Lawler, Matusky and Skelly Engineers, Inc. (1988)

Reference Number: RFW-02-0248
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is a statistical analysis of existing (up to 1986) fish, invertebrate, and
sediment data. Statistical analysis seems to be of high quality and detail. Details of sample
collection and analysis are presented in the report.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation indicates samples were collected and analyzed using several
methods. The report addresses possible effects of the use of multiple methods. Reports cited
may contain specific information on procedures

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: Specific documentation of analytical procedures is not present in report. Historical
reports are cited and my contain information on methods and detection limits. Appendix B
contains report of PCB analysis of one-inch increments of sediment from core samples of the
Housatonic River in 1986 (100 total samples) from York Lab. Detection limits are listed in the
appendix; method of analysis is described in laboratory narrative.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: Detailed information on the data review is not present.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Data quality indicators not reported. No further information on data quality exists.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Portions of the data appear to be of questionable quality due to age, changes in
methods and/or failure to follow current standards for scientific investigation.

General comments: Analytical data presented in the report are of acceptable quality. Appendix
B seems to have data not previously reported.

Overall Score: Level C - Conditionally acceptable for limited use.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-70

-------
Title: Ambient Trend Monitoring and PCB Fate and Transport.

Authors: Lawler, Matusky and Skelly Engineers 1991
Reference Number: RFW-02-0090
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: Accompanying report provides complete description of study design with
justification and rationale.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Documentation exists for most areas but is insufficient or lacking in a few areas
considered non-critical.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: Document has a QA/QC manual and SOP which address all of the laboratory
QA/QC aspects. EPA Method numbers are listed in the report.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Study incorporated all or most of the full range of QA/QC procedures, e.g., blanks,
spikes, duplicates, data review, and etc.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: DQIs not established.

Criterion 6: Data history and overall apparent data quality. Level: B

General comments:

Overall Score: Level B - Acceptable, use with some caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-71

-------
Title: Housatonic River CT Cooperative Agreement TasklVB - PCB Fate and Transport Model:

Additional Monitoring and Model Verification.

Authors: Lawler, Matusky and Skelly Engineers - November 1994

Reference Number: RFW-02-0099

Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: Report provides complete description of study design with justification and
rationale.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Documentation exists for most areas but is insufficient or lacking in a few areas
considered non-critical.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: PCB and TOC analysis performed by US EPA Methods.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: This report discusses QA/QC and ensures data validation was utilized. QA/QC
involved the collection of duplicate samples, field blanks, MS/MSD samples, and COC
procedures were followed and completed.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: Data quality indicators not established, but data appear to meet minimum standards
for DQI.

Criterion 6: Data history and overall apparent data quality. Level: B

General comments:

Overall Score: Level B - Acceptable, use with some caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-72

-------
Title: Housatonic River PCB Sediment Management Study (Chapters 1-5).

Authors: Lawler, Matusky & Skelly Engineers, 1985
Reference Number: RFW 02-0247
Reviewer: D. Sagges

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: Characterize areas, screen alternative technologies for PCB removal or treatment,
and provide a comparison of current technologies. Some analytical data listed, but no specific
procedures outlined.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Support documentation was not provided including COCs, and sample attribute
information. Laboratory SOPs were not provided. No support information for Water Quality
samples collected in 1972. List of Sediment samples run for PCB concentration can be found in
Appendix A of Frick et al. (1982) report. See serial number 34, reference number 02-0016.

Criterion 3: Analytical methods used and detection limits achieved. Level: D

Comments: Approved EPA analytical methods were not cited in this report. The data source for
Water Quality was not provided and PCB data was not listed in this report.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan. No statements concerning data review,
evaluation or validation.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not established within this compiled report. Additional
reference was listed, but was not provided. QA samples were not documented.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Data is over five years old (1972 and 1982). Data history was not discussed within
the body of the report. Overall intentions of sampling effort were not specifically addressed
within this report.

General comments: Overall the report lacked support documentation; however, based on the
data sources, the integrity of the data is not suspect. The scopes of this investigation and
intended quality procedure were not provided.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-73

-------
Title: Ninety-first Christmas Bird Count, Central Berkshire, and Massachusetts.

Authors: Neumuth, EJ (1991), American Birds 45(4),621
Reference Number: RFW 99-0403
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Comments: Report consists of data tables from bird counts for a 15-mile diameter that
encompasses the towns of Lanesboro, Pittsfield, Dalton, Hinsdale, Washington, Lee, Lenox,
West Stockbridge, Richmond, and Hancock, Massachusetts. Pittsfield and Lenox fall entirely
within the 15-mile diameter and the remaining towns are represented by portions of their area.
No formal report accompanies the data tables.

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Data tables were reviewed and validated by the report author.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: A

Comments: The same person has completed the bird count in the area of interest annually for at
least 8 years.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-74

-------
Title: Ninety-second Christmas Bird Count, Central Berkshire, and Massachusetts.

Authors: Neumuth, EJ (1992), American Birds 46(4),613
Reference Number: RFW 99-0404
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Data tables were reviewed and validated by the report author.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: A

Comments: The same person has completed the bird count in the area of interest annually for at
least 8 years.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-75

-------
Title: Ninety-third Christmas Bird Count, Central Berkshire, and Massachusetts.

Authors: Neumuth, EJ (1993), American Birds 47(4),580
Reference Number: RFW 99-0405
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Data tables were reviewed and validated by the report author.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: A

Comments: The same person has completed the bird count in the area of interest annually for at
least 8 years.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-76

-------
Title: Ninety-fourth Christmas Bird Count, Central Berkshire, and Massachusetts.

Authors: Neumuth, EJ (1994), American Birds 48(4), 456
Reference Number: RFW 99-0406
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Data tables were reviewed and validated by the report author.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: A

Comments: The same person has completed the bird count in the area of interest annually for at
least 8 years.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-77

-------
Title: Ninety-fifth Christmas Bird Count, Central Berkshire, and Massachusetts.

Authors: Neumuth, EJ (1995), American Birds 49(4), 426-427
Reference Number: RFW 99-0407
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: N/A

Criterion 2: Formal documentation of procedures. Level: N/A

Comments: No formal documentation of procedures is discussed regarding the collection,
analysis, or validity of the data.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: No description of analytical methods used or detection limits achieved.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Data tables were reviewed and validated by the report author.

Criterion 5: Assessment of data quality indicators. Level: N/A

Comments: No data analyses are discussed in this document.

Criterion 6: Data history and overall apparent data quality. Level: A

Comments: The same person has completed the bird count in the area of interest annually for at
least 8 years.

General comments: No analytical data specific to Housatonic River environmental samples is
contained in this document.

Overall Score: N/A

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-78

-------
Title: Housatonic River Study, 1980 and 1982 Investigations
Authors: Stewart Laboratories, Inc. December 1982
Reference Number: RFW-02-0030
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: The data are accompanied by a narrative report that provides complete detail of the
study design and includes discussion of the underlying reasons for selecting the stated sampling
locations and methods.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Documentation exists for most areas but is insufficient or lacking in a few areas
considered non-critical. No chain of custody records, quality assurance plans, or SOPs are
contained in the report. Field sampling plans and analysis protocols are described in the
Housatonic River Study Proposal submitted to the EPA and MDEP.

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: Only chromatographic peaks verified by GC/MS to be free of interferences from
materials other than PCBs were used for the analysis. Aroclor 1242 reported only if >5% of
1242 was present in the PCB total. ASTM methods were used for determinations of sediment
specific gravity. Particle size determinations were made using ASTM Standard Method D21.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: The report does contain a section pertaining to sample handling and quality
controls for each of the types of matrix sampled. Appendix nine is contains split data for river
sediment splits, fish composite splits and replicates, aquatic plant splits, and information on the
EPA split sample program.

Criterion 5: Assessment of data quality indicators. Level: C

Comments: It does not appear that DQIs were established as part of the planning process.

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Some of the data may be questionable due to the age of the data. Sample locations
are poorly marked

General comments:

Overall Score: Level C: Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-79

-------
Title: Preliminary Report - Wetland Characterization and Function - Value Assessment.

Housatonic River from Newell Street to Woods Pond.

Author: Woodlot Alternatives (1998).

Note: Report prepared for Tech Law, Inc.

Reference Number: RFW 00-0309

Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s).	Level: B	

Comments: The report provides a description of study design, justification, or rationale of the
wetlands characterization.

Criterion 2: Formal documentation of procedures.	Level: B

Comments: Procedures of the wetlands characterization were documented.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A
Comments: This study does not contain any analytical data.

Criterion 4: Data review, validation, and quality assurance.	Level: C	

Comments: Field verification of photo-interpretation was performed by walking the identified
sites.

Criterion 5: Assessment of data quality indicators.	Level: N/A

Comments: Data quality indicators not reported or discussed.

Criterion 6: Data history and overall apparent data quality.	Level: A	

Comments: Data appear to be of acceptable quality. Methods are judged to have produced data
equivalent to current methodologies.

General comments: Report is a wetlands characterization report, and contains no analytical
data.

Overall Score: N/A - contains no analytical data.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC	C 1 80	07/10/03

-------
Title: Ambient Air Monitoring for PCBs, August 20, 1991 through August 14, 1992.

Author: Zorex Environmental Engineers, Inc. (1992).

Reference Number: RFW 01-0135
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: The report provides complete description of study design and sample locations, with
justification and rationale.

Criterion 2: Formal documentation of procedures. Level: A

Comments: Scope of Work (SOW), and Quality Assurance Project Plan (QAPP) exist. SOW
approved by MDEP 12/1990, and QAPP submitted 8/1991.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: Analytical procedures follow documented standard methods (EPA Compendium
Method 608) are not well documented.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Data review and quality assurance was described in detail. All quality objectives
seem to have been met, and invalid samples were identified in the report with descriptions of the
problems.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: Study had established DQI and data substantially meet all acceptability criteria for
completeness comparability, representativeness, precision, and accuracy.

Criterion 6: Data history and overall apparent data quality. Level: B

General comments: Report prepared in cooperation with the Connecticut DEP. Sample
locations well documented, analytical protocol and review information weak.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC C 1 81 07/10/03

-------
Title: Ambient Air Monitoring for PCBs - May 4, 1993 to August 17, 1993 Book 1 of 3.
Authors: Zorex Environmental
Reference Number: RFW-01-0137
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: Report provides complete description of study design with justification and
rationale.

Criterion 2: Formal documentation of procedures. Level: A

Comments: Report has an associated QAAP. Instrument calibrations were noted.

Criterion 3: Analytical methods used and detection limits achieved. Level: A, caution
Comments: Samples were collected according to EPA methods. GE developed their own
CG/MS method to study PCB degradation. Any data associated with the degradation study
should be used cautiously, as there may be a conflict of interest in the GE study.

Criterion 4: Data review, validation, and quality assurance. Level: A, caution

Comments: Data was validated, field blanks were collected and COC procedures were
incorporated into sample handling. This study incorporated all or most of QA/QC procedures.
Data anomalies were observed, which are documented and the reader is cautioned.

Criterion 5: Assessment of data quality indicators. Level: A

Comments: Study had established DQI indicators and data substantially met acceptability
criteria for completeness, comparability, representativeness, precision, and accuracy.

Criterion 6: Data history and overall apparent data quality. Level: B

General comments: Books 1 to 3 evaluated as one report. Air sampling data appears to be of
good quality. PCB degradation data collected by GE should be used with caution.

Overall Score: Level B - Acceptable, use with caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-82

-------
Title: Ambient Air Monitoring for PCBs - May 10, 1995 to August 24, 1995.

Authors: Zorex Environmental - January 1996
Reference Number: RFW 02-0254
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: Report provides complete description of study design with justification and
rationale.

Criterion 2: Formal documentation of procedures. Level: A

Comments: Report has an associated QAAP. Laboratory and field notes are included in the
report for review. Instrument calibrations were noted.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: Samples were collected according to EPA methods. High volume samples were
collected in accordance with EPA Compendium Method TO-10 and analyzed for PCBs using
EPA Method 608. Low volume samples were collected in accordance with EPA Method TO-10
and analyzed for PCBs using EPA Method 608.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: Data was validated, field blanks were collected and COC procedures were
incorporated into sample handling. This study incorporated all or most of QA/QC procedures.

Criterion 5: Assessment of data quality indicators. Level: A

Comments: Study had established DQI indicators and data substantially met acceptability
criteria for completeness, comparability, representativeness, precision, and accuracy.

Criterion 6: Data history and overall apparent data quality. Level: A

Comments: Data are recent, reported in standard units, and are reasonable and internally
consistent. Methods followed meet current standards for scientific investigation and were
followed consistently.

General comments: This document appears to be well written.
Overall Score: Level A - Acceptable, unrestricted use.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-83

-------
Title: Length-age Relations andPCB Content of Mature White Suckers from the Connecticut
and Housatonic River Basins.

Note: Samples are composites

Author: Coles, J.F. 1999. Northeastern Naturalist 6(3): 263-275.

Reference Number: RFW 02-0255
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: The report provides a description of study design with justification and rationale.
Sample location information given, but may not be sufficient to allow determination of
individual sample locations.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Samples were collected as part of the USGS National Water Quality Assessment
Program. Field procedures are described in good detail. Samples were shipped frozen, but no
Chain-of-Custody information is provided.

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: Specific documentation of analytical procedures is not provided, however, the
USGS laboratory in Arvada Co. analyzed samples for organochlorine concentrations, including
Total PCBs.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: No information provided on the review and validation of the data. Replicate
samples were taken at two sites but results are not discussed.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Data quality indicators not reported or discussed.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality. Methods may not meet current standards
but are judged to have produced data equivalent to current methodologies.

General comments: The USGS organochlorine, including total PCB methods are judged to have
been produced data equivalent to current methodologies. Data quality not discussed in article,
but samples analyzed by a known laboratory.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-84

-------
Title: Endocrine Biomarkers, OC-Pest, PCBs, In Large Mouth Bass from Woods Pond,
Housatonic River, MA 9/1994 and 5/1995.

Authors: USGS in cooperation with US EPA, Reston, Virginia 1997
Reference Number: RFW-02-0121
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report provides complete description of study design with justification and
rationale.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Detailed study methods are given for sample extraction and laboratory analysis.

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: Analytical methods are described and protocols are NAWQA standards. Positive
controls were used and the coefficient of variation was calculated for each duplicate sample.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: Unknown if data validation was performed on the laboratory results. The report
does make some references to positive controls and duplicate samples.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Not possible to evaluate DQIs.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Data appear to be of acceptable quality but some of the data potentially derives
from studies conducted prior to 1995. Methods may not meet current standards but are judged to
have produced data equivalent to current methodologies. PCB and congener results are
presented without units.

General comments: This report contains useful PCB and Congener data from fish samples
collected from Woods Pond; however, no result units are presented in the data tables.

The only location information for the fish is Woods Pond. No location information is provided
with each sample location.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-85

-------
Title: Geographical Distribution and Potential Adverse Biological Effects of Selected Trace
Elements and Organic Compounds in Streambed Sediment in Connecticut, Housatonic, and
Thames River Basins, 1992-1994

Authors: USGS Water Resources Investigation Report 97-4169
Reference Number: RFW-02-0209
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Streambed-sediment samples were collected during three sampling periods; June-
November 1992, July-September 1993, and August-September 1994. A total of 87 samples were
collected from 43 sampling sites. Sample collection techniques are described and analytical
methods are not described but references are provided.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Some documentation may be insufficient in some areas deemed non-critical.
References are provided in the report for the USGS Quality assurance manual, USGS National
Water Quality Laboratory, etc.

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: Analytical methods are described and protocols are NAWQA standards. Samples
for organic-constituent analysis were wet sieved through a 2-mm stainless sieve and analyzed for
32 organochlorine compounds including PCBs, and 64 semi-volatile organic compounds at the
USGS National Water Quality Laboratory.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: Unknown if data validation was performed on the laboratory results. Report
references RFW 02-0210, RFW 02-0025, RFW 02-0212.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Not possible to evaluate DQIs. Data are not available in this report to determine
DQIs.

Criterion 6: Data history and overall apparent data quality. Level: D

General comments: This report contains useful data from sediment samples collected from the
Connecticut, Housatonic, and Thames river basins. Analytical results are not listed in data tables
and no lat/long are provided for sample locations. Report references RFW 02-0210, RFW 02-
0025, RFW 02-0212.

Additional information would most likely increase the overall score for this report.

Overall Score: Level C - Conditionally acceptable, for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC C 1 86 07/10/03

-------
Title: Organochlorine Compounds and Trace Elements in Fish Tissue and Ancillary Data for

the CT, Housatonic, and Thames River Basin Study Unit 1992-1994.

Authors: USGS - 1992 to 1994

Reference Number: RFW-02-0210

Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is generally complete and well written but lacks sufficient details in a few
areas.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation generally not available but sufficient information is known or
available via other sources to establish validity of field and analytical procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: A
Comments: Analytical procedures follow documented standard methods such as EPA or
ASTM.

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: Spiked samples, surrogates and blind samples and internal reference sample are
documented; however, the level of data validation is unknown.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Not possible to evaluate DQIs.

Criterion 6: Data history and overall apparent data quality. Level: B

General comments: This report is a public release document developed by the USGS. The
format of the report differs from scientific reports associated with the Housatonic River Project.
Although some criterions are evaluated with a Level D rating, the source of the data (USGS) will
be considered for the overall level of confidence in the data.

Report contains results and analysis for 42 metals, 32 PCB fish, and approximately 250 wt,
lengths, sex, and age data for fish. There are also 32 locations with lat/long information.

Overall Score: Level B - Acceptable, use with some caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC C 1 87 07/10/03

-------
Title: Water Quality in the CT, Housatonic, and Thames River Basin.

Authors: USGS - 1992 to 1995
Reference Number: RFW-02-0213
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: Report is incomplete but does provide sufficient information for one or more
parameters of interest.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation generally not available but sufficient information is known or
available via other sources to establish validity of field and analytical procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: D

Comments: Insufficient information provided or available via other sources to establish validity
of data.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No QA/QC procedures employed or documented. References RFW 02-0209, RFW
02-0210.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Not possible to evaluate DQIs.

Criterion 6: Data history and overall apparent data quality. Level: B

References RFW 02-0209, RFW 02-0210.

Overall Score: Level B - Acceptable, use with some caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-88

-------
Title: Inorganic and Organic Constituents and Grain Size Distributions in Streambed Sediment
and Ancillary Data for the CT, Housatonic, and Thames River Basin Study Unit 1992-1994.
Authors: USGS - 1992 to 1994
Reference Number: RFW-02-0212
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: Report is generally complete and well written but lacks sufficient details in a few
areas.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation generally not available but sufficient information is known or
available via other sources to establish validity of field and analytical procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: A
Comments: Analytical procedures follow documented standard methods such as EPA or
ASTM.

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: Spiked samples, surrogates and blind samples and internal reference sample are
documented; however, the level of data validation is unknown. A total of 43 locations were
sampled for organic and inorganic constituents.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: Not possible to evaluate DQIs.

Criterion 6: Data history and overall apparent data quality. Level: B

General comments: This report is a public release document developed by the USGS. The
format of the report differs from scientific reports associated with the Housatonic River Project.
Although some criterions are evaluated with a Level C and D rating, the source of the data
(USGS) will be considered for the overall level of confidence in the data.

Overall Score: Level B - Acceptable, use with some caution, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-89

-------
Title: Report on Lower Housatonic River Sediment PCB Sampling.

Authors: BBL (1996)

Reference Number: RFW-02-0149

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: This report is a compilation of sampling efforts performed in the Lower Housatonic
on behalf of either USEPA Region I or Connecticut DEP. It does not specifically reference
Workplans, study designs or QAPPs that were implemented on behalf of these investigations;
however, the reliability of the data sources would indicate the existence of support
documentation. In most cases, sampling locations were provided in the form of Lat/Log or on
maps.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Documentation was not provided, as discussed in criterion 1. However sampling
support documentation was provided in most cases, including COCs, and sample attribute
information. Laboratory SOPs were not provided, but some of the CTDEP analyses were
performed at the USACE New England Environmental Laboratory.

Criterion 3: Analytical methods used and detection limits achieved. Level: A
Comments: In all cases, approved EPA analytical methods were cited and implemented.

Criterion 4: Data review, validation, and quality assurance. Level: A/B

Comments: No indication of a Quality Assurance Plan; however the EPA data set contained
data validation summaries. The CTDEP data sets contained varying degrees of QC
procedures/analyses, including surrogate, spike and blank analyses. The EPA data set level for
criterion 4 is A. The CTDEP data sets level for criterion 4 is B, since there were no statements
concerning data review.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: DQI indicators were not established within this compiled report. Additional
references were not provided. QA samples were documented and in some cases validated
against Region I Functional Guidelines. Based on the information provided, the reviewer could
not fully evaluate the assessment of DQIs, but the data does appear to meet minimum standards.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data is over five years old (1981-1985). Data History was not discussed within the
body of the report. Overall intentions of sampling effort were not specifically addressed within
this report; however approved methodologies were performed.

General comments: BB&L has compiled this report of sampling efforts performed on the
Lower Housatonic on behalf of either USEPA Region I or CTDEP. Overall the report lacked
support documentation; however, based on the data sources, the integrity of the data is not
suspect. The scopes of the various investigations and intended quality procedures were not
provided.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-90

-------
Title: Task 220: Exploratory Site Investigation - Replacement of the Stevenson Dam Bridge

(CTRT34) Monroe/Oxford, CT.

Authors: GEI Consultants (1999)

Reference Number: RFW-02-0202

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Comments: A complete description of the study design, scope of work and rationale are
provided. This investigation was performed on behalf of ConnDOT. The sight history, local
environment, field investigation methods and potential receptors were all addressed. Sampling
procedures, boring logs, and sample handling were discussed. In addition, sampling locations
and depths were well presented.

Criterion 2: Formal documentation of procedures. Level: B

Comments: Samples were collected utilizing procedures established within this report. A
formalized QAPP was not generated for this investigation, and no field QA/QC samples were
collected. The complete laboratory reports were provided, which included COCs, and sample
attribute information. However, laboratory SOPs were not supplied.

Criterion 3: Analytical methods used and detection limits achieved. Level: A
Comments: In all cases, approved EPA analytical methods were cited and implemented.

Criterion 4: Data review, validation, and quality assurance. Level: B

Comments: No indication of a Quality Assurance Plan. Some internal laboratory QA/QC
analyses were performed. There were no statements concerning data review, evaluation or
validation.

Criterion 5: Assessment of data quality indicators. Level: B

Criterion 6: Data history and overall apparent data quality. Level: A

Comments: Data is new (1999). Data History was discussed within the body of the report.
Overall intentions of sampling effort were specifically addressed and approved methodologies
were performed.

General comments: Overall the report was complete. The investigation did lack a QA plan and
QA sampling scope, but it is believed that the integrity of the sample results were not
compromised.

Overall Score: Level A - Acceptable, unrestricted use.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-91

-------
Title: CTDEP, Bureau of Water Management, "InterdepartmentalMessage, USGS1992

Housatonic River Sediment PCB Data. "

Authors: Charles Fredette (CTDEP-1992)

Reference Number: RFW-02-0206

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: A complete description of the study design and rationale were not provided. The
overall intent was not clear and no specific procedures were outlined. Results may be presented
in their entirety in the original source document by USGS.

Criterion 2: Formal documentation of procedures. Level: D

Comments: Formal documentation and procedures were not addressed. Reviewer unable to
make accurate assessment of procedures performed.

Criterion 3: Analytical methods used and detection limits achieved. Level: D
Comments: Analytical methods were not cited within the text of this memorandum.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan. There were no statements concerning
data review, evaluation or validation.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not established within this report. Additional references were
not provided.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Data is over five years old (1992). Data History was not discussed within the body
of the report. Overall intentions of sampling effort were not specifically addressed.

General comments: Overall the information provided did not support the sample results
presented. This memorandum relayed information internally with CTDEP. The background of
the USGS investigation was not discussed, the data user should rely on source investigation for
analytical result presentation and use.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-92

-------
Title: State of Connecticut, Department of Health Services: Housatonic River PCB Fish Log

Book, 1979 Samples.

Authors: CTDOH(1979)

Reference Number: RFW-02-0204

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: A complete description of the study design and rationale were not provided. The
overall intent was not clear and no specific procedures were outlined.

Criterion 2: Formal documentation of procedures. Level: D

Comments: Formal documentation and procedures were not addressed, with the exception of
the procedure for recording tissue attribute information (i.e. species, data sampled, body of
water, area, method of capture, length, weight, sex, age, comments and PCB Level). Reviewer
unable to make accurate assessment of other procedures performed.

Criterion 3: Analytical methods used and detection limits achieved. Level: D
Comments: Analytical methods were not cited, but analyses were performed at the DOH
Laboratory Division.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan. There were no statements concerning
data review, evaluation or validation.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not established within this report. Additional references were
not provided.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Data is over five years old (1979). Data History was not discussed within the body
of the report. Overall intentions of sampling effort were not specifically addressed.

General comments: Overall the information provided did not support the sample results
presented. Report consists of copies of hand-written fish logbook pages, rough maps and a brief
set of laboratory notes. No support documentation or procedures were referenced or provided.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-93

-------
Title: PCBsin the Housatonic River: Determination and Distribution
Authors: Sawhney, Frink and Glowa (1981)

Reference Number: RFW-02-0205
Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: A complete description of the study design, scope of work and rationale were
provided. Brief background was addressed. Sampling protocol and preservation were discussed.
Specific locations and depths were not recorded.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Samples were collected utilizing procedures established within this report. There
was no evidence of a formalized QAPP and no indication that field QA/QC samples were
collected.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: A detailed description of the analytical procedure was presented in the study text.
Did not reference approved EPA method, but mirrored EPA protocol.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: No indication of a Quality Assurance Plan. There were no statements concerning
data review, evaluation or validation.

Criterion 5: Assessment of data quality indicators. Level: C

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Data is over five years old (1981). Data History was discussed within the body of
the report. Overall intentions of sampling effort were specifically addressed.

General comments: This study examined the determination and distribution of PCBs. The
sample data were presented for examination and evidence purposes only; therefore, no specific
sampling locations were provided. The paper was well written and organized, but the data was
not intended for overall investigation use.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-94

-------
Title: Higher Level ofPCBs in Housatonic Feared
Authors: Connecticut Post (1993)

Reference Number: RFW-02-207
Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: The article discussed the potential release ofPCBs as a result of the pond drainage
at the Rising Paper Dam Project. Brief site history was discussed, including the potential effect
on the levels in fish specimens

Criterion 2: Formal documentation of procedures. Level: D

Comments: Formal documentation was not presented; associated studies and/or data sources
were not referenced.

Criterion 3: Analytical methods used and detection limits achieved. Level: D

Comments: Specific sample results were not presented; therefore an analytical procedure was
not presented in the text of the article.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: There were no statements concerning data review, evaluation or validation.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not established within this report. Additional references were
not provided. Based on the information provided, the reviewer could not fully evaluate the
assessment of DQIs

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Article is over five years old (1993). Data History was discussed within the body of
the text.

General comments: This article does not present specific sample results. It discusses the
potential problems that may exist as a result of the pond drainage in relation to the Rising Paper
Dam Project.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-95

-------
Title: CandlewoodPCB Data

Authors: Environmental Research Institute, UCONN (2000)

Reference Number: RFW-02-0208
Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: Series of result summary pages. It does not specifically reference Workplans,
study designs or QAPPs that were implemented on behalf of these analyses; however, the
reliability of the data sources would indicate the existence of support documentation.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Support documentation was not provided including COCs, and sample attribute
information. Laboratory SOPs were not provided. However the analyses were performed on
behalf of CTDEP. Samples were shipped frozen, but Chain-of-Custody information is provided.

Criterion 3: Analytical methods used and detection limits achieved. Level: B

Comments: Approved EPA analytical methods were not cited; however the data source would
indicate implementation of appropriate methods. The USGS laboratory in Arvada Co. analyzed
samples for organochlorine concentrations, including Total PCBs.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: No indication of a Quality Assurance Plan and no statements concerning data
review. Replicate samples were taken at two sites but results are not discussed.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: DQI indicators were not established within this compiled report. Additional
references were not provided. QA samples were not documented. Based on the information
provided, the reviewer could not fully evaluate the assessment of DQIs, but the data does appear
to meet minimum standards.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data is new (2000). Data History was not discussed within the body of the report.
Overall intentions of sampling effort were not specifically addressed within this report.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-96

-------
Title: Letter to Mr. Richard Thibedeau, Massachusetts Department of Environmental

Management, From Michael J. Harder

Authors: CTDEP (1994)

Reference Number: RFW-02-0218

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: The cover letter is followed by a series of PCB result summary pages and graphical
distributions for fish and insect larvae tissue. It does not reference Workplans, study designs or
QAPPs that were implemented on behalf of these analyses; however, the reliability of the data
sources would indicate the existence of support documentation. Sampling locations were not
reported.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Support documentation was not provided including COCs, and sample attribute
information. Laboratory SOPs were not provided. However the analyses were performed on
behalf of CTDEP

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: EPA analytical methods were not cited; however the data source would indicate
implementation of appropriate methods.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: No indication of a Quality Assurance Plan and no statements concerning data
review.

Criterion 5: Assessment of data quality indicators. Level: C

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Data is over five years old (1994). Data History was not discussed within the body
of the report. Overall intentions of sampling effort were not specifically addressed within this
report.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-97

-------
Title: Letter to D. Westone from R. W Berliner, Yale University, Review of Health Services

Study of PCB-contaminated Fish from the Housatonic River.

Authors: Berliner, Yale Univ. (1981/82)

Reference Number: RFW-02-0221

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: The letter discussed the general interpretation of study results in PCB-contaminated
fish. Specific sample results were not included. It does not reference Workplans, study designs
or QAPPs that were implemented on behalf of these analyses.

Criterion 2: Formal documentation of procedures. Level: D

Comments: Support documentation was not provided including COCs, and sample attribute
information. Laboratory SOPs were not provided.

Criterion 3: Analytical methods used and detection limits achieved. Level: D

Comments: Letter specifically addresses the concern of the review committee in regards to lack
of information on the analytical method.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan and no statements concerning data
review.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not established within this compiled report. Additional
references were not provided. Based on the information provided, the reviewer could not fully
evaluate the assessment of DQIs.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Data is over five years old (1981). Data History was not discussed. Overall
intentions of sampling effort were not specifically addressed within this report.

General comments: Overall the report lacked documentation. The scopes of this investigation
and intended quality procedure were not provided.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-98

-------
Title: PCBs in the Housatonic River Fish Fact Sheet
Authors: CTDOH(1981)

Reference Number: RFW-02-0222
Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: The fact sheet lists warnings against human consumption of Housatonic River fish.
It does not reference specific sample results, workplans, study designs or QAPPs that were
implemented on behalf of these analyses.

Criterion 2: Formal documentation of procedures. Level: D

Comments: Support documentation was not provided including COCs, and sample attribute
information. Laboratory SOPs were not provided.

Criterion 3: Analytical methods used and detection limits achieved. Level: D

Comments: Fact sheet does not specifically address the analytical methods utilized in the
analysis of the fish tissues.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan and no statements concerning data
review.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not discussed in this fact sheet. Additional references were not
provided. Based on the information provided, the reviewer could not fully evaluate the
assessment of DQIs.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Sheet is over five years old (1981). Data History was not discussed. Overall
intentions of sampling effort were not specifically addressed within this report.

General comments: The fact sheet did not reference support documentation. The scopes of this
investigation and intended quality procedure were not provided.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-99

-------
Title: Memorandum toM. Harder from C.G. Fredette, Summary of 1992 CTDEP Housatonic

PCB Monitoring RE: Rising Dam, Great Barrington, MA.

Authors: CTDEP (1993)

Reference Number: RFW-02-0220

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: C

Comments: The memorandum is followed by a series of PCB result summary pages and
graphical distributions for sediment, water and trout sampling in conjunction with Rising Dam
during 1992. It does not reference Workplans, study designs or QAPPs that were implemented
on behalf of these analyses; however, the reliability of the data sources would indicate the
existence of support documentation. Sampling lat/long was reported.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Support documentation was not provided including COCs, and sample attribute
information. Laboratory SOPs were not provided. However the analyses were performed on
behalf of CTDEP

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: EPA analytical methods were not cited; however the data source would indicate
implementation of appropriate methods.

Criterion 4: Data review, validation, and quality assurance. Level: C

Comments: No indication of a Quality Assurance Plan and no statements concerning data
review.

Criterion 5: Assessment of data quality indicators. Level: C

Criterion 6: Data history and overall apparent data quality. Level: C

Comments: Data is over five years old (1992). Data History was not discussed within the body
of the report. Overall intentions of sampling effort were not specifically addressed within this
report.

Overall Score: Level C - Conditionally acceptable for limited uses.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-100

-------
Title: Organochlorine Compounds in Fish Tissue from the Connecticut, Housatonic, and

Thames River Basins Study Unit, 1992-94.

Authors: USGS (1998)

Reference Number: RFW-02-0211

Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Criterion 2: Formal documentation of procedures. Level: B

Comments: Samples were collected utilizing procedures referenced within this report. There
was no evidence of a formalized QAPP and no indication that field QA/QC samples were taken,
but existence of quality control procedures were indicated. Significant statistical analyses were
presented.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: A detailed description of the analytical procedure was presented in the study text.
Did not reference approved EPA method, but mirrored EPA protocol.

Criterion 4: Data review, validation, and quality assurance. Level: B

Criterion 5: Assessment of data quality indicators. Level: C

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data is over five years old (1994). Data History was discussed within the body of
the report. Overall intentions of sampling effort were specifically addressed.

General comments: This study examined the extent of contamination of organochlorine
compounds. The sample data were well presented, extensive background information was
provided and the intent of scope was discussed. The paper was well written and organized.
Report references RFW 02-0050, RFW 02-0210, RFW 02-0025, RFW 02-0212, report also
references unpublished data from Kenneth Carr, USFW Region Office, Concord, NH.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.l-101

-------
Title: PCB Concentrations in Fishes from the Housatonic River, Connecticut in 1984, 1986 and
1988.

Authors: Academy of Natural Sciences (1990)

Reference Number: RFW-02-0214
Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Criterion 2: Formal documentation of procedures. Level: A

Locations were discussed in-depth and background sampling was presented. Extensive statistical
analyses were performed.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: A detailed description of the analytical procedure was presented in the study text.
Did not reference approved EPA method, but mirrored EPA protocol.

Criterion 4: Data review, validation, and quality assurance. Level: B

Criterion 5: Assessment of data quality indicators. Level: B

Comments: DQI indicators were not established within this report. Various QA/QC samples
were documented, including spikes, standards, replicates and off-site splits.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data is over five years old (1988). Data History was discussed within the body of
the report. Overall intentions of sampling effort were specifically addressed.

General comments: This study examined the extent of PCB contamination in fish, water and
sediment samples. Age determination was performed along with the other analytical parameters.
The sample data were well presented, extensive background information was provided and the
intent of scope was discussed. The paper was well written and organized.

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.1-102

-------
Title: PCB Concentrations in Fishes and Benthic Insects from the Housatonic River,
Connecticut in 1984, to 1998.

Authors: Academy of Natural Sciences of Philadelphia, prepared for General Electric Company

15 November 1999

Reference Number: RFW-02-0216

Reviewer: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: A

Criterion 2: Formal documentation of procedures. Level: A

Locations were discussed in-depth and background sampling was presented. Extensive statistical
analyses were performed.

Criterion 3: Analytical methods used and detection limits achieved. Level: A

Comments: A detailed description of the analytical procedure was presented in the study text.
Did not reference approved EPA method, but mirrored EPA protocol.

Criterion 4: Data review, validation, and quality assurance. Level: A

Comments: A quality assurance statement is provided which states that the study was performed
under the general provisions of the Patrick Center Quality Assurance Implementation Plan. The
assurance also indicates that the final report was reviewed and it was determined that the report
is an accurate reflection of the data obtained. There were no specific statements concerning data
review, evaluation or validation. The raw data and the final report are filed in the Patrick Center
archives. Based on the source of the sample results and extensive content within text, the quality
of the analytical results is not in question.

Criterion 5: Assessment of data quality indicators. Level: A

Comments: Various QA/QC samples were documented, including spikes, standards, replicates
and off-site splits. According to the text, the median coefficient of variation of TPCB for 1998
replicate analysis was 0.10 and the median standard deviation of LNTPCB was 0.10. A standard
reference material (carp) obtained fro Canada was used to evaluate analytical accuracy. The
values select PCB congeners were within an average of 15% of the certified values. Detected
limits for congeners were estimated by repeated analysis of a spiked sample of fish with
undetectable PCB concentration, with the detection limit calculated in accordance with EPA
protocols (40 CFR 136). Detection limits were relatively similar among congeners, with an
average value of 0.31 ppb wet weight. Text in the report indicates that the sample sizes for
benthic insect groups were to small to permit meaningful test of significance with regard to PCB
concentrations.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC ^ i 07/10/03

-------
Criterion 6: Data history and overall apparent data quality.	Level: B	

Comments: Data History was discussed within the body of the report. Overall intentions of
sampling effort were specifically addressed. Due to the age of the older, some results may not be
of the same quality; however, the study most likely produced data that are equivalent to what
would have been produced using current methodologies.

General comments: The sample data were well presented, extensive background information
was provided and the intent of scope was discussed. The paper was well written and organized.
Three results for individual benthic insect collected in 1998 are contained within the report.
According to the report 80 fish samples were collected in 1998; however, no individual PCB
results are presented. Report references RFW 02-0050 and PCB Concentrations in Fishes from
the Housatonic River, Connecticut in 1984, to 1996 (Academy of Natural Sciences of
Philadelphia).

Overall Score: Level B - Acceptable, some use restrictions may apply.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.1-104

-------
Title: Housatonic River FloodPCBs
Authors: CTDOH(1984)

Reference Number: RFW-02-0217
Reviewer: K. Spittler

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: Consists of a series of sample result summaries. It does not reference a specific
workplan, study design or QAPP that were implemented on behalf of these analyses.

Criterion 2: Formal documentation of procedures. Level: D

Comments: Support documentation was not provided including COCs, and sample attribute
information. Laboratory SOPs were not provided.

Criterion 3: Analytical methods used and detection limits achieved. Level: D

Comments: Result summaries, maps and transport records do not specifically address the
analytical methods utilized in the analysis of the samples.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No indication of a Quality Assurance Plan and no statements concerning data
review.

Criterion 5: Assessment of data quality indicators. Level: D

Comments: DQI indicators were not discussed. Additional references were not provided.
Based on the information provided, the reviewer could not fully evaluate the assessment of
DQIs.

Criterion 6: Data history and overall apparent data quality. Level: D

Comments: Data summaries, maps and transport records are over five years old (1984). Data
History was not discussed. Overall intentions of sampling effort were not specifically addressed
within this report.

General comments: The result summaries did not reference support documentation. The scopes
of this investigation and intended quality procedure were not provided.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.1-105

-------
Title: Background Soil Data Assessment for the GE-Pittsfield/Housatonic River Site.

Author: Blasland and Bouck & Lee, Inc. (December 15, 2000)

Reference Number: RFW 00-0489
Reviewed by: Doug Godfrey

Criterion 1: Overall quality of and level of detail in report(s). Level: B

Comments: The report provides background soil data. All data has been previously reported.
Sampling locations specified.

Criterion 2: Formal documentation of procedures. Level: C

Comments: Documentation generally not available, but sufficient information is known or
available, but sufficient information is known or available via other sources to establish validity
of field and analytical procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: Detection limits reported in report, but information about methods and required
detection limits is not present.

Criterion 4: Data review, validation, and quality assurance. Level: D

Comments: No information on the validity or quality of the data is present in this report.

Criterion 5: Assessment of data quality indicators. Level: B

Comments: Data Quality indicators not established, but data appear to meet minimum standards
for DQIs.

Criterion 6: Data history and overall apparent data quality. Level: B

Comments: Data appear to be of acceptable quality but some data are derived from studies
conducted prior to 1995. Methods may not meet current standards, but are judged to have
produced data equivalent to current methodologies.

General comments: No new analytical data provided in this report.

Overall Score: C - Conditionally acceptable for limited uses. Data previously provided.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.1-106

-------
Title: USGS Database Query.

Authors: USGS - No author names appear on document
Reference Number: RFW-99-0397
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s).	Level: N/A	

Comments: Report consists of three large sheets of paper with Station Names, latitude and
longitude, date, time, and Aroclor and PCB results presented in ug/L and ug/kg. The samples
range in date from 1971 to 1994.

Criterion 2: Formal documentation of procedures.	Level: N/A

Comments: No documentation procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: N/A	

Comments: No information contained in document to indicate analytical methods or detection
limits.

Criterion 4: Data review, validation, and quality assurance.	Level: N/A

Comments: No indication of data review, validation, or quality assurance.

Criterion 5: Assessment of data quality indicators.	Level: N/A

Comments: No data quality indicators noted in this report.

Criterion 6: Data history and overall apparent data quality.	Level: N/A

Comments: Unable to determine data history and overall apparent data quality.

General comments:

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC

C.1-107

-------
Title: Fish Data - PCB.

Authors: Connecticut State Agency Files
Reference Number: RFW-99-0396
Reviewed by: Scott Campbell

Criterion 1: Overall quality of and level of detail in report(s). Level: D

Comments: Information contained in folder is copies of CT State files from 1976 to 1985. Data
is presented on hand written tables, laboratory sheets, typed summary tables, and in memos and
correspondences.

Criterion 2: Formal documentation of procedures. Level: D

Comments: Procedures are not clearly outlined. The data sets generally exist independent of a
written narrative report and are therefore not be dependable with regard to design and
methodology. In some cases it is unclear as to the procedures used for sample analysis and
laboratory procedures.

Criterion 3: Analytical methods used and detection limits achieved. Level: C

Comments: The data sets do not list an analytical method, however, there is documentation for
some data sets indicating the CT Dept. of Health and the FDA laboratories were involved in

sample analysis. The reviewer did not observe any reference to detection limits achieved for any
of the sets of data. However due to the high qualification of the laboratories involved, a Level C
will be applied to this criterion.

Criterion 4: Data review, validation, and quality assurance.

Level: D

Comments: The level of validation, data review, and quality assurance is unknown.

Criterion 5: Assessment of data quality indicators.

Level: D

Comments: No data quality indicators noted in this report.

Criterion 6: Data history and overall apparent data quality.

Level: D

Comments: Unable to determine data history and overall apparent data quality. The data sets in
general exist independent of a written narrative report and are therefore not be dependable with
regard to design and methodology.

General comments: Report consists mainly of data summary tables or laboratory sheet. Many of
the tables are handwritten and difficult to read. A number of the tables do not provide any
background information, reporting method (wet weight, dry weight), percent lipids.

Overall Score: Level D - Conditionally acceptable, use with caution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC1_3_PB.DOC q j 07/10/03

-------
APPENDIX C.2

APPROACH FOR TREATING NON-DETECTS IN
DATA ANALYSIS FOR THE HOUSATONIC RIVER ERA

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

APPENDIX C.2

APPROACH FOR TREATING NON-DETECTS IN DATA ANALYSIS FOR

THE HOUSATONIC RIVER ERA

C.2.1 Background

Analysis of chemical concentrations in water, soil, sediment or tissue samples often result in
some samples where the concentrations are reported as "not detected." The result, "not
detected," is reported when the value derived from the analysis is below sample quantification
limits (hereafter referred to as non-detects, or NDs). Non-detect values represent one of two
situations. Either the contaminant is not present in the sample, or the actual concentration is not
quantifiable and below the level that the laboratory can reliably measure using a particular
method on a particular sample.

Samples for which a value of ND is reported pose a challenge when calculating summary
statistics for data sets containing these values. The traditional approach for dealing with non-
detects in many Superfund assessments has been to assign values equal to half the sample
quantification limit (hereafter referred to as the detection limit, or DL) when estimating
population parameters such as the mean and variance. In some cases, this approach can severely
skew summary statistics, particularly when non-detects are frequent in the data set. For example,
if the DL does not differ significantly from the concentrations at which the contaminant is
detected, the sample mean calculated using estimates of half DL is likely to be biased low. Non-
detects can also present a unique set of problems when averaging duplicate sample results,
because, unlike NDs representing separate samples in environmental data sets, the values for
duplicates are supposed to represent the exact same sample. In this paper, methods for treating
non-detects in statistical analyses are reviewed, and methods that will be used for the Housatonic
River Ecological Risk Assessment are discussed. A case study is included to illustrate these
methods.

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-1

-------
1 C.2.2 Available Approaches

2 In statistical terms, data sets with non-detects are referred to as Type I left censored. Because the

3 DLs typically vary among samples, the data are multiply censored, as opposed to singly

4 censored. This is a common situation in environmental data analysis, and many techniques have

5 been developed to deal with it. The severity of the censoring, the types of analyses to be

6 performed, distributional assumptions, and the size of the data set are factors to be considered

7 when selecting the appropriate method. It should be noted that none of the methods discussed

8 here are intended to estimate concentrations for individual samples. Instead, they are used to

9 facilitate appropriate estimates for population parameters.

10 The three methods most frequently used for handling non-detects are listed below.

11 ¦ Method 1: Substitution methods assign a constant value to non-detects. Three

12 commonly used conventions are: (1) assume non-detects are equal to 0 (the lowest

13 possible value, and least conservative approach); (2) assume non-detects are equal to

14 the DL (the highest possible value, and most conservative approach); or (3) assume

15 non-detects are equal to one-half the DL (a compromise between the two). See

16 Figure 1.

17 ¦ Method 2: Maximum likelihood methods, sometimes referred to as distributional

18 methods, use an assumed distribution to estimate values for non-detects. Given a

19 distribution, estimates of summary statistics are computed that best match the

20 observed concentrations above the DL and the percentage of data below the limit.

21 See Figure 2.

22 ¦ Method 3: Probability plot methods, sometimes referred to as filling in with expected

23 values, use probability plots to impute values for non-detects. The following synopsis

24 describes how this technique may be performed (see Gleit 1985; Millard 1997 for

25 details):

26 1. Select an initial estimate for the mean and variance using a normal or lognormal

27 probability plot.

28 2. Using the assumed distribution and the current value of the mean and variance

29 estimate, calculate (impute) expected values for the first m values, where m is the

30 number of censored observations.

31 3. Calculate the mean and standard deviation as usual from the constructed data set.

32 4. Repeat Steps 2 and 3 until the mean and variance are stable.

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-2

-------
1

There are variations on this procedure for multiply censored data and for different
distributions. See Figure 3.

C.2.3 Discussion of Approaches

A review of the published literature on censored data indicates that all of the methods have some
advantages as well as limitations. However, one common theme is that when a large proportion
of the data is non-detects or the sample size is small, no method works particularly well. From a
statistical standpoint, the evaluation of a method for dealing with non-detects is based on
accuracy and precision, measured by bias and mean square error, respectively. Because the
proportions of data above the DLs will vary among chemicals, any bias in the concentration
estimates will be applied unevenly across a data set from an environmental investigation where
multiple chemicals are examined. Results for chemicals that are detected less frequently will be
more biased than results for chemicals detected more often.

Method 1 is most often used in the analysis of environmental data sets, mainly for simplicity.
Examples of agencies and programs that use this method include:

¦ U.S. Environmental Protection Agency (EPA) Superfund Program: The guidance for
risk assessment at Superfund sites recommends the use of one-half the DL for non-
detects (EPA 1989a); however, guidance for determining cleanup standards (EPA
1989b) often uses DL substitution.

¦ EPA Office of Water, Fish Contaminant Section: The guidance for assessing
chemical contaminant data for use in fish advisories recommends the use of one-half
the DL for non-detects (EPA 1995).

¦ U.S. Fish and Wildlife Service, National Contaminant Biomonitoring Program: This
program, which measured chemical concentrations in fish throughout the United
States during the 1970s and 1980s, recommended the use of one-half the DL for non-
detects when conducting statistical computations (Schmitt et al. 1990).

Replacement of non-detects by a constant leads to inconsistent and biased estimates of the mean
(El-Shaarawi and Esterby 1992). Helsel (1990) remarks that there is no theoretical basis for
using these estimates, and that their use is not defensible. However, substitution by one-half the
DL has been found to provide acceptable results for hypothesis testing when the censoring is not
excessive (Clarke and Brandon 1996). Also, if conclusions are unaffected by the choice of
method, there is no need to perform a more rigorous evaluation of the non-detects.

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-3

-------
1

Method 1 is the simplest method to implement, but other techniques perform better with respect
to accuracy and precision (see, for example, Gleit 1985; Haas and Scheff 1990; Helsel 1990;
Kushner 1976). With a proper software program, these other methods are relatively easy to
perform (for example, Environmental Stats for S-PLUS [Millard 1997] has an array of methods
for estimating parameters from censored data sets, including several variations of Methods 2 and
3 described above). If the distribution of the data is known, or if there are enough data to
reliably estimate the distribution, maximum likelihood estimation (Method 2) provides superior
estimates of percentiles. However, Method 2 still produces bias in the estimates of means and
variances for lognormally distributed data (Helsel 1990), and typical data sets will have
insufficient data to reliably estimate the distribution. Although some correction for this bias is
possible (Hass and Scheff 1990), the techniques have not been widely used and may be difficult
to implement for data with multiple detection limits.

Gleit (1985) used simulations to compare the performance of the three methods listed above for
small sample sizes (n=5) and singly censored normal or transformed lognormal data. Based on
observed performance, he recommended Method 3, probability plot methods. Without further
research and a good simulation study, it is assumed for the purposes of this discussion that this
method is also preferable for multiply censored data. Helsel (1990) also recommends the
probability plot method (referred to as a robust method because the reliance on a particular data
distribution only affects the non-detected data). Care must be taken when using this method,
however, to ensure that the imputed values are within the interval from 0 to the detection limit.
If not, the estimates will not be useful.

C.2.4 Approach for the Housatonic River ERA

There may be situations in using the data at this site where there is sufficient mechanistic
knowledge to determine that non-detect values should be considered zero. For example, if
Aroclor 1260 has been the only mixture released into the environment, and the congener patterns
have not been altered by degradation or dechlorination reactions, the concentrations of congeners
not present in Aroclor 1260 (e.g., congeners 76 to 82) reported as non-detects could then be
considered zero in the samples. While these types of situations may be limited, mechanistic
knowledge should be used where possible to set certain non-detect values to zero.

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-4

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

The primary concern for the ERA is whether the different treatments for non-detects will affect
risk assessment conclusions. Because the individual methods perform marginally in most "real
data" situations (where there is limited information about the distribution), a tiered approach for
treating non-detects will be used. First, the simplest approach (Method 1) will be utilized and the
outcome evaluated. If risk conclusions are not affected by applying the extreme values from
Method 1 (substitution by zero and substitution by the DL), then further evaluation of the non-
detects is unnecessary. Bounded estimates can be calculated for the true chemical concentration,
which will in turn establish bounded estimates for the calculated risk levels (i.e., a "Bounds
Analysis"). If both the upper and lower risk estimates allow the conclusion that there is no risk,
or that the risk is very high, then there is little reason to try to refine the concentration estimate
further. For example, if hazard quotients (HQs) were used for quantifying risk, and the risk
bounds comprised of the lower HQ (calculated with zero substitution) and the upper HQ
(calculated with DL substitution) were both less than 1 or both much greater than 1, there would
be no need to consider other methods for non-detects.

If the bounds analysis yields unacceptable uncertainty in the estimate of risk, or if the bounds are
so wide to be effectively useless (e.g., encompassing the complete realm of no risk to very high
risk), an alternative approach for refining the estimated concentrations is needed. For adequate
sample sizes with at least a moderate detection frequency, Method 3 is the generally preferred
approach. The resulting refined estimates of summary statistics will be based on assumptions
about the data that may or may not be valid. The quantity and quality of the available
information and the validity of assumptions must be considered when evaluating risk
calculations developed using Method 3. If the sample size is very small (n<5) or the detection
frequency is low (<25%), no method is likely to be wholly satisfactory as noted above. In this
case, summary statistics using all three options from Method 1 (nd=0, nd=l/2 DL, and nd=DL)
will be calculated to demonstrate the range of possible risk values.

When appropriate, some of the bias problems associated with the estimation of means and
variances from censored data sets can be avoided by using summary statistics based on
percentiles (e.g., the median instead of the mean, and the interquartile range instead of the
variance). Percentile-based summary statistics are more robust in the treatment of non-detects,

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-5

-------
1 because the statistics are less influenced by extreme values, such as those in the truncated left

2 tail.

3 Any decision on which method to use for non-detects should be made on the basis of the

4 characteristics of each statistically defined population. For example, the distribution of PCBs in

5 sediments might be expected to vary widely across reaches, so the concentrations in each reach

6 should be considered to be a single population. Thus, decisions should be made on a reach-by-

7 reach basis.

8 Figure 4 displays the decision tree for estimating summary statistics of Type I left-censored data

9 sets for use in the Housatonic River Ecological Risk Assessment. The bottom half of the

10 decision tree can also be used for hypothesis testing, correlation analyses and possibly other

11 statistical analyses. In summary, the decision process is:

12 1. When available, apply mechanistic knowledge to "zero-out" possible non-detects.

13 2. Calculate risk estimates using 0 and the full DL.

14 3. If results from Step 2 are consistent (i.e., both upper and lower bounds conclude

15 no risk or extremely high risk) then report the risk bounds.

16 4. If results from Step 2 are inconclusive (e.g., the calculated bounds straddle risk

17 levels), then use the bounded results in the uncertainty discussion. Modify the

18 risk estimate by refining the concentration summary statistics, using Step 5 or 6 as

19 appropriate.

20 5. If the detection frequency is moderate (>25%), with a sufficient sample size

21 (n>5), use Method 3 to summarize the data and report the estimated risk.

22 6. If the detection frequency is low (<25%) or the data set is very small (<5), then no

23 method will work well. In this case, only the percentage of data below the

24 detection limit and the upper percentile values (e.g., 90th or 95th percentiles) of the

25 detected data will be reported. Other summary statistics will not be reported. If a

26 point estimate of risk is absolutely needed, Method 1 with V2 DL option (per EPA

27 guidance) will be used, and the upper and lower bounds will be included in the

28 uncertainty discussion.

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-6

-------
1

C.2.5 Duplicate Samples with Non-Detects

For this discussion, it is assumed that differences between duplicate samples are due to
laboratory variability, and that each of the duplicates is equally representative of the true
analytical concentration of the sample. In statistical terms, it is assumed that both values are
unbiased estimates of the concentration in the sample. Another assumption is that the samples
have been screened for anomalies, i.e., grossly different concentrations in duplicate samples
should be evaluated on an individual basis.

In general, the options for treating non-detects in duplicate samples are to either ignore the
duplicate or to substitute the DL, one-half the DL, or zero for non-detects and average the
samples. Varying detection limits and varying patterns of detection complicate the substitution
methods. The following conservative approach will be used for the treatment of duplicate
samples:

1. If both samples have a concentration above the detection limit, use the sample
with the greatest concentration.

2. If one sample has a value above the detection limit and the other sample value is
below the DL, use the detected concentration.

3. If both values are non-detects, report the value as a non-detect at the lower DL.

When summing PCBs, duplicate concentrations for congeners will be separately evaluated.
Next, mechanistic knowledge should be applied as discussed previously, so that congeners
unlikely to be present and not detected should be assigned zero. Non-detects for congeners that
are present in other samples at the site should be included in the sum using V2 DL substitution.
Care will be taken to ensure that non-detected values for congeners are not driving toxicity
calculations.

C.2.6 Case Study

To demonstrate the approach, some of the data for total PCB concentrations in sediments are
evaluated. These data are meant for illustration purposes only and are not presented for use in

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-7

-------
1 actual risk calculations. The sample sizes and detection frequencies by Facility ID1 (HO through

2 H6) are displayed in Table 1. The bounds of the sample arithmetic mean are found by

3 substituting the DL and zero for the non-detected values. The low frequency of detection for HO

4 leads to a decision to report only the detection frequency, the range of detected data (0.02,12.7),

5 and the bounded estimates for the mean. If a risk estimate were needed for this Facility ID, V2DL

6 substitution would be used. For Facility IDs HI, H2, H3, and H4, differences in the calculated

7 bounds are unlikely to affect the risk estimates, so no further analysis is required. Differences in

8 bounds for H5 and H6 are slightly larger, and may affect risk estimates. For the purposes of

9 illustration, it is assumed that the two mean estimates for Facility ID H6 caused different

10 (acceptable vs. unacceptable) risk conclusions, and the example illustrates how the sample

11 arithmetic mean would be calculated using Method 3.

12 The data in area H6 do not appear to be skewed, so it is assumed that the data are approximately

13 normally distributed in this case. First, a quantile-quantile regression on the detected

14 observations was used to estimate potential values for the non-detects. The sample mean of the

15 observed and imputed values was calculated (Iteration 1 in Table 1) by predicting them from the

16 regression equation, transforming the log-scale imputed values back to the original scale, then

17 computing the method of moments estimates of the mean and standard deviation based on the

18 observed and imputed values. The regression was re-fit on this data set, and the process was

19 repeated for a total of four iterations. The iteration was stopped when the mean of the observed

20 and imputed values did not change beyond 2 significant figures. Figure 5 displays the bounding

21 estimates and the estimates imputed using Method 3.

22 C.2.7 References

23 Clarke, J.U. and D.L. Brandon. 1996. Applications Guide for Statistical Analyses in Dredged

24 Sediment Evaluations. U.S. Army Corps of Engineers Technical Report D-96-2, Waterways

25 Experiment Station, Vicksburg, MS.

26 El-Shaarawi, A.H. and S.R. Esterby. 1992. "Replacement of Censored Observations by a

27 Constant: An Evaluation." Water Research, 26(6): 835-844.

1 Facility ID is a database location code that is generally equivalent to a river reach designation.

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC C.2- 8 7/10/2003

-------
1

2	Table 1

3

4	Summary of tPCB Concentrations (ppm) for Sediments by Facility ID



Facility ID

HO

HI

H2

H3

H4

H5

H6

Total Sample Size

271

172

166

916

65

178

45

Detection Frequency

6%

63%

81%

92%

88%

44%

31%

Sample Median

Method 1 - DL Substitution

0.58

1.58

2.77

7.49

11.9

0.50

0.50

Method 1-0 Substitution

0

1.54

2.77

7.45

11.9

0

0

Sample Arithmetic Mean

Method 1 - DL Substitution

0.64

88.2

11.4

18.6

39.9

1.87

0.46

Method 1-0 Substitution

0.07

88.0

11.3

18.5

39.8

1.58

0.13

Method 3 (H6 only):
Iteration 1













0.40

Iteration 2













0.41

Iteration 3













0.42

Iteration 4













0.42

5

6	EPA (U.S. Environmental Protection Agency). 1989a. Risk Assessment Guidance for

7	Superfund-Volume I. Human Health Evaluation Manual (Part A). Interim Final.

8	EPA/540/1-89/002. Chapter 5, page 5-10. U.S. Environmental Protection Agency,

9	Washington, DC.

10	EPA (U.S. Environmental Protection Agency). 1989b. Methods for Evaluating the Attainment

11	of Cleanup Standards, Volume 1. Soils and Solid Media. EPA/230/02C89-042. Chapter 2,

12	page 2-1. U.S. Environmental Protection Agency, Washington, DC.

13	EPA (U.S. Environmental Protection Agency). 1995. Guidance for Assessing Chemical

14	Contaminant Data for Use in Fish Advisories. Volume 1, Fish Sampling and Analysis. EPA

15	823-R-95-007. U.S. Environmental Protection Agency, Office of Water, Washington, DC.

16	Gleit, A. 1985. "Estimation for Small Normal Data Sets with Detection Limits." ES&T19:

17	1201-1206.

18	Haas, C.N. and P. A. Scheff 1990. "Estimation of Averages in Truncated Samples." ES&T, 24:

19	912-919.

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC	C.2-9	7/10/2003

-------
1	Helsel, D.R. 1990. "Less than Obvious: Statistical Treatment of Data Below the Detection

2	Limit." ES&T.\ 24(12): 1766-1774.

3	Kushner, E.J. 1976. "On Determining the Statistical Parameters for Pollution Concentration

4	from a Truncated Data Set." Atmospheric Environ., 10: 975-979.

5	Millard, S.P. 1997. EnvironmentalStats for S-Plus Users Manual, Version 1.0. Probability,

6	Statistics, and Information, Seattle, WA.

7	Schmitt, C.J., J.L. Zajicek and P.H. Peterman. 1990. "National Contaminant Biomonitoring

8	Program: Residues of Organochlorine Chemicals in U.S. Freshwater Fish, 1976-1984."

9	Arch. Environ. Contam. Toxicol., 19: 748-781.

10

MK01 |O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-10

-------
& &



V

s

>•

U >-

£ g
.» 8

„ /#
i



#o
//

/

J® #

#



Concentration

Figure 1. Histograms for simple substitution methods for
handling data reported below laboratory detection limits

MK0110:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-11

-------
1

Fit 'best" lognormal distribution to the data...

O

52

3 3

cro

ED Q

it 3

Detection limit



















Filled "besi"
"" iogrwmal
dislribuEion























O qj

3 3

ai o

Concentration

...and then delermine summary sta!is!ics of Eha fitted distribution to
reprG&em the data

Detection limit

Median
¦ Maan

Concentration

AAWtStoti. HttiMH I I'lSUI

3

4

Figure 2, Distributional (maximum likelihood) method for computing

summary statistics

MK0110:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-12

-------
1

100

c
o

s

e

O
o

10-

t



























• Date abow* SrtwStion lira*
— = HffgrpssKjn llmf

+ "WB-ii»" * 8»icr» (WcclHjn litrw*























	m	









/•





m





















A















J-*"1"





























-2	-10	12

Normal scores

* 'FiW-irts' are relransfcrnied to ordinal units and combined with dala at-ove the detection limit
Id compu» estimates di summary slalisfics

Figure 3a. Probability plot; Regression of log of concentration vs.
normal score

£ £¦

a> £

.s>8

Detection limit

"Fill-ins"
from

probability pk>1

/

Observed
dala

Goficentralion

Figure 3b. Robust (probability plot} method of estimating summary
statistics

2

3

¦WjUmJ tTrmv	jlWH

MK0110:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-13

-------
Figure 4. Decision tree for treating noodetects

2

MK01|O:\20123001.096\ERA_PB\ERA_APC2_PB.DOC	C.2-14	7/10/2003

-------
1

2

3

4

E
a

Q.

»
c

se

I

I

rr.

E

M?1
<5

Tf

¦71

OJ

Ci

n *

QQOOOOOOOOOOOOOOOOOQOOOOo|2^^

		

xx 		

+ + +	. a * *

o

BOindlnf HllmalH

A

Fiiil ilTif.jM

-1-

SiKDiid impjhi

*

Thrrl imfmitn

O

Fourth impute

O

d

Individual Sample

6	Figure 5. Scatter plot showing ordered values estimated for non-detects.

7	Bounding estimates are estimates for non-detects based on substitution by the

8	DL and zero. Imputed values are estimated using Method 3 with four iterations.

MK0110:\20123001.096\ERA_PB\ERA_APC2_PB.DOC

C.2-15

-------
APPENDIX C.3

APPROACH TO SPATIAL WEIGHTING OF tPCB
CONCENTRATIONS IN FLOODPLAIN SOIL

O:\20123001.096\ERA PB\ERA APC3 PB.doc

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

APPENDIX C.3

APPROACH TO SPATIAL WEIGHTING OF
tPCB CONCENTRATIONS IN FLOODPLAIN SOIL

As described where appropriate in Appendices D through K, exposure concentrations in the
floodplain for some receptors were calculated using spatially weighted data rather than raw data.
The spatially weighted surface of total PCB (tPCB) concentrations in the Reaches 5 and 6
floodplain was generated from the measured concentrations in floodplain soil samples using the
inverse distance weighting (IDW) procedure contained in the ArcView Spatial Analyst
extension. The basic IDW approach was modified to include information on the habitat types
delineated in the floodplain as part of the ecological characterization (see Appendix A). PCBs
were carried onto the floodplain during storms and other high-water events over the last 70 years.
The frequency and extent of such inundations at a particular location in the floodplain is
governed by the same topographic and hydrologic factors that also control the distribution of
biological habitats. Accordingly, it was appropriate to only consider data from similar habitat
types in conducting the spatial weighting exercise. The use of habitat-restricted spatial
weighting also reduced the effect of nonrandom sampling and the clustering of samples in areas
of known high levels of contamination.

Spatial weighting was limited to the floodplain adjacent to Reaches 5 and 6 because of the
limited sampling and habitat mapping data for downstream reaches. Data on PCB concentrations
in floodplain soils from the 0- to 6-inch depth interval derived from EPA and GE recent
sampling studies were extracted from the project datamart and exported to a dBase file (dbf),
which was imported to ArcView Version 3.2 as an event theme. Values reported as non-detects
were replaced with one-half the reported sample quantitation limit (detection limit). This
treatment of non-detects, which was used for reasons of cost-effectiveness, differs from the
general approach to non-detects for the ERA described in Appendix C.2, but in test runs on small
areas of the floodplain was found to have no practical effect on the results of the spatial
weighting analysis. The habitat boundary theme previously developed for the ecological
characterization was also imported and modified to group similar habitats into six "super
habitats." This grouping step was necessary to avoid large numbers of habitat polygons without

MK01 |O:\20123001.096\ERA_PB\ERA_APC3_PB.DOC	^ ^ ->	7/10/2003

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

sampling data, and also had the advantage of reducing computational time. The final super
habitat groupings were as follows:

¦	Shallow emergent marsh, deep emergent marsh, and wet meadow.

¦	Transitional floodplain forest, black ash-red maple-tamarack calcareous seepage
swamp, and red maple swamp.

¦	High terrace floodplain forest, northern hardwoods-hemlock-white pine forest, red
oak-sugar maple transition forest, rich mesic forest, successional northern hardwoods,
cultural grassland, and agricultural field.

¦	Shrub swamp.

¦	Low-gradient stream, medium-gradient stream, high-gradient stream, riverine
point bar.

¦	Moderately alkaline lake/pond.

This grouping reduced the number of habitat polygons in the theme from the original 870 to 744
and greatly decreased the number of polygons without data. A series of test runs was used to
establish that a 3-square-meter (3 m2) grid produced spatially weighted surfaces that were
essentially identical to those generated with a much more computationally intensive 1-m2 grid
and was sufficient to adequately resolve concentration boundaries for the purposes of
determining exposures.

The 3-m2 grid was populated from the sample data using the standard IDW algorithm in
Arc View Spatial Analyst:

G(x,y) = ^Wif(xi,yi)

i

Where:

G(x,y)	=	IDW estimation at (x,y).

Wi	=	1/dip.

d\	=	Distance from (x,y) to (xjj'j),

p	=	Power, a real number.

/(xjj'j)	=	Observed value at (xj,j'j).

This interpolation assumes that each input point has a local influence that diminishes with
distance. Thus, the interpolated points (the new surface) will be more influenced by nearby

MK01 |O:\20123001.096\ERA_PB\ERA_APC3_PB.DOC

C.3-2

-------
1 points than more distant points. The weights are inversely related to distance and are scaled such

2 that the sum of all the weights will add to 1. The number of points or "neighbors" (n) used in the

3 interpolation and the power term (p) are user-specified.

4 The FIELDS cross-validation procedure was used to optimize the values of n and p for this

5 application. Cross validation is an iterative technique in which a datum at a particular location is

6 temporarily discarded from the sample data set. The value at that location is then estimated

7 using the remaining samples (Isaaks and Srivastava 1989). The difference between these two

8 values is the cross-validated residual. Cross validation is performed for each unique interpolation

9 permutation (e.g., neighbors = 1, power = 1; neighbors = 2, power = 1) for the IDW interpolator,

10 and the combination of n and p that produces the lowest sum of residuals is used for calculation

11 of the final surface.

12 These recommended variables from the cross-validation process were passed to the IDW

13 processor and the interpolated grid surface was created for that habitat polygon. The grid surface

14 was stored temporarily and the next polygon in the list of boundary polygons (habitat) was

15 processed. Once all surface grids were created, they were merged to form one continuous grid

16 covering the entire floodplain within Reaches 5 and 6.

17 Because the IDW interpolation was not allowed to cross the habitat boundaries, if there was only

18 one data point in a particular polygon, each 3-m2 cell in the polygon was assigned the value of

19 that single point. If there were no samples in a particular polygon, the entire area of that polygon

20 was assigned the value "no data." In such cases, the polygon was examined and manually linked

21 to the nearest most similar habitat with sampling data, then recalculated. Because of the

22 grouping of individual habitats into the six "super habitats" described, this final adjustment was

23 necessary for only a very small amount of the total area in the floodplain.

24 REFERENCES

25 Isaaks, E.H. and R.M. Srivastava. 1989. An Introduction to Applied Geostatistics. Oxford

26 University Press, New York, NY.

MK01 |O:\20123001.096\ERA_PB\ERA_APC3_PB.DOC

C.3-3

-------
APPENDIX C.4

APPROACH FOR USING MONTE CARLO AND PROBABILITY BOUNDS

ANALYSES

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

-------
1

APPENDIX C.4

APPROACH FOR USING MONTE CARLO AND PROBABILITY BOUNDS

ANALYSES

C.4.1 Introduction

The purpose of probabilistic risk analysis is to comprehensively characterize not just the best
estimate or a conservatively biased estimate of risk, but the entire statistical distribution of the
values it might take on. This includes the "tail risks" associated with relatively rare but serious
extreme events such as species extinction, or a large number of animals receiving very large
doses of a toxicant. A probabilistic risk analysis explains the effect of the intrinsic variability
associated with natural populations and expresses the reliability of quantitative estimates by
accounting for uncertainty in the inputs it uses. Such an approach is more comprehensive and
informative than an analogous deterministic approach because it can make use of virtually all the
relevant empirical data. There is a growing consensus that probabilistic risk analyses serve an
invaluable role in ecological risk assessments. After considerable discussion, EPA guidance on
how to conduct such analyses in Superfund and other programs has been produced (EPA 1997,
1999). The remainder of this appendix will present and discuss the two approaches to
probabilistic risk analysis that will be used in the GE/Housatonic River ERA.

C.4.2 Monte Carlo Simulation

Morgan and Henri on (1990) review the use of probabilistic techniques in risk analysis. In this
context, statistical distributions are used to represent the variability and uncertainty in general
that accompany some or all of the input variables. These distributions may be combined in an
operation that generalizes statistical convolution. In general, the calculations needed for a risk
assessment constitute a complex problem for which direct analytical solutions are intractable.
Consequently, approximate methods are almost always used in practice. These include the delta
method (Seber 1973; Kirchner 1992), Laplace and Mellin transformations (Springer 1979), and
various strategies based on discretization (Ingram et al. 1968; Kaplan 1981; Ahmed et al. 1982;
Williamson and Downs 1990). By far the most commonly used method is Monte Carlo

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC ^ a -i 7/10/2003

-------
1

simulation (Brately et al. 1983; Morgan 1984; Ripley 1987) in which a computer randomly
generates specific values for the input variables according to their respective distributions and
evaluates the mathematical expression. This process is repeated many times and the distribution
of the results obtained over the many replications reveals the range of possible outcomes as well
as the probability associated with each magnitude.

When there are many input variables in a simulation, a simple sampling strategy may be
insufficient to overcome the "curse of dimensionality" in which the number of replicates is too
small to adequately explore the high-dimensional space of possible variation. Structured
sampling strategies such as Latin hypercube sampling (Iman and Conover 1980, 1982; Iman et
al. 1981a, 1981b; Iman and Shortencarier 1984) may be employed to improve estimates when the
number of replicates must be limited. Several commercially available software packages offer
their users a choice of simple or Latin hypercube sampling.

Analysts must select the statistical distributions to be used to model the variability and
uncertainty in each of the input variables, and this is often not a trivial task. For some variables,
there may be enough empirical information to fit parametric distributions or even specify
empirical histograms. More commonly, there may be little empirical evidence to support the
distributions selected as inputs. In this circumstance, best judgment or a convention among
practitioners often suggests which distributions should be used. As a result, the analysis usually
requires assumptions that cannot be justified by appeal to evidence. The consequences of this
may be substantial because the results of probabilistic risk analyses are known to be sensitive to
the choice of distributions used as inputs (Bukowski et al. 1995), an effect which can be even
stronger for the tail probabilities. The difficulties of developing and justifying input distributions
are well known in the field of risk analysis and have been the subject of considerable attention
(e.g., Haimes et al. 1994; Finley et al. 1994). Although there is a great volume of literature on
the subject of estimating probability distributions from empirical data (see Morgan and Henri on
1990, page 78ff; Cullen and Frey 1999, page 122ff), standard approaches are of limited practical
effectiveness when few data exist. In situations where data are severely limited, several
strategies have been suggested to meet the challenge.

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-2

-------
1

C.4.2.1 Extrapolations and Surrogate Data

In many circumstances, directly relevant empirical information is sparse or completely absent,
but related data are available from which it is possible to make extrapolations. For instance,
evidence from similar cases such as a congeneric species or a comparable site can be used to
inform the selection of a distribution. These extrapolations, whether they are across species,
from laboratory results to predictions about the field, or bridging some other form of surrogacy,
usually require theoretical justifications. In deterministic risk assessments, rules of thumb have
become conventional in some regulatory contexts about how to account for the uncertainty
inherent in such extrapolations. In the probabilistic assessments, however, the principles that
govern such extrapolations are still poorly defined, and the approach although often absolutely
indispensable, is in some respects more of an art than a science.

C.4.2.2 Elicitation from Experts

There are various approaches to eliciting information about input variables from experts or
otherwise knowledgeable persons. These approaches range from simply asking them in informal
and uncontrolled settings to rather elaborate formal schemes (Morgan and Henri on 1990;
Warren-Hicks and Moore 1998, page 3 Iff). Formal elicitation schemes can be very costly to
implement. Sometimes it is reasonable to let experts define the shapes of the input distributions
subjectively, but this is not always a workable strategy and sometimes leads to even more
controversy about the inputs.

C.4.2.3 Maximum Entropy Criterion

An objective approach to the problem of selecting input distributions when empirical information
is sparse is based on the maximum entropy criterion (Jaynes 1957). This criterion has its
intellectual roots in Laplace's Principle of Insufficient Reason that uses a uniform distribution to
represent a variable about which only the potential range is known. The maximum entropy
criterion generalizes this strategy by uniquely defining the distribution of a variable using
whatever information is available. The criterion specifies the distribution having the largest
Shannon entropy (Shannon and Weaver 1949) consistent with the available knowledge.
Therefore, what is known empirically about a distribution constitutes a set of constraints on the

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC n A 1. 7/10/2003

-------
1

possible shapes of the distribution. This approach makes no assumptions about distribution
shape directly. Proponents argue that the criterion allows one to select the input distribution in a
way that makes optimal use of the limited information about the variates. Although the
derivations of which distributions should be used in particular circumstances can be rather
complex, the conclusions are easy to summarize as shown below (Lee and Wright 1994; cf.
Tilwari and Hobbie 1976):

When you know Use this shape

{minimum, maximum} uniform

{mean, standard deviation} normal

{minimum = 0, mean} exponential
{minimum, maximum, mean, standard deviation} beta

{minimum = 0, some quantile} exponential

{minimum > 0, some quantile} gamma

{minimum, maximum, mean} beta

In principle, the maximum entropy approach could be extended to cover the problem of not
knowing correlations. Any continuous multivariate distribution can be decomposed into the
marginal distributions of each dimension and a copula function (Schweizer 1991; Schweizer and
Sklar 1983; Schweizer and Wolff 1981; Nelsen 1991, 1995) that specifies how the marginals are
brought together to form the joint distribution. Not knowing the correlations and dependencies
among the variables amounts to not knowing this copula function. The idea is that, under
ignorance about correlations, the Principle of Insufficient Reason can be used to justify using the
copula with the largest entropy to tie together the marginal distributions. The maximally
entropic copula is the independence copula.

The use of the maximum entropy criterion in selecting input distributions for Monte Carlo
analysis may be better than other strategies in common use and is considered by many to be the
state-of-the-art. However, the maximum entropy criterion is controversial (Jaynes 1979; Levine
and Tribus 1979; Grandy and Schick 1991; Moore 1996). Among several different kinds of
criticisms, perhaps the most serious problem with the approach is that the model of uncertainty it
uses is inconsistent through changes of scale. For instance, suppose that all one knows about a
particular variable A is its range. The maximum entropy criterion would suggest using a uniform
distribution over this range to represent this state of knowledge. Now consider the related

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-4

-------
1

variable A2. If all we knew about A was its range, then surely all we know about A2 is its range,
which is just the interval between (left bound of A)2 and (right bound of A)2. This means that we
should pick another uniform distribution to model A2 too. But, given the uniform distribution for
A, we can compute the distribution it implies for A2, and it turns out not to be the same as a
uniform distribution over the squared range. Similar problems occur when log and other
transformations are used or when a variable is arithmetically combined with other variables. The
inconsistencies mean that analysts must arbitrarily pick a scale on which to express their
uncertainty and resist comparing it across different scales.

C.4.2.4 Correlations and Dependencies

When the correlations among input variables can be estimated empirically, standard
computational techniques can be used to simulate the dependence relationships in a Monte Carlo
analysis. Scheuer and Stoller (1962) describe a numerical method to generate normal deviates
with a specified matrix of Pearson product-moment correlation coefficients in the general
multivariate case (i.e., more than two variables). Iman and Conover (1982) give an ad hoc but
perhaps robust technique for simulating deviates from distributions of more general shapes and
(Spearman) rank correlation structure. Note, however, that the traditional Pearson correlation
coefficient and rank correlation can differ substantially. Nelsen (1986, 1987) gives methods to
simulate bivariate deviates from distributions having arbitrary marginal shapes and arbitrary rank
correlation (measured with either Spearman's rho or Kendall's tau). Nelsen's method is based
on the mathematical notion of copula, and therefore, should have a straightforward multivariate
generalization. When the marginals are normal, the correlation can be specified with the Pearson
coefficient, but because copulas can be freely wedded to arbitrary marginals, the approach
immediately applies to all other distribution shapes too. The resulting correlations are no longer
the specified Pearson's coefficients, but the transformation leaves rank correlations unchanged.
Whatever simulation technique is employed, an analyst must ensure that the matrix of planned
correlation coefficients is feasible (positive definite). The analyst should also check that the
realized dependency patterns are consistent with the input specifications. Although these
methods for simulating correlations are often very useful, they are not flexible enough to account
for all types of dependence. In general, other strategies may be needed, either because the

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-5

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

relationship is intrinsically nonlinear or because the available empirical information is
insufficient to justify a particular structure of correlation coefficients.

Despite the availability of these methods, many analysts assume independence among the
variables in a risk assessment because they have no specific empirical information available to
specify the correlations. In many cases, however, this assumption is inappropriate, especially in
ecological risk analyses. For instance, it is not reasonable to assume that skin surface area and
body mass are independent. Nor is it plausible that water intake and breathing rate are
independent. Assuming independence counterfactually may lead to under-estimating the tail
probabilities (Ferson 1994; cf. Smith et al. 1992). In contrast to interval analysis, which takes no
account of dependencies at all, an approach that uses independence assumptions as a
convenience makes the opposite mistake.

Morgan and Henrion (1990) recommend that the risk expression be constructed to resolve all
dependencies, which means selecting input variables that are mutually independent. This is a
difficult problem and unknown dependencies among input variables that violate independence
assumptions probably exist in many assessments. The peer review of the ecological risk
assessment guidance from the U.S. Environmental Protection Agency (EPA) (1996) identified
evaluation of the effect of independence assumptions on the propagation of errors as a primary
research priority.

C.4.3 Interval Probability or Probability Bounds

Many researchers have argued that lack of empirical information implies that probability
distributions of input variables cannot be precisely specified (Wolfenson and Fine 1982; Walley
and Fine 1982; Pearl 1988; Walley 1991; Tessem 1992; Ferson and Ginzburg 1996; Kyburg
1999). Probability bounds analysis represents an uncertain input distribution with an entire class
of probability distributions that conform to the available empirical information about the
variable. Sometimes this class is very small, and might be a single distribution when information
is abundant. Other times, the class can be large, reflecting a poor state of knowledge about the
variable. Classical results in probability theory such as the Chebyshev and Markov inequalities
can be used to derive optimal bounds on the cumulative distribution functions in these classes.
The maximum entropy approach selects a single distribution out of this class to use as an

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC	C* A &	7/10/2003

-------
1

exemplar for the entire class, but there is usually no justification for such a selection in the
context of risk analysis. By using probability bounds analysis, an analyst can propagate the
entire class through the risk expression.

With efficient numerical algorithms (e.g., Williamson and Downs 1990; Berleant 1993, 1996), it
is possible to compute bounds on the cumulative distribution function from the arithmetic
combinations of probability distributions when the input distributions are represented by
probability bounds. In risk assessment, the use of probability bounds analysis reveals how much
larger (or smaller) the probability of a result of a given magnitude might be. Probability bounds
analysis offers a means for determining the reliability of risk estimates that is more
comprehensive than what-if or interval approaches and yet computationally cheaper than Monte
Carlo methods.

When each variable in the risk expression appears only once, the results from a probability
bounds analysis are guaranteed to be optimally narrow. In other words, the bounds could not be
any tighter unless there was more information (or more assumptions) about the inputs. Like
interval analysis and multistage Monte Carlo simulations, repeated occurrences in the risk
expression can complicate the analysis, and may necessitate the use of brute force numerical
methods to obtain optimal tight results.

The problem of not knowing the dependencies among input variables can be addressed using the
probability bounds approach with dependency bounds analysis (Williamson and Downs 1990;
Glaz and Johnson 1984; Frank et al. 1987; Ferson and Long 1995; Ferson and Burgman 1995;
Ferson 1996; Berleant and Goodman-Strauss 1998). For each arithmetic operation, an analyst
may assume independence between the operands, or the analyst may choose to make no
assumption at all about this dependence. The method considers all of the possible copula
functions (Nelsen 1991, 1995) that could conjoin the marginal distributions. The algorithms
compute bounds on the resulting operation even though only the marginal distributions can be
specified. Specific information about the dependencies among variables can also be used to
tighten the bounds on the resulting cumulative distribution function, although the algorithms to
make the calculations are more complicated than those for the independent and no-assumption
cases.

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-7

-------
1 Probability bounds analysis is much more efficient than sensitivity analysis when there are many

2 variables involved and, for many estimation problems, it is quite easy to implement. However, it

3 provides only upper and lower bounds without any indication about the relative likelihoods

4 within the range. It is more comprehensive than what-if or sensitivity studies and vastly cheaper

5 computationally and often easier to interpret than analogous second-order Monte Carlo methods.

6 C.4.4 References

7 Ahmed, S., R.E. Clark, and D.R. Metcalf, 1982. A method for propagating uncertainty in

8 probabilistic risk assessment. Nuclear Technology 59:238-245.

9 Berleant, D. 1993. Automatically verified arithmetic with both intervals and probability density

10 functions. Interval Computations 2:48-70.

11 Berleant, D. 1996. Automatically Verified Arithmetic on Probability Distributions and Intervals.

12 In Applications of Interval Computations, R.B. Kearfott and V. Kreinovich (eds.), Kluwer,

13 Dordrecht, The Netherlands.

14 Berleant, D. and C. Goodman-Strauss. 1998. Bounding the results of arithmetic operations on

15 random variables of unknown dependency using intervals. Reliable Computing 4:147-165.

16 Bratley, P., B.L. Fox, and L.E. Shrage. 1983. A Guide to Simulation. Springer Verlag, New

17 York, NY.

18 Bukowski, J., L. Korn, and D. Wartenberg. 1995. Correlated inputs in quantitative risk

19 assessment: The effects of distributional shape. Risk Analysis 15:215-219.

20 Cullen, A.C. and C.H. Frey. 1999. Probabilistic Techniques in Exposure Assessment: A

21 Handbook for Dealing with Variability and Uncertainty in Models and Inputs. Plenum Press,

22 New York, NY.

23 EPA (U.S. Environmental Protection Agency). 1996. Proposed guidelines for ecological risk

24 assessment. Federal Register 61(175):47552-47631.

25 EPA (U.S. Environmental Protection Agency). 1997. Guiding Principles for Monte Carlo

26 Analysis. EPA/630/R-97/001. Office of Research and Development, U.S. Environmental

27 Protection Agency, Washington, DC.

28 EPA (U.S. Environmental Protection Agency). 1999. Risk Assessment Guidance for Superfund:

29 Volume 3 - Part A, Process for Conducting Probabilistic Risk Assessment. Draft. Office of Solid

30 Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, DC.

31 Ferson, S. 1994. Naive Monte Carlo methods yield dangerous underestimates of tail

32 probabilities. Proceedings of the High Consequence Operations Safety Symposium, Sandia

33 National Laboratories, SAND94-2364, J.A. Cooper (ed.). pp. 507-514.

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-8

-------
1	Ferson, S. 1996. What Monte Carlo methods cannot do. Human and Ecological Risk Assessment

2	2:990-1007.

3	Ferson, S. and M. Burgman. 1995. Correlations, dependency bounds and extinction risks.

4	Biological Conservation 73:101-105.

5	Ferson, S. and L.R. Ginzburg. 1996. Different methods are needed to propagate ignorance and

6	variability. Reliability Engineering and Systems Safety 54:133-144.

7	Ferson, S. and T.F. Long. 1995. Conservative uncertainty propagation in environmental risk

8	assessments. Environmental Toxicology and Risk Assessment, Third Volume, ASTM STP 1218,

9	J.S. Hughes, G.R. Biddinger, and E. Mones (eds.), American Society for Testing and Materials,

10	Philadelphia, PA. pp. 97-110.

11	Finley, B., D. Proctor, P. Scott, N. Harrington, D. Pasutenbach, and P. Price. 1994.

12	Recommended distributions for exposure factors frequently used in health risk assessment. Risk

13	Analysis 14:533-553.

14	Frank, M.J., R.B. Nelson, and B. Schweizer. 1987. Best-possible bounds for the distribution of a

15	sum—a problem of Kolmogorov. Probability Theory and Related Fields 74:199-211.

16	Glaz, J. and B.M.K. Johnson. 1984. Probability inequalities for multivariate distributions with

17	dependence structures. Journal of the American Statistical Association 79:436-440.

18	Grandy, W.T., Jr. and L.H. Schick. 1991. Maximum Entropy and Bayesian Methods. Kluwer

19	Academic Publishers, Dordrecht, The Netherlands.

20	Haimes, Y.Y., T. Barry, and J.H. Lambert. 1994. When and how can you specify a probability

21	distribution when you don't know much? Risk Analysis 14:661-706.

22	Iman, R.L. and W.J. Conover. 1980. Small sample sensitivity analysis techniques for computer

23	models, with an application to risk assessment. Communications in Statistics A9:1749-1842.

24	Iman, R.L. and W.J. Conover. 1982. A distribution-free approach to inducing rank correlation

25	among input variables. Communications in Statistics B11:311-334.

26	Iman, R.L., J.C. Helton, and J.E. Campbell. 1981a. An approach to sensitivity analysis of

27	computer models, Part 1. Introduction, input variable selection and preliminary variable

28	assessment. Journal of Quality Technology 13:174-183.

29	Iman, R.L., J.C. Helton, and J.E. Campbell. 1981b. An approach to sensitivity analysis of

30	computer models, Part 2. Ranking of input variables, response surface validation, distribution

31	effect and technique synopsis. Journal of Quality Technology 13:232-240.

32	Iman, R.L and M.J. Shortencarier. 1984. A Fortran 77 Program and User's Guide for the

33	Generation of Latin Hyper cube and Random Samples for Use with Computer Models.

34	NUREG/CR-3624, SAND83-2365, Sandia National Laboratories, Albuquerque, NM.

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-9

-------
1	Ingram, G.E., E.L. Welker, and C.R. Herrmann. 1968. Designing for reliability based on

2	probabilistic modeling using remote access computer systems. Proceedings of the 7th Reliability

3	and Maintainability Conference, American Society of Mechanical Engineers, pp. 492-500.

4	Jaynes, E.T. 1957. Information theory and statistical mechanics. Physical Review 106:620-630.

5	Jaynes, E.T. 1979. Where do we stand on maximum entropy? The Maximum Entropy Formalism.

6	R.D. Levine and M. Tribus (eds.), MIT Press, Cambridge, MA. pp. 15-118.

7	Kaplan, S. 1981. On the method of discrete probability distributions—application to seismic risk

8	assessment. Risk Analysis 1:189-198.

9	Kirchner, T.B. 1992. QS-CALC: An Interpreter for Uncertainty Propagation. Quaternary

10	Software, Fort Collins, CO.

11	Kyburg, H.E. Jr. 1999. Interval-valued probabilities, Imprecise Probability Project.

12	G. de Cooman (ed.), http://ippserv.rug.ac.be/documentation/interval_prob/interval_prob.html.

13	Lee, R.C. and W.E. Wright. 1994. Development of human exposure-factor distributions using

14	maximum-entropy inference. Journal of Exposure Analysis and Environmental Epidemiology

15	4:329-341.

16	Levine, R.D. and M. Tribus. 1979. The Maximum Entropy Formalism. MIT Press, Cambridge,

17	MA.

18	Moore, D.R.J. 1996. Using Monte Carlo analysis to quantify uncertainty in ecological risk

19	assessment: Are we gilding the lily or bronzing the dandelion? Human and Ecological Risk

20	Assessment 2:628-633.

21	Morgan, J.T. 1984. Elements of Simulation. Chapman & Hall, London, United Kingdom.

22	Morgan, M.G. and M. Henrion. 1990. Uncertainty: A Guide to Dealing with Uncertainty in

23	Quantitative Risk and Policy Analysis. Cambridge University Press, Cambridge, United

24	Kingdom.

25	Nelsen, R.B. 1986. Properties of a one-parameter family of bivariate distributions with specified

26	marginals. Communications in Statistics (Theory andMethods) A15:3277-3285.

27	Nelsen, R.B. 1987. Discrete bivariate distributions with given marginals and correlation.

28	Communications in Statistics (Simulation and Computation) B16:199-208.

29	Nelsen, R.B. 1991. Copulas and association. Advances in Probability Distributions with Given

30	Marginals. G. Dall'Aglio, S. Kotz, and G. Salinetti (eds.), Kluwer Academic Publishers,

31	Dordrecht, The Netherlands, pp. 51-74.

32	Nelsen, R.B. 1995. Copulas, characterization, correlation, and counterexamples. Mathematics

33	Magazine 68:193-198.

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-10

-------
1	Pearl, J. 1988. On probability intervals. International Journal of Approximate Reasoning 2:211-

2	216.

3	Ripley, B.D. 1987. Stochastic Simulation. John Wiley, New York, NY.

4	Scheuer, E.M. and D.S. Stoller. 1962. On the generation of normal random vectors.

5	Technometrics 4:278-281.

6	Schweizer, B. 1991. Thirty years of copulas. Advances in Probability Distributions with Given

7	Marginals. G. Dall'Aglio, S. Kotz, and G. Salinetti (eds.), Kluwer Academic Publishers,

8	Dordrecht, The Netherlands, pp. 13-50.

9	Schweizer, B. and A. Sklar. 1983. Probabilistic Metric Spaces. Elsevier Science Publishing,

10	New York, NY.

11	Schweizer, B. and E.F. Wolff. 1981. On nonparametric measures of dependence for random

12	variables. The Annals of Statistics 9:879-885.

13	Seber, G.A.F. 1973. The Estimation of Animal Abundance. Griffin, London, United Kingdom.

14	Shannon, C.E. and W. Weaver. 1949. The Mathematical Theory of Communication. University

15	of Illinois Press, Urbana, IL.

16	Smith, A.E., P.B. Ryan, and J.S. Evans. 1992. The effect of neglecting correlations when

17	propagating uncertainty and estimating the population distribution of risk. Risk Analysis 12:467-

18	474.

19	Springer, M.D. 1979. The Algebra of Random Variables. John Wiley, New York, NY.

20	Tessem, B. 1992. Interval probability propagation. International Journal of Approximate

21	Reasoning 7:95-120.

22	Tilwari, J.L. and J.E. Hobbie. 1976. Random differential equations as models of ecosystems. II.

23	Initial condition and parameters specifications in terms of maximum entropy distributions.

24	Mathematical Biosciences 3:37-53.

25	Walley, P. and T.L. Fine. 1982. Towards a frequentist theory of upper and lower probability.

26	Annals of Statistics 107:41-761.

27	Walley, P. 1991. Statistical Reasoning with Imprecise Probabilities. Chapman and Hall, New

28	York, NY.

29	Warren-Hicks, W.J. and D.R.J. Moore, (eds.). 1998. Uncertainty Analysis in Ecological Risk

30	Assessment. SETAC Press, Pensacola, FL. 277 p.

31	Williamson, R.C. and T. Downs. 1990. Probabilistic arithmetic I: Numerical methods for

32	calculating convolutions and dependency bounds. International Journal of Approximate
3 3	Reasoning 4:89-158.

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-11

-------
1	Wolfenson, M. and T.L. Fine. 1982. Bayes-like decision making with upper and lower

2	probabilities. Journal of the American Statistical Association 77:80-88.

MK01 |O:\20123001.096\ERA_PB\ERA_APC4_PB.DOC

C.4-12

-------
APPENDIX C.5

APPROACH FOR CALCULATING EXPOSURE POINT
CONCENTRATIONS

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

-------
1

APPENDIX C.5

APPROACH FOR CALCULATING EXPOSURE POINT
CONCENTRATIONS

This appendix describes the methods applied to obtain the exposure point concentrations (EPCs)
that were used throughout the human health and ecological risk assessments for the
GE/Housatonic River, Rest of River site. It is organized into four main sections. The
introduction briefly explains the essential concepts considered in developing this approach, and
provides the context with regard to relevant EPA guidance.

The second section includes the flowchart used to determine which of the various methods for
computing upper confidence limits (UCLs) on distribution means would be used in the
assessment, and describes each of these methods, their statistical properties, and their limitations.
For each method, instructions for computing the UCLs and an example calculation are provided.

The third section describes the calculation of UCLs under the spatial weighting strategy used in
the human health risk assessment for floodplain soil-related exposures. Minor modifications to
the UCL formulas were made that permit their application to data sets that have been spatially
interpolated from the sample data. The last section discusses the effect of outliers and non-
detects on the calculation of UCLs.

C.5.1 Introduction

An important element of the risk assessment process for any site is the estimation of the
concentration of each contaminant of potential concern (COPC) included in the assessment. This
concentration, commonly referred to as the exposure point concentration (EPC), represents a
conservative estimate of the contaminant concentration in environmental media. The EPC is
used to assess the average concentration of a contaminant in an exposure unit that a receptor may
come in contact with over time. An exposure unit is the area throughout which a receptor moves
and encounters a medium of concern for the duration of the exposure. To estimate potential
exposure within an exposure unit for human health risk assessment, EPA guidance recommends
using the reasonable maximum exposure (RME), a measure that combines an estimate of the
average exposure concentration with reasonably conservative default values for intake and
duration (EPA 1989). However, there are many ways to estimate an average concentration.
EPA guidance suggests that, "Because of the uncertainty associated with estimating the true
average concentration at a site, the 95 percent upper confidence limit (UCL) of the arithmetic
mean should be usedfor this variable" (EPA 1992; emphasis in original text).

The mean of the concentration distribution is the focus of attention (rather than some extreme
percentile of the distribution or the distribution as a whole) because receptors integrate many
exposures through time and space to contaminated media. The choice of the UCL, rather than
the sample mean, is an acknowledgment of the fact that the estimates of the mean necessarily
depend on sample data, which unavoidably contains statistical uncertainty. The 95% UCL on the
mean is the 95th percentile of a hypothetical distribution of means from samples, all of size n,
collected at random from the underlying distribution of concentrations. This value accounts for

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-1

-------
1

the sampling uncertainty in the estimate of the mean of the underlying distribution that results
from collecting only n samples of concentrations. Figure 1 illustrates how the 95% UCL is a
conservative estimate of the mean of the underlying distribution of measured concentrations in
an environmental medium. The larger graph in the figure is a probability density function. The
abscissa is the value of the concentration, often in units such as milligrams of contaminant per
kilogram of medium (e.g., soil or water). The graph represents the variation in concentration
values experienced by receptors during individual exposure events. Typically, this is the
variation across space within the exposure area. In many cases with environmental data, this
underlying concentration distribution is positively skewed, as shown in the figure. Because of
this skewness, the mean of the distribution is somewhere to the right of the modal value. The
smaller distribution at the top of the figure is also a probability density function. Its graph has
the same horizontal axis and scale as the larger graph. It also has the same vertical axis
(probability), but the scale is obviously much different, because by definition, the areas of both
curves equal one. The distribution in the upper graph is the distribution of means that one would
obtain by repeatedly taking independent samples of some size n from the concentration
distribution and computed their means. Given the standard deviation 5 of the distribution, then
sNn is the (distribution-free) formula for the standard deviation of the upper distribution. If the
underlying distribution is skewed, the upper distribution will be skewed too, although it will be
less skewed and more symmetric than the lower distribution. In the limit as n approaches
infinity, the upper distribution will approach a perfectly symmetric shape of a normal
distribution. For finite sample sizes, however, the skewness will persist to some degree. The
UCL is the 95% percentile of this latter distribution. As the sample size grows, the UCL
approaches the true mean. However, when n is small, the UCL must be larger than the true mean
to account for the statistical uncertainty introduced by empirical sampling.

distribution of means

UCL95o/o

Figure 1 Interpretation of the UCL on the Mean

Until recently, there were only two methods widely used to compute UCLs. These were the
classical UCL associated with the theory of normal distributions, and the generalization of this
approach for lognormal distributions (Land 1972). Although the EPA guidance (1992) implicitly
assumed that one or the other of these two methods was sufficient for practical calculations,
many statistical distributions encountered in practice are neither normal nor lognormal in shape.
Moreover, as has been widely acknowledged, the Land method can often produce unreasonably
high values for the UCL (Gilbert 1987; Schmoyer et al. 1996; Singh et al. 1997; Schulz and
Griffin 1999). Statistical issues such as non-detects can have the effect of further significantly

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC C 5 2 7/10/2003

-------
1

increasing the value of the UCL. In the past several years, several other UCL methods have been
proposed for the situation in which the risk assessor cannot reasonably make assumptions about
the distributional type. Singh et al. (1997, 2000) and Schulz and Griffin (1999) reviewed and
suggested several alternate methods for calculating a UCL for non-normal data distributions.
Both papers give examples and comparisons of the UCLs calculated by various methods. Newly
drafted EPA guidance reviews a variety of alternative methods for calculating UCLs (EPA
2002). None of the known methods is ideal in all circumstances. Thus, some choices about
which method or methods to use must be made on a case-by-case basis.

C.5.1.1 Use of the Maximum Observed Concentration

Because some of the UCL methods outlined here can produce very high estimates of the UCL,
EPA guidance (EPA 1992, 2002) allows the maximum observed concentration to be used as the
EPC rather than the calculated UCL in cases where the UCL appears to be overly conservative.
It is important to note, however, that defaulting to the maximum observed concentration may not
provide a good estimate of the upper bound on the mean when sample sizes are very small,
because the observed sample maximum is generally not the true maximum of the population.
The observed maximum is usually lower, sometimes substantially so if sample error is high
(such as it is when n is small). The observed maximum may even be smaller than the population
mean. For instance, if n = 2, the observed maximum has a 25% chance of being less than the
median, which itself will under-estimate the mean of positively skewed distributions such as
concentrations. The use of the maximum as the default EPC is reasonable only when the data
samples have been collected at random from the exposure unit and the sample size is larger than
5 or 10.

C.5.2 UCL Estimation Methods

The two most commonly used methods for computing UCLs are based on assumptions about the
shape of the underlying distribution of concentrations. When the distribution is normal, the
classical approach based on Student's t statistic is preferred. When the distribution is lognormal,
the Land method based on the H statistic is used. There are distribution-free or non-parametric
methods available if the risk assessor cannot reasonably make assumptions about the
distributional type. Singh et al. (2000) describe several available methods. For very large data
sets, an approach based on the Central Limit Theorem (CLT) with a correction for positive
skewness may be used. If there are not very many samples collected in an exposure area, then
this method is not reliable with strongly positively skewed concentration distributions. For data
sets that are not large enough for this approach, multiple approaches have been suggested in the
statistical and environmental engineering literature (Singh et al. 2000; Schulz and Griffin 1999).
However, no method is ideal in all circumstances. Of the general methods suggested, Hall's
method based on a bootstrap* resampling procedure that corrects for skewness was selected for
use as the default method when the Student, Land, or CLT methods were not appropriate. This

*The Hall UCL should not be confused with a simple bootstrapping approach. There are many bootstrap
formulations for computing the UCL, and there is a great deal of variation among these formulations. The Hall
UCL has features that make it appropriate for use with the GE/Housatonic River risk assessment.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-3

-------
1

method was selected primarily because of its correction for skewness, which is often strong in
contaminant concentration data sets, but also because of its freedom from distributional
assumptions, its practicality with relatively small sample sizes, and its ease of computation.
Limited simulation studies assessing the performance of this method are described later in this
section. Figure 2 summarizes the decision process leading to the choice of a UCL method.

Are data
normal?

Are data
lognormal?

v«

Is n large, and
skewness small?

Yes

Use Student's t method

Yes

Use Land's// method

Use Central Limit Theorem method

Use Hall's bootstrap method

Figure 2 Flow Chart for Selecting UCL Method

Schulz and Griffin (1999) emphasize that fitting distributions to the data is a part of exploratory
data analysis that is crucial for a correct analysis. As recommended in EPA (1992) "where there
is a question about the distribution of the data set, a statistical test should be used to identify the
best distributional assumption for the data set." Because no single distribution type fits all
environmental data sets, it is best to perform some sort of comprehensive analysis to identify the
distribution. Risk assessors may encounter environmental data sets that appear normally
distributed, and others that appear lognormally distributed. They also may encounter data sets
that do not fit either normal or lognormal distributions. Distributions can be analyzed by a
variety of methods, many of which are described in Gilbert (1987). EPA (2000, Section 4.2)
gives guidance for testing distribution assumptions. Data plotting can also help identify a useful
distributional assumption.

C.5.2.1 Student UCLs for Normal Distributions

If the data are normally distributed, then the one-sided (1-a) upper confidence limit UCL\-a on
the mean should be computed in the classical way using the Student's t statistic (EPA 1992;
Gilbert 1987, page 139; Student 1908). Box 1 gives the procedure for computing the UCL of the
mean when the underlying distribution is normal. Box 2 gives a numerical example of an
application of the method.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-4

-------
1

Box 1: Computing the Student UCL for Normal Distributions

Let X\, X2 ,..., X„ represent the n randomly sampled concentrations.

— 1 "

STEP 1: Compute the sample mean X = —^Xi .

i=1

STEP 2: Compute the sample standard deviation 5 =.

1 i=i

[ n -

STEP 3: Use a table of quantiles of the Student's t distribution to find the (l-a)th quantile of
the Student's t distribution with n-1 degrees of freedom. For example, the value at the
0.05 level with 40 degrees of freedom is 1.684. A table of Student's t values can be
found in Gilbert (1987, page 255, where the values are indexed by p= 1-a, rather than
a level). The t value appropriate for computing the 95% UCL can be obtained in
Microsoft Excel with the formula TINV((l-0.95)*2, n-1).

STEP 4: Compute the one-sided (1-a) upper confidence limit on the mean

t/CZ1_a =X + t,

a,n-V

' / yfn .

Box 2: Example Calculation of the Student UCL

Concentrations were measured for 25 samples collected at random from an exposure unit. The
values observed were 228, 552, 645, 208, 755, 553, 674, 151, 251, 315, 731, 466, 261, 240, 411,
368, 492, 302, 438, 751, 304, 368, 376, 634, and 810 ng/L. It seems reasonable that the data are
normally distributed, and the Shapiro-Wilk W test for normality failed to reject the hypothesis
that they were (W= 0.937). The UCL based on Student's t will be computed.

STEP 1: The sample mean of the n=25 values is X = 451.

STEP 2: The sample standard deviation of the values is s = 197.5.

STEP 3: The lvalue at the 0.05 level for 25-1 degrees of freedom is ^0.05,25-1 = 1.71088.

STEP 4: The one-sided 95% upper confidence limit on the mean is therefore

UCL950/o =X + toma4s/-Jn =451 + 1.71088 xl97.5/V25 = 519 .

Caveats About this Method. In principle, this formulation will be correct, even when the sample
size is small, so long as the concentrations are normally distributed. However, caution is
warranted when there is a possibility that portions of the site with high contaminant
concentrations have not been adequately sampled. In such a case, the observed mean and
standard deviation may not be truly representative of the underlying population of concentrations
and the estimated UCL could fall below the true mean.

Testing for Normality. The Student UCL is robust to non-normality if the sample size is
sufficiently large. However, for moderate or small //, this method of computing the UCL can be
incorrect if the underlying data are not normally distributed. Therefore, it is important to test the
data for normality. EPA (2000, Section 4.2) gives guidance for several approaches to testing

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-5

-------
1

normality. The tests described therein are available in DataQUEST, which is a convenient
software tool (EPA 1997).

Accounting for Non-Detects. The use of substitution methods to account for non-detects is
recommended only when a very small percentage of the data is censored (e.g., < 15%), under the
presumption that the numerical consequences of censoring will be minor in this case.
Substitution methods tend to violate the assumption of normality. Moreover, the effect of the
various substitution rules is difficult to predict. Replacing non-detects with half the detection
limit can under-estimate the UCL, and replacing them with zeros may over-estimate the UCL
(because doing so inflates the estimate of the standard deviation). When censoring is moderate,
it is preferable to account for non-detects with Cohen's method (Gilbert 1987). EPA (2000,
beginning on page 4-43) provides guidance on the use of Cohen's method, which is a maximum
likelihood method for correcting the estimates of the sample mean and the sample variance to
account for the presence of non-detects among the data. This method requires that the detection
limit be the same for all the data and that the underlying data are normally distributed, conditions
that often are not met.

C. 5.2.2 Land UCLs for Lognormal Distributions

It is inappropriate to extend the Student UCL method to lognormally distributed samples by log-
transforming the data, computing a UCL, and then back-transforming the results. For
concentration data sets that appear to be lognormally distributed, it is preferable instead to use
Land's method to compute the UCL on the mean (Land 1971, 1975; Gilbert 1987; EPA 1992;
Singh et al. 1997). This method requires the use of the H statistic, tables for which were
published by Land (1975) and Gilbert (1987, Tables A10 and A12). Box 3 gives step-by-step
directions for this method, and Box 4 gives a numerical example of its application.

Box 3: Computing the Land UCL for a Lognormal Distribution

Let X\, X2 ,..., X„ represent the n randomly sampled concentrations.

STEP 1: Compute the arithmetic mean of the log-transformed data In X = — ^ ln(X;).

STEP 3: Look up the H\-a statistic for sample size n and the observed standard deviation of the
log-transformed data. Tables of these values are given by Gilbert (1987, Tables A-10
and A-12) and Land (1975).

STEP 4: C ' '' ' ' ' N er confidence limit on the mean

STEP 2: Compute the associated standard deviation sln =

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-6

7/10/2003

-------
1

Box 4: Example Calculation of the Land UCL

Concentrations were measured for 31 samples collected at random from an exposure unit. The
observed values were 2.8, 22.9, 3.3, 4.6, 8.7, 30.4, 12.2, 2.5, 5.7, 26.3, 5.4, 6.1, 5.2, 1.8, 7.2, 3.4,
12.4,0.8, 10.3, 11.4,38.2,5.6, 14.1, 12.3,6.8,3.3,5.2,2.1, 19.7, 3.9, and 2.8 mg/kg. Because of
their skewness, the data may be lognormally distributed. The Shapiro-Wilk Wtest for normality
rejected the hypothesis, at both the 0.05 and 0.01 levels, that the distribution is normal. The
same test failed to reject at either level the hypothesis that the distribution is lognormal. The
UCL on the mean based on Land's H statistic will be computed.

STEP 1: Compute the arithmetic average of the log-transformed data h\X = 1.8797.

STEP 2: Compute the standard deviation of the log-transformed data ,V|n = 0.8995.

STEP 3: The H statistic for n = 31 and sin =0.90 is 2.31.

STEP 4: The one-sided 95% upper confidence limit on the mean is therefore
UCL950/o = exp(l.8797 + 0.89952 12 + 2.3lx 0.8995/V31-l)= 14.35 .

Caveats About this Method. Land's approach is known to be sensitive to deviations from
lognormality. The formula may commonly yield estimated UCLs substantially larger than
necessary when distributions are not truly lognormal if variance or skewness is large (Gilbert
1987). Singh et al. (1997) assert that when sample sizes are less than 30, the method can yield
impractically large UCLs even when the underlying distribution is lognormal. Also, because the
method log-transforms the concentrations, it cannot be applied without modification to data sets
containing sure zeros.

Testing for Lognormality. Because the Land method is sensitive to violations of the assumption
of lognormality, it is very important to test this assumption. EPA (2000, Section 4.2) gives
guidance for several approaches to testing distribution assumptions. The tests are also available
in the DataQUEST software tool (EPA 1997).

Accounting for Non-Detects. Gilbert (1987, page 182) suggests extending Cohen's method to
account for non-detect values in lognormally distributed concentrations. Cohen's method (EPA
2000, page 4-43f) assumes the data are normally distributed, thus it must be applied to the log-
transformed concentration values. If {iy and or are the corrected sample mean and standard

£ = exp(^ +6J/2)

6 = ^ex p(6')-l

deviation, respectively, of the log-transformed concentrations produced by Cohen's method, then
the expressions give the corrected estimates of the mean and standard deviation of the underlying
lognormal distribution. This method requires that there be a single detection level for all the data
values.

C.5.2.3 UCLs for Other Specific Distribution Types

Methods for computing UCLs on the means of other distribution shapes in addition to the normal
and lognormal have appeared in the statistical literature. For example, Johnson (1978) described
a method for computing the UCL for asymmetrical distributions such as the exponential. Schulz

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC ^ 1 7/10/2003

-------
1

and Griffin (1999) described Wong's method (1993) for obtaining confidence limits on the mean
of a gamma distribution. In general, if there are mechanistic arguments that suggest a population
of concentrations should follow a particular distribution shape, and if statistical testing confirms
the expected shape is in reasonable conformance with available data, then the UCL computed by
a method developed specifically for the distribution shape, if one exists, is likely to be most
appropriate for the data set. However, UCL methods for other specific distribution types were
not used in this risk assessment, because there were no known theoretical or empirical arguments
that would suggest any of these shapes of concentration distributions.

C.5.2.4 Central Limit Theorem UCLs When Sample Sizes are Large

There are also distribution-free approaches to computing UCLs on the mean that do not make
specific shape assumptions about the underlying distribution of concentrations. If the sample
size is sufficiently large, the Central Limit Theorem (CLT) implies that the mean will be
normally distributed, no matter how complex the underlying distribution of concentrations might
be. This is the case even if the underlying distribution is strongly skewed, has outliers, or is a
mixture of different populations, so long as it is stationary (not changing over time), has finite
variance, and the samples are collected independently and at random. Then the distribution of
means converges to normality as the sample size increases. However, the theorem does not say
how many samples are sufficient for the assumption of normality to hold. When the sample size
is moderate or small, the means will not generally be normally distributed, and this non-
normality is aggravated by the skewness of the underlying distribution. Given the skewness
observed in the Housatonic data sets, even several hundred samples may be insufficient to ensure
that the UCL computed with the CLT will be reliable. Chen (1995) suggested an approach that
partially accounts for positive skewness. Singh et al. (1997, 2000) call this approach the
"adjusted CLT" method. They suggest it is an appropriate alternative to Land's method for
lognormal distributions when the standard deviation is less than 1 and the sample size is larger
than 70, or when the standard deviation is less than 1.5 and the sample size is larger than 100.
Box 5 describes the steps for this method, and Box 6 gives a numerical example.

Box 5: Computing the Adjusted Central Limit Theorem UCL

Let X\, X2 ,..., X„ represent the n randomly sampled concentrations.

— 1 "

STEP 1: Compute the sample mean X = —^Xi .

i=1

STEP 2: Compute the sample standard deviation 5 = .' " x ' " "2

n n (x -xV

STEP 3: Compute the sample skewness (3 = V — . This quantity can be

(n-\)(n-2)^\ s )

calculated in Microsoft Excel with the SKEW function.

STEP 4: Let za be the (l-a)th quantile of the standard normal distribution. For the 95%

confidence level, za = 1.645.

STEP 5: Compute the one-sided (1-a) upper confidence limit on the mean

UCLl_a=X +

fz . p '

6-Jn

(l + 2zl) 5 / yfn .

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-8

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

Box 6: Example Calculation of the Adjusted Central Limit Theorem UCL

Concentrations were measured for 360 samples collected at random from an exposure unit. The
values observed were 45, 209, 109, 25, 18, 21, 16, 14, 14, 1, 39, 17, 17, 33, 30, 32, 97, 15, 33,
32, 4, 15, 51, 22, 29, 15, 25, 53, 12, 187, 24, 172, 21, 25, 47, 34, 31, 27, 25, 21, 20, 22, 32, 37,
55, 40, 2, 29, 36, 8, 33, 4, 40, 5, 23, 41, 39, 42, 31, 117, 62, 174, 50, 30, 31, 30, 33, 22, 21, 59, 3,
6, 46, 33, 47, 44, 192, 12, 2, 34, 25, 11, 50, 45, 21, 36, 20, 38, 29, 32, 23, 52, 28, 30, 40, 5, 4, 22,
21, 24, 16, 39, 10, 36, 64, 2, 40, 7, 51, 19, 7, 13, 14, 39, 15, 39, 35, 6, 24, 33, 9, 83, 46, 40, 38, 0,

19,	18, 44, 11, 29, 29, 4, 38, 53, 35, 80, 8, 37, 30, 0, 12, 25, 22, 4, 19, 24, 47, 7, 72, 44, 121, 24,

20,	47, 22, 5, 195, 13, 35, 9, 0, 51, 31, 8, 21, 26, 34, 54, 7, 41, 8, 20, 33, 26, 5, 20, 22, 47, 122,

21,	13, 2, 11, 22, 12, 12, 16, 13, 19, 41, 16, 22, 47, 48, 26, 15, 18, 6, 3, 30, 31, 18, 7, 11, 21, 35,
26, 33, 193, 26, 78, 34, 2, 35, 0, 10, 37, 38, 6, 26, 12, 22, 7, 55, 2, 21, 5, 23, 37, 21, 16, 0, 23, 28,
1, 38, 20, 44, 91, 5, 120, 121, 25, 36, 5, 9, 30, 0, 11, 16, 10, 38, 28, 14, 12, 89, 16, 18, 4, 14, 6,
47, 36, 17, 40, 2, 26, 41, 16, 3, 29, 26, 25, 21, 4, 25, 157, 41, 13, 12, 41, 10, 15, 49, 47, 39, 47, 5,
35, 36, 57, 40, 6, 0, 41, 33, 21, 1, 38, 25, 155, 125, 11, 2, 39, 28, 9, 7, 23, 6, 20, 51, 34, 40, 31,
159, 8, 8, 18, 33, 2, 48, 6, 24, 2, 17, 31, 2, 99, 20, 57, 39, 2, 39, 22, 31, 169, 44, 27, 8, 35, 19, 25,
4, 62, 12, 25, 55, 23, 29, 37, 38, 22, 20, 37, 3, 45, 19 and 195 \ig/L. Filliben's test shows that
this distribution is significantly different (at the 1% level) from both a normal and a lognormal
distribution. The UCL based on the Central Limit Theorem will be computed.

STEP 1:	The sample mean of the n = 360 values is X = 32.90.

STEP 2:	The sample standard deviation of the values iss = 34.86.

STEP 3:	The sample skewness P = 2.881.

STEP 4:	The z statistic is 1.645.

STEP 5: The one-sided 95% upper confidence limit on the

f	T «81 /	_ 0\

mean is therefore

UCL95 0/o =32.90-

2 881 /	\)		

1.645+ —= (l + 2xl.6452) 34.86/V360 = 34.84.
6V360	,

Caveats About this Method. A sample size of 30 is sometimes prescribed as sufficient for an
appeal to the CLT, but when the data are highly skewed (as many concentration data sets are), a
much larger number of samples may be needed to approximate normality (cf. Gilbert 1987, page
140). Although this method is distribution-free in principle, it does not produce UCLs that will
be reliably larger than the true mean unless the sample size is large. The necessary number of
samples increases as the skewness of the underlying population of concentrations becomes
greater.

C.5.2.5 Hall UCLs Based on Bootstrap Resampling

Bootstrap procedures (Efron 1982) are an assortment of generally robust nonparametric
statistical methods that can be used to construct approximate confidence limits for the population
mean. While these procedures assume the samples are representative of the underlying
distribution of concentrations, they require no assumptions about the shape of that distribution
and are applicable to a variety of situations of arbitrary complexity. Although a parametric
statistical method that depends on a distributional assumption is usually more efficient when it is

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-9

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

appropriate, its assumption is often difficult to justify empirically. In such cases, methods based
on statistical resampling techniques are often preferred. An elaborated bootstrap procedure
developed by Hall (1988, 1992; Manly 1997; Schulz and Griffin 1999; Zhou and Gao 2000) that
takes account of bias and skewness is described in Box 7 and an example of its application to
data is described in Box 8.

Box 7: Computing the Hall UCL

Let X\, X2 ,..., X„ represent the n randomly sampled concentrations.

— 1 "

STEP 1: Compute the sample mean X = —'Y,Xi.

n i=i

STEP 2: Compute the (biased) standard deviation 5 =

n '/=!

STEP 3: Compute the sample skewness k = -rl (*.-4

ns I=l

STEP 4: For b = 1 to B (a very large number) do the following six steps:

4.1: Generate a bootstrap sample data set; i.e., for i = 1 to n let j be a random integer

between 1 and n and add observation X, to the bootstrap sample data set.
4.2: Compute the arithmetic mean Xb of the data set constructed in Step 4.1.
4.3: Compute the associated standard deviation Sb of the constructed data set.
4.4: Compute the skewness kb of the constructed data using the formula in Step 3.

4.5: Compute the Studentized mean W= (Xb~X)/sb..

4.6: Compute Hall's statistic Q = W + kbW2/3 + k^W3121 + kb /(6n).

STEP 5: Sort all the Q values computed in Step 4 and select the lower ath quantile of these B
values. It is the (aB)th value in an ascending list of Q's. This value is from the left tail
of the distribution.

( /	, xxl/3A

STEP 6: Compute W(Qa) = —

k

1-



STEP 7: Compute the one-sided (1-a) confidence limit on the mean UCLx_a = X + W(Qa)s .

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-10

-------
1

11 Caveats About Resampling. In the application of bootstrap procedures, the only assumption

12 made is that the sample data are representative of the underlying population. However, this

13 assumption has profoundly more import than those of many other statistical procedures. In

14 parametric methods, only the estimated parameters need to be reasonably close to the population

15 parameters to obtain accurate inferences. Because bootstrap methods exploit more of the

16 information in a sample, that sample must be a statistically accurate characterization of the

17 underlying population in all respects (not just in its mean and standard deviation). In practice, a

18 random sampling procedure satisfies the representativeness assumption. Therefore, the data

19 must be random samples of the underlying population. Bootstrapping procedures are

20 inappropriate for use with data that were non-randomly sampled, or targeted to a particular

21 objective, such as delineating areas of high contamination. When sample sizes are small, the

22 ways for a sample to deviate from its parent distribution in a way that affects the calculation will

23 become more numerous than for analogous parametric approaches. Because of this sensitivity to

24 small sample sizes, it is not altogether clear that the lack of a distribution assumption always

25 makes the bootstrap method preferable to a parametric approach.

26 C.5.2.6 Coverage Performance of the Hall UCL

27 There is no known distribution-free method for reliably computing UCLs. That is, for each

28 suggested method there are distributions such that, when a 95% UCL is computed on random

29 samples from the distribution, it is less than the true distribution mean more than 5% of the time.

30 How often the computed UCL is larger than the true mean is called the method's coverage rate,

31 which measures how it performs its function as an upper estimate of the mean. A method's

32 coverage rate depends on the nature of the distributions to which it is applied. Therefore, it is

33 important to study the expected performance of a method by applying it to distributions that

34 might be typically encountered, to assess whether the method achieves or approximates its

35 nominal coverage rate.

36 Using a Monte Carlo simulation study with 2,000 iterations, the actual coverage rates were

37 assessed for the 95% UCL computed by Hall's bootstrap method for 12 hypothetical

38 distributions (for which true means are known). The 12 distributions were:

Box 8: Example Calculation of the Hall UCL

Using the same concentration values given in Box 4, the UCL can also be computed based on the
bootstrap resampling method.

STEP 1: The sample mean of the n= 31 values is X = 9.59.

STEP 2: The standard deviation (using n as divisor) of the values iss = 8.946.

STEP 3: The skewness k= 1.648.

The Pascal-language software shown in Appendix B of EPA (2002) can be used to make the
calculation. (See also the slightly modified code for Hall UCLs on page C.5-28 of this
appendix.) The one-sided 95% UCL on the mean is 13.3. Because this value depends on random
deviates, it can vary somewhat upon recalculation.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-11

-------
1

An even (i.e., 50:50) mixture of two normal distributions, the first with a mean of 20 and
a standard deviation of 2.0, and the second with a mean of 40.0 and a standard deviation
of 4.0.

B. A 3-to-l mixture of the same two normal distributions mentioned in (A).

C. A 9-to-l mixture of the same two normal distributions.

D. An empirical distribution function based on data collected in one of the exposure areas of
the site.

E. The same empirical distribution with an appended tail 5 times longer than the distance
between the largest observed value in the distribution from the second largest value.

F. The same empirical distribution with an appended tail 10 times longer than the distance
between the largest observed value in the distribution from the second largest value.

G. An exponential distribution with mean 50.

H. A mixture model composed of an exponential distribution with mean 35 and weight 60%
and a delta function of zeroes with 40% weight.

I. A mixture composed of an exponential distribution with mean 2 and weight 2/3 and a
normal distribution with mean 50 and standard deviation 15 and weight of 1/3.

J. A triangular distribution with minimum and mode of zero and maximum of 99.

K. A Gumbel (extreme value) distribution with mean 50 and standard deviation 25.

L. A chi-squared distribution with one degree of freedom and multiplication by 20.

Figure 3 shows a plot of the distribution function for each of the 12 hypothetical distributions A-
L used in the simulation. The ordinate for each plot is cumulative probability, ranging from 0 up
to 1. The abscissa is concentration in arbitrary units (but typically mg/kg). Also displayed for
each plot is the theoretical mean |j,. All of these distributions are models of the kinds of
distributions that can arise in contamination situations. Distributions D, H, I, and L were fitted
visually so as to mimic specific concentration distributions observed among the Housatonic data
sets. Note that most of these study distributions have theoretically infinite ranges, including A,
B, C, G, H, I, K, and L.

In the Monte Carlo simulation study, random numbers were generated from and according to
each of these distributions. All numbers were truncated at zero to reflect the non-negativity of
concentration values. For each distribution, the theoretical mean was corroborated by testing
that an average of 20 million random numbers drawn from the distribution approached the
planned value. In all cases, the observed mean was well within one thousandth of the theoretical
value. Sample data sets of 30 values were generated for each distribution and the UCL was
computed from the values. If the UCL was equal to or larger than the theoretical mean, a tally
was incremented. This process was repeated 2,000 times, and the tallies recorded the

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-12

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

10

20

30

40

50

60

10

20

30

40

50

60

10

20

30

40

50

60

60

120

180

50

100

150

200 250

50

100 150 200 250 300

50

100 150 200 250 300

50

100

150

200 250

20

40

60

80

100

20

40

60

80

100

40

80

120

160 200

50

100

150

200 250

Figure 3 Cumulative Distribution Functions for the 12 Concentration Distributions
Used in the Coverage Study for the Hall UCL

performance of the Hall UCL for the various distributions. The observed coverage rates are
supposed to be at or above the nominal rate of 0.950. If there were values substantially below
this level, it would indicate that the method is deficient for that kind of distribution. The
coverage rate was calculated as the frequency with which the computed UCL exceeded the
theoretical mean for the distribution from which the sample was drawn at random. The observed
coverage rates were:

Distribution

Coverage

A

50:50 normals

0.958

B

75:25 normals

0.970

C

90:10 normals

0.913

D

Housatonic (EA 29-2)

0.949

E

Housatonic, 5xtail

0.934

F

Housatonic, 1 Oxtail

0.937

G

Exponential

0.946

H

Exponential/zero

0.939

I

Exponenti al / normal

0.982

J

Triangular

0.958

MK0110:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-13

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

Distribution

Coverage

K Gumbel	0.943

L Chi-squared	0.931

Figure 4 graphically depicts the same information as a scattergram of the Hall UCL coverage
rates compared to the target nominal performance level of 95%, which is shown as a horizontal
line. The 12 squares correspond to the 12 distributions in the coverage study. (The horizontal
positions of the squares are meaningless.) It is apparent that the rates are not all above 95%,
although all are above 90%. Nevertheless, it is clear that the observed coverages roughly
approximate the nominal coverage rate of 95% for the distributions studied.

0.99

0.98
"§ 0.97

n

2 0.96

a.

0) 0.95

re

i_

<1) 0.94
U)

re

a>
>
o
o

0.93 -
0.92 -
0.91 -
0.9 -

Figure 4 Scattergram of Hall UCL Coverage Rates

Because coverage performance was examined only for a single sample size and a limited variety
of distribution shapes, these results do not constitute a thorough study of the performance of the
Hall UCL method, but they do provide relative confidence in the appropriateness of this
approach for use with the site data. As previously mentioned, there is no general statistical
method that can produce a true 95% UCL in the "distribution-free" case when the underlying
distribution of concentrations is not assumed to be from a special or restricted family of
distributions. Nevertheless, this modest simulation study demonstrates that the Hall UCL
performs reasonably well on the kinds of distributions that might be encountered at the site, even
including cases of distributions with theoretically infinite tails. It approximates the nominal
coverage rate of 95% without being excessively conservative.

MK0110:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-14

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

C.5.2.7 ChebyshevUCL

Singh et al. (1997, 1999, 2000) suggested the use of the Chebyshev inequality to estimate UCLs
for a variety of distributions so long as the skewness is not very large. The one-sided version of
the Chebyshev inequality (Allen 1990, page 79; Savage 1961, page 216) is appropriate in this
context (cf. Singh et al. 1997, 2000). A one-sided (1-a) UCL on the mean is

—-1 o/yfn.

UCLx_a = (J, +

V a

where \x is the arithmetic mean of the underlying distribution of concentrations and a is the
standard deviation of the distribution, and n is the sample size. This formula can be used to
compute a distribution-free estimate of the UCL for the population mean when the population
mean and standard deviation are known. In practice, however, these values are not known and
must be estimated from data. Singh et al. (1997, 2000) suggest two ways to make these
estimations. The first is to estimate the population mean and standard deviation by the sample
mean and sample standard deviation, respectively. The second way is to use the minimum-
variance unbiased estimators (MVUE) for the mean and variance derived for the lognormal
distribution. The MVUE of the population mean for a lognormal distribution is

M-ln =exp(j)g„(Sy/2),

where

y=-fi ln(X;)

and

4 =

^(ln (Xt)-y)\

and where gn denotes a function for which tables are available (Aitchison and Brown 1969, Table
A2; Koch and Link 1980, Table A7). The MVUE for the variance of the estimate jiLN for
lognormal distributions is

62(|lLN) = exp(2j)

{Sn(sy /2))2

n-2 7

—Tsy

n-1

The square root of this quantity is used in place of oNn to compute the UCL. Singh et al. (2000)
claim that using the MVUE formulas with the Chebyshev inequality will often yield an estimated
UCL that is smaller than that obtained from the Land method under an assumption that the
underlying distribution of concentrations is lognormal. Both the MVUE Chebyshev and
Chebyshev based on the sample mean and standard deviation have been implemented in the
software tool ProUCL (Lockheed Martin 2000).

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-15

-------
1

Caveats About the Chebyshev Method. Although the Chebyshev inequality method makes no
distributional assumptions, it does assume that the parametric mean and standard deviation of the
underlying distribution are known. As Singh et al. (1997) acknowledge, when these parameters
must be estimated from data, the estimate of the UCL is not guaranteed to be larger than the true
mean with the prescribed frequency implied by the a level. In fact, using only an estimate of the
standard deviation can substantially under-estimate the UCL when the variance or skewness is
large, especially for small sample sizes. Nevertheless, for many distributions, the Chebyshev
method does more often meet or exceed the nominal 95% coverage rate than other UCL
methods; i.e., 95% or more of the time that the Chebyshev UCLs are computed, they will be
larger than the true mean. But so would a method that simply added a positive constant to any
UCL estimate. As high as it is, the Chebyshev UCL is still not guaranteed to have the nominal
coverage rate for all distributions. Thus, it is more conservative than necessary in many cases,
and yet as a general method it is not reliable.

C.5.2.8 Kolmogorov-Smirnov UCL

It is possible to construct a UCL on the mean of a distribution from the classical result of
Kolmogorov (1933, 1941) that gives confidence limits on empirical distribution functions (Sokal
and Rohlf 1981, page 721; Crow et al. 1960, page 90f; Feller 1948). Analogous to simple
confidence intervals around a single number, these are bounds on a statistical distribution as a
whole. Kolmogorov showed that if samples were drawn at random from a continuous
distribution, then confidence limits on the distribution can be constructed by adding and
subtracting a constant that depends only on the sample size (and not the underlying distribution
itself) from the empirical distribution function representing the observed data. If the value of the
function becomes negative or greater than 1, it is truncated at 0 or 1 respectively. Smirnov
(1939) derived the value of this constant; Miller (1956; see also Birnbaum 1952) gives an
improved formula and extensive tables for the constant. This result is very rich and it provides
practical methods for comparing empirical distributions in addition to producing confidence
limits. The UCL method proceeds by computing an empirical distribution function for the
sample data, finding the 95% confidence limits on the distribution, and then computing the
bound as the mean associated with the right such limit. The resulting UCL will surely be as
large or larger than the mean of the underlying distribution with the required probability.

One issue with this approach is that the right limit is a distribution function that places some non-
zero mass at infinity. Consequently its mean is also infinity. If, however, the analyst can specify
on grounds other than those of the empirical sampling what the largest possible value of the
underlying distribution is, then the bounding distribution can be truncated at this value and its
mean becomes finite. This assumption of knowing the maximum possible concentration value
seems to be somewhat weaker and therefore more acceptable than the assumption by the
Chebyshev UCL that the true mean and standard deviation parameters were known. Given the
assumption, this approach produces a true confidence limit, i.e., one that is guaranteed to bound
the true mean with at least the prescribed frequency. The method is statistically well behaved,
even for very small data sets, and it is truly distribution-free in that one does not to make shape
assumption (other than the maximum) about the underlying distribution.

This approach has several useful properties. So long as the maximum possible value can be
established, the Kolmogorov-Smirnov (KS) UCL is not at all sensitive to the degree of skewness

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-16

-------
1 of the underlying distribution. It is also possible to account for the censoring associated with

2 laboratory non-detects, the resulting UCL estimate is never strongly inflated by such censoring,

3 and it often is not affected at all by censoring. The approach can be extended to account

4 comprehensively for measurement error, which would allow it to handle the plus-or-minus

5 ranges reported by laboratories. Finally, it is compatible with spatial or other forms of weighting

6 that might be applied to the data.

7 Caveats About the Kolmogorov-Smirnov Method. Although the KS UCL method makes no

8 distributional assumptions, the method is guaranteed to have the nominal coverage rate for all

9 distributions only if the maximal value can be specified. In practice, identifying the maximum

10 possible value may be as contentious as computing the UCL itself. Because the method yields

11 reliable results for all possible distributions, the answers may in practice be conservative given

12 reasonable constraints on the nature of the concentration distributions. Thus, the KS UCL will

13 likely be more conservative than necessary in many cases.

14 C.5.2.9 Performance Comparisons Among Nonparametric UCLs

15 This section describes further simulations that compare several potential UCL methods for the

16 case when neither an assumption of normality nor one of lognormality is appropriate. The

17 simulations used the same 12 distributions depicted in Figure 3. Three situations were

18 considered. In the first, summarized in Figure 5, there were 100 sample points available to the

19 analyst. There were 30 samples in the second situation, which is summarized in Figure 6, and, in

20 the third shown in Figure 7, there were only 10 samples.

1 -I

0.99-

0.98 -

0.94-

0.93-

0.92-

0.91 -

0.9-

Sample size =100

» m A A A A A^-

¦Chebyshev

¦ Hall

¦ nominal (95%;

¦ adjusted CLT
CLT

ABCDEF GHIJ KL

Figure 5 Performance of Six UCL Methods with 100 Samples

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-17

-------
Sample size = 30

0.85

0.95

a>
+¦>

2


O

o

0.85

0.75

0.65

KS

^^Chebyshev
-¦-Hall

^—nominal (95%)
—^ adjusted CLT
CLT

ABCDEFGHIJ KL

Figure 6 Performance of Six UCL Methods with 30 Samples

Sample size =10

KS

¦Chebyshev
¦Hall

¦nominal (95%)
¦adjusted CLT
CLT

CDEFGHIJ
Figure 7 Performance of Six UCL Methods with 10 Samples

MK0110:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-18

-------
1

Each point in each graph represents the results of a simulation involving 10,000 Monte Carlo
replications in which a hypothetical data set of the given size is randomly sampled from the
distribution, and a UCL is computed and compared with the theoretically known mean for the
distribution. The frequency with which the UCL was larger than the mean is given on the
ordinate. The rates for the 12 different distributions are arrayed along the abscissa, although the
horizontal position has no significance. The three situations for sample size roughly span what is
observed among Housatonic data sets. For a single exposure area, 100 would be a fairly large
number of samples, although there are a few exposure areas that have more than a couple
hundred samples. The average number of samples per exposure area was about 25, and the most
sparsely sampled exposure areas had only a few samples. For each sample data set, five UCLs
were computed: the KS UCL with the observed maximum value of each data set used as the
theoretical maximum, the non-MVUE version of the Chebyshev UCL, the Hall UCL, the
adjusted CLT UCL, and the traditional CLT UCL.

These are general findings:

¦ The KS UCL and the Chebyshev UCL of Singh et al. (2000) are often excessively
conservative.

¦ The CLT UCL does not achieve a nominal coverage rate of 95% even when the
sample size is fairly large.

¦ The Hall UCL yields balanced results with a coverage rate that approximates the
nominal rate except for some extreme distributions when the sample size is low.

¦ The adjusted CLT is a close second, behind the Hall UCL, in terms of coverage
performance.

The Hall UCL does not appear to achieve the nominal coverage rate in the situation where the
sample size is low, leading to the conclusion that either the KS or Chebyshev UCL should be
used for data sets when the sample sizes are this low. It should be noted, however, that for n =
10, the coverages generally approximated 90%, which is another confidence level that is
sometimes used in assessments. The exception is distribution C. In fact, distribution C seems
especially problematic for all of the methods when the sample size is low. The reason for this is
that 10 samples are simply inadequate to detect the second mode of the bimodal distribution,
which contains only 10% of the mass to start with. In fact, an elementary application of the
theory of extreme values* can be used to show that, even if one were to use the maximum
observed value as the UCL, the nominal coverage rate of 95% could not be achieved with only
10 sample points for such a distribution. Indeed, it might be considered problematic if some
method actually performed well for this case, because, to do so, the method must be
fundamentally insensitive to the data. In any case, however, it may not often matter which UCL
method is used when sample sizes are very low, simply because the EPC may often be
determined in such cases by the maximum observed value rather than the computed UCL.

*The 5th percentile of the tenth power of the cumulative distribution function (which represents the distribution of
the largest of 10 random samples from that distribution) is smaller than the theoretical mean of the underlying
distribution C.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-19

-------
1

C.5.3 Spatial Weighting

Because exposures occur as receptors move within an exposure area, it is reasonable to use
spatial averaging to estimate the mean exposure. The receptors themselves are effectively doing
such averaging if they are wandering around the exposure area in an essentially random way.
However, such averaging would only be appropriate if the receptors are actually experiencing
these exposures randomly. For instance, if some receptors were preferentially attracted to a
subarea where concentrations are very high, then averaging could under-estimate exposures
received by these receptors. Alternatively, receptors such as adults and children in large
recreational areas may use portions of an exposure area less frequently because they are under
water part of the time, or the shrubs and other growth are particularly dense and limit the use of
some areas. These relative uses are accounted for in the risk assessment with area-use weighting
factors.

Another reason to employ spatial weighting is to partially correct for non-random sampling. The
UCL methods assume that the observed concentrations were randomly distributed. This means
that the samples are independent and identically distributed. The phrase "identically distributed"
means that the values come from the same statistical population. This assumption is satisfied by
a reasonable definition of the exposure area. The independence assumption is more problematic.
The existence of spatial structure in contaminant concentration usually means that samples taken
close together in space are likely to be positively correlated and not independent. For the
purposes of statistical inferences, it might be best to locate all samples by randomly assigning
sample locations within the exposure area without respect to the broad patterns of contamination,
river flow, or habitat use by receptors. A truly random sampling strategy is often impractical,
however, and can also be inefficient given the data quality objectives of the site investigation and
risk assessments, beyond statistical inferences. As a result, sampling strategies often include
various transects across geographic clines, increased density around potential "hot spots" of high
contamination, and special attention to areas of intense habitat use by receptors. Because the
available methods to compute a UCL all assume random sampling, it is important to account for
or somehow lessen the biasing effect of any non-randomness in the sampling strategy.

One way to do this is to spatially weight the concentration data by the proportion of the exposure
area that it represents. In this risk assessment, an interpolation strategy based on inverse distance
weighting (IDW) was used as developed by EPA Region 5 (www.epa.gov/region5fields/). In
this approach, a spatial grid of predicted concentrations over the entire exposure area is
computed. The interpolated value at each node in the grid represents an estimate for the
concentration at that location, given the gross features of the landscape and the concentration
values actually observed nearby. The grid of concentration values then is taken as the
representation of the spatial variation of concentration in the exposure area. An important
purpose of this spatial weighting scheme is to limit the statistical impact of non-random
sampling.

Spatial weighting introduces complications into the statistical calculations. In the absence of
spatial weighting, the degrees of freedom are determined by the number of samples. However,
IDW interpolation yields a grid of concentrations that has many more values than were originally
observed. The number is determined by the geographical area of the exposure area and the
resolution of the grid used in the interpolation. It would be statistically improper to use this
artificially high number of values to determine the degrees of freedom for computational

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC q ^ 2Q 7/10/2003

-------
1

purposes. This is clear because the UCL calculation will produce a value arbitrarily close to the
sample mean simply by increasing the number of grid points. This would thwart the intended
use of the UCL to reflect sampling uncertainty about the mean, without the expense of collecting
the empirical data that could legitimately reduce that sampling uncertainty.

What is needed is a way to make use of all the extra data values from the interpolation without
forgetting the true degrees of freedom justified by the original sampling effort. To do both, the
UCL formulas were modified slightly to differentiate between the true degrees of freedom
resulting from the original empirical effort and the number of data values used to compute the
various statistics used in the mathematical expression. Consider, for instance, the calculation of
the Student UCL,

UCL = X + tn_lS/4n ,

which involves the sample mean, the sample standard deviation, the sample size, and a critical t
statistic from a table that depends on the sample size. In computing the mean and standard
deviation from interpolated data, it makes sense to use the number of grid values actually
combined in the divisors (otherwise the characterization of the concentration population would
be obviously biased). The t statistic and the V// divisor, on the other hand, modify the estimate of
the dispersion of the underlying population to reflect sample uncertainty in the (upper)
distribution of sample means illustrated in Figure 1. These values should be based on 20, the
true degrees of freedom.

This strategy distinguishes the two senses of "sample size" into vector length and degrees of
freedom. Similar considerations can be applied to modify the formulas for the Land UCL and
the Hall UCL. Explicit instructions for computing UCLs are given in Boxes 9, 10, and 11.
Below the boxes, a series of Monte Carlo simulation studies is described that was employed to
study the behavior of these formulas and to confirm that they produce reasonable results.

Box 9: Computing the Student UCL for Spatially Weighted Data

Let X\, X2 ,..., X„ represent the n interpolated concentrations that were obtained by spatially
weighting the m originally sampled concentrations Yh Y2Ym.

— 1 "

STEP 1: Compute the mean of the interpolated data X = —^Xj .

i=1

STEP 2: Compute their standard deviation 5 =

x-xf

STEP 3: Use a table of quantiles of the Student's t distribution to find the (l-a)th quantile of
the Student's t distribution with m-1 degrees of freedom. For example, the value at
the 0.05 level with 40 degrees of freedom is 1.684. A table of Student's t values can
be found in Gilbert (1987, page 255, where the values are indexed by p= 1-a, rather
than a level). The t value appropriate for computing the 95% UCL can be obtained in
Microsoft Excel with the formula TINV((l-0.95)*2, m-1).

STEP 4: Compute the one-sided (1-a) upper confidence limit on the mean accounting for
spatial weighting UCLx_a = X + ta m_xs/yjm .

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-21

-------
1

12 C.5.3.1 Demonstration that the UCL Formulas Modified for Spatial Weighting are

13 Reasonable

14 The strategy of differentiating the two senses of "sample size" into vector length and degrees of

15 freedom may seem innocuous or even self-evident, but it deserves some attention to ensure that it

16 yields reasonable results in the practical situations encountered with the Housatonic River

17 floodplain soil data sets.

18 An alternative strategy of randomly sampling from the grid of interpolated concentrations to

19 make a synthetic collection of values having the same size as the original sampling could be

20 considered. For instance, suppose there were 30 samples originally collected and processed to

21 obtain 1,024 interpolated values. A random sample (with replacement) of size 30 from the 1,024

22 points could be collected and the UCL computed. This value would be an estimate of the

23 population UCL. This estimate has the correct degrees of freedom, and it was based on data

24 values that were independent and identically distributed (i.e., random). This process could be

25 repeated many times, each time sampling 30 points and computing the UCL from these values.

26 Because each of the resulting UCLs would be based on limited sampling, no single estimate

27 would be maximally reliable, but perhaps the collection of such UCLs would appropriately

28 reflect the underlying population.

29 As a numerical experiment, a series of Monte Carlo simulation studies were conducted that

30 implemented this repeated random sampling strategy. The simulation involved random samples

31 from a hypothetical data set similar to those produced by the grid interpolations. The spatial

32 variation in the hypothetical data used in this numerical experiment is depicted in the two

33 graphical displays in Figure 8. The graph on the left shows the variation in a gray-scale contour

34 plot. Contours lines are shown for concentration values of 50, 75, and 100 units. The graph on

35 the right depicts the same spatial variation in the hypothetical data set as a three-dimensional,

36 perspective view of a surface. The x and y coordinates on the base below the surface correspond

Box 10: Computing the Land UCL for Spatially Weighted Data

Let Xh X2 ,..., X„ represent the n interpolated concentrations that were obtained by spatially
weighting the m originally sampled concentrations 7i, Y2 Ym.

I n

STEP 1: Compute the mean of the log-transformed interpolated data In X = — ^ ln(X ).

z=l

STEP 2: Compute the associated standard deviation sln = J—"-j"X('n(^/) - Inx f .

STEP 3: Look up the H\-u statistic for sample size m (rather than n) and the observed standard
deviation of the log-transformed interpolated data. Tables of these values are given by
Gilbert (1987, Tables A-10 and A-12) and Land (1975).

STEP 4: Compute the one-sided (1-a) upper confidence limit on the mean accounting for
spatial weighting UCLx_a = exp(ln X + sln2 / 2 + H] tt.vln /-Jtn -1).

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-22

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

Box 11: Computing the Hall UCL for Spatially Weighted Data

Let Xh X2 X„ represent the n interpolated concentrations that were obtained by spatially
weighting the m originally sampled concentrations 7i, Y2 Ym.

— 1 "

STEP 1: Compute the mean of the interpolated data X = — IX-

n

Z=1

STEP 2: Compute the associated (biased) standard deviation 5 =



STEP 3: Compute the associated skewness k = —

ns I=1

STEP 4: For b = 1 to B (a very large number) do the following six steps:

4.1: Generate a bootstrap sample data set of size m by random sampling with
replacement from the interpolated data; i.e., for i = 1 to m let j be a random
integer between 1 and n and add the datum X, to the bootstrap sample data set.
4.2: Compute the arithmetic mean Xb of the data set constructed in Step 4.1.
4.3: Compute the associated standard deviation Sb of the constructed data set.
4.4: Compute the skewness kb of the constructed data using the formula in Step 3.
4.5: Compute the Studentized mean W= (Xb-X)/sb..

4.6: Compute the statistic Q = W + kbW213 + kbW3 /27 + kb /(6m).

STEP 5:

Sort all the Q values computed in Step 4 and select the lower ath quantile of these B
values. It is the (o(B)th value in an ascending list of Q's. This value is from the left tail
of the distribution.

STEP 6: Compute W(Qa) = —

k

3 ^ '

1-

1 + * 0«" —

6m

1/3 N

STEP 7: Compute the one-sided (1-a) confidence limit on the mean accounting for spatial
weighting UCLx_a = X + W (Qa )s .

to the spatial dimensions over which the weighting algorithm was applied. The surface appears
as 1,024 intersections of 32-by-32 grid lines. The height of the surface at each x,j'-location is the
concentration value associated with that point in space. Contour lines at the values of 50 and 100
are also shown. The largest value of concentration was 121.4 (at the upper, left-hand corner in
the graph on the left and at the corner farthest from the viewer in the graph on the right). The
smallest value was 30.7 (at the lower, right-hand corner or the nearest corner). For reference, the
formula used to create this hypothetical surface was z = 1,900 - 0.0025[(x-10)3 + (y-20)2 +
3(x-5)2 + 3(y—15)3], where x and_y both ranged among the integers [1, 32],

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-23

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17











































































































































-i































t-





























































/-





























-L

7-





















--































--































--i































-t





L.

























/-





























—



-i































\i































m

150

2§.-§@ a §§--7§ a

Figure 8 Two Displays of Spatial Variation in the Hypothetical Data Set Used to
Test the Modified UCL Formulas for Spatial Weighting

A sample of 30 values was taken at random from the data set depicted in Figure 8. The Student
UCL was computed for this sample, and this process was repeated 2,000 times. Figure 9 depicts
the resulting variation of the Student UCL values obtained from 2,000 Student UCLs, each based
on random samples of size 30. As evidenced by the dispersion of the histogram, any particular
result from a calculation involving a single random sampling of 30 points from the grid would be
relatively unreliable, ranging roughly between 87 and 103. The average of these 2,000 UCLs
was 95.09 and is shown by an arrow on the graph. Using the modified Student UCL formula that
accounts for spatially weighted data, with m = 30, the Student UCL based on all 1,024 values
was 95.12. This value is graphically indistinguishable from the mean of the 2,000 estimates.
Thus, this modification to the formula for the Student UCL appears to produce the same result as
would a more elaborate computation averaging many UCLs based on random sampling from the
surface of spatially weighted concentration estimates.

mean

0.16-

> 0.12
o
c
a)

o- 0.08 H
a)

0.04

4

0 1 1-1
85

r-n rH~



90	95

Student UCL

100

Figure 9 Histogram of Student UCL Values

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-24

-------
1

Figure 10 shows the convergence of the running average of estimated Student UCLs based on
random samples of size 30 as the number of estimates averaged together increases from 1 to
2,000. The running average converges to the value obtained from the Student UCL formula
modified to handle spatial weighting (shown as a horizontal line) based on all 1,024 values. If
the entire numerical experiment were repeated, the left side of the trajectory of this convergence
would likely be quite different, but the convergence on the right side of the figure would almost
surely persist. Thus, it is expected that making one calculation of the Student UCL with the
formula in Box 9 as modified for spatially weighted data will produce the same result as finding
the average of many calculations of the Student UCLs for randomly sampled subsets with the
appropriate degrees of freedom from the surface of concentrations. Therefore, the formula
modified for spatially weighted data may be thought of as a shortcut for conducting elaborate
sampling studies from the grid of interpolated concentrations.

c
a)
-a

3
(/)

1 100 10000

Number of estimates averaged together

Figure 10 Convergence of Running Averages of Student UCLs

To assess the effect of the modifications to handle spatial weighting on the Hall UCL formulas,
another simulation study was conducted using the same data set of 1,024 grid values. There
were 4,000 random samples of size 30 taken (with replacement) from the grid of 1,024 values in
the hypothetical data depicted in Figure 8. The Hall UCL was computed for each of these
samples. Figure 11 displays a histogram of the resulting 4,000 Hall UCLs. Again, the histogram
is rather wide, with a range of over 20 units. The mean of the 4,000 estimates, shown by a black
arrow, was 94.67. This mean agrees with the value obtained from the Hall UCL formula
modified to handle spatial weighting that was applied directly to the 1,024 values, which was
94.66. The correspondence of the mean of the Hall UCLs and the Hall UCL from the formula
modified to handle spatial weighting again suggests that the latter may be used as an acceptable
alternative for the former approach.

MK0110:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-25

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

0.2
0.16

>»

g 0.12 i
I" 0.08 i

Ll_

0.04

0

mean

I

i I I I l~T

i=m

~h-m

i—i

85

90

100

95
Hall UCL

Figure 11 Histogram of Hall UCL Values

Figure 12 illustrates the convergence of the running average of estimated Hall UCLs based on
random samples of size 30 as the number of estimates averaged together increases from 1 to
4,000. The running average converges to the value obtained from the Hall UCL formula
modified to handle spatial weighting (shown as a horizontal line) based on all 1,024 values. As
was true for the modification of the Student UCL formula for spatially weighted data, the
similarly modified version of the Hall UCL also yields a value to which the average of iterated
applications of the Hall UCL on random samples of size 30 apparently eventually converge,
although the convergence appears to be slower. The slowness may be a result of the fact that the
bootstrap calculations themselves involve random numbers.

o

1	100	10000

Number of estimates averaged together

Figure 12 Convergence of Running Averages of Hall UCLs

MK0110:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-26

-------
1

When the modified formula for the Land UCL was also applied to the spatial data set of Figure
8, very similar quantitative results were observed, in terms of the dispersion of the sample
estimates and convergence of their average, although the Land UCL estimate increased slightly
to 96.2. This simulation is unrealistic, however, in that the distribution of grid values (which is
negatively skewed) would not be considered by the analyst to be amenable to a lognormal model.
Indeed, given the IDW interpolation scheme, which computes the concentration at each grid
intersection as a weighted sum of nearby concentrations, it seems likely to tend to produce grid
values that are more normally distributed than the original data. However, to assess the effect of
the modifications to handle spatial weighting on the Land UCL formulas, another simulation
study was employed.

In this study the spatially distributed data were generated from a lognormal distribution specified
by the median e = 2.718 and geometric standard deviation exp(1.5) = 4.482. Most of the values
of this distribution are less than about 130. Figure 13 depicts the variability of the Land UCL
values observed in this simulation. The depicted histogram represents the results from 2,000
random samples of size 30 taken from the hypothetical data set of 1,024 lognormally distributed
concentrations. The geometric average of these Land UCLs was 19.69 (indicated by the black
arrow). Note that this is a geometric average of Land UCLs, which are themselves estimates of
an arithmetic average. This is relevant and reasonable because of the skewness of the
distribution of UCLs. The Land UCL from the modified method for spatially weighted data with
m = 30 was 19.37. This value is graphically indistinguishable from the geometric mean.

0.1-

iric

geometric mean

0.5 ~i

0.4-
0.3-
0.2-

~l I I—I—I—N

i i I ) i i i—i=i—i—i—i i i i—i-

O
CN

o
o

Land UCL

Figure 13 Histogram of Land UCL Values

Figure 14 shows the convergence of the running geometric average of estimated Land UCLs
based on random samples of size 30 as the number of estimates averaged together increases from

MK0110:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-27

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

1 to 2,000. The running geometric average converges to the value obtained from the modified
Land UCL formula (shown as a horizontal line) based on all 1,024 values. Thus, the
modification to the Land UCL formula to account for spatial weighting yields results that
apparently correspond to the central tendency, as estimated by the geometric mean, of the more
elaborate computational scheme based on random sampling from the surface of spatially
weighted concentration estimates.

50

40

g 30

T3

_§ 20

10
0

1	100	10000

Number of estimates averaged together

Figure 14 Convergence of Geometric Averages of Land UCLs
C.5.3.2 Pascal Software for Calculating the Hall UCL

The Pascal code for computing the Hall UCL is presented below (it requires a Pascal compiler).
To use it, place the data in the vector x. These data are the concentration values, or, if spatial
weighting is used, the interpolated values from the grid of concentrations obtained by spatial
weighting. Specify the original number of concentration values df the number n of values in the
vector x, the vector x, and the alpha4evel in the call to the function Hall UCL. To obtain a 95%
UCL, let alpha be 0.05. The code can also be used for raw data without spatial weighting by
letting df and n be the same value. In any case, df should be no smaller than n. The number of
bootstrap iterations can be changed by modifying the value of the constant bmax.

const

max = 300;
bmax = 100000;
type

float = extended;

index = 1..max;

bindex = l..bmax;

vector = array[index] of float;

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC	£ ^ 2g	7/10/2003

-------
1

bvector = array[bindex] of float;

function getmean(n:integer; const x:vector):float;
var s : float; i : integer;
begin
s := 0.0;

for i := 1 to n do s := s + x[i];

getmean := s / n;

end;

function getstddev(n:integer; xbar:float; const x:vector):float;
var s : float; i : integer;
begin
s := 0.0;

for i := 1 to n do s := s + (x[i] - xbar) * (x[i] - xbar);

getstddev := sqrt(s / (n-1));

end;

function getQstddev(n:integer; xbar:float; const x:vector):float;
var s : float; i : integer;
begin+ s := 0.0;

for i := 1 to n do s := s + (x[i] - xbar) * (x[i] - xbar);

getQstddev := sqrt(s / n);

end;

function getskew(n:integer; xbar, stddev:float; const x:vector):float;
var s,s3 : float; i : integer;
begin
s := 0.0;

s3 := stddev * stddev * stddev;

for i:=l to n do s:=s+(x[i] - xbar)*(x[i] - xbar)*(x[i] - xbar)/s3;

getskew := s / n;

end;

procedure qsort(var a: bvector; lo,hi: integer);
procedure sort(1,r: integer);
var i,j : integer; x,y: float;
begin

i:=1; j:=r; x:=a[(l+r) div 2];
repeat

while a[i]j;
if l
-------
1

for i := 1 to some do
begin

j := random(n) + 1;
y [i] : = x [ j ] ;
end;
end;

function Hall UCL(n, df:integer; const x:vector; alpha:float) : float;
var

xbar, stddev, skew, bxbar, bstddev, bskew, k, w, q, a : float;
b : integer; bx : vector; qq : bvector;
begin

for i := 1 to bmax do qq[i] := 0;
xbar := getmean(n,x);
stddev := getQstddev(n,xbar,x);
skew := getskew(n,xbar,stddev,x);
for b := 1 to bmax do
begin

bootsample(df,n, x, bx) ;
bxbar := getmean(df, bx) ;
bstddev := getQstddev(df,bxbar,bx);
k := getskew(df,bxbar,bstddev, bx) ;
w := (bxbar - xbar) / bstddev;

q := w + skew*w*w / 3 + k*k*w*w*w / 27 + k / (6*df{n});

qq[b] := q;

end;

qsort(qq,1,bb);
q := qq[round(alpha * bb)];
a := (1 + skew*(q - skew/(6 * df{n})));
if aa=0.0

then w := 3 / skew

else w := 3 * (1 - exp(ln(a) / 3)) / skew;

Hall UCL := xbar + w * stddev;
end;

C.5.4 Outliers

Outliers are values in a data set that are not representative of the set as a whole, usually because
they are very large relative to the rest of the data. When outliers arise from unreliable
observations, they may need to be removed before the data set is subjected to statistical analysis.
On the other hand, outliers may represent true extreme values from a distribution of
concentrations that vary widely across space. Removing accurate data with large values and
failing to remove outliers that arise from mismeasurement are opposite kinds of errors that will
both lead to a distorted estimate of the EPC.

There are a variety of statistical tests for determining whether one or more observations are
outliers in a statistical sense (EPA 2000, Section 4.4). These tests should be used judiciously,
however. It is common in concentration data that the distribution is strongly skewed so that it
contains a few very high values corresponding to local hot spots of contamination. To estimate
the average EPC correctly it is important to take account of these values. Therefore, one should
be careful not to exclude values merely because they are large. However, sometimes outliers
arise from instrumentation or other laboratory errors, typographical or transcription errors,
computational mistakes, or other breaches in the sample collection or measurement protocols.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-30

-------
1 Scientific review of the generation, manipulation, and documentation history of a data value may

2 uncover clear or possible problems that suggest a value is unreliable. (Good statistical practice

3 would subject all data to such review and not just those whose values are large.) EPA guidance

4 suggests that, when outliers are suspected of being unreliable and statistical tests show them to

5 be unrepresentative of the underlying data set, any subsequent statistical analyses should be

6 conducted both with and without the outlier(s) and that the entire process, including

7 identification, statistical testing, and review of outliers, be fully documented.

8 C.5.5 References

9 Aitchison, J. and J.A.C. Brown. 1969. The LognormalDistribution. Cambridge University Press,

10 Cambridge.

11 Allen, A.O. 1990. Probability, Statistics and Queueing Theory with Computer Science

12 Applications. Second edition. Academic Press, Boston, MA.

13 Birnbaum, Z.W. 1952. Numerical tabulation of the distribution of Kolmogorov's statistic for

14 finite sample size. Journal of the American Statistical Association 47:425.

15 Chen, L. 1995. Testing the mean of skewed distributions. Journal of the American Statistical

16 Association 90:767-772.

17 Crow, E.L., F.A. Davis and M.W. Maxfield. 1960. Statistics Manual with Examples Taken from

18 Ordnance Development. Dover Publications, New York.

19 Efron, B. 1982. The Jackknife, the Bootstrap and Other Resampling Plans. SIAM, Philadelphia,

20 PA.

21 EPA (U.S. Environmental Protection Agency). 1989. Risk Assessment Guidance for Superfund,

22 Volume I - Human Health Evaluation Manual (Part A). Interim Final. EPA/540/1-89/002.

23 http://www.epa.gov/superfund/programs/risk/ragsa/. Office of Emergency and Remedial

24 Response, U.S. Environmental Protection Agency, Washington, DC.

25 EPA (U.S. Environmental Protection Agency). 1992. A Supplemental Guidance to RAGS:

26 Calculating the Concentration Term. http://www.deq.state.ms.us/newweb/opchome.nsf/pages/

27 HWDivisionFiles/$file/uclmean.pdf. Publication 9285.7-081. Office of Solid Waste and Emergency

28 Response, U.S. Environmental Protection Agency, Washington, DC.

29 EPA (U.S. Environmental Protection Agency). 1997. Data Quality Evaluation Statistical

30 Toolbox (DataQUEST) User's Guide, http://www.epa.gov/quality/qs-docs/g9d-final.pdf, EPA

31 QA/G-9D QA96 Version, U.S. Environmental Protection Agency, Office of Research and

32 Development, Washington, DC. [The software is available at http://www.epa.gov/quality/qs-

33 docs/dquest96.exe]

34 EPA (U.S. Environmental Protection Agency). 2000. Guidance for Data Quality Assessment:

35 Practical Methods for Data Analysis, http://www.epa.gov/rl0earth/offices/oea/epaqag9b.pdf,

36 EPA QA/G-9, QAOO Update, U.S. Environmental Protection Agency, Office of Environmental

37 Information, Washington, DC.

38 EPA (U.S. Environmental Protection Agency). 2002. Calculating Upper Confidence Limits for

39 Exposure Point Concentrations at Hazardous Waste Sites. Draft. OSWER 9285.6-10 Office of

40 Emergency and Remedial Response, Washington, DC. December 2002.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-

-------
1 Feller, W. 1948. On the Kolmogorov-Smirnov limit theorems for empirical distributions. Annals

2 of Mathematical Statistics 19:177-189.

3 Gilbert, R.O. 1987. Statistical Methods for Environmental Pollution Monitoring. Van Nostrand

4 Reinhold, New York.

5 Hall, P. 1988. Theoretical comparison of bootstrap confidence intervals. Annals of Statistics

6 16:927-953.

7 Hall, P. 1992. On the removal of skewness by transformation. Journal of the Royal Statistical

8 Society B 54:221-228.

9 Johnson, N.J. 1978. Modified t-tests and confidence intervals for asymmetrical populations. The

10 American Statistician 73:536-544.

11 Koch, G.S., Jr., and R.F. Link. 1980. Statistical Analyses of Geological Data. Volumes I and II.

12 Dover, New York.

13 Kolmogorov [Kolmogoroff], A. 1933. Sulla determinatione empirica di una legge di

14 distributione. Giornale dell'IstitutoItaliano deli Attuari 4:83-91.

15 Kolmogorov [Kolmogoroff], A. 1941. Confidence limits for an unknown distribution function.

16 Annals of Mathematical Statistics 12:461-463.

17 Land, C.E. 1971. Confidence intervals for linear functions of the numeral mean and variance.

18 Annals of Mathematical Statistics 42:1187-1205.

19 Land, C.E. 1972. An evaluation of approximate confidence interval estimation methods for

20 lognormal means. Technometrics 14:145-158.

21 Land, C. E. 1975. Tables of Confidence Limits for Linear Functions of the Normal Mean and

22 Variance. Selected Tables in Mathematical Statistics, Volume III. p 385-419.

23 Lockheed Martin. 2000. ProUCL. [software for Windows 95, accompanied by draft

24 documentation "Background and a brief description of the program PROUCL," Installation

25 Guide: Program PROUCL, and User's Guide: Program PROUCL.]

26 Manly, B.F.J. 1997. Randomization, Bootstrap, and Monte Carlo Methods in Biology. Second

27 edition. Chapman and Hall, London.

28 Miller, L.H. 1956. Table of percentage points of Kolmogorov statistics. Journal of the American

29 Statistical Association 51:111-121.

30 Savage, I.R. 1961. Probability inequalities of the Tchebycheff type. Journal of Research of the

31 National Bureau of Standards -B. Mathematics and Mathematical Physics 65B :211 -222.

32 Schmoyer, R.E., J.J. Beauchamp, C.C. Brandt, and F.O. Hoffman. 1996. Difficulties with the

33 lognormal model in mean estimation and testing. Environmental and Ecological Statistics 3:81-

34 97.

35 Schulz, T.W. and S. Griffin. 1999. Estimating risk assessment exposure point concentrations

36 when the data are not normal or lognormal. Risk Analysis 19:577-584.

37 Singh, A.K., A. Singh, and M. Engelhardt. 1997. The Lognormal Distribution in Environmental

38 Applications. EPA/600/R-97/006.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-

-------
1 Singh, A.K., A. Singh, and M. Engelhardt. 1999. Some Practical Aspects of Sample Size and

2 Power Computations for Estimating the Mean of Positively Skewed Distributions in

3 Environmental Applications. EPA/600/s-99/006.

4 Singh, A.K., A. Singh, M. Engelhardt, and J. Nocerino. 2000. On the Computation of the Upper

5 Confidence Limit of the Mean of Contaminant Data Distributions. Submitted for publication.

6 Smirnov [Smirnoff], N. 1939. On the estimation of the discrepancy between empirical curves of

7 distribution for two independent samples. Bulletin de I'Universite de Moscou, Serie

8 internationale (Mathematiques) 2: (fasc. 2).

9 Sokal, R.R. and F.J. Rohlf. 1981. Biometry. Freeman, San Francisco, CA.

10 Student [W.S. Gossett], 1908. On the probable error of the mean. Biometrika 6:1-25.

11 Wong, A. 1993. A note on inference for the mean parameter of the gamma distribution. Stat.

12 Prob. Lett. 17:61-66.

13 Zhou, X.-H. and S. Gao. 2000. One-sided confidence intervals for means of positively skewed

14 distributions. The American Statistician 54:100-104.

MK01 |O:\20123001.096\ERA_PB\ERA_APC5_PB.DOC

C.5-

-------
APPENDIX C.6
APPLICATIONS OF A WEIGHT-OF-EVIDENCE APPROACH

O:\20123001.096\ERA PB\ERA APC6 PB.doc

-------
1

APPENDIX C.6
APPLICATIONS OF A WEIGHT-OF-EVIDENCE APPROACH

C.6.1 Introduction

Conclusions reached in the risk characterization component of an ecological risk assessment are
usually reached by weighing multiple lines of evidence rather than by using traditional scientific
standards of proof that typically are not available to the risk assessor (EPA 1992). The weight of
evidence approach, as defined by Menzie et al. (1996), is a process by which measurement
endpoints are related to an assessment endpoint to evaluate whether significant harm or risk is
posed to the environment.

The approach taken can range from a simple qualitative assessment to a highly quantitative
evaluation; however, no matter what form the weight of evidence process takes, it should provide
clear and transparent documentation of each step of the thought process used by the risk
assessors in characterizing risk. In addition, the process should include open communication
with stakeholders to ensure that the end result is understandable by all.

The term "line of evidence" as used in this discussion follows the definition provided in
Guidelines for Ecological Risk Assessment (EPA 1998): "Information derived from different
sources or by different techniques that can be used to describe and interpret risk estimates."
Unlike the term "weight of evidence," a line of evidence does not imply assignment of
qualitative or quantitative weights to information, but simply different types of measurement
endpoints. According to Glenn Suter and colleagues, there are four general lines of evidence
under which most measurement endpoints fall (Hull and Suter 1994; Suter et al. 1995, 2000):

¦ Biological survey data that indicate the conditions observed in the environment,

¦ Media toxicity data that indicate whether the contaminated media are toxic (e.g.,
laboratory or in situ toxicity testing),

¦ Single chemical toxicity data that indicate the toxic effects of the concentration
measured in site media, and

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 1

-------
1 ¦ Biomarkers and pathologies that may be indicative of exposure to or effects of

2 contaminants.

3 The use of quantitative or qualitative weighing of evidence and/or professional judgment about

4 which lines of evidence are most reliable is based on the implicit assumption that the lines of

5 evidence are logically independent (Suter et al. 2000). In this paper, the approaches outlined in

6 the Massachusetts (MA) Weight of Evidence Special Report (Menzie et al. 1996) and the process

7 followed by the ecological work group at Oak Ridge National Laboratory (ORNL) (Suter et al.

8 1995, 2000; Suter 1996, 1997) are reviewed. The approach that will be incorporated in the

9 Housatonic River ecological risk assessment is then described.

10 C.6.2 Available Approaches

11 A brief overview of the MA and ORNL approaches for conducting a weight of evidence analysis

12 is provided below. A more detailed presentation of these methodologies is provided in the

13 source literature.

14 C.6.2.1 Option 1 - Massachusetts Weight of Evidence Workgroup Approach

15 (Menzie et al. 1996)

16 The Massachusetts Weight-of-Evidence Workgroup Approach (hereafter, MA Approach)

17 evaluates three characteristics of measurement endpoints: (a) the weight assigned to each

18 measurement endpoint; (b) the magnitude of response observed in the measurement endpoint;

19 and (c) the concurrence among outcomes of multiple measurement endpoints, for a given

20 assessment endpoint. The following discussion provides further detail on these three key

21 components. In this discussion, the terms "measurement endpoint" and "line of evidence" are

22 considered synonymous.

23 Determination of the Weight Assigned to a Measurement Endpoint

24 Measurement endpoints are organized for each assessment endpoint to be evaluated. An

25 evaluation is then conducted of each measurement endpoint based on eleven attributes (Table 1)

26 that can be grouped into the following three categories: (a) strength of association between

27 assessment and measurement endpoints; (b) data quality; and (c) study design and execution.

28 Attributes are characteristics of measurement endpoints that can be used to assess the overall

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 2

-------
1 strength of measurement endpoints. There are three steps in assigning a weight to measurement

2 endpoints: (1) individual attribute scaling; (2) measurement endpoint attribute scoring; and (3)

3 overall measurement endpoint weighting.

4 Attribute Scaling — Determining the relative importance of individual attributes is a subjective

5 process involving evaluation of available data and professional judgment; this process is referred

6 to as attribute scaling. Attribute scaling involves assigning quantitative (e.g., numerical, 1-5) or

7 qualitative (e.g., high, medium, or low) weights to each attribute. While it is expected that the

8 eleven attributes will be scaled consistently throughout the ERA, if deemed appropriate, the scale

9 applied to an attribute may be adjusted for different endpoints. One simplifying approach

10 considers the specific attributes in assigning a weight to each of the three categories rather than

11 weighting the eleven attributes individually. If all attributes are scaled equally, then the first step

12 does not influence the weight assigned to a measurement endpoint.

13 Attribute Scoring — The next step involves establishing the weight (quantitative or qualitative)

14 for each measurement endpoint based upon an evaluation of each of the scaled attributes; this

15 process is called attribute scoring. The MA Approach recommends that consideration to

16 attribute weights be given during the problem formulation process so that appropriate

17 measurement endpoints are selected. The final measurement endpoint weighting is conducted as

18 part of the risk characterization and includes a thorough evaluation of how well the measurement

19 endpoint performed in relation to each attribute or attribute category.

20 Overall Weight of the Measurement Endpoint — This step of the process determines the

21 overall quantitative or qualitative value assigned to the measurement endpoint.

22 Table 2 provides a condensed example of a qualitative approach to assigning weights to

23 measurement endpoint attributes.

24 Magnitude of Response

25 The magnitude of response is considered together with the weight in judging the overall weight

26 of evidence for a measurement endpoint. The evaluation of the magnitude of response is divided

27 into two questions:

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 3

-------
1

1. Does the measurement endpoint indicate the presence of harm (yes, no, or
undetermined)?

2. Is the magnitude of response low or high (where the response to question one is yes or
no)?

While the options presented for these questions are discrete functions, it is recognized that
responses are more likely to occur along continuous gradients and should be presented as such in
the discussion portion of the risk characterization. However, it was agreed to by the Massachusetts
Weight-of-Evidence Workgroup that discrete categories accompanying a detailed analysis would
more clearly communicate results to risk managers and other interested parties.

When evaluating the magnitude of response, each measurement endpoint should be accompanied
by a discussion detailing the basis of the determination. The MA Approach recommends one or
more of the following metrics to be used to evaluate the magnitude of response:

¦ A change or difference in the response variable that is considered potentially
ecologically relevant.

¦ Spatial scale of the change or difference, as related to the assessment endpoint.

¦ Temporal scale of the change or difference, as related to the assessment endpoint.

The level of response considered indicative of environmental harm or risk, with respect to the
assessment endpoint, should be determined prior to assessing the magnitude of effects.

Table 3 presents a summary matrix for a given measurement endpoint that provides a simple
communication tool and documents the risk assessor's conclusions regarding the magnitude of
response along with the weight of the measurement endpoint.

Concurrence Among Outcomes

Concurrence among measurement endpoints is evaluated by plotting the findings of the two
preceding steps on a matrix for all measurement endpoints associated with a given assessment
endpoint (see Figure 1). The matrix allows easy visual examination of agreements or
divergences among measurement endpoints, facilitating interpretation with respect to the
assessment endpoint. Logical connections, interdependence, and correlations among endpoints
should also be considered when evaluating concurrence.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 4

-------
1

C.6.2.2 Option 2 - ORNL Approach (Suteret al. 1995, 2000; Suter 1996, 1997)

The ORNL approach is summarized in Figure 2. This approach begins by assembling the
available lines of evidence for each assessment endpoint. At this point an estimation of the risk
has been established for each line of evidence, so the process of weighting the evidence amounts
to determining what estimate of risk is most likely. If all the lines of evidence are consistent
(i.e., all lines indicate effects or all lines indicate no effects), the result of the weighting of
evidence is clear. Provided there is no bias associated with one or more lines of evidence related
to a given assessment endpoint, agreement among several lines of evidence indicates strong
support for a conclusion (Suter et al. 2000).

Table 4 is an example of a simple summary of results using the ORNL weight of evidence
process. The "evidence" column provides a brief description of the line of evidence being
evaluated. In the "results" column, symbols are used to indicate significant effects: "+" if the
evidence is consistent with significant effects to the assessment endpoint; if the evidence is
consistent with insignificant or no effects; and "±" if the evidence is too ambiguous to assign to
either category. The "explanation" column presents a summary of the results of the risk
characterization for that line of evidence. The last line of the table presents the weight of
evidence based conclusion concerning whether significant effects are occurring and a brief
statement on the basis for this conclusion.

If there are inconsistencies in the results, a more in-depth weighting of evidence must occur. If
required, weights are determined based on six attributes: relevance, exposure/response, temporal
scope, spatial scope, quantity, and quality. These attributes are similar to those proposed in the
MA Approach and are described in Table 5. To present this weight of evidence method, a
column for "weight" is added to a table such as Table 4, to judge the relative reliability and
credibility of the conclusions for that line of evidence (see Table 6). While a qualitative
weighting approach is presented in Table 6, a more formalized numerical scoring system could
be developed.

The ORNL approach, while structurally similar to the MA Approach, differs in that the ORNL
approach reserves the whole weight of evidence process for the risk characterization phase,
rather than giving consideration and potential weights in the problem formulation phase.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 5

-------
1 C.6.3 Approach

2 The MA Approach will be used as the primary vehicle for conducting the weight of evidence

3 analysis for the Housatonic River ERA. In general, a qualitative approach will be used for

4 assigning weight to measurement endpoints. More quantitative approaches (e.g., meta analysis)

5 may be used where warranted (e.g., to help sort out complex scenarios or to better understand

6 risk).

7 C.6.4 References

8 EPA (U.S. Environmental Protection Agency). 1998. Guidelines for Ecological Risk Assessment.

9 Risk Assessment Forum. Washington, DC. EPA/630/R-95/002F.

10 EPA (U.S. Environmental Protection Agency). 1992. Framework for Ecological Risk

11 Assessment. Risk Assessment Forum. Washington, DC. EPA/630/R-92/001.

12 Hull, R.N. and G.W. Suter II. 1994. Using the Weight-of-Evidence Approach for Ecological Risk

13 Assessment at a DOE Facility. SET AC Denver, Colorado. November 1, 1994.

14 Menzie, C., M.H. Henning, J. Cura, K. Finkelstein, J. Gentile, J. Maughan, D. Mitchell, S.

15 Petron, B. Potocki, S. Svirsky, and P. Tyler. 1996. "Special Report of the Massachusetts Weight-

16 of Evidence Workgroup: A Weight-of Evidence Approach for Evaluating Ecological Risks."

17 Hum. Ecol. Risk. Assess. 2(2):277-3 04.

18 Suter, G.W., II, R.A. Efroymson, B.E. Sample, and D. S. Jones. 2000. Ecological Risk

19 Assessment for Contaminated Sites. Lewis Publishers. Boca Raton, Florida.

20 Suter, G.W., II. 1997. A Framework for Assessing Ecological Risks of Petroleum-Derived

21 Materials in Soils. Oak Ridge National Laboratory. Oak Ridge, Tennessee. Environ. Sci. Div.

22 Publ. No.4666.

23 Suter, G.W., II. 1996. Risk Characterization for Ecological Risk Assessment of Contaminated

24 Sites. Oak Ridge National Laboratory. Oak Ridge, Tennessee. ES/ER/TM-200.

25 Suter, G.W., II, B.E. Sample, D.S. Jones, T.L. Ashwood, and J.M. Loar. 1995. Approach and

26 Strategy for Performing Ecological Risk Assessments for the U.S. Department of Energy's Oak

27 Ridge Reservation. Oak Ridge National Laboratory. Oak Ridge, Tennessee. ES/ER/TM-33/R2.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 6

-------
Table 1

Menzie et al. (1996) Attributes for Evaluating Measurement Endpoints

Category/Attribute

Description

I. Relationship Between Measurement and Assessment Endpoint

¦ Degree of Association

The extent to which the measurement endpoint is representative of, correlated with, or applicable to the assessment endpoint; in particular, with
respect to similarity of effect, target organ, mechanism of action, and level of ecological organization.

¦ Stressor/Response

The ability of an endpoint to demonstrate effects from chronic exposure to the stressor and to correlate the effects with the degree of exposure,
susceptibility, and magnitude of effect.

¦ Utility of Measure

The ability to judge results of the study against well-accepted standards, criteria, or objective measures (e.g., sediment quality criteria and
toxicity thresholds). As such, the attribute describes the applicability, certainty, and scientific basis of the measure, as well as the sensitivity of
a benchmark in detecting environmental harm.

II. Data Quality

¦ Quality of Data

The degree to which data quality objectives (DQOs) and other recognized characteristics of high quality studies are met. The appropriateness
of data collection and analysis practices, as well as the implementation of the experimental design and the minimization of confounding factors
strongly influence the data quality.

III. Study Design

¦ Site Specificity

The extent to which chemical and/or biological data, media, species, environmental conditions, and habitat types used in the study reflect the
site of interest.

¦ Stressor Specificity

The degree to which the measurement endpoint is associated with the specific stressor(s) of concern. Some measurement endpoints respond to
a broad range of stressors, complicating interpretation of results, while others are more specific to a particular stressor.

¦ Sensitivity

The ability to detect a response in the measurement endpoint, expressed as a percentage of the total possible variability that the endpoint is able
to detect. Additionally, this attribute reflects the ability of the measurement endpoint to discriminate between responses to a stressor and those
resulting from natural or design variability and uncertainty.

¦ Spatial Representativeness

The degree of compatibility or overlap between the study area and locations of measurements or samples, stressors, ecological receptors, and
potential exposure points.

¦ Temporal

Representativeness

The compatibility or overlap between when data were collected or the period for which data are representative and the period during which
effects of concern would be likely to be detected. Also linked to this attribute is the number of measurement or sampling events and the
expected variability over time.

¦ Quantitativeness

The degree to which numbers can be used to describe the magnitude of response to the stressor. Some measurement endpoints yield qualitative
or hierarchical results, while others are more quantitative. In addition, this attribute encompasses the extent to which biological significance
can be interpreted from statistical significance.

¦ Use of a Standard Method

The extent to which the study follows protocols recommended by a recognized scientific authority (e.g., study designs or chemical measures
published in the Code of Federal Regulations, developed by ASTM, or repeatedly published in the peer-reviewed scientific literature). This
attribute also reflects the suitability and applicability of the method to the endpoint and the site, as well as any modifications made to the
method.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 7 7/10/2003

-------
Table 2

Example Measurement Endpoint Weighting

Score each measurement endpoint from low to high

Assessment Endpoint:

Category/Attribute

Attribute Scaling
Factor* (weight)

Measurement Endpoint A

Attribute
Scoring

Rationale

I. Relationship between Measurement and Assessment Endpoints

¦ Degree of Association

High

Medium

Measurement and assessment endpoints directly linked; however, level of organization
different.

¦ Stressor/Response

Medium

High

Correlation between stressor and response is statistically significant.

¦ Utility of Measure

Medium

Measure was developed by a third party and is well accepted; however, benchmark is
relatively insensitive.

II. Data Quality

¦ Quality of data

High

DQOs are rigorous and comprehensive. All DQOs met.

III. Study Design

¦ Site-specificity

Medium

High

Data, media, species, environmental conditions, benchmarks, and habitat type are derived
from or reflect the site.

¦ Stressor-specificity

Medium

Similar stressors known to affect measurement endpoint in the manner expected.

¦ Sensitivity

Medium

Changes of 10-100 times can be detected.

¦ Spatial representativeness

Low

Medium

The study area, sampling/measurement site, stressors overlap well; receptor and points of
potential exposure overlap is limited.

¦ Temporal representativeness

Low

Data during one season only and is not collected during breeding season - the period
during which effects are expected.

¦ Quantitativeness

Low

High

Results are quantitative and can be tested for statistical and biological significance.

¦ Use of a standard method

Low

Medium

The standard method was modified to be more site-specific.

Total Score

—

Medium

...

*Attribute scaling factor to be determined by ecorisk team consensus. Adapted from scaling presented in Menzie et al. 1996.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 8 7/10/2003

-------
Table 3

Scoring Sheet for Evidence of Harm and Magnitude

Assessment Endpoint:

Measurement Endpoints

Weighting Score
(high, medium or low)

Evidence of Harm
(Y es/N o/Undetermined)

Magnitude
(High/Low)

Endpoint A

Endpoint B

Endpoint C

Adapted from: Menzie et al. 1996.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 9

-------
Table 4

Example of a Simple Summary of a Risk Characterization by Weight of Evidence for a Soil Invertebrate Community

Evidence

Results*

Explanation

Biological Surveys

Soil microarthropod taxonomic richness is within the range of reference soils
and is not correlated with concentrations of petroleum components.

Ambient Toxicity Tests

No reduction in the survival of the earthworm Eisenia foetida. Sublethal
effects were not determined.

Organism Analyses

Concentrations of PAHs in depurated earthworms were elevated relative to
worms from reference sites, but toxic body burdens are unknown.

Soil Analyses/Single
Chemical Tests

If the total hydrocarbon content of the soil is assumed to be composed of
benzene, then earthworm mortality is expected. Toxicity data for other
detected contaminants are unavailable.

Weight-of-Evidence

Although earthworm tests may not be sensitive, they and the biological surveys
are both negative and are both more reliable than the comparison of single
chemical toxicity data with soil analytical results.

* + Evidence is consistent with the occurrence of a 20% reduction in species richness or abundance of the invertebrate community.
- Evidence is inconsistent with the occurrence of a 20% reduction in species richness of abundance of the invertebrate community.
+ Evidence is ambiguous.

Adapted from: Suter 1996.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 10

-------
Table 5

ORNL Attributes for Judging Measurement Endpoints

Attribute

Weighting Considerations

Relevance

Evidence is given more weight if the measure of effect is more directly related to (i.e., relevant to) the assessment endpoint. Is the
measure of effect a direct estimate of the assessment endpoint? Have validation studies demonstrated that the measurement endpoint is
predictive of the assessment endpoint? Is the mode of exposure similar to the site media? Is the chemical form tested similar to that found
on-site?

Exposure/Response

Is there a relationship between the magnitude of exposure and the effects?

Temporal Scope

Do the data account for the relevant range of temporal variance in conditions?

Spatial Scope

Do data adequately represent the area to be assessed, including directly contaminated areas, indirectly contaminated areas, and indirectly
affected areas?

Quality

Data quality is evaluated in terms of the study protocols for and execution of sampling, analysis, and testing. Are the methods appropriate
and implemented well?

Quantity

Is the number of observations adequate relative to the natural variation, variations in analysis, and to potential sampling biases?

Uncertainty

Lines of evidence with less uncertainty are given greater weight. Uncertainty in a risk estimate may be a function of the data quality and
quantity; however, the major source of uncertainty is the extrapolations between the measures of effect and the assessment endpoint. In
addition, the uncertainty in extrapolating from the measures of exposure to actual receptors may be considerable with regards to
bioavailability and temporal dynamics.

Adapted from: Suter et al. 2000.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 1 1 7/10/2003

-------
Table 6

Example Summary of a Risk Characterization for a Fish Community

Evidence

Result3

Weightb

Explanation

Biological Surveys

Fish community productivity and species richness in reach 2 are high relative
to reference reaches. Data are abundant and of high quality.

Ambient Toxicity
Tests

High lethality to fathead minnow larvae in a single test at Site 3.3, but
variability is too high for standard statistical significance. No other aqueous
toxicity was observed.

Water Analyses/Single
Chemical Tests

Only Zn is believed to be potentially toxic in water and only to highly sensitive
species. Few water samples were analyzed.

W eight-of-Evidence

Reach 2 supports a high quality fish community. Other evidence that suggests
toxic risks is much weaker (single chemical toxicology) or inconsistent and
weak (ambient toxicity tests).

a + Indicates that the evidence is consistent with the occurrence of the endpoint effect.
- Indicates that the evidence is inconsistent with the occurrence of the endpoint effect.
+ Indicates that the evidence is too ambiguous to interpret.
b Weights assigned to individual lines of evidence: high (H), moderate (M), and low (L).

Adapted from: Suter 1996.

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 12

-------
Assessment Endpoint: Maintenance of a benthic community that can serve as a prey base for local fish

Harm/Magnitude

Low Weight

Medium Weight

High Weight

Yes/High

Yes/Low

Undetermined

No/Low

No/High

Notes:

Use letter designations to place measurement endpoints in boxes.

~ Indicates increasing confidence or weight.

Adapted from: Menzie et al. 1996.

Figure 1 Example of Qualitative Assessment

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 1 3 7/10/2003

-------
Figure 2 Risk Characterization Based on Weighting
Multiple Lines of Evidence

O:\20123001.096\ERA_PB\ERA_APC6_PB.doc C 6 14 7/10/2003

-------
APPENDIX C.7
ANALYSIS OF PCB CONGENER COMPOSITION

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

-------
1

APPENDIX C.7

CONGENER EVALUATION

4 C.7.1 Introduction

5 PCB mixtures in environmental samples consist of combinations of up to 209 distinct chemical

6 structures (i.e., congeners). Because congener patterns affect the environmental fate and

7 toxicological properties of the total PCB mixture, the degree to which congener patterns change

8 over space, time, and/or media is of interest for ecological and human health risk assessment.

9 There are multiple processes that can result in alterations of the composition of PCB mixtures.

10 Two broad categories of fate processes are defined below:

11 Degradation — This term refers to an active process of dechlorination, metabolism, or other

12 breakdown of parent materials. The environmental persistence of PCBs means that chemical

13 breakdown of PCB parent materials is very slow. However, under certain conditions, PCBs can

14 be progressively dechlorinated or metabolized by organisms. Therefore, in the context of this

15 report, degradation refers mainly to biological processes that attack the PCB molecular structure,

16 resulting in formation of daughter products.

17 Transformation — This term refers to selective enrichment or depletion of congeners in PCB

18 mixtures, as governed by chemical properties and physical transport of the material.

19 Transformation is therefore controlled by environmental partitioning behavior and does not alter

20 the structure of individual PCB molecules.

21 Alterations in PCB profiles can occur over time, over space (i.e., with distance from source of

22 contamination), or across environmental media. The term "weathering" is often used to describe

23 alterations in composition of parent materials over time. In this report, weathering is used in a

24 broad context, meaning that it encompasses both abiotic and biotic forces affecting the congener

25 distributions. In this context, weathering accounts for both degradation and transformation

26 processes described above.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-1

7/10/2003

-------
1

C.7.1.1 Purpose

The overall purpose of this appendix is to evaluate the composition of the PCB congener
mixtures in Housatonic River environmental media to help determine the extent that the
congener mixtures in the samples are similar over space and/or various environmental media
(reflecting similar sources and similar potential risks) or conversely to demonstrate that the
congener fingerprints differ significantly among the samples.

Specifically, this appendix addresses the following questions:

¦ Are there differences in congener distributions across environmental media (both
abiotic and biotic), and if so, how large are the differences?

¦ Are there differences in congener distributions with distance from the source material
(i.e., GE facility), and if so, are these differences consistent with expected
environmental degradation of the samples?

¦ Do the differences in analytical techniques/laboratories used in the measurement of
PCB congeners affect the interpretations of the above?

C.7.1.2 Categories of Analysis

Three levels of analysis were applied, with metrics chosen specific to each type:

¦ Overall congener profile - Multivariate methods were used to assess whether the
overall congener profile of the PCB mixture differed across space and/or media. Due
to differences in PCB quantitation across samples, such as variations in the lists of
congeners quantified by laboratories, it was necessary to restrict the analysis to a
subset of the 209 congeners. It was also necessary to group congeners, due to co-
elution of congeners in the laboratory analysis.

¦ Concentrations of individual (indicator) congeners - Individual congeners (or
coeluting groups) were selected for a more refined analysis. The selection of
indicator congeners was made based on available literature on weathering of PCBs
(Attachment 1).

¦ Concentrations and patterns of homologs - Because both the multivariate and
individual congener approaches relied on assessment of subsets of the total PCB
mixture, an analysis was conducted on the homolog profiles to ensure that the
evaluation was not biased by the choice of congener subset. The term homolog refers
to a group of PCBs for which each congener contains the same "level" of
chlorination, ranging from one (i.e., monochlorobiphenyl) to ten (i.e.,
decachlorobiphenyl).

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-2

-------
1

The weight-of-evidence provided by these three approaches provides a high degree of confidence
that differences in congener profiles, if present, will be detected.

C.7.2 Methods

C. 7.2.1 Overview of Statistical Methods

A variety of qualitative and quantitative techniques were applied, including:

¦ Graphical Analysis - Profile plots are bar charts showing the individual congeners (or
congener groups) on the x-axis and the percent composition of each on the y-axis.
All groups are shown on the same x-axis scale so that pattern differences can easily
be discerned. PCB concentrations were normalized to the calculated or estimated total
PCB concentration (i.e., sum of all congeners).

¦ Multivariate Methods - Euclidean distances are an objective measure of the
magnitude of difference in composition across sampling locations and species.
Euclidean distance is a common tool for describing the similarities between samples
in exploratory multivariate techniques such as cluster analysis or multi-dimensional
scaling (see for example, Johnson and Wichern 1992). These distances were placed
into context by comparing the samples to common Aroclor formulations (1254, 1260)
and site non-aqueous phase liquid (NAPL). Centroids were calculated for each group
as the average composition of the group (i.e., a composition vector where each
element is the arithmetic average of the congener percentages averaged over all
samples for the group). Principal components analysis (PCA) was also conducted on
the congener mixtures.

¦ Univariate Statistics - evaluation of concentrations of homologs or indicator
congeners. Measures of central tendency (e.g., mean, median) and univariate
hypothesis tests (e.g., ANOVA) were used to assess potential trends.

C.7.2.2 Data Sets Evaluated

C.7.2.2.1 Ambient Environmental Media

The analysis considered available congener data in the PSA and upstream reference areas, from
multiple media and laboratories. Because laboratories use different methods resulting in
different sets of congeners and co-eluting congener sets, the analysis was limited to the three
primary laboratories used by EPA and GE for congener analysis:

¦ Geochemical and Environmental Research Group (GERG).

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC 7/10/2003

-------
1

¦ Pacific Analytical Chemistry (PAC).

¦ Northeast Analytical, Inc. (NEA).

In order to compare congener patterns in data from all three labs, it was necessary to reduce the
set of congeners to those that were common to all three labs, combining individual congeners
into co-eluting sets of congeners as necessary. This resulted in a final set of 23 congener groups,
as displayed in Table 1.

Media were selected to provide a broad range of sample types and trophic levels (Table 2).
Surface water data were not included in the congener evaluation; instead porewater was the only
aqueous medium considered. Aqueous PCB congener concentrations reflect primarily
differences in congener composition related to the physical properties of liquid versus solid
media, rather than indications of environmental transformation of the samples. QEA/BBL
(2003) demonstrated that the average chlorination level is greater in solid media (sediments)
relative to porewater. The mean number of chlorines per biphenyl ring ranged from 5.8 to 6.2 in
sediments, compared to 5.2 to 5.5 in porewater (using reach averages from the 2001 EPA/GE
partitioning data). The congener evaluation herein included porewater samples, but focused
mainly on solid media (soil, sediment, and tissue samples). The congener evaluation also
emphasized data collected from 1998 to 2002; temporal trends were not explicitly evaluated
within this time frame, but are not expected to be significant given the high environmental
persistence of the PCB mixtures observed.

Soils and Sediments — Samples were screened to include only those representing the top six
inches of the sediment. Therefore, the currently biologically active layer was assessed, while
also avoiding confounding effects of depth in the analysis. Samples included:

¦ Soil samples collected March 1999 to February 2000, analyzed by PAC (n = 123).

¦ Soil samples collected April 2000 to April 2002, analyzed by GERG (n = 47).

¦ Sediment samples collected from March 1999 to October 1999, analyzed by PAC (n
= 94).

¦ High-resolution floodplain samples collected in March 2002 to April 2002, analyzed
by GERG (n = 10).

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-4

-------
1 ¦ Sediment samples collected for the porewater study from September 2001 to October

2 2001, analyzed by NEA (n = 45).

3 Porewater — Included 45 porewater samples collected in the PSA from September to October

4 2001 and analyzed by NEA, as part of the GE/EPA porewater partitioning study.

5 Amphibians — 29 frog tissue samples were collected between 1998 and 2000 for which

6 congener data are available:

7 ¦ Bullfrog "offal" samples collected in August of 1999 (n = 20).

8 ¦ Bullfrog whole body sample collected in Reach 5C in October 1998 (n = 1)

9 ¦ Leopard frog "offal" samples collected in April and May of 2000 (n = 5)

10 ¦ Wood frog composite samples collected in July 2000 (n = 3)

12 Crayfish — 40 Crayfish whole body samples were collected in September to October of 1999.

13 Terrestrial Invertebrates — 36 samples were collected in August 2000, including:

14 ¦ Earthworm samples (n = 30) collected for the earthworm study.

15 ¦ Litter invertebrate samples (n = 6) collected for the invertebrate study (2 samples

16 from Reach 5C were not included due to small sample size).

17 Fish — Fish tissue congener data for the period 1998-2000 were considered in this study:

18 ¦ Whole body, composite, filet, and offal samples collected Sept - Oct 1998 (n = 625).

19 Species included brown bullhead (n = 107), bluegill (n = 2), bluntnose minnow (n =

20 2), common shiner (n = 1), fallfish (n = 5), goldfish (n = 41), golden shiner (n = 12),

21 largemouth bass (n = 124), pumpkinseed (n = 121), smallmouth bass (n = 2), and

22 yellow perch (n = 208).

23 ¦ Largemouth bass offal samples (n = 6) collected from Woods Pond in May 1999.

24 ¦ Whole body samples (n = 66) collected in August-October 2000. Species included

25 common carp (n = 8), goldfish (n = 1), and white sucker (n = 57).

26 Vegetation — Grass samples (n = 10) were collected in Reach 5B (DeVos Farm) during July

27 2001.

28 Small mammals — Samples with PCB congener data were collected in August-September

29 1999:

30 ¦ Short-tailed shrew samples (n = 24)

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-5

-------
1

¦ White-footed mouse samples (n = 52)

Waterfowl — Samples with PCB congener data were collected in August-September 1998 from
Woods Pond and backwaters:

¦ Mallard samples (n = 10)

¦ Wood duck samples (n = 40)

Passerine birds — Samples with PCB congener data collected in 1998-2000 include:

¦ Tree swallow samples collected in May-June 1998 (n = 67).

¦ Tree swallows (n = 106), 5 House wrens (n = 5), and chickadees (n = 3) collected in
May- June 1999.

¦ Tree swallow samples collected in May-June 2000 (n = 89)

Benthic invertebrates — Samples with PCB congener data collected in 1998-2000 include:

¦ Kick net samples of predators/shredders (n = 14) collected in July 1999.

¦ Composite tree swallow diet samples (n = 17) from three separate collections (June
1998; May-June 1999; June 2000)

C.7.2.2.2 PCB Source Materials

Six pure NAPL samples were collected from recovery wells located on GE property, two each
from East Street Area 2 (E2), Lyman Street Area (LS), and Newell Street Area 2 (N2). These
samples were used for comparison to congener profiles of other media, and the distance among
the three separate well locations was used as yardstick for distance comparisons.

The GE facility is a large and complex site with numerous historical waste sources. Several
types of product releases to the environment were associated with the transformer and coal
gasification operations along East Street. The primary contaminant sources in the East Street
Area include sites where spills, leaks and the dumping of four types of non-aqueous phase
liquids (NAPLs, types of oil) occurred. The four NAPLs are Pyranol, Aroclor 1260, 10C
insulating oil and coal tar liquids. Pyranol is a mixture of PCBs and 1,2,4-trichlorobenzene; the
PCB portion consisted of either Aroclor 1260 (1932-1971) or Aroclor 1254 (1971-1977). The
East Street Area 2 NAPL materials contain a PCB mixture dominated by Aroclor 1260, whereas

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-6

-------
1

the other two recovery wells have congener profiles more similar to Aroclor 1254, consistent
with the composition of the localized NAPL sources found at the Lyman Street and Newell
Street areas.

C.7.2.2.3 Aroclor Standards

Two sources for Aroclor 1254 and Aroclor 1260 congener concentrations were considered for
comparison to the PCB congener profiles in the Housatonic River.

¦ Beliveau (2002, personal communication) provided the results of high-resolution
PCB congener analyses conducted by AXYS Analytical Services. The results
consisted of analysis from SPB-OCTYL and DB1 chromatography columns, and
included analysis of Aroclor 1254 and Aroclor 1260 as well as other Aroclors.

¦ Frame et al. (1996) evaluated a complete set of congener distributions for 17 Aroclor
mixtures, using high-resolution gas chromatography techniques. The analyses
included assessment of Aroclor lots obtained from three sources (AccuStandard [A],
Supelco [S], and neat Aroclors used by GE Corporate R&D as calibration standards
from Monsanto [G],

Two methods were used to assess the magnitude of differences between AXYS and Frame
congener profiles for Aroclors 1254 and 1260. First, each Aroclor was assessed individually by
comparing the distance among duplicate laboratory analyses to the distance between the average
profile from AXYS and the average profile from Frame. For this analysis, 75 congeners or co-
eluted congener groups (comprising 88-94% of the total PCB) were used for Aroclor 1254, and a
different set of 76 congeners or co-eluted congener groups (81-84% of total PCB) were used for
Aroclor 1260.

The second method of comparison was to narrow the set of congeners to the congeners or co-
eluted congener groups that were common to both Aroclor 1254 and 1260 (55, comprising 57-
70%) of total PCB). With this method, the distance between 1254 and 1260 could be compared
to differences between the two Aroclors, as well as to the distance among duplicates.

C.7.2.3 Data Processing Assumptions

In order to achieve consistency in the analysis, it was necessary to exclude several congeners for
which there were incompatibilities in measurement across laboratories. This occurred due to

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-7

-------
1

inconsistent co-elutions across laboratory methods, or lack of reporting for some congeners. In
some instances, professional judgment was used to determine which congeners could be
confidently assumed to be negligible in concentration. For example, the concentration of PCB-
61 was assumed to be zero for the GERG data, since this congener was not detected in any
largemouth bass samples collected from the PSA (Buckler 2002), and was not identified as a
constituent of Aroclor 1254 or 1260 source materials (Frame et al. 1996). Other congeners were
assumed to be negligible using similar rationale, as summarized in Table 1.

Sample Selection and Assignment — The sample sizes of data used for the analyses (total =
1603) are summarized by laboratory and media type in Table 2. Detailed taxonomic distinctions
were made only for small mammals (i.e., short-tailed shrew and white-footed mouse) and
terrestrial invertebrates (i.e., earthworms and litter invertebrates), because these taxa exhibited
strong profile differences in some reaches. For tissue data, filet samples, offal samples (defined
as the remainder of the body after selected tissues, usually filet, were removed), and whole body
samples were used individually because they were separate chemical analyses. However, ovaries
or other body parts analyzed separately were not included. Other data processing details
included:

¦ Due to small sample sizes in Reach 5D (backwaters), this reach designation was
combined with Woods Pond (Reach 6), for comparisons among reaches in the
multivariate assessment.

¦ Samples from the West Branch (Facility ID = HW) were included only if they were
located in the Southwest branch. Samples with a Facility ID of HO (East Branch)
were included only if they were located above Unkamet Brook, in order to restrict the
analysis to low concentration reference samples upstream of the GE facility. With the
exception of bird samples, concentrations in these reaches greater than 1 mg/kg PCB
were also excluded.

¦ Samples that were non-detects for total PCBs were removed from the analysis.

¦ Amphibian samples H3-T012RP39-0-F001 and H3-TW11BF01-0-8C20 were
assigned to Reach 5C.

¦ Fish sample H3-TW07PSC1-0-8S29 taken from the WWTP impoundment was
removed from the analysis.

¦ There were only two benthic invertebrate samples (H3-TMI13001-0-P001, H3-
TMI13001-0-S001) in Reaches 5D/6; these samples were excluded from the analysis

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-8

-------
1 due to insufficient sample size. There were only two litter invertebrate samples in

2 Reach 5C (H3-TW15LI01-0-0G09, H3-TW15LI02-0-0G10); these samples were

3 excluded from the analysis for the same reason.

4 Non-detects — The similarity analyses using Euclidean distance were performed with both zero

5 and method detection limit substitution for non-detects. For this bounding analysis, no major

6 differences were observed in the Euclidean distance results. Thus, only the zero substitution

7 results are reported in this Appendix.

8 Calculating Percent Compositions — The PCB percent composition of a sample was

9 calculated as the concentration of each congener (or congener group) divided by the sum of

10 estimated total PCBs for that sample, and multiplied by 100. In some cases, the concentration of

11 total PCBs required estimation, because the sum of the congeners measured by each laboratory

12 represented different proportions of total PCB mass. In order to estimate total PCB, the

13 proportion of each laboratory's congener list in Aroclor 1260 was calculated. For example, the

14 congeners measured by PAC accounted for 66% by weight of Aroclor 1260 (the average of the

15 three Aroclor 1260 results reported by Frame et al. 1996). For each sample, the sum of all

16 congeners reported was then adjusted by this amount to estimate total PCBs. For example, the

17 sum of congeners from a PAC sample was divided by 0.66 to estimate total PCBs for that

18 sample. The adjustment for GERG samples was 0.96, and the NEA samples did not need to be

19 adjusted. With the exception of two samples, the set of 23 congeners comprised 20-68% of the

20 total PCB concentration in samples used for this analysis.

21 Laboratory Duplicates — The PCB congener concentrations for laboratory duplicate samples

22 were averaged for this analysis. In addition, the Euclidean distances between duplicate samples

23 for the 23 congener groups were used as a basis for comparison with distances among media and

24 reach groups.

25 C.7.2.4 Multivariate Analysis Methods

26 The PCB percent composition of a sample was calculated as the concentration of the individual

27 congeners divided by the estimated total PCB concentration in that sample, and multiplied by

28 100. No other data transformation was performed. The level of similarity between two percent

29 composition profiles was measured using Euclidean distance. The Euclidean distance between

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-9

-------
1

two samples in multiple dimensions is an extension of the familiar concept of distance between
two points on a Cartesian plane, extended to calculation of distance between two points in a
hyperspace with dimensions equal to the number of variables. The Euclidean distances were
used to develop a quantitative and objective measure of the similarity of the PCB compositions
among individuals within sample groups, as well as the level of similarity of PCB compositions
among different groups defined by laboratories, media types, or locations.

C.7.2.4.1 Euclidean Distance

Euclidean distance is used as an objective measure of the magnitude of difference in congener
profiles. It is a common tool for describing the similarities among samples in exploratory
multivariate techniques such as cluster analysis or multi-dimensional scaling (Johnson and
Wichern 1992). The method reported here was also applied to PCB congener data in fish and
sediments collected from the Hudson River to evaluate similarities between congener patterns
(McGroddy et al. 1997). Litten et al. (1993) used Euclidean distances to compare PCB homolog
concentrations in water samples between river locations.

Euclidean distance was calculated between two samples a and b using the following equation

where at and h, are the percent compositions for congener (/ = 1,2, ... 23) in samples a and b,
respectively. Two samples may generate a large distance value due to large-magnitude
differences in one or two congeners, or due to smaller-magnitude differences in many congeners.
While both of these situations could result in a large distance value, the type of deviation cannot
be determined from the univariate distance value alone. Pattern differences and principal
congeners responsible for these differences must instead be identified from congener profile
plots.

Tissue sample groups were defined on the basis of laboratory, media, and location. Centroids
were calculated as the average composition of the group (i.e., a composition vector where each
element is the arithmetic average of the congener percentages averaged over all samples for the

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC C* 1 '\ (\ 7/10/2003

[1]

-------
1

group). Using these definitions, Euclidean distance was used to measure similarity within
groups and between groups. The within-group distances were calculated between each sample
and the group centroid. The between-group distances were the distances between the group
centroids.

There are no absolute criteria for scaling the differences among congener patterns indicated by
the value of the Euclidian distances. In order to judge the significance of the magnitude of
distances between groups, three different measures were used:

1. Within-group distances. Between-group distances that are larger than most or all within-
group distances can be considered significantly different. This is not a measure of
statistical significance, but could be viewed as potentially environmentally significant.

2. Distances among laboratory duplicates. These distances are a measure of laboratory
variability. Between-group distances that are within the range of laboratory variability
are judged to be very similar.

3. Distances between Aroclors 1254 and 1260. These "goalposts" may be used as a
measure of a large distance. Two congener profiles with this magnitude of difference or
larger are quite different, since these Aroclor formulations have different levels of
chlorination and associated congener patterns.

The distance between the Aroclors was also used as a indication of potential outliers (unusual
congener profiles) for individual samples within groups. Any sample with a distance from the
group centroid greater than the distance between the two Aroclors was considered for exclusion
from the statistical analysis.

C.7.2.4.2 Principal Components Analysis

Principal Components Analysis (PCA) was performed on the percent composition data for the
principal congeners. PCA reduces the dimensionality of a multivariate dataset to a subset of
latent variables (i.e., principal components) that can be used to describe the correlation structure.
Plotting the first two components or major axes of the PCA analysis provides a useful visual
interpretation of distances among samples. The PCA was based on the correlation matrix for the
percent composition data; no other standardization of the data was performed. Percent
composition data are already standardized on the same scale (percent of total); this analysis

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-11

-------
1 evaluates the nature of the relationships based on relative contribution to the total PCB

2 concentration.

3 C.7.2.5 Indicator Congener Analysis Methods

4 The main objective of the "indicator congener" evaluation was to determine whether spatial or

5 among-media trends existed for individual congeners that were not evident in the analysis of

6 whole PCB mixtures. It is possible that weathering of select PCBs could be masked by the

7 variation introduced by considering the entire mixture of PCB congeners in the multivariate

8 analysis. Therefore, select individual congeners were assessed, focusing on those congeners

9 selected a priori as having high potential for weathering. A few congeners with low anticipated

10 weathering were also evaluated for comparative purposes.

11 C.7.2.5.1 Selection of Indicator Congeners and Congener Groups

12 A literature review of PCB degradation and transformation patterns in environmental media was

13 conducted to determine a subset of representative congeners (Attachment 1). The purpose of the

14 literature review was to identify specific congeners that may exhibit different environmental

15 partitioning and degradation properties relative to homologs or other individual congeners with

16 similar degree of chlorination. For example, some PCBs were identified based on a tendency to

17 be enriched or depleted by anaerobic bacterial action, while others were identified based on

18 potential for metabolism by fish and mammals.

19 Individual congeners and congener groups selected included:

PCB-49/43
PCB-110/77/148
PCB-138/163/164/158/160/186
PCB-151

PCB-153/132/105/161
PCB-170/190

PCB-180
PCB-194
PCB-77
PCB-126

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-12

7/10/2003

-------
1

PCB-77 and PCB-126 were assessed using slightly different methods because concentrations of
these non-ortho PCB congeners were much lower; resulting detection limit issues required a
more limited analysis. The above list includes five congener "groups"; in these cases, the
congener listed first (in bold) represents the majority of the PCB mass within the group. For
these five co-eluting groups, Table 3 indicates the proportion of each congener based on the
percentages found in Aroclor 1260 standards (Frame et al. 1996) and in largemouth bass tissues
collected in the Housatonic River (Buckler 2002). Table 3 shows that the proportions of
individual congeners within the groups remain fairly constant across media, and that the
"dominant congener" comprises two-thirds or more of the mass of each group. Although it is
possible that differential partitioning of congeners within a group could confound subtle profile
differences across space and/or media, the dominance of mass for the principal congeners in each
group was sufficient to identify major differences in the congener profile.

C.7.2.5.2 Analysis of Indicator Congeners and Groups

A bounding analysis, similar to that described above for the multivariate assessment, was
performed using assumptions of: (a) non-detect values replaced by the detection limit, and (b)
non-detect values replaced by zero. For PSA samples, the differences between reach averages
(% total PCB) were typically <20%. Because differences between substitution methods were
small, the analysis presented here is based on a zero DL substitution for non-detect values.

To account for differences in PCB measurement across laboratories, sample proportions were
adjusted depending on the measuring laboratory, as described in Section C.7.2.3. To adjust for
the laboratory differences, sample proportions were multiplied by one of the following factors:
GERG samples (0.96); PAC samples (0.66); NEA samples (1.0).

The differences in individual congener concentrations across locations and media were evaluated
comparing measures of central location (i.e., median and mean). This was performed both
numerically and graphically. Medians and arithmetic means were calculated and compared to
each other and between reaches for each media for each congener group. To assess potential
differences among reaches, reach means were compared to the mean for the entire PSA. When
the reach mean was outside the range of +/-50% of the PSA mean, then the reach was flagged as

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-13

-------
1 potentially having different congener concentrations. Congener profile plots were also used to

2 identify if there was either an increasing or decreasing trend across reaches.

3 In some cases, formal statistical tests for differences (e.g., test for differences among media)

4 were conducted using a nonparametric test (Kruskal-Wallis test) at a significance level of 0.05.

5 The null hypothesis for these tests was that the congener percentage is the same for all media

6 groups.

7 Some additional data treatment considerations were necessary for the PCB-77 and PCB-126

8 evaluations. Summary statistics and graphs were created with and without one benthic

9 invertebrate sample (MCDIET) since the values of these congeners were anomalously high. A

10 bounding analysis was also conducted with this sample included. For these non-ortho congeners,

11 only samples from GERG and PAC labs were included since NEA samples measured PCB-77

12 and PCB-126 as co-eluting with other higher concentration congeners. Therefore, the analyses

13 for these congeners excluded 45 sediment samples and excluded porewater as a media group.

14 Note that percentages of PCB-77 and PCB-126 are recorded as 100X the measured percentage

15 for presentation purposes.

16 The first step in the evaluation of indicator congeners was to investigate differences across

17 reaches for each media type. Summary statistics (minimum, mean, median, 90th percentile,

18 maximum) for the congener percentages, boxplots, and barplots were used to display the

19 differences between the concentrations of these congeners in the different media for both zero-

20 DL and half-DL substitution methods. Where no significant trends across space were observed,

21 the data were pooled across all reaches to provide greater statistical power for cross-media

22 comparisons.

23 C.7.2.6 Homolog Analysis Methods

24 Homolog concentrations were calculated by summing the congener concentrations within each

25 homolog group. For co-eluted congeners in different homolog groups, assumptions were made

26 concerning the relative content of congeners based on the Frame et al. (1996) and AXYS

27 (Beliveau 2002, personal communication) Aroclor standards. The assumptions necessary for the

28 PAC samples were as follows:

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-14

-------
9
10

PCB-22/51: assumed all present as PCB-22
PCB-42/59/37: assumed equal shares of the three congeners
PCB-91/55: assumed all present as PCB-91
PCB-95/80: assumed all present as PCB-95
PCB-179/141: assumed equal shares
PCB-123/149: assumed all present as PCB-149
PCB-171/202: assumed all present as PCB-171
PCB-137/176: assumed equal shares
PCB-157/173/201: assumed equal shares

11 No assumptions were necessary for the GERG samples. NEA samples were not included in

12 these graphs.

13 As was done for the congener analysis, the homolog concentrations were converted to percent of

14 total PCB by dividing by the total PCB concentration and multiplying by 100. Plots of homolog

15 profiles by media and reach were evaluated qualitatively for trends.

16 Tests for differences among mean chlorination levels in sediments by reach were conducted

17 using analysis of variance (ANOVA) with alpha = 0.05. There was one sample (H4-SE000871-

18 0-0000) in Reach 5D/6 with an unusually low average chlorination level (4.97). This sample

19 caused a violation of the assumptions of parametric ANOVA (i.e., normal distribution and equal

20 variance among cells). Therefore, this sample was removed from the ANOVA analysis. Tests

21 were also conducted with the outlier included (using non-parametric ANOVA), and conclusions

22 were qualitatively similar.

23 C.7.3 Results

24 C. 7.3.1 Multivariate Analysis of Overall Congener Profile

25 C.7.3.1.1 Comparisons of Aroclors

26 Comparisons of Aroclor standard profiles were made to assess the variations in congener profiles

27 among reference materials, and to establish the "distance" between Aroclor 1254 and 1260 for

28 use as a yardstick for comparing other profile differences.

29 The results of the two methods for comparing the AXYS (Beliveau 2002, personal

30 communication) and Frame (Frame et al. 1996) Aroclor data are displayed in Table 4. The

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC ^71^ 7/10/2003

-------
1 distance between the Frame Aroclor 1254 standards is 12 (or 11 with 55 congeners), which is

2 relatively large in comparison to other distances. The Frame "G" standard exhibited a distance

3 of only 1.7 from both AXYS duplicates, whereas the Frame "A" standard is a distance of 13

4 from the AXYS duplicates. Therefore, the Frame "A" standard is an unusual profile for 1254;

5 Frame et al. (1996) note that differences in congener profiles for Aroclor 1254 are common.

6 Because the Frame "G" duplicate and the AXYS results were similar, the Frame "G" standard

7 was used as the default Aroclor 1254 standard in this Appendix.

8 The Aroclor 1260 congener profiles for AXYS and Frame were similar (distances = 0.89 - 2.69).

9 The average congener profile from the three Frame 1260 samples was used as the Aroclor 1260

10 standard for this analysis.

11 For the reduced set of 23 congeners used to compare Housatonic River samples, the distance

12 between Aroclor 1254 and Aroclor 1260 was 17. Distances of 17 or greater are indicative of

13 large alterations in congener profiles between samples.

14 Laboratory Duplicates

15 Out of the 1603 samples included in this analysis, duplicate laboratory analyses were available

16 for both GERG and NEA analyses. A quantitative comparison of laboratory duplicates was

17 conducted for the 37 GERG samples. The Euclidean distances between laboratory duplicate

18 congener profiles for these samples are summarized in Table 5. The 90th percentile of the

19 duplicate distances (2.4) is used for comparisons to between-group distances, and suggests that

20 Euclidean distances of 2-3 indicate insignificant differences in congener profiles.

21 Outliers

22 Soil sample H3-FL000571-0-0000 had a Euclidean distance of 21 from the soil group centroid

23 for PAC results in Reach 5C. This sample had a very low total PCB concentration (0.0037

24 mg/kg for zero substitution), and only two congeners (PCB-101 and PCB-138) were detected.

25 This sample was removed from the analyses as an outlier.

26 Leopard frog sample H3-T012RP39-0-F001 had a Euclidean distance of 18 from the amphibian

27 centroid for Reach 5C. This sample was collected from an area west of the railroad tracks and

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-16

-------
1 northwest of the backwaters, which is not officially part of Reach 5C. The total PCB

2 concentration was low (0.038 mg/kg), with only 11 congeners detected. This sample was

3 removed from the analyses as an outlier.

4 There were several white-footed mouse samples with distances greater than 17 from their

5 respective group centroids. However, this group was highly variable (see media results below),

6 and there was no strong basis for considering the more distant samples to be outliers. Therefore,

7 these samples were retained.

8 The sediment samples in the reference reach HO (above Unkamet Brook) were highly variable,

9 with median distance to the group centroid of 40. There was no basis for considering any of the

10 samples to be outliers, so all samples were retained.

11 Comparisons among Laboratories

12 Inter-laboratory comparisons were performed by comparing congener profiles among

13 laboratories within a given reach and media. This comparison was only performed where there

14 were three or more samples from multiple labs within a media/reach group. The Euclidean

15 distances within and among labs are displayed in Table 6, and summarized in Figure 1. The bar

16 lengths in Figure 1 are average distances between labs across reach/media groups, and the 90th

17 percentile values for laboratory duplicate and within-lab/group distances.

18 GERG profiles are similar to both the PAC and NEA congener profiles for sediment and soil,

19 with Euclidean distances of 2.1 and 2.2, respectively. The distances between the PAC and NEA

20 congener profiles (Euclidean distance = 3.9) are slightly greater than the laboratory duplicates

21 (Euclidean distance = 2.4) for the same sample, but are similar to distances among samples

22 within the same laboratory, media, and reach. This result suggests that combining laboratory

23 results for comparing media and locations will not inflate the within reach/media group

24 distances. Figure 2 displays the profile plots comparing the 23 congeners across laboratories for

25 each reach. The profile plots indicate a high degree of similarity among laboratories for both

26 sediment and soil samples. Therefore, results from different laboratories were combined for the

27 among-reach and among-media comparisons.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-17

-------
1

C.7.3.1.2 Comparisons Among Media

Media comparisons were conducted among all media with 3 or more samples within a study
reach. The results are displayed in Table 7. Bold and highlighted values indicate distances that
exceed the distance between Aroclor 1254 and 1260 (17), and unbolded highlighted values
indicate a smaller difference greater than the 90th percentile of the within-group distances for all
reaches (4.7). The table indicates a wide range of Euclidean distances between groups (1.3 -
25), indicating that while some media exhibit similar congener profiles, others are quite different.
Congener profiles for both mammal species are the most unlike other media, with the strongest
disparity in Reaches 5A and 5B. Congener profiles for amphibians, porewater, and terrestrial
invertebrates are also different from most other media in most reaches, with distances typically
in the 5 to 12 range. However, benthic invertebrates, passerine birds, crayfish, fish, sediment,
and soil samples have very similar profiles in all reaches.

The general similarity of the congener profiles in sediment, benthic invertebrates, crayfish and
fish suggests that significant weathering of PCBs is not occurring within the main channel
sediment food web. Passerine birds (i.e., tree swallows) may have similar congener patterns to
the aquatic media due to their diet consisting primarily of aquatic insects. Porewater samples
likely have different congener patterns due to the difference in physical properties of these
media.

Congener profiles are plotted for selected groups of media in Figures 3 to 7:

¦ Figure 3 shows that fish and crayfish congener profiles are very similar, with no
systematic pattern of differences. The percentages for individual congeners are
typically within one standard deviation, suggesting that the small differences
observed may be due to sampling variation.

¦ Figure 4 indicates the high degree of similarity between aquatic media (sediment,
benthic invertebrates, fish). In some cases, there appear to be small alterations in
percentages across these media. For example, PCB-187 shows a pattern of
sediment
-------
1

this congener is quite high in all reaches (i.e., higher mean value driven by a few
extreme values).

¦ Figure 6 compares the congener profiles for the two small terrestrial mammal species
to other media. The first bar for each congener represents the average percentage for
six media types (sediment, soil, benthic invertebrates, crayfish, fish, passerine birds)
that were demonstrated to be highly similar. The profile plots show the shift in
congener composition for small mammals. The shrew samples indicate large
elevations in PCB-187 within reaches 5A and 5B, with reductions in most other
congeners relative to aquatic media. However, these differences were not as
pronounced in Reach 5C. The white-footed mouse tissues exhibited a profile that was
different from both the aquatic media and the shrew samples. Specifically, the mice
exhibited significant increases in PCB-180 and PCB-194, with concentrations of
lower chlorinated congeners that were similar to the shrew but lower than the aquatic
media. Again, the larger profile differences observed in Reaches 5A and 5B were not
observed to the same extent in 5C. In summary, small mammals appear to exhibit
selective congener enrichment/depletion that is highly congener-specific and that
varies somewhat between reaches.

¦ Figure 7 compares the congener profiles for amphibians to the "average" profiles for
aquatic media. The differences are not nearly as large as for small mammals;
however, some differences were observed that are consistent across reaches and
therefore do not appear to be explainable based on sampling variation alone.
Specifically, the proportions of the higher molecular weight PCBs (e.g., 180, 183,
187, 194) are greater in amphibian samples relative to the aquatic media. This pattern
is observed in all four reaches examined. Conversely, the concentrations of several
lower chlorinated PCBs (e.g., 101/90, 110/77, 149/123/118, 151) are consistently
lower in amphibian samples. Overall, the amphibian profiles are indicative of an
overall shift toward more highly chlorinated compositions, relative to aquatic media.

The congener profiles for environmental media were compared to Aroclor standards to determine
which of the commercial mixtures most closely approximates the PCB mixtures sampled
throughout the PSA. In Figure 8, the average profile for the six "similar" media groups (benthic
invertebrates, sediment, soil, crayfish, fish, and birds) was compared to the Aroclor 1254 and
Aroclor 1260 standard profiles. The profile plots suggest a general similarity between the
environmental media and the Aroclor 1260 standard, across all four reaches. Some minor
differences were observed, such as a modest reduction in PCB-180 in environmental samples
relative to Aroclor 1260; however, the overall patterns suggest that aquatic Housatonic River
media exhibit a profile similar to unweathered Aroclor 1260. In contrast, the Aroclor 1254
profile was quite different from the other two profiles. The majority of the Aroclor 1254 mass
was found in the lower chlorinated congeners (PCB-44, PCB-70, PCB-87/81/115, PCB-101/90,

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-19

-------
1 PCB-110/77) instead of the higher chlorinated congeners (PCB-180, PCB-187) found at high

2 proportions in the Housatonic River media and Aroclor 1260.

3 C.7.3.1.3 Comparisons Among Reaches

4 Reach comparisons were conducted for each media type with samples in multiple reaches. The

5 reference reaches were also included in these analyses. The Euclidean distance results are

6 displayed in Table 8. Because the within-group profile distances may vary with media type, the

7 between-reach distances are compared to the 90th percentile of the media-specific within group

8 distances (exceedances shown in shaded cells in Table 8). For example, the 90th percentile of

9 within-group distances for crayfish is 1.8, so the distance of 2.5 between the mean profile for

10 Reach 5A and the mean profile for Reach 5C indicates a modest difference in congener profile.

11 Reach differences were also compared to the Euclidean distance between Aroclors 1254 and

12 1260 (indicative of a large profile difference).

13 Overall, the analysis indicated few large profile differences among PSA media. For benthic

14 invertebrates, passerine birds, fish, porewater, sediment, and soil, no significant differences in

15 congener profiles were observed across reaches 5A, 5B, 5C, or 5D/6. Significant profile

16 differences were observed for these media when comparisons were made to the reference reaches

17 however (for benthic invertebrates, fish, and sediment), presumably reflecting reduced exposure

18 of the reference areas to PCBs originating from the GE facility.

19 The largest differences were observed for upstream East Branch (HO) sediments relative to PSA

20 sediments (distances of 24 - 26). This reflects the shift in PCB concentration resulting from

21 PCB sources at the GE facility. Significant, but smaller-scale, differences were also observed for

22 West Branch reference sediments relative to PSA sediments.

23 Large differences were also observed for small terrestrial mammals (short-tailed shrew). Large

24 differences (15 - 18) were observed when comparing profiles in Reaches 5A and 5B to Reach

25 5C. This reflects the profile differences discussed in Section C.7.3.1.2, and depicted in Figure 6.

26 Significant differences in congener profiles were observed across reaches for soil invertebrate

27 samples (earthworms and litter invertebrates). However, the Euclidean distance values were not

28 large (3.4 to 5.7), indicating that such profile differences are relatively subtle.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC q j 2Q 7/10/2003

-------
1 Overall, the comparisons among reaches indicate that, with a few exceptions, the congener

2 profiles remain consistent moving from upstream to downstream within the PSA. The aquatic

3 media have a higher level of congener profile consistency relative to terrestrial fauna (soil

4 invertebrates and small mammals). Figure 9 depicts the among-reach congener profile

5 differences for three representative aquatic media (sediment, fish, and amphibians). The high

6 degree of consistency in congener profile is evident in this figure.

7 C.7.3.1.4 Principal Components Analysis

8 Because differences among reaches were not substantial for most media (Section C.7.3.1.3)

9 within the PSA, PC A analyses were run using all PSA reaches combined. The plot in Figure 10

10 shows the best representation of the distance among the congener profiles in two dimensions.

11 For simplicity, only the group centroids are shown in this plot; Aroclor standards and NAPL

12 samples are also included. These two principle components represent only 38% of the total

13 variability among samples; therefore, this 2-dimensional representation is only an approximate

14 portrayal of the 23-dimensional space.

15 Figure 10 indicates that the Aroclor 1254 standard has a congener profile that is quite different

16 from most other samples. Only the Newell Street (N2) and Lyman Street (LS) NAPL samples

17 have a congener profile similar to Aroclor 1254. Most environment media exhibit a congener

18 profile similar to Aroclor 1260 and similar to the East Street (E2) NAPL samples. A few media

19 (white-footed mouse, short-tailed shrew, and to a lesser extent amphibians) exhibited higher PCI

20 scores than other media, reflecting the higher chlorination level observed in these media (i.e.,

21 greater contribution of more highly chlorinated ["heavier"] PCB congeners).

22 C.7.3.2 Indicator Congener Analysis

23 C.7.3.2.1 Comparisons Among Reaches

24 Figure 11 shows the means and standard deviations for all indicator congeners except the non-

25 ortho PCB-77 and 126, which were too low in concentration to be shown on the same scale. Few

26 significant differences were observed across reaches within a given media type; no consistent

27 trend in the PCB concentrations among the reaches was observed. This similarity across reaches

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC C 7 21 7/10/2003

-------
1

applied not only to the aquatic media, but also to the terrestrial invertebrates and small mammals.
Therefore, cross-reach differences were considered to be small, and samples from all PSA
reaches were combined for testing differences across media.

C.7.3.2.2 Comparisons Among Media

Table 9 presents the mean percentages for each indicator congener group, using all PSA samples
combined. The means and standard deviations are plotted in Figure 12 for all but the non-ortho
PCB congeners. Alterations in the profiles of the PCB mixtures were observed for all media.
Hypothesis testing based on the Kruskal-Wallis test indicated statistically significantly
differences across media for each congener grouping based on alpha = 0.05.

The PCB observed profiles differed among the various indicator congeners. For some low- to
moderately-chlorinated congeners (PCB-49, PCB-110, PCB-151), there was a tendency for
lower percent composition at higher trophic levels and/or for organisms in contact with terrestrial
habitat. Small mammals, waterfowl, and amphibians contained lower percentages of these
congeners relative to "aquatic" media such as sediment, benthic invertebrates and fish (Figure
12). The reverse pattern was observed for the higher chlorinated congeners; PCB-153, 170, 180,
and 194 all exhibited the highest percentages in mammal, waterfowl, and amphibian tissues.
PCB-153 and 180 exhibited an approximate doubling of concentration from the bottom of the
food web (sediment, soil) to small mammals. The PCB-194 concentration increased
substantially in small mammals relative to other media. These findings indicate that biologically
mediated transformation of PCB mixtures has occurred in the Housatonic River PSA.

The comparisons across media for the non-ortho PCBs (77 and 126) are presented in Figures 13
and 14. The bar graphs of mean concentration in Figure 13 suggest that waterfowl accumulate
higher proportions of these toxic congeners relative to other biota. However, the analysis for
these congeners is confounded by a few extreme values, particularly in soil and sediment. Figure
14 provides a boxplot showing the range, quartiles and median concentrations of these
congeners. Given the variation introduced by extreme values, it is not possible to identify clear
cross-media differences for PCB-77 or 126, with the possible exception of elevated waterfowl
percentages.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-22

-------
1

The congener profiles in biota were compared to the compositions found in Aroclor 1260
standards (Frame et al. 1996). The following differences were observed for congener
percentages:

¦ PCB-49/43, PCB-151, PCB-180 and PCB-194 percentages were found to be elevated
in tissue samples relative to Aroclor 1260.

¦ PCB congener group 110/77/148 exhibited percentages that were generally higher in
biota relative to Aroclorl260, with the exceptions of the amphibian, waterfowl and
mammal groups, for which this group was relatively depleted.

9
10

¦ PCB congener group 138/163/164/158/160/186 was depleted in site media relative to
Aroclor 1260, with the exception of waterfowl.

¦ PCB congener group 153/132/105/161 was depleted in lower trophic media, but
enriched in higher tropic media, relative to Aroclorl260.

¦ PCB congener group 170/190 was depleted generally in site media relative to Aroclor

1260.

15 In summary, statistically significant differences in PCB congener compositions were observed

16 across media within the Housatonic River PSA. The differences in composition appear to be

17 related to the degree of chlorination of the congener and the position in the food web. In most

18 cases, the average percent compositions are not large (i.e., within a factor of two); however some

19 congeners (e.g., PCB-194) and species (e.g., terrestrial mammals and waterfowl) has profiles that

20 differ from aquatic media (e.g., sediment, aquatic invertebrates, fish) by a larger amount.

21 C.7.3.3 Homolog Analysis

22 Figure 15 present profile plots for homologs, organized by media and reach. Some media

23 exhibited consistent homolog profiles across reaches (e.g., soil, terrestrial invertebrates, small

24 mammals); therefore there does not appear to be a progressive decrease in chlorination level over

25 space for terrestrial-based media. However, the aquatic media exhibited a modest transition in

26 homolog profile, in which the overall degree of chlorination decreased somewhat with distance

27 downstream. The transition is typified by a loss of hepta- and octa- PCB congeners, with a

28 corresponding increase in tetra- and penta- PCB congeners. This general pattern was observed in

29 sediment, benthic invertebrate, amphibians, and to a lesser extent in crayfish, fish, and passerine

30 birds (tree swallows that consume aquatic insects).

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC C 7 23 7/10/2003

-------
1 A simple measure of the level of PCB chlorination is the average number of chlorines per

2 biphenyl ring (CBP). This metric represents the "average" chlorination or central tendency of

3 the homolog histograms presented in Figure 15. Table 10 presents a summary of the CBP data

4 organized by media and river reach. To test whether the patterns of homolog differences

5 observed in Figure 15 are statistically significant, a parametric ANOVA was conducted. The

6 results of the ANOVA followed by Tukey multiple comparisons, with the single outlier

7 removed, are displayed in Table 11. Pairwise comparison tests indicated that there was a

8 significant difference in chlorination level between the upper reaches and the lower reaches.

9 The analysis presented above was conducted including PAC and GERG sediment data only.

10 QEA and BBL (2003) conducted an evaluation of the NEA data associated with the 2001

11 porewater study, and documented similar findings (i.e., modest reduction in overall chlorination

12 with distance downstream in the PSA).

13 C.7.4 Conclusions

14 This Appendix addressed the following three questions related to the PCB congener profiles in

15 Housatonic River environmental media:

16 ¦ Are there differences in congener distributions across environmental media (both

17 abiotic and biotic), and if so, how large are the differences?

18 ¦ Conclusion: Some differences in congener profiles were observed across media. Both

19 the multivariate analyses and the evaluation of individual congeners/groups

20 documented differences in the PCB compositions. In general, aquatic media

21 (sediments, crayfish, benthic invertebrates, fish) exhibited very similar congener

22 profiles. However, small mammals, soil invertebrates (and to a lesser extent

23 amphibians) exhibited somewhat different PCB congener profiles. Although still

24 similar to Aroclor 1260, these media had a higher proportion of some high molecular

25 weight PCB congeners. The patterns in PCB percentages for individual congeners

26 depended on the level of chlorination. High molecular weight congeners were often

27 enriched in mammals, waterfowl, and amphibians, whereas lower chlorinated PCBs

28 were relatively depleted in these biota (Figure 12).

29 ¦ Are there differences in congener distributions with distance from the source material

30 (i.e., GE facility), and if so, are these differences consistent with expected

31 environmental degradation of the samples?

32 ¦ Conclusion: In general, congener patterns within a given medium were relatively

33 stable across the PSA. In the evaluation of congener profiles and specific individual

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-24

-------
9

congeners, major pattern differences across the PSA were not typically observed.
Concentrations and proportions of PCB congeners from aquatic media (sediment,
benthic invertebrate, fish tissue) were consistent across the PSA. However, the
analysis of homologs suggested that a modest reduction in overall level of
chlorination is occurring in the downstream portions of the PSA, and this subtle but
statistically significant alteration in homolog profile is observed in several aquatic
media. This finding is consistent with assessments of sediments conducted by
QEA/BBL (2003) as well as Bedard and May (1996), Bedard et al. (1996, 1997) and
Bedard and Quensen (1995). Some reach differences in terrestrial fauna congener
compositions were also observed (small mammals and soil invertebrates) but these
were not consistently observed. Overall, the spatial variation in PCB congener
profiles is sometime statistically significant but generally small in magnitude.

¦ Do the differences in analytical techniques/laboratories used in the measurement of
PCB congeners affect the interpretations of the above?

¦ Conclusion: An assessment of the three main laboratories (NEA, GERG, PAC) did
not indicate significant interlaboratory differences that would have a major bearing on
the conclusions of this study.

18 Overall, the congener evaluation indicated that differences in profiles are sometimes evident, but

19 that most media exhibit congener profiles similar to Aroclor 1260 across all reaches.

20 Modifications to the congener profiles were sometimes observed both spatially and across media,

21 with the latter differences larger than the former. However, the degree of chlorination of

22 Housatonic River PCB mixtures was not usually large, especially in comparison to PCB sites

23 with lower chlorination levels, such as the Hudson River (Chen et al. 1988).

24 C.7.5 References

25 Bedard, D.J. and R.J. May. 1996. Characterization of the polychlorinated biphenyls in the

26 sediments of Woods Pond: Evidence for microbial dechlorination of Aroclor 1260 in situ.

27 Environ. Sci. Technol. 30(l):237-245.

28 Bedard, D.L. and J.F. Quensen III. 1995. Microbial Reduction Dechlorination of Polychlorinated

29 Biphenyls. Chapter 4 (pp. 127-216) in L.Y. Young (Ed.), Microbial Transformation and

30 Degradation of Toxic Organic Chemicals. Wiley-Liss, 1995.

31 Bedard, D.L., H.M. Van Dort, R.J. May, and L.A. Smullen. 1997. Enrichment of

32 microorganisms that sequentially meta, para-dechlorinate the residue of Aroclor 1260 in

33 Housatonic River sediment. Environ. Sci. Technol. 31:3308-3313.

34 Bedard, D.L., S.C. Bunnell, and L.A. Smullen. 1996. Stimulation of microbial para-declorination

35 of polychlorinated biphenyls that have persisted in Housatonic River sediment for decades.

36 Environ. Sci. Technol. 30:687-694.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC CI 2,5 7/10/2003

-------
1 Beliveau, A. 2002. Personal communication. PCB congener data for various Aroclor mixtures

2 produced by AXYS Analytical Services. Analysis from SPB-OCTYL and DB1 chromatography

3 columns (AXYS METHOD CL-EX-1668A/Ver.l 1668AD7). Contract No.: 4052. Excel

4 spreadsheet provided by Andy Beliveau (US EPA) to Richard Beach (WESTON Solutions).

5 November 2002.

6 Buckler, D.R. 2002. Fish reproductive health assessment in PCB contaminated regions of the

7 Housatonic River, Massachusetts, USA: Investigations of causal linkages between PCBs and fish

8 health. Interim Report of Phase II Studies. Prepared for U.S. Fish and Wildlife Service, Concord,

9 New Hampshire and U.S. Environmental Protection Agency, Boston, Massachusetts.

10 Chen, M., C.S. Hong, B. Bush, and G.-Y. Rhee. 1988. Anaerobic biodegradation of

11 polychlorinated biphenyls by bacteria from Hudson River sediments. Ecotoxicology and

12 Environmental Safety. 16:95-105.

13 Frame, G.M., J.W. Cochran, and S.S. Bowadt. 1996. Complete PCB congener distributions for

14 17 Aroclor mixtures determined by 3 HRGC systems optimized for comprehensive, quantitative,

15 congener-specific analysis. J. High Resol. Chromatogr. 19:657-668.

16 Johnson, Richard A., and Dean W. Wichern. 1992. Applied Multivariate Statistical Analysis.

17 Third Edition. Prentice Hall, N.J.

18 Litten, S., B. Mead, and J. Hassett. 1993. "Application of passive samplers (PISCES) to

19 locating a source of PCBs on the Black River, New York." Environ. Toxicol. Chem. Vol. 12, pp.

20 639-647.

21 McGroddy, S. E., L.B. Read, L.F. Field, C.G. Severn, and R.N. Dexter. 1997. Data Summary

22 and Analysis Report — Draft, April 1997. Prepared for NOAA, Damage Assessment Center,

23 Silver Springs, MD. Prepared by EVS Environment Consultants, Seattle, WA.

24 QEA (Quantitative Environmental Analysis, LLC) and BBL (Blasland, Bouck & Lee, Inc). 2003.

25 Housatonic River - Rest of River RCRA Facility Investigation Report. Volume 1. Report.

26 Prepared for General Electric Company, Pittsfield, MA. January 2003.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-26

-------
Table 1

Congeners and Co-eluting Congener Groups Used in the Analysis Combining
Across Labs (Shaded Groups Retained)

IIJPAC Numbers

GERG

NEA

PAC

Congener Used for
Analysis

Congener Assumed Not
Present in Co-Elutions

PCB-1

PCB-28/50

PCB-28

PCB-50

PCB-104/44

PCB-44

PCB-104

PCB-70

PCB-74/61

PCB-74/94/61

PCB-74

PCB-61,94

PCB-81

PCB-87

PCB-87/115

PCB-115

PCB-81/87/117/125/111/115/145

PCB-87/81/115

PCB-117,125,111,145

PCB-79/99/113

PCB-99

PCB-79,113

PCB-101

PCB-90

PCB-101/90

PCB-110

PCB-77

PCB-77/110/148

PCB-110/77

PCB-148

PCB-119/150

PCB-119

PCB-150

PCB-12 8/162

PCB-128

PCB-162

PCB-118

PCB-123

PCB-149

PCB-149/123

PCB-106/118/139/149

PCB-149/118/123

PCB-106,139

PCB-151

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC q j 7/10/2003

-------
Table 1

Congeners and Co-eluting Congener Groups Used in the Analysis Combining
Across Labs (Shaded Groups Retained)

(Continued)

IIJPAC Numbers

GERG

NEA

PAC

Congener Used for
Analysis

Congener Assumed Not
Present in Co-Elutions

n is-r~

VflHI

flHl

flflH

IS"

VHi

PCB-182/187

I'l li-|x~

flHl

ri m-is:

I'i |;-|S"

flHl

flflH

¦HHi

rci;-i"i

IHI

IHi

flHi

IHriHHH!

I^BBBIIMIMIB

I'i is-:".-

¦¦¦

IHR

111 "

flHB

III

flHi

IHNHHHI

Wmmtmmmm

| i'|

Mil

ill

Hli

IHHHHHI

aOnly measured in small number of high-resolution samples.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC q j 7/10/2003

-------
Table 2

Counts of Samples Used for Congener Analysis

Media

Lab

Reference
(HO)

5ABa

5BCb

5D/6

Reference
(HW)

Amphibians

GERG

Benthic Invertebrates

GERG

Passerine Birds

GERG

Crayfish

GERG

Fish

GERG

252

249

Mammal -Short Tailed Shrews

GERG

Mammal - White Footed Mice

GERG

Porewater

NEA

Sediment

GERG

Sediment

NEA

Sediment

PAC

Soil

GERG

Soil

PAC

Terrestrial Invertebrates - EW

GERG

Terrestrial Invert - LI

GERG

Vegetation - GR

GERG

Waterfowl

GERG

a Amphibians in Reaches 5A and 5B were combined due to low sample size.

b Fish samples were not collected separately in Reaches 5B and 5C.

MK01|O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC £ rj ^C) 7/10/2003

-------
Table 3

Congener breakdown of five congener groups selected for individual analysis,
using Aroclor 1260 standards (Frame et al. 1996) and Housatonic River
largemouth bass tissue PCB data (Buckler 2002)

Congener Group

PCB153/132/105/161

PCB110/77/148

PCB170/190

Congener

PCB153

PCB132

PCB105

PCB161

PCB110

PCB77

PCB148

PCB170

PCB190

A1260A

9.09

2.84

0.22

no values

1.38

no values

3.97

0.82

A1260S

9.17

2.96

0.21

no values

1.36

no values

4.01

0.80

% ot Congener
from total PCBs
measured in

A1260G

9.91

2.91

0.23

no values

1.25

no values

4.36

0.85

Average of
A,S,G samples

9.39

2.90

0.22

no values

1.33

no values

4.11

0.82

Aroclor 12601

Proportion of

congener
within

75%

23%

100%

83%

17%

congener group

Woods Pond A

13.19

1.88

0.40

no value

1.62

-------
Table 3

Congener breakdown of five congener groups selected for individual analysis,
using Aroclor 1260 standards (Frame et al. 1996) and Housatonic River
largemouth bass tissue PCB data (Buckler 2002)

(Continued)

Congener Group

PCB138/163/164/158/160/186

PCB49/43

Congener

PCB138

PCB163

PCB164

PCB158

PCB160

PCB186

PCB49

PCB43

% ot Congener
from total PCBs
measured in
Aroclor 12601

A1260A

6.47

2.41

0.72

0.57

no values; co-
eluted with
129/138/160/163

no values

0.01

no values

A1260S

6.41

2.43

0.70

0.58

no values

0.02

no values

A1260G

6.73

2.44

0.66

0.60

no values

0.01

no values

Average of A,S,G
samples

6.54

2.43

0.69

0.58

no values

0.01

no values

Proportion of
congener within
congener group

64%

24%

100%

% of Congener
from total PCBs

measured in
Largemouth Bass

captured in the
Housatonic River

Woods Pond A

7.54

2.26

0.75

0.86

no value

0.70

0.01

Woods Pond B

7.56

2.22

0.78

0.86

no value

0.73

0.01

Woods Pond C

7.63

2.23

0.79

0.84

no value

0.71

0.01

Average of Woods
Pond samples

7.58

2.24

0.78

0.86

0.00

0.71

0.01

Proportion of
congener within
congener group

66%

20%

99%

Deep Reach A

7.75

2.14

0.45

0.86

no value

0.51

0.01

Deep Reach B

7.74

2.14

0.68

0.86

no value

0.51

0.01

Deep Reach C

8.14

2.24

0.51

0.89

no value

0.51

0.01

Average of Deep
Reach samples

7.88

2.17

0.54

0.87

0.00

0.51

0.01

Proportion of
congener within
congener group

69%

19%

98%

1. Frame et al. 1996

2. Buckler 2002

MK01|O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-31

-------
Table 4

Results of Comparisons for Aroclor Standards

Analysis
Method3

Number of
Congeners in
Analysis

Aroclor

Lab

Type of Distance

Distance

AXYS

Duplicate

0.20

1254

Frame

Duplicate

Both

Between Lab

6.7

AXYS

Duplicate

0.25

1260

Frame

Duplicate

0.34-1.7

Both

Between Lab

2.6

AXYS

Duplicate

0.18

1254

Frame

Duplicate

Both

Between Lab

5.8

AXYS

Duplicate

0.24

1260

Frame

Duplicate

0.21-1.1

Both

Between Lab

0.89

Both

Between Aroclor

aSee Section C.7.2.2.3 for explanation of statistical methods

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

-------
Table 5

Euclidean Distance Results for Laboratory Duplicates

Group

Sample Size

Minimum

Median

90th Percentile

Maximum

Amphibian

0.34

0.45

0.53

0.56

Fish

0.23

1.1

2.5

4.9

Sediment

0.26

1.2

2.3

Soil

0.93

1.8

2.9

3.6

Total

0.23

1.1

2.4

4.9

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

-------
Table 6

Euclidean Distance Results of Laboratory Comparisons

Between Group Distance

Within Group Distance Range

Medium

Reach

Lab

Sample
Size

GERG

NEA

Min

Median

90th
Percen-
tile

Max

GERG

0.86

1.9

2.2

2.3

Sediment

NEA

2.2

0.37

1.2

2.7

3.6

PAC

1.9

3.0

0.67

2.0

4.0

7.3

Total

0.37

1.9

3.4

7.3

GERG

1.9

3.2

3.8

3.9

Sediment

NEA

2.2

0.88

1.4

2.0

5.3

PAC

2.5

3.3

0.62

2.2

3.1

4.6

Total

0.62

2.0

3.5

5.3

NEA

0.56

1.5

4.2

5.4

Sediment

PAC

4.3

1.0

1.4

8.6

9.5

Total

0.56

1.4

6.1

9.5

NEA

0.53

1.3

2.3

3.1

Sediment

PAC

5.0

1.8

4.5

Total

0.53

2.5

9.2

Total Sediment

0.37

1.9

4.5

GERG

1.7

2.7

6.0

6.9

Soil

PAC

2.3

0.79

2.2

4.6

7.1

Total

0.79

2.4

5.4

7.1

GERG

0.51

0.9

2.3

4.0

Soil

PAC

1.6

0.61

1.5

2.2

3.1

Total

0.51

1.3

2.3

4.0

GERG

1.5

2.4

4.5

5.0

Soil

PAC

29a

2.4

0.8

1.7

3.8

Total

0.8

1.8

4.2

Total Soil

0.51

1.8

4.2

Total (All)

0.37

1.9

4.4

a Sample ID H3-FL000571-0-0000 removed as an outlier.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

-------
Table 7

Euclidean Distance Results for Media Comparisons0

Reach

Group

Sample
Size

Between Group Distance

Within Group Distance Range

Min

Median

90th Pctl

Max

Amphibian (A)

5.9

4.9

5.5 6.0 15 17 11 6.3 7.3 11 11

4.6

5.6

8.5

9.3

Benthic Invert (BI)

5.9

2.1

2.3

2.4

5.0

3.0

2.9

9.3

6.3

1.7

3.5

4.5

9.2

Bird (B)

4.9

2.1

1.2

1.6

6.5

3.4

3.5

8.5

6.8

1.3

2.9

5.7

7.3

Crayfish (C)

5.5

2.3

1.2

1.8

6.2

2.9

3.3

8.7

6.7

0.39

1.2

2.1

2.8

Fish (F)

6.0

2.4

1.6

1.8

5.7

3.7

3.3

7.4

5.7

0.44

2.4

4.1

9.4

Mammal - Short
Tailed Shrew (ST)

0.81

1.6

2.8

3.8

Mammal - White
Footed Mouse (WF)

6.8

Porewater (P)

5.0

6.5

6.2

5.7

5.8

4.7

4.1

0.7

1.4

2.9

3.9

Sediment (SE)

6.3

3.0

3.4

2.9

3.7

5.8

2.3

9.7

6.8

1.0

2.2

4.3

7.8

Soil (S)

7.3

2.9

3.5

3.3

4.7

2.3

0.0

8.8

5.0

0.9

2.4

6.0

8.0

Earthworm (EW)

9.3

8.5

8.7

7.4

9.7

8.8

0.0

8.0

0.86

1.8

2.6

2.7

Litter Invertebrate (LI)

6.3

6.8

6.7

5.7

4.1

6.8

5.0

8.0

1.4

2.0

2.9

3.1

Total 5A

0.4

2.4

6.5

M K0110:\20123001.096\ERA_PB\ERA_APC7_PB. DOC

C.7-35

-------
Table 7

Euclidean Distance Results for Media Comparisons0

(Continued)

Reach

Group

Sample
Size

Between Group Distance

Within Group Distance Range

Min

Median

90th Pctl

Max

Amphibian (A)

7.1

6.2

6.5

6.9

7.4

9.3

4.6

5.6

8.5

9.3

Benthic Invertebrate
(BI)

7.1

1.3

2.6

4.7

2.7

2.2

3.6

4.3

1.4

2.3

3.0

3.2

Bird (B)

6.2

1.3

2.4

5.6

2.9

2.6

4.6

5.4

0.50

1.5

2.3

6.2

Fish (F)

252

6.5

2.6

2.4

5.7

4.0

3.9

3.8

5.5

0.85

2.6

4.3

9.9

Mammal - Short
Tailed Shrew (ST)

0.0

1.4

2.8

4.9

5.3

Mammal - White
Footed Mouse (WF)

0.0

5.5

Porewater (P)

4.7

5.6

5.7

0.0

4.6

5.5

3.5

3.4

0.9

1.5

2.7

4.4

Sediment (SE)

6.9

2.7

2.9

4.0

4.6

0.0

2.1

4.6

4.9

1.6

2.5

7.8

9.5

Soil (S)

7.4

2.2

2.6

3.9

5.5

2.1

0.0

4.7

4.8

0.70

1.4

2.7

4.7

Earthworm (EW)

0.0

0.89

2.1

2.7

3.6

Terrestrial Litter
Invertebrate (LI)

9.3

3.6

4.6

3.8

3.5

4.6

4.7

0.0

3.1

1.0

1.2

1.5

1.6

Vegetation (V)

4.3

5.4

5.5

3.4

4.9

4.8

3.1

0.9

1.5

3.9

4.1

Total 5B

0.5

2.2

4.7

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC n 1 7/10/2003

-------
Table 7

Euclidean Distance Results for Media Comparisons0

(Continued)

Reach

Group

Sample
Size

Between Group Distance

Within Group Distance Range

Min

Median

90th Pctl

Max

Amphibian (A)

12b

3.1

4.4

1.8

3.3

5.9

Benthic Invert (BI)

2.2

2.3

3.3

2.8

2.4

2.7

1.3

2.0

4.1

4.9

Bird (B)

2.2

1.4

2.9

4.2

2.8

2.7

0.60

1.4

2.4

6.6

Crayfish (C)

2.3

1.4

2.4

4.5

3.2

3.1

0.24

0.93

1.3

3.0

Fish (F)

252

3.3

2.9

2.4

5.2

4.1

4.0

OO
OO

0.85

2.6

4.3

9.9

Mammal - Short
Tailed Shrew (ST)

3.1

3.3

0.5

0.8

1.2

Mammal - White
Footed Mouse (WF)

4.4

3.3

3.2

10.8

Porewater (P)

2.8

4.2

4.5

5.2

3.2

3.9

1.1

1.6

2.5

6.1

Sediment (SE)

2.4

2.8

3.2

4.1

1.4

1.3

2.5

3.7

6.3

Soil (S)

37a

2.7

3.1

4.0

3.9

1.4

0.91

2.1

4.6

Earthworm (EW)

OO
OO

1.2

2.5

3.3

3.6

Total 5C

0.24

2.3

4.3

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC n 1 "11 7/10/2003

-------
Table 7

Euclidean Distance Results for Media Comparisons0

(Continued)

Reach

Group

Sample
Size

Between Group Distance

Within Group Distance Range

Min

Median

90th Pctl

Max

5D/6

Amphibian (A)

5.0

1.8

2.6

4.8

7.4

Fish (F)

249

0.0

4.6

3.0

3.2

8.5

0.74

2.4

3.9

Porewater (P)

4.6

0.0

3.9

5.8

0.70

2.2

3.6

4.3

Sediment (SE)

3.0

3.9

0.0

2.8

9.5

1.7

2.9

11.1

Soil (SE)

3.2

5.8

2.8

0.0

9.3

1.3

1.9

4.6

5.9

Waterfowl - Duck (D)

5.0

8.5

9.5

9.3

0.90

2.7

5.3

Total 5D/6

0.70

2.5

4.5

Total All PSA Reaches

0.24

2.3

4.8

a Sample ID H3-FL000571-0-0000 removed as an outlier
b Sample H3-T012RP39-0-F001 removed as an outlier

0 Highlighted cells withhold text indicate distances greater than the distance between Aroclors 1254 and 1260 (17). Highlights without bold indicate distances
greater than the 90th percentile of the within group distances for all reaches (4.7).

MK01|O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-38

-------
Table 8

Euclidean Distance Results for Among Reach Comparisons0

Group

Reach and
Sample Size

Between Group Distance

Within Group Distance Range

Amphibian

Reach

5AB

5D/6

Min

Median

90th Pctl

Max

5AB

6.0

6.5

4.6

5.6

8.5

9.3

12b

6.0

1.2

1.8

3.3

5.9

5D/6

6.5

1.2

0.0

1.8

2.6

4.8

7.4

Total

1.8

3.7

7.3

Benthic Invertebrates

Reach

Min

Median

90th Pctl

Max

2.9

flHH

3.9

3.5

4.2

6.6

7.2

2.9

3.9

3.5

2.6

3.0

4.0

4.3

4.4

3.9

2.1

3.0

1.7

3.5

4.5

9.2

5,0

3.5

2.1

0.0

1.9

1.4

2.3

3.0

3.2

3.9

2.6

3.0

1.9

1.3

2.0

4.1

4.9

Total

1.4

3.2

4.5

9.2

Passerine Birds

Reach

Min

Median

90th Pctl

Max

0.0

1.1

2.3

1.6

1.1

1.0

2.2

4.5

7.9

1.1

0.0

1.7

0.83

1.4

1.0

3.0

6.1

8.3

2.3

1.7

0.0

1.7

2.7

1.3

2.9

5.7

7.3

1.6

0.83

1.7

0.0

1.4

0.50

1.5

2.3

6.2

1.1

1.4

2.7

1.4

0.0

0.60

1.4

2.4

6.6

Total

0.50

1.8

4.8

8.3

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC ^ q 7/10/2003

-------
Table 8

Euclidean Distance Results for Among Reach Comparisons0

(Continued)

Group

Reach and
Sample Size

Between Group Distance

Within Group Distance Range

Crayfish

Reach

Min

Median

90th Pctl

Max

2.5

0.39

1.2

2.1

2.8

2.5

0.24

0.93

1.3

3.0

Total

0.24

1.0

1.8

3.0

Fish

Reach

5BC

5D/6

Min

Median

90th Pctl

Max

5.1

4.9

3.9

1.1

2.4

4.4

6.8

5.1

1.0

2.4

0.44

2.4

4.1

9.4

5BC

252

4.9

1.0

2.0

0.85

2.6

4.3

9.9

5D/6

249

3.9

2.4

2.0

0.0

0.74

2.4

3.9

Total

0.44

2.5

4.2

Small Mammal - Short
Tailed Shrew

Reach

Min

Median

90th Pctl

Max

3.1

0.81

1.6

2.8

3.8

3.1

0.0

1.4

2.8

4.9

5.3

0.0

0.46

0.85

1.2

Total

0.46

1.8

3.9

5.3

Small Mammal - White
Footed Mouse

Reach

Min

Median

90th Pctl

Max

4.9

6.8

4.9

0.0

8.9

5.5

8.9

0.0

3.2

Total

3.2

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC n 1 AC\ 7/10/2003

-------
Table 8

Euclidean Distance Results for Among Reach Comparisons0

(Continued)

Group

Reach and
Sample Size

Between Group Distance

Within Group Distance Range

Porewater

Reach

5D/6

Min

Median

90th Pctl

Max

2.3

1.2

2.2

0.75

1.4

2.9

3.9

2.3

1.0

0.88

1.5

2.7

4.4

1.2

2.3

2.1

1.1

1.6

2.5

6.1

5D/6

2.2

1.0

2.1

0.0

0.70

2.2

3.6

4.3

Total

0.70

1.8

3.5

6.1

Sediment

Reach

5D/6

Min

Median

90th Pctl

Max

0.0

7.3

7.2

5.6

6.0

4.0

5.3

7.3

0.0

1.7

2.9

3.3

1.0

2.2

4.3

7.8

7.2

1.7

0.0

1.9

2.2

1.6

2.5

7.8

9.5

5.6

2.9

1.9

0.0

1.0

1.3

2.5

3.7

6.3

5D/6

6.0

3.3

2.2

1.0

0.0

1.7

2.9

11.1

Total

0.97

2.5

7.8

Total (without reference sites)

1.0

2.4

4.7

Soil

Reach

5D/6

Min

Median

90th Pctl

Max

0.53

1.0

1.2

0.88

2.4

6.0

8.0

0.53

1.1

0.70

1.4

2.7

4.7

37a

1.0

1.1

1.3

0.91

2.1

4.6

5D/6

1.2

1.1

1.3

0.0

1.3

1.9

4.6

5.9

Total

0.70

1.9

4.7

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC H A\ 7/10/2003

-------
Table 8

Euclidean Distance Results for Among Reach Comparisons0

(Continued)

Group

Reach and
Sample Size

Between Group Distance

Within Group Distance Range

Terrestrial Invertebrates
- Earthworms

Reach

Min

Median

90th Pctl

Max

3.4

5.7

0.86

1.8

2.6

2.7

3.4

0.0

4.9

0.89

2.1

2.7

3.6

5.7

4.9

0.0

1.2

2.5

3.3

3.6

Total

0.86

2.2

3.1

3.6

Terrestrial Invertebrates
- Litter Invertebrates

Reach

Min

Median

90th Pctl

Max

3.8

1.4

2.0

2.9

3.1

3.8

0.0

1.0

1.2

1.5

1.6

Total

1.0

1.5

2.6

3.1

Total All Media/Reaches

0.24

2.3

5.0

Total All Media/Reaches (without sediment reference areas)

0.24

2.3

4.8

a Sample ID H3-FL000571-0-0000 removed as an outlier.
b Sample H3-T012RP39-0-F001 removed as an outlier.

c Shaded cells indicate distances greater than the 90th percentile for the respective media. Shaded cells with bold text indicate distances greater than the distance between
Aroclors 1254 and 1260 (17).

MK01|O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-42

-------
Table 9

Summary of Average Percentages for Selected Congener Groups within the PSA
(Percent of Estimated Total PCB Concentration)

Dominant
Congener

PCB 49

PCB 194

PCB 180

PCB 170

PCB 153

PCB 151

PCB 138

PCB 110

Coeluters

PCB 43

PCB 190

PCB 132/
105/161

PCB 163/
164/158/
160/186

PCB 77/
148

Soil

0.58

1.87

8.04

3.03

9.71

3.00

9.02

2.22

Sediment

1.31

1.92

7.99

2.93

7.67

3.04

7.09

1.92

Porewater

2.37

0.89

4.68

2.09

6.54

3.01

6.32

3.06

Benthic Invertebrates

1.48

1.37

7.31

2.92

13.52

2.81

8.56

1.97

Vegetation

1.67

0.71

4.95

2.08

13.14

3.32

8.54

3.18

Terrestrial
Invertebrates

0.72

1.11

5.69

2.12

11.68

3.56

9.46

1.93

Crayfish

1.77

1.25

8.92

2.50

11.61

3.10

10.08

1.74

Amphibian

0.12

2.99

14.85

5.20

16.75

0.30

9.75

0.07

Fish

1.66

1.27

7.61

2.82

12.42

2.95

9.87

2.31

Waterfowl

0.12

2.32

12.02

4.81

19.19

0.31

12.83

0.21

Bird

1.25

1.42

8.76

2.85

14.78

2.54

9.98

1.69

Mammal

0.04

9.39

15.72

4.55

15.88

0.23

8.42

0.15

Min

0.04

0.71

4.68

2.08

6.54

0.23

6.32

0.07

Max

2.37

9.39

15.72

5.20

19.19

3.56

12.83

3.18

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-43

-------
Table 10

Average Chlorines per Biphenyl Ring for Housatonic River Media, by Reach and

Media Type

Media

Upstream
Reference

Reach
5A

Reach
5A/B

Reach
5B

Reach
5BC

Reach
5C

Reach
5D/6

West
Branch

Amphibian

6.60

6.21

6.48

Benthic
Invertebrate

5.21

5.95

6.09

5.86

5.63

Passerine
Birds

6.06

6.21

6.15

6.04

6.13

Crayfish

6.18

6.00

Fish

5.66

6.16

6.07

5.98

Mammals

6.86

6.85

6.80

Sediments

4.37

6.35

6.32

6.11

5.91

5.90

Soils

6.28

6.31

6.21

6.22

Terrestrial
Invertebrates

6.23

6.20

6.18

Vegetation

5.76

Waterfowl

6.40

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-44

-------
Table 11

ANOVA for Sediment PCB Chlorination Level Across Reaches within the

Housatonic River PSA

Note: Outlier sample H4-SE000871-0-0000 with average chlorination level of 4.97 from Reach
5D/6 removed from analysis - causing ANOVA assumption violations

Chlorination Levels by Reach

Reach 5A

Reach 5B

Reach 5C

Reach 5D/6

Mean

6.35

6.32

6.11

6.00

Sample Size

ANOVA Output

Sum of Sq

Mean Sq

F Value

Pr(F)

Reach

1.708

0.569

20.029

6.20E-10

Residuals

2.444

0.028

Multiple Comparisons Tests (95% Simultaneous Confidence Intervals for Specified
Linear Combinations, by the Tukey method)

Estimate

Std.Error

Estimate

Lower
Bound

Upper
Bound

Signif-
icant?

H3A-H3B

0.0279

0.0581

0.480

-0.124

0.18

H3A-H3C

0.242

0.0413

5.860

0.134

0.351

YES

H3A-H3D4

0.352

0.0562

6.263

0.205

0.5

YES

H3B-H3C

0.215

0.0592

3.632

0.0596

0.37

YES

H3B-H3D4

0.325

0.0704

4.616

0.14

0.509

H3C-H3D4

0.11

0.0573

1.920

-0.0402

0.26

Conclusion: Reach 5A and 5B are not significantly different, Reach 5C and Reach 5D/6 are not significantly
different but all other reaches are significantly different.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

-------
0 5 10 15

Euclidean Distance

2 Figure 1 Summary of Euclidean Distances for Laboratory Comparisons.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-46

-------
20
18
16
14

Sediment - Reach 5A

m
o

0)
Q.

IS 12
o

l~ 10

M— 1W

¦e 8
6
4
2
0

IGERG ~ NEA ~ PAC

J1^:_ J-Iff I

ItJIiUi

00
CM

o ^r

h- h-

C0 CO CO CO

o o o o

CL CL CL CL

h-
00

CD O
CD CD
CO
O
CL

h-
h-

CO CO

o o

CL CL

CD 00
CM

CO CO

o o

CL CL

N- N-
LO (Oh--

O CO
00 00

CO CO CO CO CO CO

™ o o o o o

CL CL CL CL CL

"sf

cri
o

N CD ^ (D N CD

CO CO CD O O O

r; r; r (\j (Nj (Nj

CO CO CO CO CO CO

o o o o o o

CL CL CL CL CL CL

Sediment - Reach 5B

20
18

m 16

2 14

1 12
o

l~ 10

M— 1W

£ 8 I-

a 6

4
2
0

~ NEA ~ PAC

"sf

l^-

n ffi

¦M.

j£-

N-

t—

N-

"sf

N-

t—

N-

T—

t—

N-

CO
O
CL

o o

CL CL

"sf

cd
O
CL

Figure 2 Profile Plots for Laboratory Comparisons by Media and Reach. Error

bars are + 2 standard deviations.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-47

-------
20
18

m 16

t 14

5 12
o

l~ 10

M— 1 w

* 8
6
4
2
0

Sediment - Reach 5C

0)
Q.

IGERG ~ NEA ~ PAC

A- j

,.m ..it

Iff

-3"

(J)

(»

c )

C )

t—

CD
O
CL

co o ^r

cn ^ r*- r*-

di CQ CD CD

O O O O

CL CL CL CL

00
CD

CD CD
CD ^
O
CL

h-
h-
O

CD CO CO ¦<- h- h-
T- CN T- LO CD h-

O CO h-
00 00 00 00

CD CD

o o

CL CL

C0C0COC0C0C0C0C0
OO^OOOOO
CLCL^CLCLCLCLCL
CD

Sediment Reach 5D/6

T—

¦*—

00
CD
O
CL

CD CD
O O
CL CL

T~ h- h-
LO CD h-

CQCQCOCQCQCQCQCQCQCQCQ
OO^OOOOO
CLCL^CLCLCLCLCL
CD

1 1 1 rC

(»

C )

t—

Figure 2 (cont.) Profile Plots for Laboratory Comparisons by Media and Reach.

Error bars are + 2 standard deviations.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-48

-------
Soil - Reach 5A

COC0C0C0C0C0C0CO

^ooooooo

"*0.0.0.0.0.0.1
O)

co
o

Soil - Reach 5B

re
•*->

IGERG nPAC

oo
CM

o
r*-

r*-

CO
O
CL

co
O
0.

co
O
0_

co
O
o_

t r

(J)

(»

c )

C )

t—

cocococococococo

Figure 2 (cont.) Profile Plots for Laboratory Comparisons by Media and Reach.

Error bars are + 2 standard deviations.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-49

-------
Soil - Reach 5C

¦ GERGnPAC

CD
O
CL

co o ^r

cm ^ h- h-

ni ni di ni

o o o o

CL CL CL CL

00
CD
O
CL

CD CD
CD
O
CL

O) CO CO
T- CN T-

LO CD

00 00 00 00

"3-

ti
O
CL

CD CD CD CD
O O O O
CL CL CL CL

CD CD
O O
CL CL

CD Is-

O O

CN CN

CD CD

O O

CL CL

a>
o
CN
CD
O
CL

Figure 2 (cont.) Profile Plots for Laboratory Comparisons by Media and Reach.

Error bars are + 2 standard deviations.

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-50

-------
Reach 5A

CO ^ O

c\j h- h-

m m m m

o o o o

Q_ Q_ Q_ Q_

\—

t—

r-~

t—

C\J

T—

m
O
Q_

Reach 5C

t—

T~~

T—

¦*-

h-"

^h-h-ocoh-cD^tcQh-cD
LOCQh-COCOCOCOCDOOO
TTTTTTTTCNCNCNl
CQCQCQCQCQCQCQCQCQCQCQ

m
O
Q_

Q_ Q_ Q_ Q_

Q_ Q_ Q_ Q_ Q_

Figure 3 Comparison of Congener Profiles for Fish and Crayfish. Error bars

are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-51

-------
16

Reach 5A

m 12
o

3 10
o

<1)

o
5

~ Sediment ~ Benthos ~ Fish

^Pt'tFtTT,

DO
O
Q_

LO CD O h- CD 00 00
G) O) h- (N t-

CD O ^ ^ "?~

t-qOt-CQCQCO_
Q_ T T O O ™ O
"" "" CL Q_ T Q_

CO
00

CD
O
Q_

cq co

o o

Q_ Q_

I^I^OCOI^CD^CDI^CD
CDI^OOOOOOOOOOOO

CQCQCQCQCQCQCQCQCQCQCQ

O O O
Q_ Q_ Q_

O O O O O O O
Q_ Q_ Q_ Q_ Q_ Q_ Q_

O
Q_

co 12

75 10

Q) v

1 4

Reach 5B

~ Sediment ~ Benthos nFish

-K

5e&

f*fFLL

r-~

T—

N-

r-~

T—

r-~"

CQ
O
CL

Figure 4 Comparison of Congener Profiles for Sediment, Benthic
Invertebrates, and Fish. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-52

-------
16

Reach 5C

m 12

3 10
o

a)
o

~ Sediment ~ Benthos ~ Fish

ffi

DO
O
Q_

T—

O
Q_

Reach 5D/6

4—

-4-»

~ Sediment ~ Fish

Jg r

1 I

T n

O
Q_

T—

t—

T—

CLi

cri

CLi

T—

CLi

cri

T—

CLi

cri

O
LL

cri
o

O CO h- CD
LO CO h- CO CO CO CO

Figure 4 (cont.) Comparison of Congener Profiles for Sediment, Benthic
Invertebrates, and Fish. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-53

-------
16

Reach 5A

GO -in

« 10
+¦»

\z s

2
0

0)
£
0)
Q.

~ Sediment nSoil ~ Passerine Bird

i^rLa. —i&ii'ffllTI Inn

aFrr

T—

t—

r-~

T—

t—

T—

"vl"

CO
O
CL

16
14

GO -|o

« 10
•*->
o

l~ 8

6
4
2
0

Reach 5B

0)
£
0)
Q.

~ Sediment nSoil ~ Passerine Bird

¦=2=x

5j2k«

00
CM

\—

t—

r-~

T—

t—

T—

t—

h-*

"^1"

O
Q_

Figure 5 Comparison of Congener Profiles for Sediment, Soil, and Birds. Error

bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-54

-------
Reach 5C

r-~

\—

r-~

t—

r-~

T—

r-~

"<—

T—

¦*""

r-~"

,!:d"

CO
O
CL

m 12

1 10

!: s

Reach 5D/6

0)
o

0)
Q.

6
4
2
0

~ Sediment ~ Soil

-I L: [ft

iffiTi fa.rii,-1, [Tl, ml U11J&

CQ CQ

O O

Q_ Q_

00
CM

h-
00

CD CD

CD 00 00

¦*— c\j •*—

r-~ r-~
LO CO h-

00 00 00 00

co m

o o

Q_ Q_

CD O O

C\j CNJ

cdcdcdcdcdcdcdcdcdcdcdcdcd
oo^oooooooooo
clclTclclclclclclclclclcl

m
O
Q_

h- cd
o
CM

cd
O
Q_

Figure 5 (cont.) Comparison of Congener Profiles for Sediment, Soil, and Birds.

Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-55

-------
40

Reach 5A

-------
Reach 5C

m m

o o

Q_ Q_

T—

t—

¦*-

T—

"vl"

O
Q_

Figure 6 (cont.) Comparison of Congener Profiles for Mammals and the Average
Profile for Sediment, Soil, Benthic Invertebrates, Crayfish, Fish, and Birds. Error

bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-57

-------
Reach 5A

0Q
O
Q.

-------
Reach 5C

DO
O
Q_

00
CM

,:3-
,:3-

CD
CD

O
Q_

CD
O
Q_

O
Q_

1^
oo

CD
O
Q_

Is**;
o

cq co

o o

Q_ Q_

N CD CO CO

h- h-

CO N O

LOCDI^OOOOOOOOCD

0 N O)

(XI CD CO _

O O ™ o

Q_ Q_ ^ Q_

cri
o

o o o

CM CM CM

CDCDCDCQCQCQCQCDCQCDCD

oooooooooo

Q_Q_Q_Q_Q_Q_Q_Q_Q_Q_

Reach 5D/6

CQ
O
D.

5 15
o

~ Average of Six Media ¦ Amphibians

20 ¦

•£ 10

5 ¦

"T"

¦=^i

O
Q_

1^
oo

cri
o

CD CD

cri
o

h- CD 00 00

1^
o

m cd h-

O CO N CD
00 00 00 00

CO h- CD

cq co

o o

Q_ Q_

o o o

CM CM CM

cricricocricricricricricricricricricricri
oo^ooooooooooo

LLLL^LLLLLLLLLLLLLLLLLLLLLL

cd
o

Figure 7 (cont.) Comparison of Congener Profiles for Amphibians and the
Average Profile for Sediment, Soil, Benthic Invertebrates, Crayfish, Fish, and
Birds. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-59

-------
Reach 5A

O 10

* ft
O o

e 4

~ Average of Six Media ~ Aroclor 1260 ~ Aroclor 1254

LL)

iDi.

Lfin

JSl

O 'vf

h- h-

h-
oo
m
O
Q_

CD CD

CD 00 00
¦*- CM <<-

m m

o o

Q_ Q_

^h-h-OCOh-CD^CQh-CT)
LOCOh-COCOCOCOCDOOO

CQCQCOCQCQCQCQCQCQCQCQCQCQCQ
00^)00000000000
Q_Q_^Q_Q_Q_Q_Q_Q_Q_Q_Q_Q_Q_

m
O
Q_

Reach 5B

T—

t—

r-~

t—

T—

t—

h-*

O
Q_

Figure 8 Comparison of Congener Profiles for Aroclor Standards and the
Average Profile for Sediment, Soil, Benthic Invertebrates, Crayfish, Fish, and
Birds. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-60

-------
14

o 10

o 8

° 6

e 4

Reach 5C

~ Average of Six Media ~ Aroclor 1260 ~ Aroclor 1254

illi

IZfco.

n rr i I

¦nT

r-~

t—

r-~

t—

T—

r-~

T—

x—

r-~"

O
Q_

Reach 5D/6

¦nT

r-~

¦nT

r-~

N-

T—

m m

o o

Q_ Q_

^r^r^ocor^cD^rcDr^a)
LncDr^ooooooooa)ooo

CQCQCOCQCQCQCQCQCQCQCQCQCQCQ
OO^OOOOOOOOOOO
Q_Q_^Q_Q_Q_Q_Q_Q_Q_Q_Q_Q_Q_
CD
¦nT

Figure 8 (cont.) Comparison of Congener Profiles for Aroclor Standards and the
Average Profile for Sediment, Soil, Benthic Invertebrates, Crayfish, Fish, and
Birds. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-61

-------
Sediment

m
o
a.

5 15
o

~ 10

Q- c

¦ Reach 5A

~ Reach 5B

~ Reach 5C

¦ Reach 5D/6

il J—BfcalffMiftiJ ILi

t- co o

Q-j CN h- h~

o 0Q CO 0Q 0Q

CL o o o o

Q_ Q_ Q_ Q_

LO (J) O
t- O) G)
CO

O
Q_

CQ CQ
O O
Q_ Q_

oooooT-r^r^ocor^o^rcor^o)

T-CNT-lDCOr^COOOOOOOOOOO

CQCQOOCQCQCQCQCQCQCCiCQCQCQCQ
OO^OOOOOOOOOOO
Q_Q_^:Q_Q_CLCLQ_Q_Q_Q_Q_Q_Q_

Fish

OQ
O
CL

5 15
o

-£ 10

-------
25

Amphibians

3 15
o

¦ Reach 5AB
~ Reach 5C

¦ Reach 5D/6

¦£ 10

o- c:

is.

fe l

j&k

_^nJ

tit.

N O) 00 00

CD
O
Q_

LOCOh-COCOCOCOO)

ID N O)

CD CD CD CD

O O O O

CD CD

O O

CQCQCOCQCQCQCQCQCQCQCQCQCQ
OO^OOOOOOOOOO

Q_Q_ Q_Q_CLCLCLCLCLCLCLCL

Ct)

cci
O

Figure 9 (cont.) Comparison of PCB congener profiles across reaches, for
selected media. Error bars are 2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-63

-------
o
o
CO
CM

O
Q_

PC 1 Score

Figure 10 Plot of Scores for first two principal components for all media in the
PSA combined. Only the centroids of the PCA scores for each group are shown

on the plot for simplicity.

Symbols:

1254 = Aroclor 1254 Standard; 1260 = Aroclor 1260 Standard; E2 = NAPL Samples from East Street Area 2; LS =
NAPL Samples from Lyman Street Area; N2 = NAPL Samples from Newell Street Area 2; A = Amphibians; B =
Birds; BI = Benthic Invertebrates; C = Crayfish; D = Waterfowl (ducks); EW = Earthworms; F = Fish; LI = Litter
Invertebrates; P = Porewater; S = Soil; SE = Sediment; ST = Short-tailed Shrew; V = Vegetation; WF = White-
footed Mouse

MK01 |O:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-64

-------
Sediments

o 15

ra
o

~ East Branch ~ West Branch ¦ Reach 5A

~ Reach 5C ~ Reach 5D ¦ Reach 6A

~ Reach 5B

~ Reach 6B

Porewater

m
o
o.

15
o

Figure 11 Concentrations of indicator congeners and congener groups across
PSA reaches. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-65

-------
20

Terrestrial Invertebrates

¦ Reach 5A ~ Reach 5B ~ Reach 5C

m
o
o.

ra 10

ili

xj.

¦Tl

Soils

Figure 11 (cont.) Concentrations of indicator congeners and congener groups
across PSA reaches. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-66

-------
Benthic Invertebrates

Crayfish

m
o
o.

8 10

¦ Reach 5A

~ Reach 5C

1st

An.

Figure 11 (cont.) Concentrations of indicator congeners and congener groups
across PSA reaches. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-67

-------
Amphibians

Fish

~ East Branch ¦ Reach 5A

~ Reach 5BC

~ Reach 6

8 15

tit

XT.T,

Figure 11 (cont.) Concentrations of indicator congeners and congener groups
across PSA reaches. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-68

-------
Passerine Birds

Waterfowl

30
25
20

m
o
o.

2 15
o

5,„

5
0

~ Reach 5D

¦ Reach 6

x_L

£6

Figure 11 (cont.) Concentrations of indicator congeners and congener groups
across PSA reaches. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-69

-------
Small Mammals

Figure 11 (cont.) Concentrations of indicator congeners and congener groups
across PSA reaches. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-70

-------
35

¦ Porewater

~ Vegetation

~ Fish

~ Sediment

¦ Terrestrial invert

~ Waterfowl

~ Soil

¦ Crayfish

~ Passerine

~ Benthic invert.
¦ Amphibian

~ Mammal

£ 20

re
+¦»
o

PCB 49/43

PCB 110/77/148

PCB

138/163/164/158/160/186

PCB 151

" 20

re
+-»
o

¦ Porewater ~ Sediment nSoil ~ Benthic invert.

~ Vegetation ¦ Terrestrial invert ¦Crayfish ¦ Amphibian

~ Fish ~ Waterfowl ~ Passerine ~ Mammal

PCB 153/132/105/161

PCB 170/190

PCB 180

PCB 194

Figure 12 Patterns of selected PCB congener concentrations across media, all
PSA reaches combined (zero substitution for
-------
o
o

re
o

18
16
14

o 12

re
o

10
8
6
4
2
0

¦ Porewater

~ Soil

~ Vegetation

¦ Crayfish

~ Fish

~ Bird

¦ Porewater
n Soil

~ Vegetation

¦ Crayfish

~ Fish

~ Bird

~ Sediment

~ Benthic Invertebrates

¦ Terrestrial Invertebrates

¦ Amphibian

~ Waterfowl

~ Mammal

PCB77

~ Sediment

~ Benthic Invertebrates

¦ Terrestrial Invertebrates

¦ Amphibian

~ Waterfowl

~ Mammal

PCB 126

Figure 13 Concentrations of non-ortho PCB congeners 77 and 126 across
media, with single outlier benthic invertebrate sample (MCDIET) removed, and
non-detect set to one-half detection limit. Error bars are +2 standard deviations.

Note: boxplots are based on the minimum, 1st Quartile, median, 3rd Quartile and maximum values for each group

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-72

-------
PCB-126

PCB 126 (< DL set to 0)

100

80 "

m
O
CL

60 "

? 40-

20 "

PCB 77 (
-------
Soil

0 40

\z 30
o

-4-»

" 20

¦ Reach 5A

~ Reach 5B

~ Reach 5C

¦ Reach 5D/6

T T

mono

T—'

Xt,T-

tri

tetra penta hexa hepta

octa

nona

deca

Sediment

O 40

3
o

¦ Reach 5A

~ Reach 5B

~ Reach 5C

¦ Reach 5D/6

jLL.

-CfiEa

mono

tri

tetra

penta

hexa

hepta

octa

nona

deca

Figure 15 Homolog composition plots for Housatonic River media, subdivided
by reach. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-74

-------
60

Terrestrial Invertebrates

50 ¦

0 40

\z 30
o
•*->

<1)

" 20

¦ Reach 5A
~ Reach 5B
n Reach 5C

10 ¦

I T

XXI,

mono di

tri tetra penta hexa hepta octa nona deca

Benthic Invertebrates

50 ¦

0 40

" 20

¦ Reach 5A

~ Reach 5B

~ Reach 5C

30 ¦

10 ¦

T .J

L x La

mono di tri tetra penta hexa hepta octa nona deca

Figure 15 (cont.) Homolog composition plots for Housatonic River media,
subdivided by reach. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-75

-------
Amphibians

50 ¦

0 40

\z 30
o
•*->

<1)

" 20

10 ¦

¦ Reach 5A
~ Reach 5C

¦ Reach 5D/6

I XI*

¦ML

mono di

tri tetra penta hexa hepta octa nona deca

Crayfish

50 ¦

¦ Reach 5A
~ Reach 5C

0 40

30 ¦

" 20

10 ¦

mono di

tri tetra penta hexa hepta octa nona deca

Figure 15 (cont.) Homolog composition plots for Housatonic River media,
subdivided by reach. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-76

-------
Fish

0 40

\z 30
o

-4-»

" 20

¦ Reach 5A
~ Reach 5BC

¦ Reach 5D/6

5ft

mono

tri tetra penta hexa hepta octa nona deca

Small Mammals

O 40

3
o

" 20

¦ Reach 5A

~ Reach 5B

~ Reach 5C

mono

tri

tetra penta hexa hepta octa nona deca

Figure 15 (cont.) Homolog composition plots for Housatonic River media,
subdivided by reach. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-77

-------
60

Passerine Birds

50 ¦

0 40

\z 30
o
•*->

<1)

" 20

¦ Reach 5A

~ Reach 5B

~ Reach 5C

10 ¦

XL T

¦ ¦I

mono di

tri tetra penta hexa hepta octa nona deca

Figure 15 (cont.) Homolog composition plots for Housatonic River media,
subdivided by reach. Error bars are +2 standard deviations.

MK0110:\20123001.096\ERA_PB\ERA_APC7_PB.DOC

C.7-78

-------
APPENDIX C.7, ATTACHMENT 1

RATIONALE FOR SELECTION OF INDIVIDUAL CONGENERS AND

CONGENER GROUPS

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

-------
1 APPENDIX C.7, ATTACHMENT 1

3 RATIONALE FOR SELECTION OF INDIVIDUAL CONGENERS AND

4 CONGENER GROUPS

5 1. INTRODUCTION

6 The purpose of this attachment is to provide the rationale for the selection of "indicator"

7 congeners for detailed assessment of congener profiles (Appendix C.7). Several broad criteria

8 were used to assess whether individual congeners are suitable indicators of PCB transformation;

9 these are detailed in Sections 2 through 5. The criteria were:

10 ¦ Consideration of chemical properties, such as degree of chlorination and

11 hydrophobicity, that mediate the differential biomagnification of congeners (Section

12 2).

13 ¦ Potential for clearance by aquatic macrofauna, such as biotransformation/metabolism

14 by fish (Section 3).

15 ¦ Potential for clearance by terrestrial organisms, such as biotransformati on/metabolism

16 by mammals (Section 4).

17 ¦ Potential for dechlorination by sediment microorganisms (Section 5).

18 A summary of the selected congeners and the rationales for their selection are presented in

19 Section 6 and Table 1.

20 2. CHEMICAL PROPERTIES OF CONGENERS

21 The chemical properties of PCB congeners help determine their environmental and biological

22 fate in aquatic environments. Because differences in chlorination confer different physical and

23 chemical properties upon the congeners (molecular weight, octanol-water partition coefficient

24 [K0w]), alterations in congener profiles can be expected based on differential partitioning of the

25 congeners. This may affect environmental distribution, toxicity, and potential for accumulation

26 of PCB congeners (Hill and Napolitano 1997).

MK01\O:\20123001.096\ERA_PB\era_at1C7_pb.doc

-------
1 The degree of biomagnification of PCBs congeners is influenced primarily by the degree of

2 chlorination, as reflected in the log octanol-water partition coefficients (Kow). The more

3 chlorinated congeners (heavier homolog groups) with greater K0w values are therefore expected

4 to increase in concentration with increasing ecosystem compartment. Differences in

5 biomagnification factors across congeners could result in congener profiles changes.

6 ¦ Russell et al. (1999) investigated the role of chemical and ecological factors in the

7 bioaccumulation of chlorinated organic contaminants in the Detroit River aquatic

8 ecosystem. Biomagnification, measured as the increase in lipid-based chemical

9 concentrations in predator over that in prey, was observed for high Kow chemicals

10 (log Kow > 6.3), but was less evident in moderate Kow substances (log Kow 5.5-6.3),

11 and was not evident in lower Kow substances.

12 ¦ Koslowski et al. (1994) found that in a Western Lake Erie food web, PCB

13 biomagnification was Kow dependent and was more predominant for substances with

14 1 og Kow greater than 6.0.

15 "In Lake Ontario, not only does total PCB content increase with trophic level, but the

16 chlorine content of the PCBs increases at the higher trophic levels (Oliver and Niimi

17 1988).

18 ¦ Parkerton (1993) showed that the degree of biomagnification shows a parabolic

19 relationship to Kow, with maximum biomagnification near Kow of 7, and lower

20 degrees of biomagnification above or below this value.

21 ¦ Niimi (1996) evaluated the trend in PCB congener concentrations over multiple

22 trophic levels in the Lake Ontario food web. They observed a trend of increasing

23 PCB concentrations with trophic level consistent with similar studies. Further

24 examination among homologs indicated mean biomagnification factors (BMFs)

25 increased with chlorine content among four fish species.

26 In addition to Kow, contaminant biomagnification is influenced by the stereochemistry of

27 congeners (Koslowski et al. 1994).

28 To address potential alterations in congener profiles that result from differences in chlorination

29 level, the following approaches were applied in the congener evaluation:

30 ¦ Congeners were selected to span a wide range of chlorination (i.e., tetra- through

31 octa-chlorinated groups), spanning the range of congeners that comprise the majority

32 of Aroclor 1260.

33 ¦ The congener evaluation considered the homolog profiles (i.e., considered whether

34 groups of congeners with the same chlorination level and similar Kow behaved

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

-------
1

similarly). For example, Willman et al. (1997) observed a systematic enrichment of
penta-, hexa- and heptachloro congeners and a systematic depletion of trichloro
congeners with increasing compartment (trophic level) in a Green Bay Wisconsin
food web.

5 3. CLEARANCE BY AQUATIC ORGANISMS

6 PCB clearance from aquatic biota can occur through non-metabolic routes or through

7 metabolism of these components to readily excretable daughter products (Metcalfe and Metcalfe

8 1997). A number of PCB congeners are biotransformed by fish (Sijm 1992, Niimi 1996). The

9 initial reaction that degrades the compounds is probably an oxidation reaction, catalyzed by the

10 cytochrome P-450 monooxygenase system (Sijm 1992). However, fish appear to be less capable

11 of metabolizing PCBs than other animal groups.

12 In contrast to mammals, fish have only a limited ability to biotransform PCBs containing a

13 higher number of chlorine atoms (Lech and Peterson 1987). Mixed function oxidase enzyme

14 systems that could transform PCBs are present in fish but activity levels are lower than in

15 mammals (Niimi and Oliver 1983). In general, while mono-, di-, and tri-chlorobiphenyl

16 congeners are readily biotransformed by several species of fish, congeners with higher amounts

17 of chlorination tend to be more slowly biotransformed. Fish do not readily metabolize or excrete

18 PCBs with four or more chlorines (Coristine et al. 1996). The overall result of this reduced

19 capacity to biotransform many PCB congeners contributes to the persistence of PCB congeners

20 that are less persistent in mammals (Lech and Peterson 1987).

21 Biotransformation differences between aquatic organisms and mammals may result in congener

22 profile differences across media. In addition to these broad differences, specific PCB congeners

23 often exhibit different levels of biomagnification, related to differences in rates of metabolic

24 clearance of these compounds (Metcalfe and Metcalfe 1997).

25 3.1 RAPIDLY CLEARED PCB CONGENERS

26 Metabolic clearance of PCBs is partially dependent upon the structure of individual congeners,

27 specifically the chlorine substitution pattern at ortho-, meta-, and para- positions of the biphenyl

28 rings. In order to be metabolized, the PCB molecule must have at least one region with an

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

7/10/2003

-------
1

adjacent pair of carbons without chlorine substitution at either the meta-para or ortho-meta
positions (Metcalfe and Metcalfe 1997). It has been observed that coplanar PCBs, such as PCBs
77, 126, and 169, exhibit different partitioning behavior from other PCB congeners. The ability
of fish and crustaceans to metabolize PCBs 77 and 126 has been suggested by several
researchers based on the lower proportions of these congeners compared with persistent PCB
congeners such as PCB 153 (Kannan et al. 1995; Metcalfe and Metcalfe 1997). Although there
is conflicting evidence regarding the role of ortho-substitution in the uptake of PCB congeners,
most studies suggest that accumulation of toxic non- and mono-ortho substituted congeners is
different from other PCB congeners:

¦ Metcalfe and Metcalfe (1997) investigated the persistence of numerous congeners and
found that PCBs 77, 126, 151, and others were depleted relative to persistent
congeners such as PCB 153. Biomagnification factors for these congeners were
consistently lower than compounds of similar K0w-

¦ Willman et al. (1997) found that many coplanar congeners, notably PCB 77, are
depleted with increasing compartment. No systematic enrichment in mono- or non-
ortho- substituted congeners was observed in their study, in spite of higher K0w
values for these congeners relative to their homolog groups. The non-ortho congener
77 decreased in relative abundance relative to both tPCB and within its homolog
group with increasing compartment.

¦ Willman et al. (1997) reports results from studies in the Northwest territories, the
Baltic Sea, and the Netherlands, in which PCB 77 was found to be depleted in fish
tissues, relative to other PCB congeners.

¦ PCB 77 had the fastest rate of elimination of non-ortho PCBs from experimental
trout; the literature has consistently reported that PCB 77 is more susceptible to
elimination that would be predicted from its K0w (Coristine et al. 1996). PCB 126
had relatively slow elimination compared to other coplanar congeners in this study.

¦ Niimi (1996) observed several congeners (PCB 77, 126, 169) that had lower PCB
concentrations in mysids and fish species than other congeners with a similar level of
chlorination and/or K0w value. Overall, the author found that while PCB partitioning
is driven primarily by chlorine content and K0w, certain congeners, particularly
specific non-ortho substituted congeners (77, 126, 169) exhibit reduced uptake.

¦ Koslowski et al. (1994) studied the distribution of 42 PCBs, including three coplanar
congeners (77, 126, 169) in the food web of Western Basin of Lake Erie. A
multivariate assessment indicated that although most congeners had similar
distributions, the coplanar congeners exhibit different distributions. There was no
evidence of biomagnification of congener 77 in either the pelagic or benthic food

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

-------
1

chains. The exposure dynamics of PCB 77 appeared to be different from other PCB
congeners.

¦ Willman et al. (1997) showed that PCB 77 was significantly depleted in fish tissues
relative to Green Bay sediment concentrations, in contrast to most other PCB
congeners which exhibited enrichment. The depletion of congener 77 in fish
suggested that metabolic degradation might be occurring in Green Bay.

7 3.2 SLOWLY CLEARED PCB CONGENERS

8 Congeners that have two adjacent positions occupied by chlorine atoms, such as PCB 153 and

9 PCB 187, resist metabolic degradation by monooxygenases found in fish livers (Hill and

10 Napolitano 1997). Combined with hydrophobicity and biomagnification, this can result in

11 disproportionately high concentrations in upper trophic levels, especially game fish and other top

12 predators.

13 Morrison et al. (1999) applied a food web PCB bioaccumulation model to Lake Ontario aquatic

14 biota. The calibration of the model considered PCB congeners with both high and low metabolic

15 potentials. The latter category included PCBs 31/28, 52, 66, 90/101, 118, 105, 153, 138, and

16 180, which were considered to be environmentally persistent.

17 Hill and Napolitano (1997) investigated relative PCB congener patterns in an eastern Tennessee

18 contaminated stream (East Fork Poplar Creek, a site with degree of chlorination similar to the

19 Housatonic River). PCB congeners 153 and 187, purported to be resistant to metabolic

20 degradation by monooxygenases found in fish liver, were enriched in fish tissue relative to

21 periphyton. In contrast, PCB congeners 101, 105, 110, and 138 were slightly more abundant in

22 periphyton relative to fish. The latter congeners are metabolizable by P-450 enzymes. However,

23 overall relative concentrations of various congeners were roughly similar, suggesting that

24 metabolic transformation processes were subtle. All of these congeners have low rates of

25 metabolism relative to the coplanar congeners described in Section 3.1.

27 Biotransformation of PCBs in various organisms, such as mammals and birds, starts with an

28 arene intermediate and results in mono and dihydroxylated products (Sijm 1992; Coristine et al.

26 4. CLEARANCE BY TERRESTRIAL ORGANISMS

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

7/10/2003

-------
1 1996). The main mechanism of PCB metabolism in vertebrates is monooxygenase-catalyzed

2 binding of singlet oxygen on biphenyl rings, ultimately leading to degradation to a more polar

3 hydroxylated PCB (Coristine et al. 1996). Not all compounds are metabolizable, however. In

4 order to be metabolized by this mechanism, PCB congeners must have at least one region with

5 an adjacent pair of unsubstituted carbons at either the meta-para or ortho-meta positions. For

6 example, 2,2',4,4',5,5'-hexachlorobiphenyl (PCB 153) did not produce metabolites in the excreta

7 of rats, pigeons, or trout (Sijm 1992).

8 Studies on mammals indicate that many tetra- to hexa-chlorobiphenyls can induce microsomal

9 enzyme activity. While there is considerably interspecies variability in half-life times of

10 elimination, the overall elimination rates are higher in mammals than aquatic animals (Niimi and

11 Oliver 1983). The higher elimination rates of terrestrial biota can result in modified congener

12 profiles, as intermediate K0w PCBs are cleared to a greater extent from mammals than from

13 aquatic biota. However, even in birds and mammals, metabolism of highly chlorinated PCBs is

14 extremely slow (Coristine et al. 1996).

15 Metcalfe and Metcalfe (1997) note a specific formation (congeners with ortho-meta vicinal

16 hydrogens and one ortho- chlorine) that is non-persistent in marine mammals and birds relative

17 to their prey. Congeners with this formation include PCBs 66, 105, 118, and 156.

18 5. BIOTRANSFORMATION I METABOLISM BY MICROORGANISMS

19 Microorganisms usually degrade PCBs via a dioxygenase attack that results in the formation of

20 chlorobenzoic acids, chloroacetophenone or further biodegradation products. However, the rate

21 of aerobic biodegradation of PCBs tends to decrease as the chlorination level increases

22 (Hartkamp-Commandeur et al. 1996). Due to the highly chlorinated source materials (i.e.,

23 Aroclor 1260) found in the Housatonic River, anaerobic degradation is the most important

24 degradation process. Anaerobic dechlorination of PCBs can also occur followed by oxidation

25 reaction (Sijm 1992).

26 In situ dechlorination of PCBs attributed to sediment microorganisms has been documented in

27 the Hudson River (New York), Silver Lake, Woods Pond, and New Bedford Harbor

28 (Massachusetts), the St. Lawrence River (New York), as well as several other locations (Bedard

MK01 |O:\20123001,096\ERA_PB\ERA_AT 1 C7_PB.DOC 6 7/10/2003

-------
1

and May 1996, Bedard and Van Dort 1998, Bedard and Quensen 1995). The dehalogenation of
PCBs by anaerobic microorganisms can occur via several different processes, some of which
preferentially remove meta- and para- chlorines (Bedard and Quensen 1995), and some of which
reduce ortho chlorines (Hartkamp-Commandeur et al. 1996). The precise nature of the anaerobic
degradation is related to a number of environmental factors including PCB chlorination level,
temperature, nutrients, bioavailability, and co-contaminants (Wu et al. 1997).

Bedard and Quensen (1995) outlined individual dechlorination processes observed in the
laboratory, each designated by a letter code (M, Q, H', H, P, and N). Although various
combinations of processes may occur in nature, the series of mother-daughter products for each
allows for the fingerprinting of the congener alterations that occur in nature. Dechlorination
processes M, Q, and H' involve dechlorination of lower molecular weight congeners (i.e.,
Aroclor 1242 congeners) that are not abundant in Housatonic River media. Process H has been
reported for Aroclor 1260, but not within the Housatonic River system. The latter two processes
(N, P) have been observed in the Housatonic River system (Silver Lake, Woods Pond) and
elsewhere, and are relevant to the transformation of Aroclors 1254 and 1260 (Bedard and
Quensen 1995). These processes are summarized briefly:

¦ Process N - This process selectively removes meta- chlorines, but only those that are
flanked by at least one chlorine in the para- or ortho- position. Mother PCB
congeners 170, 180, and 194 are progressively dechlorinated, resulting in
accumulations of numerous daughter products including PCBs 47 and 49.

¦ Process P - This process removes only para- chlorines that are flanked by at least one
meta- chlorine (Bedard and May 1996). The largest decreases in this process are
observed for PCB congeners 85, 87, 153, 138, and 149, and large increases of PCBs
44, 49, 52 and 92 are observed.

Bedard and May (1996) and Bedard et al. (1996, 1997) investigated in detail the evidence for
microbial dechlorination of PCBs in Woods Pond sediments. A quantitative mother-daughter
analysis of congener distributions was conducted that indicated that select tri-, tetra-, and penta-
PCBs were enriched in Woods Pond sediments relative to Arcolor 1260. Conversely, select
hexa- and hepta- PCBs were decreased. The authors concluded that the congener distribution of
sediment PCBs could not be explained by combinations of unaltered Aroclors, and that in situ
dechlorination has occurred. The largest losses of individual congeners were for PCBs 149, 153,

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

-------
1 138, 180, and 170. The changes observed in the Woods Pond sediment profiles were consistent

2 with the dechlorination processes N and P. The authors concluded that the extent of

3 dechlorination in Woods Pond is modest compared to dechlorination of PCB mixtures at other

4 sites. In contrast to the Hudson River and Silver Lake sites, for which nearly all meta and para

5 chlorines were dechlorination targets, even the most dechlorinated Woods Pond samples had lost

6 "only 13.7% of the meta- and para- chlorines" (Bedard and May 1996). Therefore, the overall

7 degree of "PCB dechlorination activity under these intrinsic conditions was limited" (Deweerd

8 and Bedard 1999).

9 In summary, anaerobic dechlorination of Housatonic River sediments has been observed. This

10 process is expected to result in modifications of the congener profiles, as well as shifts in the

11 homolog profiles, relative to Aroclor 1260 parent materials. However, the process is expected to

12 result in profile changes that are relatively subtle. In some cases, depletion of a congener (e.g.,

13 PCB 180 to PCB 153 via Process N) would be compensated by enrichment along a different

14 portion of the dechlorination series (i.e., PCB 194 to PCB 180 via Process N). Selection of

15 individual congeners along these "Bedard series" (both mother and daughter products) allows for

16 a rigorous evaluation of whether significant congener profile changes have occurred.

18 Based on the rationales presented in Sections 2 through 5, the following PCB congeners were

19 selected for detailed evaluation:

31 Table 1 summarizes the environmental partitioning attributes of the selected congeners and the

32 rationale for their selection. Some of the congeners were selected based on their position within

17 6. SUMMARY OF INDICATOR CONGENERS SELECTED

PCB 49 (2,2',4,5'-tetrachlorobiphenyl)
PCB 77 (3,3'4,4'-tetrachlorobiphenyl)

PCB 110 (2,3,3',4',6-pentachlorobiphenyl)
PCB 126 (3,3'4,4',5-pentachlorobiphenyl)
PCB 138 (2,2',3,4,4',5'-hexachlorobiphenyl)
PCB 151 (2,2'3,5,5',6-hexachlorobiphenyl)
PCB 153 (2,2',4,4',5,5'-hexachlorobiphenyl)
PCB 170 (2,2',3,3',4,4'5-heptachlorobiphenyl)
PCB 180 (2,2',3,4,4',5,5'-heptachlorobiphenyl)
PCB 194 (2,2',3,3',4,4',5,5'-octachlorobiphenyl)

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

7/10/2003

-------
1 anaerobic sediment dechlorination processes, some were selected based on their tendency to be

2 metabolized or cleared by aquatic and/or terrestrial biota, and some were selected to provide a

3 cross-section of physical properties of the Aroclor 1260 congener profile. In most cases,

4 multiple selection criteria were applied.

5 Some of the above indicator congeners coelute with other congeners in the laboratory methods

6 applied for most Housatonic River samples (i.e., PAC, GERG, NEA laboratory methods). For

7 this reason, it was necessary to combine the coeluted congeners for assessment in Appendix C.7.

8 However, where coelutions occurred, the congener listed above was found to comprise the

9 majority of the PCB mass within the group.

10 7. REFERENCES

11 Bedard, D.J. and R.J. May. 1996. Characterization of the polychlorinated biphenyls in the

12 sediments of Woods Pond: Evidence for microbial dechlorination of Aroclor 1260 in situ.

13 Environ. Sci. Technol. 30(l):237-245.

14 Bedard, D.L. and H.M. Van Dort. 1998. Complete reductive dehalogenation of brominated

15 bipenyls by anaerobic microorganisms in sediment. Applied and Environmental Microbiology.

16 64(3):940-947.

17 Bedard, D.L. and J.F. Quensen III. 1995. Microbial Reduction Dechlorination of Polychlorinated

18 Biphenyls. Chapter 4 (pp. 127-216) in L.Y. Young (Ed.), Microbial Transformation and

19 Degradation of Toxic Organic Chemicals. Wiley-Liss, 1995.

20 Bedard, D.L., H.M. Van Dort, R.J. May, and L.A. Smullen. 1997. Enrichment of

21 microorganisms that sequentially meta, para-dechlorinate the residue of Aroclor 1260 in

22 Housatonic River sediment. Ijiviron. Sci. Technol. 31:3308-3313.

23 Bedard, D.L., S.C. Bunnell, and L.A. Smullen. 1996. Stimulation of microbial para-declorination

24 of polychlorinated biphenyls that have persisted in Housatonic River sediment for decades.

25 Environ. Sci. Technol. 30:687-694.

26 Coristine, S. G.D. Haffner, J.J. Ciborowski, R. Lazar, M. Nanni, and C.D. Metcalfe. 1997.

27 Elimination rates of selected di-ortho, mono-ortho, and non-ortho substituted polychlorinated

28 biphenyls in rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 15: 1382-1387.

29 Deweerd, K.A. and D.L. Bedard. 1999. Use of halogenated benzoates and other halogenated

30 aromatic compounds to stimulate the microbial degradation of PCBs. Environ. Sci. Technol.

31 33:2057-2063.

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

-------
1 Hartkamp-Commandeur, I.C.M., J. Gerritse, H.A.J. Govers, and J.R. Parsons. 1996. Reductive

2 dehalogenation of polychlorinated biphenyls by anaerobic microorganisms enriched from Dutch

3 sediments. Chemosphere. 32(7): 1275-1286.

4 Hill, W.R., and G.E. Napolitano. 1997. PCB congener accumulation by periphyton, herbivores,

5 and omnivores. Arch. Environ. Contam. Toxicol. 32:449-455.

6 Kannan, N., T.B.H. reusch, D.E. Schulz-Bull, G. Petrick, and J.C. Dulnkner. 1995.

7 Chlorobiphenyls: model compounds for metabolism in food chain organisms and their potential

8 use as ecotoxicological stress indicators by application of the metabolic slope concept. Environ.

9 Toxicol. Chem. 29:1851-1859.

10 Koslowski, S.E., C.D. Metcalfe, R. Lazar, and G.D. Haffner. The distribution of 42 PCbs,

11 including three coplanar congeners, in the food web of the Western Basin of Lake Erie. J. Great

12 Lakes Res. 20(l):260-270.

13 Lech, J. and R.E. Peterson. 1983. Biotransformation and persistent of polychlorinated biphenyls

14 (PCBs) in fish. In: PCBs: Human and Environmental Hazards (Eds. F.M. D'ltri, M.A. Kamrin

15 and M.A. Butterworth), Woburn, Massachusetts, pp. 187-201.

16 Metcalfe, T.L. and C.D. Metcalfe. 1997. The trophodynamics of PCBs, including mono- and

17 non-ortho congeners, in the food web of central Lake Ontario. Sci. Total Environ. 201:245-272.

18 Morrison, H.A., D.M. Whittle, C.D. Metcalfe, and A.J. Niimi. 1999. Application of a food web

19 bioaccumulation model for the prediction of polychlorinated biphenyl, dioxin, and furan

20 congener concentrations in Lake Ontario aquatic biota. Can. J. Fish. Aquat. Sci. 56:1389-1400.

21 Niimi, A.J. 1996. Evaluation of PCBs and PCDD/Fs retention by aquatic organisms. Sci. Total

22 Environ. 192:123-150.

23 Niimi, A.J. and B.G. Oliver. 1983. Biological half-lives of polychlorinated biphenyl (PCB)

24 congeners in whole fish and muscle of rainbow trout (Salmo gairdneri). Can J. Fish Aquat. Sci.

25 40:1388-1394.

26 Oliver, B.G., and A.J. Niimi. 1988. Trophodynamic analysis of polychlorinated biphenyl

27 congeners and other chlorinated hydrocarbons in the Lake Ontario ecosystem. Environ. Sci.

28 Technol. 22:388-397.

29 Parkerton, T.F. 1993. Estimating toxicokinetic parameters for modeling the bioaccumulation of

30 non-ionic organic chemicals in aquatic organisms. PhD dissertation submitted to the Graduate

31 School-New Brunswick, Rutgers, the State University of New Jersey. May 1993.

32 Porte, C and J. Albaiges. 1993. Bioaccumulation patterns of hydrocarbons and polychlorinated

33 biphenyls in bivalves, crustaceans, and fishes. Arch. Environ. Contam. Toxicol. 26:273-281.

34 Russell, R.W., F.A.P.C. Gobas, and G.D. Haffner. 1999. Role of chemical and ecological factors

35 in trophic transfer of organic chemicals in aquatic food webs. Environmental Toxicology and

36 Chemistry. 18(6): 1250-1257.

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

-------
1 Sijm, T.H.M. 1992. Influence of biotransformation on bioaccumulation and toxicity of

2 chlorinated aromatic compounds in fish. PhD dissertation. Riiksuniversiteit te Utrecht. June 9,

3 1992.

4 Willman, E.J., J.B. Manchester-Neesvig, and D.E. Armstrong. 1997. Influence of ortho-

5 substitution on patterns of PCB accumulation in sediment, plankton, and fish in a freshwater

6 estuary. Environ. Sci. Technol. 31:3712-3718.

7 Wu, Q., D.L. Bedard, and J. Wiegel. 1997. Temperature determines the pattern of anaerobic

8 microbial dechlorination of Aroclor 1260 primed by 2,3,4,6-tetrachlorobiphenyl in Woods Pond

9 sediment. Applied and Environmental Microbiology. 63(12):4818-4825.

MK01 |O:\20123001.096\ERA_PB\ERA_AT1 C7_PB.DOC

-------
Table 1

Profiles of PCB Congeners Selected for Detailed Congener Evaluation

Species

Log Kqw

Properties and Rationale for Selection

PCB 49 (2,2',4,5'-
tetrachlorobiphenyl)

5.85

¦ Breakdown product of anaerobic dechlorination of sediment (formed by dechlorination of PCB 99 in Process
P, and dechlorination of PCB 87 and 101 in Process N) (Bedard and Quensen 1995).

¦ Congener is non-persistent in Lake Ontario food web, relative to PCB 153 (Metcalfe and Metcalfe 1997).

¦ Represents lower molecular weight end of Aroclor 1260.

PCB 77 (3,3'4,4'-
tetrachlorobiphenyl)

6.36

¦ Rapid elimination of this congener observed in aquatic biota. Lack of biomagnification observed in Lake Erie
food chain (Koslowski et al. 1994), and altered profile across trophic levels relative to other PCB congeners.

¦ Retained by fish at a low rate (Niimi 1996).

¦ Not biomagnified to the extent of other congeners with similar K0w, attributed to high metabolic clearance
(Metcalfe and Metcalfe 1997).

¦ Capacity for metabolic clearance greater among vertebrate species (fish, waterfowl) than among aquatic
invertebrates, because activity of cytochrome P-450 monooxygenases is lower (Metcalfe and Metcalfe 1997).

PCB 110 (2,3,3',4',6-
pentachlorobiphenyl)

6.48

¦ Potentially metabolizable by fish, but no large or consistent pattern of enrichment or depletion observed at East
Fork Poplar Creek site (Hill and Napolitano 1997).

¦ No increase in percentage composition with increased trophic level in Lake Ontario food web (Oliver and
Niimi 1988).

¦ Pattern within Lake Ontario aquatic food web is a decline with each step in the food chain (Metcalfe and
Metcalfe 1997).

¦ Congener is moderately non-persistent in Lake Ontario food web, relative to PCB 153 (Metcalfe and Metcalfe
1997).

Mk01\0:\20123001.096\ERA_PB\era_at1C7_pb.doc -> rs 7/10/2003

-------
Table 1

Profiles of PCB Congeners Selected for Detailed Congener Evaluation

(Continued)

Species

Log Kow

Properties and Rationale for Selection

PCB 126 (3,3'4,4',5-
pentachlorobiphenyl)

6.89

¦ Lack of biomagnification observed in Lake Erie food chain (Koslowski et al. 1994), and altered profile across
trophic levels relative to other PCB congeners.

¦ Not biomagnified to the extent of other congeners with similar Kow, attributed to high metabolic clearance
(Metcalfe and Metcalfe 1997).

PCB 138 (2,2',3,4,4',5'-
hexachlorobiphenyl)

6.83

¦ Mother product for anaerobic dechlorination of sediment (transformed to PCB 87 in Process P, and PCB 99 in
Process N) (Bedard and Quensen 1995).

¦ Breakdown product of anaerobic dechlorination of sediment (formed by dechlorination of PCB 170 in Process
N) (Bedard and Quensen 1995).

¦ In Bedard and May (1996), this congener was proposed as an intermediate product in a dechlorination series;
degrading from PCB 170 and into PCBs 85, 87, 97, and 99. A reduction of this congener was observed in
Woods Pond sediments relative to Aroclor 1260.

¦ Potentially metabolizable by fish, but no large or consistent pattern of enrichment or depletion observed at East
Fork Poplar Creek site (Hill and Napolitano 1997).

¦ Not observed to be substantially metabolized by aquatic biota (Morrison et al. 1999).

¦ In Detroit River ecosystem, indicated congener profile across trophic levels that resembled the
environmentally persistent PCB 153 (Russell et al. 1999).

¦ Congener is moderately persistent in Lake Ontario food web, relative to PCB 153 (Metcalfe and Metcalfe
1997).

MK01|O:\20123001.096\ERA_PB\ERA_AT1C7_PB.DOC -> ^ 7/10/2003

-------
Table 1

Profiles of PCB Congeners Selected for Detailed Congener Evaluation

(Continued)

Species

Log Kow

Properties and Rationale for Selection

PCB 151 (2,2'3,5,5',6-
hexachlorobiphenyl)

6.64

¦ Congener has meta-para vicinal hydrogens and 3 ortho chlorines; this structure is persistent in some mammal
and bird species and not in others (Metcalfe and Metcalfe 1997).

¦ Pattern within Lake Ontario aquatic food web is a decline with each step in the food chain (Metcalfe and
Metcalfe 1997).

¦ Congener is moderately non-persistent in Lake Ontario food web, relative to PCB 153 (Metcalfe and Metcalfe
1997). Biomagnification factors lower than would be predicted on the basis of Kow.

PCB 153 (2,2',4,4',5,5'-
hexachlorobiphenyl)

6.92

¦ Mother product for anaerobic dechlorination of sediment (transformed to PCB 101 in Process P, and PCB 99
in Process N) (Bedard and Quensen 1995).

¦ Breakdown product of anaerobic dechlorination of sediment (formed by dechlorination of PCB 180 in Process
N) (Bedard and Quensen 1995).

¦ In Bedard and May (1996), this congener was proposed as an intermediate product in a dechlorination series;
degrading from PCB 180 and into PCB 101 and PCB 99. A reduction of this congener was observed in Woods
Pond sediments relative to Aroclor 1260.

¦ Congener has no vicinal hydrogens at meta-para or ortho-meta positions, and therefore is highly persistent in
aquatic biota (Metcalfe and Metcalfe 1997).

¦ Considered environmentally persistent in aquatic biota, and not observed to be substantially metabolized by
aquatic biota (Morrison et al. 1999).

¦ PCB structure "particularly recalcitrant to degradation by fishes" (Porte and Albaiges 1993), and "resistant to
metabolic breakdown by monooxygenases found in fish liver" (Hill and Napolitano 1997).

¦ Some increase in percentage composition with increased trophic level in Lake Ontario food web (Oliver and
Niimi 1988).

MK01|O:\20123001.096\ERA_PB\ERA_AT1C7_PB.DOC -> a 7/10/2003

-------
Table 1

Profiles of PCB Congeners Selected for Detailed Congener Evaluation

(Continued)

Species

Log Kow

Properties and Rationale for Selection

PCB 170 (2,2',3,3',4,4'5-
heptachlorobiphenyl)

7.27

¦ Mother product for anaerobic dechlorination of sediment (transformed to PCB 130 in Process P, and PCB 138
in Process N) (Bedard and Quensen 1995).

¦ In Bedard and May (1996), this congener was proposed as starting a dechlorination series; degrading to PCB
146 and PCB 153, and subsequently to other less chlorinated congeners. Therefore, a reduction of this
congener could occur in sediments over time and space.

¦ Potentially metabolizable by fish, but no large or consistent pattern of enrichment or depletion observed at East
Fork Poplar Creek site (Hill and Napolitano 1997).

¦ Highly substituted congener with low metabolic potential in aquatic biota (Porte and Albaiges 1993).

¦ Congener is moderately persistent in Lake Ontario food web, relative to PCB 153 (Metcalfe and Metcalfe
1997).

MK01|O:\20123001.096\ERA_PB\ERA_AT1C7_PB.DOC -> r 7/10/2003

-------
Table 1

Profiles of PCB Congeners Selected for Detailed Congener Evaluation

(Continued)

Species

Log Kow

Properties and Rationale for Selection

PCB 180 (2,2',3,4,4',5,5'-
heptachlorobiphenyl)

7.36

¦ Mother product for anaerobic dechlorination of sediment (transformed to PCB 146 in Process P, and PCB 153
in Process N) (Bedard and Quensen 1995).

¦ Breakdown product of anaerobic dechlorination of sediment (formed by dechlorination of PCB 194 in Process
N) (Bedard and Quensen 1995).

¦ In Bedard and May (1996), this congener was proposed as starting a dechlorination series; degrading to PCB
138 and PCB 130, and subsequently to other less chlorinated congeners. Therefore, a reduction of this
congener could occur in sediments over time and space.

¦ Not observed to be substantially metabolized by aquatic biota (Morrison et al. 1999; Porte and Albaiges 1993).

¦ In Detroit River ecosystem, indicated congener profile across trophic levels that resembled the
environmentally persistent PCB 153 (Russell et al. 1999).

¦ Some increase in percentage composition with increased trophic level in Lake Ontario food web (Oliver and
Niimi 1988).

¦ Congener is persistent in Lake Ontario food web (Metcalfe and Metcalfe 1997).

PCB 194

(2,2',3,3',4,4',5,5'-
octachlorobiphenyl)

7.80

¦ Mother product for anaerobic dechlorination of sediment (transformed to PCB 180 in Process P) (Bedard and
Quensen 1995).

¦ No increase in percentage composition with increased trophic level in Lake Ontario food web (Oliver and
Niimi 1988).

¦ Lower biomagnification observed within Lake Ontario aquatic food web relative to congeners with slightly
lower Kow- Congener is persistent, however (Metcalfe and Metcalfe 1997).

¦ Represents high molecular weight end of Aroclor 1260.

MK01|O:\20123001.096\ERA_PB\ERA_AT1C7_PB.DOC -> ^ 7/10/2003

-------
APPENDIX C.8
SUMMARY OF DATA COLLECTION ACTIVITIES

O:\20123001.096\ERA PB\ERA APC8 PB.DOC

-------
1

APPENDIX C.8

SUMMARY OF DATA COLLECTION ACTIVITIES
C.8.1 Introduction

The Consent Decree between General Electric Company (GE) and the U.S. Environmental
Protection Agency (EPA) specified that EPA conduct the ecological and human health risk
assessments and a modeling study for the "Rest of River." The Rest of River is that portion of
the Housatonic River from the confluence of the East and West Branches in Pittsfield to where
the river discharges into Long Island Sound. EPA conducted a Supplemental Investigation (SI)
of the Lower Reach of the Housatonic River to support these efforts. Although data collected
previously by GE and other parties that met EPA's data quality objectives were considered in the
preparation of the ERA, data collected concurrently with the ecological investigations may be
given preference, if appropriate. These additional data collection and evaluation activities were
detailed in the Supplemental Investigation Work Plan (SIWP) prepared by Roy F. Weston, Inc.
(WESTON 2000) under contract to EPA and the U.S. Army Corps of Engineers (USACE).

The objectives of the SIWP were to:

¦ Provide surface water, hydrology, and sediment data to support the development of
the modeling study.

¦ Characterize and sample biological media, and ecological communities to support
human health and ecological risk assessments.

¦ Acquire sufficient information to compare soil and sediment concentrations against
screening risk-based concentrations.

¦ Develop site-specific human health and ecological risk assessments for the Rest of
River.

¦ Define the nature and extent of the soil and sediment contamination in the Rest of
River and associated floodplain by polychlorinated biphenyls (PCBs) and other
contaminants, and further delineate pathways of contaminant migration to support the
risk assessments and modeling study.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 1 7/10/2003

-------
1

The SIWP presented a detailed rationale outlining the data requirements, data quality objectives,
and data management procedures and controls. A project-specific Quality Assurance Project
Plan was also prepared (WESTON 2001) and implemented in concert with the SI activities. A
summary of the data collection activities (to meet the objectives listed previously) and of the
sampling procedures is presented in this appendix.

A detailed discussion of the sampling programs to support the SI activities is presented in the
Rest of River Site Investigation Data Report (SI Data Report) (WESTON 2002). The SI Data
Report details the number of samples and analyses, including duplicate analysis and QA/QC
analyses, that were proposed versus those collected and analyzed to support programs outlined in
the SIWP. Details regarding data selected for individual analyses and the data preparation and
analysis procedures are presented in the individual appendices for each ERA endpoint. Note that
throughout this appendix the term "sample" refers to individual samples, not locations (e.g., two
aliquots collected from the same sediment core and submitted for analysis of different COCs are
counted as two "samples").

In this summary, data collection activities conducted between 1998 and 2002 are discussed. This
was the period during which soils, sediments, water, and biota were sampled and analyzed or
measured under the SIWP. Sampling was typically restricted to the 10-year floodplain of the
Housatonic River. The 10-year floodplain is roughly equivalent to the 1-ppm isopleth
established by GE based upon their previous investigations to define the extent of PCB
contamination. Sampling was generally conducted in an iterative manner, with systematic
samples collected at regularly spaced intervals, followed by discrete sampling to address specific
objectives of the SI. Each sample was typically classified using two approaches, one defining
media type (e.g., water, sediment, soil, tissue) and the other defining physical location within the
system (geomorphological codes).

Sediment (for the purpose of the SIWP with regard to media type) was defined as any material
settled to the bottom of a water body, including vernal pools. Samples collected in vernal pools
have also been characterized as floodplain in the geomorphological classification system because
vernal pools are located within the floodplain. All other material was considered soil. The
SIWP called for sampling to be performed in both a systematic (i.e., samples collected at regular

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 2

-------
1

intervals over the study area) and a discrete fashion, with samples being collected from three
main geomorphological features: sediments from water bodies, soils from riverbanks, and soils
from the floodplain. In addition to sediment and soil samples, surface water, porewater, and
tissue samples were collected to support discrete sampling programs.

C.8.2 Systematic Sampling

Systematic sampling was conducted in Reaches 5 through 8 in most cases by use of transects
spaced at regular intervals perpendicular to the main axis of the river and associated floodplains.
Systematic sampling was not proposed in Reach 9, due to low or non-detected PCB
concentrations observed in historical studies. In addition, non-transect sampling programs (e.g.,
core locations along the length of the river placed at regular intervals) were also conducted to
systematically assess PCB contamination. The following sections provide a descriptive
summary of the proposed and completed systematic sampling programs, including the types of
samples (sediment, riverbank, and/or floodplain) proposed and the samples actually collected
and analyzed. A summary of the number of PCB samples analyzed by sample type (sediment,
riverbank, and floodplain) for each systematic sampling program is presented in Table C.8-1.

Table C.8-1
Summary of Systematic Sampling

Reach

Program

No. of
Transects

Approx.
Interval

Sediment
Samples"

Riverbank
Samples

Floodplain
Samples

Total
Samples

1,500-ft Transects

1,500 ft

424

575

1,030

West Branch Transects

200 ft

Modeling Transects

16b

Varies

188

841

1,029

Cores

23 cores

N/A

544

Modeling Transect

100 ft

123

2,500-ft Transects

2,500 ft

294

663

957

Cores

N/A

279

TOTALS

1,873

2,126

4,030

a Number of samples indicates the number of sediments, floodplain, and riverbank samples analyzed for tPCBs.
b One additional transect was completed for floodplain samples for a total of 17 floodplain transects.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 3 7/10/2003

-------
1

C.8.2.1 1,500-Foot Transects (Reach 5)

Transects were established at 1,500-foot (460-m) intervals perpendicular to the river, resulting in
a total of 38 transects from the confluence of the East and West Branches of the Housatonic
River to just upstream of Woods Pond. Each transect was to extend from the boundary of the
10-year floodplain on one side of the river to the same boundary on the other side of the river.
Sampling along each transect would then include floodplain and riverbank soil, and sediment, as
described in the following paragraphs:

¦ Sediment samples were collected from three locations across the river channel (right
side, mid-channel, left side) and from four depth intervals (0 to 6, 6 to 12, 12 to 18,
and 18 to 24 inches [0 to 15, 15 to 30, 30 to 45, and 45 to 61 cm]). Sediment samples
were analyzed for PCBs (total and Aroclors), TOC, and sediment grain size
parameters. In addition, approximately 10% of these samples were analyzed for
modified Appendix IX parameters (a group of approximately 220 organic and
inorganic hazardous contaminants listed in 40 CFR 264). Of the samples, 2% were
also analyzed for organophosphate pesticides and herbicides.

¦ Riverbank soil samples were collected only when pronounced banks were
encountered during transect sampling activities. Riverbank samples were
encountered only on five transects, for a total of 31 PCB samples: two banks per
transect, and three samples (0 to 6, 12 to 18, and 24 to 30 inches [0 to 15, 30 to 45,
and 61 to 76 cm]) per bank. Approximately 10% of the riverbank samples were
analyzed for TOC, grain size, and modified Appendix IX (40 CFR 264) parameters,
and 2% were analyzed for organophosphate pesticides and herbicides.

¦ Floodplain soil samples were collected from each of the 38 transects at three locations
on each side of the river, resulting in 575 PCB samples: 38 transects; six locations;
three depths (0 to 6, 12 to 18, and 24 to 30 inches). Approximately 10% of the
floodplain samples were analyzed for TOC, grain size, and modified Appendix IX (40
CFR 264) parameters, and 2% were analyzed for organophosphate pesticides and
herbicides.

¦ After a review of the PCB results, approximately 10% of the systematic transect
locations were sampled and analyzed for PCB congeners. (See Section C.8.3.27,
Discretionary Sampling, and Section C.8.3.7, Congener Sampling Program.)

C.8.2.2 West Branch Transects

Eight transects at 200-ft (60-m) intervals were sampled along the West Branch to determine
background PCB concentrations in sediments upstream of the confluence with the East Branch.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 4

-------
1 Only sediment sampling was completed for the West Branch transect program as described in the

2 following paragraphs:

¦ Sediments were sampled in three channel locations (right side, mid-channel, and left
side) at four depth intervals (0 to 6, 6 to 12, 12 to 18, and 18 to 24 inches), resulting
in 68 sediment samples, which were analyzed for PCBs (total and Aroclors), TOC,
and sediment grain size parameters. In addition, approximately 10% of the samples
were analyzed for modified Appendix IX (40 CFR 264) parameters, and 2% were
analyzed for organophosphate pesticides and herbicides.

¦ After review of the PCB data, approximately 10% of the total transect locations were
to be sampled and analyzed for PCB congeners. (See Section C.8.3.27, Discretionary
Sampling, and Section C.8.3.7, Congener Sampling Program.)

12 C.8.2.3 Channel Geometry/Modeling Transects (Reaches 5 and 6)

13 A total of 17 transects were sampled in Reach 5, perpendicular to the river across the entire

14 width of the 10-year floodplain, to define the channel geometry for the modeling study

15 synoptically with the PCB concentrations. In addition, one transect was sampled in Reach 6

16 (Woods Pond). The modeling transect sampling program included sediment, riverbank, and

17 floodplain locations as described below:

18 ¦ Sediment samples were collected along each transect in Reach 5 from three locations

19 across the river channel (right side, mid-channel, left side) and from four depth

20 intervals (0 to 6, 6 to 12, 12 to 18, and 18 to 24 inches), resulting in 188 sediment

21 samples collected for PCBs (total and Aroclors), TOC, and sediment grain size

22 parameters. Due to re-sampling/analysis, additional samples were retained for TOC

23 and grain size analysis, resulting in more samples analyzed for TOC and grain size

24 parameters than PCBs.

25 ¦ Sediment samples in Reach 6 (Woods Pond) were collected every 100 ft (30 m) at 6-

26 inch (15-cm) depth intervals to a depth of 2 ft (0.6 m), resulting in 76 samples

27 analyzed for PCBs (total and Aroclors), TOC, and sediment grain size parameters. In

28 addition, approximately 5% of the samples were analyzed for modified Appendix IX

29 (40 CFR 264) parameters.

30 ¦ No systematic riverbank samples were collected in Reach 5 due to the proximity of

31 floodplain samples to the river channel, which negated the need for additional

32 riverbank samples. No systematic riverbank samples were collected in Reach 6

33 (Woods Pond) because the backwater areas are characterized by a broader floodplain

34 with low or no discernable banks.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC

C.8-5

7/10/2003

-------
1

¦ Floodplain samples were collected along the 17 transects every 50 ft (15.2 m) across
the 10-year floodplain. Samples were collected at 0- to 6-inch depth intervals at
every location, and 6 to 12 inches at every other location (i.e., every 100 ft).
Approximately 841 floodplain samples were analyzed for PCBs (total and Aroclors)
along these transects in Reach 5. Of these samples, approximately 10% were also
analyzed for TOC and sediment grain size parameters, and less than 1% were
analyzed for modified Appendix IX (40 CFR 264) parameters. Within Reach 6
(Woods Pond), 11 samples were analyzed from the east and west shores for PCBs
(total and Aroclors). In addition to PCB analysis, one of the Reach 6 (Woods Pond)
samples was analyzed for TOC and sediment grain size parameters.

C.8.2.4 Sediment Characterization Cores

Sediment cores were collected from Woods Pond (Reach 6) and Rising Pond (Reach 8) to

provide information on the characteristics of deep sediment. The program is described below:

¦ A total of 23 sediment cores were collected to depth of first refusal in Woods Pond,
resulting in 319 PCB (total and Aroclors) results. In addition, each sample was
analyzed for TOC and sediment grain size parameters, and approximately 10% of the
samples were analyzed for modified Appendix IX (40 CFR 264) parameters and PCB
congeners. Due to re-sampling/analysis, additional samples were retained for TOC
and grain size analysis, resulting in more samples analyzed for TOC and grain size
than for PCBs. Organophosphate pesticides and herbicides were analyzed in 1% of
the samples, and 5% of the samples were analyzed for bulk density.

¦ A total of 20 sediment cores were collected to depth of first refusal in Rising Pond,
resulting in 279 PCB (total and Aroclors) results. In addition, each sample was
analyzed for TOC and sediment grain size parameters, and approximately 10% of the
samples were analyzed for modified Appendix IX (40 CFR 264) parameters and PCB
congeners. Due to re-sampling/analysis, additional samples were retained for TOC
and grain size analysis, resulting in more samples analyzed for TOC and grain size
than for PCBs. Organophosphate pesticides and herbicides were analyzed in 1% of
the samples.

C.8.2.5 2,500-Foot Transects

Transects were established approximately every 2,500 ft (760 m) in Reach 7, which equates to a

total of 41 transects. The proposed sampling is described as follows:

¦ Sediment samples were collected in three channel locations (right side, mid-channel,
and left side) at four depth intervals (0 to 6, 6 to 12, 12 to 18, and 18 to 24 inches) for
a total of 294 PCB (total and Aroclors) results. In addition, approximately 90% of
these sediment samples were analyzed for TOC and sediment grain size parameters.
Modified Appendix IX (40 CFR 264) constituents were analyzed in 10% of the

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 6 7/10/2003

-------
1

samples, and 2% of the samples were analyzed for organophosphate pesticides and
herbicides.

¦ Floodplain locations were established on each side of the river with depth intervals
for each location including 0 to 6 inches, 12 to 18 inches, and 24 to 30 inches. A total
of 40 floodplain transects were sampled, resulting in 663 PCB (total and Aroclors)
results. Modified Appendix IX (40 CFR 264) parameters were analyzed in 10% of
the samples, and 2% of the samples were analyzed for organophosphate pesticides
and herbicides.

C.8.2.6 Other Transect-Related Samples

Additional floodplain samples were collected in Reach 6 to further characterize the areas
adjacent to Woods Pond, as described below:

¦ For Woods Pond, 36 samples were collected for PCB (total and Aroclors) analysis.
In addition, approximately 10% of the samples were analyzed for modified Appendix
IX (40 CFR 264) parameters.

C.8.3 Discrete Sampling

Discrete sampling was conducted to supplement the systematic sampling and provide data to
support the risk assessments, modeling efforts, and spatial assessment of contamination.
Locations sampled included specific habitats, locations not sampled during the systematic
programs (e.g., depositional area behind a dam), or areas of frequent human exposure (e.g.,
residential or recreational areas). In addition, during the collection of different biota samples,
soil, sediment, or water samples may have been collected concurrently from the same area where
the biota sample was collected.

Specific areas sampled or programs during which sediment or soil samples were collected, along
with the number of samples analyzed for PCBs, are presented in Table C.8-2. A complete
summary of all samples, including those analyzed for other COCs, is provided in the Site
Investigation Data Report (WESTON 2002).

C.8.3.1 Human Health Risk Assessment Sampling

Although the primary purpose of this suite of sampling programs was to support the Human
Health Risk Assessment, data were also used as appropriate for the Ecological Risk Assessment.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 7 7/10/2003

-------
1

Program

Sediment
Samples"

Riverbank
Samples

Floodplain
Samples

Total
Samples

Recreational Areas

290

441

Commercial/Industrial

139

178

Aggrading Bars and Terraces

586

Agricultural

142

Temporary and Permanent Pool
Sampling

143

165b

308

Reaches 7 to 9 Exposure Areas

1,017

1,132

Impoundment Sampling

Sediment Cores

Residential Areas

397

527

Grain Size Fraction

322

338

Porewater Studies

Radionuclide Dating

147

Connecticut Sampling

Low-Resolution Congeners

High-Resolution Congeners

100

Benthic Invertebrate Areas

156

Sediment Macroinvertebrate
Toxicity, Bioaccumulation and
Stressor Identification

Mussel Bioaccumulation and
Growth Area

Amphibian Toxicity Areas

170

Leopard Frog Study

Wood Frog Study

Bullfrog Study

Fish Collection Areas

Tree Swallow Study Areas

258

308

Soil Invertebrate Areas

Small Mammal Areas

157

June 2000 Flood Samples

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 8 7/10/2003

Table C.8-2
Summary of Discrete Sampling

-------
Table C.8-2

Summary of Discrete Sampling
(Continued)

Sediment

Riverbank

Floodplain

Total

Program

Samples"

Samples

Discretionary Sampling

428

327

762

1 a Represent the number of PCB results for program by matrix.

2 b These samples were collected within vernal pools located on the floodplain.

3 0 Soils were collected in conjunction with this study; however, previous sampling efforts adequately

4 characterized soil PCB concentrations in the small mammal areas.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 9

-------
1 Sampling of sediment, riverbank, and floodplain soils associated with potential exposure areas

2 was conducted in an iterative manner. An initial screening of the possible exposure areas was

3 performed to determine which areas would be sampled, and to estimate the type and number of

4 samples to be collected. Due to this iterative sampling approach, the number of samples

5 collected for many of the sampling programs for the human health risk assessment were either

6 greater than, or in some cases less than, initially estimated in the SIWP.

7 C.8.3.1.1 Recreational Exposure (Reaches 5 and 6)

8 Several areas along the river in Reaches 5 and 6 were identified in the SIWP as current

9 recreational exposure areas. These areas include "Paintball Area," Canoe Meadows Audubon

10 Sanctuary, John Decker Canoe Launch, DeVos Farm (floodplain only), Lenox Sportsman Club,

11 three river access areas off October Mountain Road, Duck Blind Areas in Woods Pond

12 backwaters, and Woods Pond Boat Launch (sediment and floodplain only).

13 These areas, except for the Woods Pond Boat Launch location, are located in Reach 5. Sediment

14 samples at Woods Pond (surface to 6-inch depth interval) were collected at approximately 50-ft

15 (15-m) intervals along the shoreline of the recreational areas. Sampling was concentrated in

16 areas of easiest human access. Floodplain soils were collected from between 10 and 20 locations

17 within each recreational area and from two depths (0 to 6 inches and 6 to 12 inches). Riverbank

18 soils were collected at 50-ft intervals at two depth intervals (0 to 6 inches and 6 to 12 inches).

19 Sediment samples collected in association with the recreational area program were sampled for

20 PCBs (total and Aroclors), TOC, and sediment grain size parameters. In addition, 10% of the

21 samples were collected for PCB congeners and modified Appendix IX (40 CFR 264) parameters,

22 and 2% of the samples were analyzed for organophosphate pesticides and herbicides. Table

23 C.8-2 provides a summary of recreational soil and sediment samples and additional discrete

24 sampling program samples.

25 C.8.3.1.2 Residential Area Sampling

26 Floodplain and riverbank soils and river sediments were sampled from residential properties in

27 Reaches 5 and 6. Up to three sediment samples (to a depth of 6 inches) were collected from the

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 10 7/10/2003

-------
1

river adjacent to only those residential properties that border the river. Floodplain soils from two
depths (0 to 6 inches and 6 to 12 inches) were collected from up to five locations per residential
property or property zoned for future residential development. Riverbank soils were also
collected from those properties that border the river, at approximately 50-ft intervals.

Residential area samples were collected for PCBs (total and Aroclors) and approximately 5% of
the samples were collected for PCB duplicate analysis, PCB congeners, and Appendix IX (40
CFR 264) parameters. In addition to these analyses, approximately 10% of the samples were
analyzed for TOC sediment grain size parameters. Table C.8-2 provides a summary of the
number of samples analyzed for PCBs in association with the residential areas sampling
program.

C.8.3.1.3 Commercial/Industrial Exposure Sampling

Commercial/industrial sampling was completed in Reach 5 to assess exposure to utility workers
and groundskeepers for current commercial/industrial properties. For the utility worker exposure
scenario, it was determined that four types of utility easements intersected the study area and
required sampling: a gas pipeline, telephone lines, electrical lines, and sewer lines. An
assessment of where groundskeeping activities were being conducted in areas adjacent to the
river was completed, and it was determined that only one location (Miss Hall's School) resulted
in a current exposure scenario. Riverbank and floodplain samples were collected from each
utility easement where applicable, and floodplain sampling was completed for the groundskeeper
exposure scenario as described below:

¦ For riverbank utility easements, 16 locations at two depths (0 to 6 inches and 6 inches
to 6 ft) were sampled, for a total of 39 PCB (total and Aroclors) results. In addition,
10% of the samples were analyzed for TOC, grain size parameters, modified
Appendix IX (40 CFR 264) parameters, and PCB congeners.

¦ Groundskeeper floodplain samples were collected at two depths, 0 to 6 inches and 6
to 12 inches, resulting in 11 PCB (total and Aroclors) results.

Floodplain samples at utility easements were collected from three depths (0 to 6 inches, 6 to 12
inches, and 1 to 6 ft), resulting in 128 PCB (total and Aroclors) results. In addition,
approximately 10% of the floodplain samples were analyzed for modified Appendix IX (40 CFR

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 1 1 7/10/2003

-------
1 264) parameters and PCB congeners. Table C.8-2 provides a summary of the number of samples

2 analyzed for PCBs in association with the commercial/industrial sampling program.

3 C.8.3.1.4 Agricultural Area Sampling

4 Floodplain Sampling

5 To support the agricultural sampling program, 142 floodplain samples were collected in Reaches

6 5 and 6. Samples were collected from 0 to 6 inches and 6 to 12 inches from areas zoned for

7 agricultural uses. In addition, composite soil samples were collected from the 0 to 6-inch depth

8 and adjacent to squash, corn, and fiddlehead fern areas. All locations were sampled for PCBs

9 (total and Aroclors), with an additional 10% analyzed for PCB congeners and modified

10 Appendix IX (40 CFR 264) constituents. In addition, approximately 1% of the samples were

11 analyzed for TOC, sediment grain size parameters, and organophosphate pesticides and

12 herbicides. Table C.8-2 provides a summary of the number of samples analyzed for PCBs in

13 association with the agricultural area sampling program.

14 Corn Sampling

15 To assess the possible uptake of PCBs in corn and the potential transfer to humans or dairy cows,

16 and ultimately, consumers of dairy products, corn samples were collected from cornfields

17 extending into the floodplain where PCB contamination of soils was confirmed in the floodplain

18 sampling program.

19 In 1998, a total of nine corn samples were collected from cornstalks located in agricultural areas

20 within the floodplain and areas outside the floodplain to serve as reference locations. The stalk

21 (including leaves) and the ear were included in the analyses for PCBs (total and Aroclors). A

22 composite of three soil samples was collected in areas surrounding each corn sample for PCB

23 (total and Aroclor) analysis.

24 In 1999, 10 corn samples were collected from cornstalks located in agricultural areas within the

25 floodplain and areas outside the floodplain to serve as reference locations. The ear and stalk of

26 each corn sample was submitted for separate PCB (total and Aroclors) analysis. A composite of

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 12 7/10/2003

-------
1

three soil samples were collected in areas surrounding each corn sample for PCB (total and
Aroclor) concentrations.

Fiddlehead Fern Sampling

Surface soils (0 to 6 inches) and fiddlehead fern samples were collected by the Massachusetts
Department of Environmental Protection (MDEP) and by EPA consultants in 1999. Because all
of the 1999 fiddlehead fern data were R-qualified in validation due to low percent solids, a
second round of sampling was conducted in spring 2000. Three samples were analyzed for
PCBs (total and Aroclors).

Squash Sampling

In September 1999, four acorn squash samples were collected from a squash field on a farm in
the Reach 5 floodplain. All squash samples were analyzed for PCBs (total and Aroclors).

Grass Sampling

Reed canary grass and co-located soil samples were collected to provide data for estimation of
transfer of PCB, dioxin, and furan congeners from soil and grass to cattle. In 2001, 10 paired
pasture grass and soil samples were collected from a former dairy farm in early July, when hay
harvesting typically occurs. Samples were to be collected from areas where relatively high PCB
concentrations were detected to avoid obtaining results below detection limits.

C.8.3.1.5 Exposure Areas (Reaches 7 through 9)

Exposure areas sampled in Reaches 7 through 9 included recreational, residential, agricultural,
and commercial/industrial areas as described in the following paragraphs:

¦ Sediment samples were collected at recreational and residential areas at reduced
frequency due to lower PCB concentrations based upon historical data. A total of 95
sediment samples were collected and analyzed for PCBs (total and Aroclors), and
approximately 70% of the sediments were analyzed for TOC and sediment grain size
parameters.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 13

-------
1

¦ Riverbank samples were collected as necessary based on project protocols and
amount of riverbank present at the time of sampling, which resulted in 20 PCB (total
and Aroclors) results for Reaches 7 through 9 riverbanks.

¦ Floodplain soil samples were collected and analyzed, resulting in 1,017 PCB (total
and Aroclors) results. In addition, approximately 10% of the samples were analyzed
for TOC and grain size parameters, and 1% of the floodplain soils were analyzed for
modified Appendix IX (40 CFR 264) constituents and PCB congeners.

8 C.8.3.2 Aggrading Bars and Terraces

9 A total of 45 aggrading or point bars and river terraces were sampled within Reach 5. Locations

10 of the bars were initially based upon MDEP mapping of these features and visually confirmed in

11 the field. Two sampling locations were selected from each bar or terrace—one approximately

12 where the thickest accumulation of sediment was encountered, and the second located

13 equidistant between the first location and the farthest end of the bar or terrace feature. Samples

14 were then collected from each core to a depth of 2.5 ft (0.76 m), and for those cores where the

15 thickest accumulation of sediment occurred, samples were collected from as deep as 5.5 ft

16 (1.68 m).

17 Each of the 45 cores was divided into 6-inch sections, resulting in 586 PCB (total and Aroclors),

18 TOC, and sediment grain size results. In addition, approximately 10% of the samples were

19 analyzed for modified Appendix IX (40 CFR 264) constituents and PCB congeners, and

20 approximately 2% of the samples were analyzed for organophosphate pesticides and herbicides.

21 C.8.3.3 Temporary and Permanent Pools

22 A total of 67 vernal pools were identified during the ecological characterization. Each pool was

23 sampled at approximately five locations to a depth of 6 inches. Vernal pool samples were

24 collected for PCBs (total and Aroclors), TOC, and sediment grain size parameters. In Reaches 5

25 and 6, approximately 143 samples classified as sediments from side channel and oxbow areas

26 were collected for the temporary and permanent pool sampling program. An additional 165

27 samples were collected from areas located off the main channel (East Branch) and within the 10-

28 year floodplain.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC

C.8-14

7/10/2003

-------
1

C.8.3.4 Sediment Cores

Discrete (i.e., non-transect) sediment cores were collected from the channel of the Housatonic
River as well as from impoundments, primarily in Reaches 5 and 6. These cores were analyzed
for PCBs (total and Aroclors), TOC, and sediment grain size parameters. Selected cores were
also used for extraction and analysis of porewater and radioisotope dating. The non-transect
sediment core, grain size fractionation, porewater, and radionuclide dating programs are
described in the following sections.

C.8.3.4.1 Non-Transect Sediment Cores

To assess the location and concentration of PCBs in areas of the river not associated with the
systematic transect sampling, sediment cores were collected as part of the discrete sampling
program. Sample locations were selected based on the iterative review of chemical data as the
data were received as well as observations of river flow and sedimentation patterns. A total of
74 sediment samples were analyzed for PCBs (total and Aroclors), TOC, and grain size
parameters. Approximately 10% of these sediment samples were analyzed for modified
Appendix IX (40 CFR 264) parameters and PCB congeners. In addition to these parameters, 1%
of the samples were analyzed for organophosphate pesticides and herbicides.

C.8.3.4.2 Grain Size Fraction

A series of sediment cores were collected from the river channel, Woods Pond, and backwater
areas to provide characterization data by grain size class for use in the modeling study. The
program is described as follows:

¦ Three cores per transect were collected along 11 transects in Reaches 4 and 5. One
transect was sampled in the West Branch to provide additional grain size
characterization of sediments.

¦ Sample locations from Woods Pond and backwater areas were selected based on the
subbottom profiling survey and coincided with the cores collected as part of the
systematic sampling at Woods Pond.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 15

-------
1

¦ Each interval was sieved into three separate grain size fractions:

¦ Core samples were collected at 0 to 6-inch and 12- to 18-inch depth intervals,
resulting in 338 PCB (total and Aroclors) and TOC results. These samples were
collected from each of the three grain size fractions and bulk sediment samples from
Reaches 4 through 6. Approximately 35% of these samples were analyzed for PCB
congeners.

< 62 |j,m
62 - 250 |j,m
> 250 |j,m

9
10

11 C.8.3.4.3 Porewater and Supplemental Porewater

12 Sediments were collected for porewater (interstitial water) analysis from the 0 to 6-inch depth

13 interval at 13 locations. These data were collected to provide information on the partitioning of

14 PCBs between the sediment and water phase (porewater). In particular, porewater sampling was

15 conducted to provide data that can be used to determine the potential for sediments to be a source

16 of PCBs to surface waters. A second porewater study was conducted in 2001 to provide further

17 understanding of PCB partitioning and the effect of organic carbon on the sorptive behavior of

18 PCBs in Reaches 4 through 6.

19 Approximately 65 sediment samples were analyzed for tPCBs, TOC, grain size, bulk density,

20 percent moisture, and PCB congeners. In addition to sediment samples, approximately 61

21 porewater samples were analyzed for tPCBs, dissolved organic carbon (DOC), and PCB

22 congeners.

23 C.8.3.4.4 Radionuclide Dating

24 Nine cores were used for dating to estimate the sediment deposition rate in Woods Pond and its

25 backwaters. This study incorporated both cesium-137 and lead-210 for dating. Use of

26 beryllium-7 was limited to the surface layers (i.e., top two to three measurements) of the

27 sediment cores. A subcontract laboratory using instruments to measure the radioisotopes (i.e., to

28 measure gamma radiation) conducted the dating analyses.

29 Cores were sectioned every 2 cm for the top 15 cm, every 4 cm for the next 30 cm, every 10 cm

30 for the next 60 cm, and every 15 cm to a depth of approximately 183 cm or until refusal was

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC

C.8-16

7/10/2003

-------
1

encountered. This resulted in approximately 15 to 20 samples per core for dating analysis. In
addition to dating parameters, approximately 147 samples and 115 samples were analyzed for
PCBs (total and Aroclors) and TOC, respectively.

C.8.3.5 Impoundments (Reaches 7 through 9)

Suspended sediment transport of PCBs in the Housatonic River results in an accumulation of
these solids and their adsorbed contaminants in depositional areas, at least for the silt and larger
grain-size fractions. Concentrations of PCBs in Woods Pond are a good example of this
transport and subsequent accumulation in the sediments. Therefore, it is important to
characterize the accumulation of PCBs in other downstream depositional areas. This effort was
completed to provide data to support the risk assessments and modeling effort. Sampling
locations and numbers of samples were selected after field probing of sediment depths, site
characteristics, and review of existing data. A total of 54 samples were analyzed for PCBs (total
and Aroclors), TOC, and sediment grain size parameters.

Samples were collected upstream of the following three existing dams:

¦ Columbia Mill

¦ Willow Mill

¦ Glendale Dam

Samples were collected upstream of the following six former dams:

¦ Niagara Mills

¦ Lee/Eagle Mills

¦ Eaton-Bikeman

¦ Monument Mills No. 2

¦ Monument Mills No. 3

¦ Former Southern Berkshire Dam (Reach 9)

C.8.3.6 Connecticut Sampling

Sampling of sediments and soils in Reaches 10 through 16 in Connecticut was conducted to
evaluate spatial and temporal distribution of PCBs in river sediments and to develop a conceptual
model of the fate and transport of PCBs in the river. Samples were collected in three types of
areas:

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 17 7/10/2003

-------
1

¦ A total of 28 river channel sediment samples were collected and analyzed for PCBs
(total and Aroclors), TOC, and grain size parameters from the 0 to 6-inch and 6- to
12-inch intervals in Reaches 10 through 16. Sample intervals varied due to refusal
being encountered.

¦ Seven floodplain samples were collected from access areas along the edge of the river
in Reaches 11 and 12. Samples were collected from the 0- to 6-inch and 6- to 12-inch
intervals and analyzed for PCBs (total and Aroclors), TOC, and grain size parameters.
Sample intervals varied due to refusal being encountered.

¦ Nine sediment samples were collected in locations upstream of four impoundments
(Great Falls, Bulls Bridge, Bleachery Dam, and Derby-Shelton Dam) located in
Reaches 10, 12, 13, and 16, respectively. Sediments were collected to a maximum
depth of 3 ft (0.91 m) and from the top and bottom 6 inches of each sample core and
analyzed for PCBs (total and Aroclors), TOC, and sediment grain size parameters.

14 C.8.3.7 Congeners

15 Samples were collected for PCB congener analysis to provide data on congener profiles in

16 different media and across a range of PCB concentrations for use in the human health and

17 ecological risk assessments and the modeling study.

18 C.8.3.7.1 Low-Resolution Congener Sampling

19 Sampling locations for low-resolution PCB congener analysis (52 congeners by gas

20 chromatography/mass spectrometry [GC/MS]) were selected based on a review of the results of

21 the tPCB analyses. Sampling locations were selected to include a range of ecological habitats

22 and areas associated with various human health exposure scenarios. A total of 31 sediment

23 samples in Reach 5 and three sediment samples in reference areas were taken for low-resolution

24 PCB congener analyses. Sediment samples were also analyzed for PCBs (total and Aroclors),

25 TOC, and other parameters. A total of 54 floodplain soil samples from Reach 5 and three

26 floodplain soil samples from reference areas were analyzed for PCBs (total and Aroclors) in

27 addition to low-resolution PCB congeners.

28 C.8.3.7.2 High-Resolution Congener Sampling

29 In late 2001 and 2002, samples for high-resolution congener analysis (GC/electron-capture

30 detector [ECD] with 90+ PCB congeners or co-eluting groups) were collected to better

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC

C.8-18

7/10/2003

-------
1 characterize the dioxin-like congeners, and to supplement the congener data in specific areas and

2 habitats. The high-resolution congener sediment and floodplain samples were analyzed for the

3 same parameters as the low-resolution congener samples. In Reach 5, 33 floodplain samples and

4 15 sediment samples were collected to support this program. An additional two sediment

5 samples from Reach 6 were collected to support this program. Because congener analysis

6 provides a direct measurement for tPCBs and Aroclors and also provides values for these

7 parameters as the sum of the congeners, the actual number of data points in the project database

8 for tPCBs and Aroclors resulting from this sampling is double the number of samples.

9 C.8.3.8 Former Meanders

10 Sampling of former meanders in Reach 5 was proposed as part of the discrete sampling program.

11 At the time the SIWP was prepared, it was assumed that some of the transect sampling program

12 locations would intersect former meanders. The iterative approach used to collect data to

13 support programs within the SIWP proved effective to adequately characterize former meanders

14 and additional sampling for the former meander program was not completed.

15 C.8.3.9 Sediment Macroinvertebrate Toxicity, Bioaccumulation, and Stressor

16 Identification Study

17 This study was completed to measure water and bulk sediment toxicity in laboratory and field (in

18 situ) exposures of surrogate test organisms and to determine which class of chemicals contribute

19 to toxicity using a sediment Phase I Toxicity Identification Evaluation (TIE). Organisms for in

20 situ toxicity and bioaccumulation studies were exposed at six locations along the Housatonic

21 River in flow-through chambers for 2 to 10 days. In the laboratory, additional life-cycle

22 assessment tests were conducted on organisms for 4 to 6 weeks. The test results are presented in

23 Burton (2001).

24 C.8.3.9.1 Sediment Sampling

25 Sediment samples were initially collected and analyzed to assist in selecting locations along the

26 river for conducting the in situ toxicity studies. Following the selection of the toxicity testing

27 locations, sediment samples were collected for laboratory toxicity testing.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 19

-------
1 All sediment samples from the six areas in Reach 5 were collected to a depth of 6 inches,

2 resulting in 87 sediment samples that were analyzed for PCBs (total and Aroclors) and TOC.

3 Approximately 10% of the sediments were collected for modified Appendix IX (40 CFR 264)

4 constituents, organophosphate pesticides, herbicides, and PCB congeners. Approximately 2% of

5 the samples were analyzed for cyanide, sulfide, and sediment grain size parameters.

6 C.8.3.9.2 Laboratory Sediment Toxicity Testing

7 Laboratory sediment toxicity testing was conducted on two organisms: Hyalella azteca and

8 Chironomus tentans. Endpoints that were evaluated were survival, growth, and reproduction.

9 Hyalella azteca Life-Cycle Assessment

10 ¦ Endpoints monitored included 28-day survival and growth; 35-day survival and

11 reproduction; and 42-day survival, growth, and reproduction.

12 ¦ Water quality was measured for dissolved oxygen, temperature, conductivity,

13 hardness, alkalinity, ammonia, and pH at the beginning of the sediment exposure

14 portion of the test, weekly thereafter, then at the end of the test.

15 Chironomus tentans Life-Cycle Assessment

16 ¦ Endpoints monitored in the survival and growth portion of the study included 20-day

17 survival, dry weight, ash-free dry weight, and percent emergence.

18 ¦ Emergence data were collected for complete and partial emergence on or about Day

19 23 and continued for approximately 2 weeks. From Day 23 to the end of the test,

20 emergence of males and females, pupal, and adult mortality was recorded daily for

21 the reproductive replicates.

22 ¦ Water quality was measured for dissolved oxygen, temperature, conductivity,

23 hardness, alkalinity, ammonia, and pH at the beginning of the sediment exposure

24 portion of the test, weekly thereafter, then at the end of the test.

25 C.8.3.9.3 In Situ Toxicity and Bioaccumulation Testing

26 In situ toxicity and bioaccumulation testing was performed at six sites along the Housatonic

27 River and proposed to include two testing periods: a low-flow exposure period and a high-flow

28 exposure period. A high-flow event was not completed due to lower than anticipated flow

29 during the study period. The organisms selected for the in situ testing included the midge

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g 2Q 7/10/2003

-------
1 Chironomus tentans (8 to 12 days post hatch), the amphipod Hyalella azteca (7 to 14 days old),

2 the oligochaete worm Lumbriculus variegatus (multiple ages), and the daphnid Daphnia magna

3 (48 hours old).

4 Other specifics of the testing included:

5 ¦ L. variegatus tissue (7-day exposure) samples from six locations, a trip blank, and an

6 ambient blank to be analyzed for PCB congeners.

7 ¦ Overlying sediment and water samples from the 7-day exposure locations were

8 analyzed for PCB congeners, and modified Appendix IX (40 CFR 264) parameters.

9 In addition, overlying sediment and water column samples from six stations for the

10 48-hour and 10-day studies were analyzed for tPCBs.

11 ¦ For each field site, water quality measurements were collected at test initiation and

12 again upon test termination. Physicochemical measurements included dissolved

13 oxygen, temperature, conductivity, hardness, alkalinity, turbidity, total ammonia, and

14 pH.

15 C.8.3.9.4 Toxicity Identification Evaluation (TIE)

16 "A TIE Phase I study was performed on sediments using Ceriodaphnia dubia.

17 ¦ Porewater samples from each location were separated and analyzed for PCBs (total

18 and Aroclors), inorganics, semivolatile compounds, and dioxins/furans.

19 ¦ The TIE Phase I approach for this study involved 24-hour exposures of Ceriodaphnia

20 dubia to baseline ambient porewater on Day 1. Day 2 test manipulations included a

21 second porewater baseline test, an oxidation-reduction addition test, an ethylene-

22 diaminetetra-acetate (EDTA) test, and a pH-adjusted filtration test.

23 C.8.3.10 Benthic Invertebrate Community Evaluation

24 The macroinvertebrate community was sampled at 13 stations, four of which were located in

25 areas believed to have background concentrations of PCBs and considered reference locations.

26 The remaining nine stations were located throughout Reach 5 and were considered

27 "contaminated" locations, based on measured quantities of PCBs and other COCs in sediments.

28 Further details on the proposed program are described in the following sections.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 21

-------
1 C.8.3.10.1 Sediment Sampling

2 Sediment samples were collected as part of the benthic invertebrate community evaluation and

3 for use in the Sediment Quality Triad evaluation. The sample collection program is described as

4 follows:

5 ¦ Each of the 156 Ponar grab samples (12 replicates at each of the 13 locations) was

6 subsampled from the 0- to 5-cm depth.

7 ¦ The two subsamples were composited in a clean stainless-steel bowl and separated

8 into two aliquots of approximately 30 cm3 (for tPCBs, Aroclors, and TOC analysis)

9 and 80 cm3 (for grain-size analysis). This resulted in 156 PCB (total and Aroclors),

10 TOC, and grain size results for the benthic invertebrate community evaluation.

11 C.8.3.10.2 Benthic Infaunal Sampling

12 The benthic macroinvertebrate infaunal community at each of the stations was sampled as

13 described below:

14 "A total of 12 replicate samples for taxonomic analysis were collected from

15 depositional habitats at each of 13 locations with a Petite Ponar grab sampler. These

16 samples were sieved through a 0.5-mm sieve prior to analysis.

17 ¦ Additional macroinvertebrate samples for analysis of contaminant tissue

18 concentrations were collected at each location using a kick-net. When sufficient

19 material was collected, these samples were separated into community functional

20 groups prior to analysis.

21 ¦ All tissue samples (21) were analyzed for PCBs (total and congeners), total lipids,

22 organochlorine pesticides, and percent moisture. Sufficient material was available for

23 dioxin/furan analysis for 50% of the tissue samples.

24 "A total of 156 replicate samples were processed for taxonomy, enumeration, and

25 biomass using stereo and compound microscopes as necessary.

26 ¦ All organisms picked from the sample were to be identified to the lowest practical

27 identification level, which was expected to be the genus in most cases.

28 C.8.3.11 Mussel Bioaccumulation and Growth Area Sampling

29 Although the mussel bioaccumulation and growth study was terminated prior to completion due

30 to unacceptable mortality related to a storm event shortly after the study was initiated, physical

31 data from the study were used as appropriate in the Ecological Risk Assessment.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g 22 7/10/2003

-------
1 C.8.3.11.1 Sediment Sampling

2 A total of 63 sediment samples from six locations were collected from the sediment surface to a

3 6-inch depth to support the Mussel Bioaccumulation and Growth Study. Samples were collected

4 from a reference area in the Connecticut River, an area upstream of the GE facility, two locations

5 within Reach 5, and an additional area in Great Barrington, MA. Sediment samples were

6 analyzed for PCBs (total and Aroclors), ammonia (NH3), TOC, and sediment grain size

7 parameters. In addition, approximately 20% of these sediment samples were analyzed for

8 modified Appendix IX (40 CFR 264) constituents and 10% were analyzed for organophosphate

9 pesticides and herbicides. One sediment sample was analyzed for PCB congeners.

10 C.8.3.11.2 Water Sampling

11 Two surface water samples from the Connecticut River reference area were sampled for PCBs

12 (total and Aroclors), TOC, semivolatiles, and dioxins/furans. In addition to the parameters listed,

13 one of the water samples was also analyzed for inorganic parameters, cyanide, sulfide, NH3, and

14 PCB congeners.

15 C.8.3.11.3 Tissue Sampling

16 The Mussel Bioaccumulation and Growth Study tissue sample analyses portion of the study was

17 terminated prior to completion due to the burial of the mussel exposure cages in the Housatonic

18 River by a storm event during the exposure period.

19 C.8.3.12 Amphibian Toxicity

20 C.8.3.12.1 Sediment Sampling

21 The objective of the wood frog and leopard frog studies was to evaluate reproductive and

22 developmental success in areas with varying concentrations of sediment PCBs. To further define

23 PCB concentrations within areas, sediment sampling was completed concurrently with the

24 amphibian and reptile surveys, which were designed to assess the abundance and richness of

25 species. Based on the results from both of these studies and the historical data, sampling

26 areas/locations were established for the wood and leopard frog studies.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC Q 7/10/2003

-------
1

A total of 170 sediment samples from 33 areas in Reach 5 were collected for PCB (total and
Aroclors) analyses and nearly all of these samples were analyzed for TOC and sediment grain
size parameters. Approximately 10% of these sediments were collected for modified Appendix
IX (40 CFR 264) constituents and 1% were collected for organophosphate pesticide, herbicide,
and PCB congener analyses.

C.8.3.12.2 Frog Reproduction and Development Study (Leopard Frog)

Sediment Sampling

The objective of this study was to assess the impact of potential PCB exposure to frogs in the
Housatonic River, with specific focus on the potential impact that PCB contamination might
have on reproduction, early development, and maturation (metamorphosis) in northern leopard
frogs (Rana pipiens). This study was designed to determine the effect of PCB exposure to
sexually mature adult frogs on reproductive capacity and developmental fitness in their progeny
using both contaminated and reference sites. Similarly, to determine the extent of which
developmental effects induced in contaminated site embryos/larvae are due to maternal PCB
transfer, a separate set of experiments was run concurrently. In those studies, embryos from
reference site females were exposed to water and sediment from three selected locations within
each contaminated site containing representative concentrations of PCBs, and developmental
effects (hatching and metamorphosis) were monitored.

Sediment samples were collected in conjunction with the collection of leopard frogs for
evaluation as part of the amphibian toxicity study. One sediment sample was collected by
compositing four grab samples at each of 12 locations where leopard frogs were harvested. All
sediment samples were analyzed for PCBs (total, Aroclors, congeners, and homologs),
dioxins/furans, organochlorine pesticides, TOC, and sediment grain size parameters.

Water Sampling

Four 2.5-gallon water samples were collected per sample location and composited into a 10-
gallon sample. A portion of each composite was analyzed for PCBs (total, Aroclors, congeners,
homologs), dioxins/furans, and organochlorine pesticides.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g 24

-------
1

Tissue Sampling

Northern leopard frogs were collected from nine sampling locations (each with varying degrees
of sediment PCB concentrations), and from three reference locations (each with sediment PCB
concentrations below method detection limits). The nine sites within Reach 5 included E-5,
W-9a, W-8, W-7a, W-6, W-4, EW-3, E-l, and W-l, while the reference sites included
Washington Mountain Lake, Threemile Pond State Wildlife Management Area, and Hinsdale
Flats State Wildlife Management Area. Northern leopard frogs were not observed at any
reference locations and therefore, no samples were collected. Collected specimens were shipped
to a laboratory for culturing and monitoring. Upon completion of monitoring, specimens were
submitted for tissue analysis. Analyses included PCBs (total and congeners), dioxins/furans, and
modified Appendix IX (40 CFR 264) pesticides and metals.

Because of the lack of specimens from reference locations, reference specimens were obtained
from a commercial biological supply company. Although no specimens were collected from the
reference sites, sediment and water samples were collected for culturing the external reference
specimens. Adult male and female R. pipiens were collected from each of the nine contaminated
site sampling locations. Of the frogs collected per site, at least four frogs were used for the
reproduction and development study, with the remaining specimens used for whole-body and
tissue residue analysis.

A total of 18 female and 18 male R. pipiens (external reference specimens) were received from a
commercial supplier specializing in aquatic biological field specimens for laboratories. A total
of 57 adult female and 51 adult male R pipiens were collected from the contaminated area. At
least six specimens of each sex were found at each of the sampling locations, with the following
exceptions. No males and only two females were found at Site E-5 (Site 31). In addition, five
female and five male specimens were found at Site W-4 (Site 36). Five males were also
collected at Site W-7a (Site 34).

C.8.3.12.3 Study of Amphibian Reproductive and Developmental Success
Within Vernal Pools (Wood Frog)

Wood frogs (Rana sylvatica) were collected from the Housatonic River and reference areas to
determine whether PCB contamination adversely affected wood frog reproduction, development,

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 25 7/10/2003

-------
1

and growth in vernal pools. Wood frog egg masses, larvae, and metamorphs were collected from
selected vernal pools varying in sediment PCB contamination, cultured in the laboratory using
site water and sediment, and evaluated for development, growth, and maturation. Similarly,
cross-over exposure studies were conducted in which reference specimens were cultured with
contaminated sediment and water, and specimens from contaminated vernal pools were cultured
with reference site sediment and water.

During 2000, wood frogs were collected from nine vernal pools in Reaches 4 and 5 and three
vernal pools within a reference area. Within Reaches 4 and 5, collection efforts were conducted
from vernal pools 8-VP-l, 18-VP-2, 23B-VP-1, 23B-VP-2, 38-VP-l, 38-VP-2, 39-VP-l*,
46-VP-l, and 46-VP-5. At the Washington Mountain Lake reference area, collection efforts
were conducted from three pools (WML-1, WML-2, and WML-3). Egg mass, larvae, and
metamorph samples collected were shipped to a laboratory for culturing and monitoring. Upon
completion of monitoring, specimens were frozen and selected samples were submitted for tissue
analysis. Analyses in all media included tPCBs, PCB congeners, dioxins/furans, polycyclic
aromatic hydrocarbons (PAHs), and Appendix IX (40 CFR 264) pesticides and metals.

C.8.3.12.4 Bullfrog Tissue Sampling

The objectives of the bullfrog study were to determine the whole-body frog tissue concentrations
for use in the fate and effects model and the ecological risk assessment, and to provide bullfrog
leg muscle tissue for contaminant analysis that can be used to qualitatively evaluate the potential
risk to human health from consumption of bullfrog leg muscle tissue, if warranted.

Sampling was conducted in August 1999 during the evening using canoes/boats, spotlights, and
nets. Survey locations included Woods Pond and backwater areas within 1 mile upstream of
Woods Pond, and two reference locations: Threemile Pond State Wildlife Management Area and
Muddy Pond (borders Hinsdale Flats State Wildlife Management Area). A total of 48 bullfrogs
were retained for tissue analysis: 16 from Woods Pond, 11 from the upper mile of Woods Pond,
11 from Threemile Pond State Wildlife Management Area, and 10 from Muddy Pond. Bullfrog

: No wood frog egg masses were observed in this pool in 2000; therefore, no tissue samples were obtained.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g 2g

-------
1

whole-body and leg muscle tissue were analyzed for PCBs (total, Aroclors, congeners/
homologs), organochlorine pesticides, dioxins/furans, percent moisture, and percent lipids.

C.8.3.13 Fish Collection Areas
C.8.3.13.1 Sediment Samples

Sediment samples were collected to characterize PCB concentrations in fish collection areas to
support the ecological and human health risk assessments and other study components as
appropriate. The proposed program included seven locations (two reference locations, Goodrich
Pond, and four locations downstream of the GE facility) for sediment sampling. The four
downstream locations were sampled under the systematic sampling programs and Goodrich Pond
was sampled previously. For this program, five sediment samples from Center Pond, located on
the Housatonic River upstream of East Housatonic Street in Dalton, MA, were collected from the
0 to 6-inch interval and analyzed for PCBs (total and Aroclors), TOC, and sediment grain size
parameters. In addition, one sediment sample was analyzed for modified Appendix IX (40 CFR
264) constituents, including organophosphate pesticides and herbicides. Seven sediment samples
were collected from Threemile Pond State Wildlife Management Area and analyzed for TOC
and sediment grain size parameters and one analysis for herbicide constituents. Sediments at
Threemile Pond State Wildlife Management Area were not analyzed for PCBs because
sediments had previously been characterized for PCB concentrations during earlier sampling
efforts.

C.8.3.13.2 Fish Tissue Sampling

Fish were collected from the Housatonic River and reference areas to determine PCB and other
organic contaminant concentrations in tissues for use in both human health and ecological risk
assessments, to evaluate congener patterns by species for use in fish and mink reproduction
studies, and for use in the PCB fate and bioaccumulation model. Fish were collected for tissue
analysis during three electrofishing events by the U.S. Fish and Wildlife Service (USFWS)
within the Housatonic River and reference areas using electrofishing boats.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g

-------
1 The principal objective of the September and October 1998 fish collection effort was to

2 determine PCB and other organic contaminant concentrations in tissue for use in both human

3 health and ecological risk assessments. In addition, the fisheries community was qualitatively

4 assessed for use in the ecological characterization of the river system. Fish were collected from

5 seven locations:

6 ¦ Upper East Branch Housatonic River

7 ¦ Housatonic River shallow reach

8 ¦ Housatonic River deep reach

9 ¦ Woods Pond

10 ¦ Rising Pond

11 ¦ Goodrich Pond

12 ¦ Threemile Pond State Wildlife Management Area

24 A summary of the fish collected for tissue analysis during 1998 is presented in Table C.8-3. Fish

25 tissue samples were submitted for analysis as whole body, whole-fish composites, and fillet/offal

26 samples. A total of 773 fish were collected for analysis. Fish tissue samples were analyzed for

27 PCBs (Aroclors, congeners/homologs), dioxins/furans, and organochlorine pesticides.

28 In May 1999, electrofishing was conducted to collect largemouth bass and bluegill to support

29 fish toxicology studies. The specific study objectives were to evaluate the survival and

30 development of offspring of fish collected from the selected PCB-contaminated locations of the

31 Housatonic River, and to determine the embryonic effects of PCBs found in fish from selected

32 areas of the Housatonic River. The collection locations for this survey included Woods Pond

33 and the deep reach upstream of Woods Pond (modeling Reaches 5B and 5C), to the New Lenox

34 Road bridge. Fish were also collected from Rising Pond and Threemile Pond State Wildlife

14 Species collected for analysis included:

Largemouth bass (Micropterus salmoides)

Yellow perch (Perca flavescens)

Pumpkinseed {Lepomis gibbosus)

Bluegill {Lepomis macrochirus)

Member of family Cyprinidae (i.e., golden shiner \Notemigonus crysoleucas])
Common shiner (Luxilus cornutus or other)

Brown bullhead (Ameiurus nebulosus)

Goldfish (Carassius auratus)

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC

C.8-28

7/10/2003

-------
1

Table C.8-3

1998 Fish Tissue Collection Summary Per Area

Common
Name

Upper
Housatonic

Housatonic
(Shallow)

Housatonic
(Deep)

Woods
Pond

Rising
Pond

Goodrich
Pond

Threemile
Pond

Goldfish

Cyprinidae

Brown
bullhead

Yellow
bullhead

Largemouth
bass

Smallmouth
bass

Pumpkinseed

Bluegill

Yellow perch

Total

151

163

112

117

Management Area, a reference site. Morphometric data collected from specimens included total
length, total weight, and sex, if possible. In addition, otoliths were collected to estimate ages of
largemouth bass.

The results from the electrofishing survey are presented in Table C.8-4. A total of 182
largemouth bass and 247 bluegills were collected during the May 1999 sampling program. A
subset of these samples was submitted for tissue analysis as whole-fish, fillet, and offal samples.
The target analytes included PCBs (total and Aroclors), percent lipids, dioxins/furans, and
organochlorine pesticides.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g 2Q

-------
1

Table C.8-4

May 1999 Fish Tissue Collection
Summary Per Area

Common
Name

Housatonic
(Deep)

Woods
Pond

Rising
Pond

Threemile
Pond

Total

Male

Female

Male

Female

Male

Female

Male

Female

Largemouth
Bass

182

Bluegill

247

In October 1999, common carp (Cyprinus carpio), goldfish, and white suckers (Catostomus
commersoni) were collected from Woods Pond and one backwater north of Woods Pond to
support the mink (Mustela vison) reproduction study. This study was designed to evaluate
whether farm-raised mink fed diets containing PCB-contaminated fish from the Housatonic
River would exhibit impaired reproductive performance and/or offspring (kit) growth and
survival. Fish were collected from the Housatonic River from Woods Pond and the large
backwaters just upstream. Morphometric data collected from specimens included total length
and total weight. A total of 16 common carp, 88 goldfish, and 24 white suckers were collected
for analysis. Target analytes included PCBs, dioxins/furans, and organochlorine pesticides. Fish
tissue was submitted for analysis as whole-fish composite, fillet, and offal samples.

C.8.3.14 Tree Swallow Study

C.8.3.14.1 Sediment and Floodplain Samples

Sediment samples were collected as part of the study designed to measure the potential effects of
PCB contamination on tree swallows (Tachycineta bicolor) from ingestion of aquatic insects.
Nest boxes were erected within three separate reaches of the river, as well as at three reference
sites.

The proposed sampling was designed to characterize the quality of the sediments within the
immediate vicinity of the nest boxes and within the 400-m (1,300-ft) average foraging radius of

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 30 7/10/2003

-------
1

the adult tree swallow during breeding season (Quinney and Ankley 1985). In addition,
sampling focused on open water areas because tree swallows forage on aquatic insects.

At each nest box cluster, samples were collected at 30-m (100-ft) intervals to cover the linear
extent of the area encompassed by the nest boxes along the river. Each sediment sample was
collected at a position midway between the bank opposite the nesting box and the centerline of
the stream. Sediments from backwater areas and portions of the river greater than 30 m away
from the nest boxes were sampled according to a stratified random design conducted radially
from the box locations. Proposed sediment sampling locations were established by
superimposing a radial grid with a 400-m radius on the nest box.

Sediments were collected from the following six areas:

¦ Taconic Valley Trucking Co. area located upstream of the GE facility on the East
Branch of the Housatonic River.

¦ Locations adjacent to the Canoe Meadows Audubon Sanctuary (Reach 5).

¦ New Lenox Road Area (Reach 5).

¦ Roaring Brook Area (Reach 5).

¦ Threemile Pond State Wildlife Management Area (Reference).

¦ Southwest Branch Area (Reference).

Sediment samples were collected from 0 to 15-cm depth intervals. A relatively small number of
samples (less than 1%) were collected from intervals other than 0 to 15 cm. A total of 258
sediment samples were collected for PCBs (total and Aroclors), TOC, and sediment grain size
parameters. Approximately 10% of the samples were analyzed for modified Appendix IX (40
CFR 264) constituents, and one sample was analyzed for herbicides and organophosphate
pesticides. Additional PCB congener samples were collected (approximately 14% of the total)
subsequent to the initial core swallow box sampling effort. PCB duplicate samples were
analyzed at approximately 2% of the tPCB analysis.

In addition, approximately 50 samples were collected from floodplain locations, vernal pools,
and permanent backwater locations. These samples were analyzed for PCBs (total and
Aroclors), and nearly 80% of the floodplain samples were collected for TOC and grain size

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 3 1 7/10/2003

-------
1 parameters. Of the floodplain samples, 5% were analyzed for modified Appendix IX (40 CFR

2 264) constituents, and one sample was analyzed for PCB congeners.

3 C.8.3.14.2 Biological Sampling

4 Samples were collected for analysis of tissue concentrations of PCBs and other analytes in

5 various life stages of tree swallows. The design of the tree swallow study is described as

6 follows:

7 "For three study seasons (1998, 1999, and 2000), nest boxes were installed at four

8 sites—three sites along the Housatonic River downstream of the GE facility in Reach

9 5 (location adjacent to the Canoe Meadows Audubon Sanctuary, New Lenox Road,

10 and Roaring Brook) and 1 located along the Southwest Branch, a tributary of the

11 Housatonic River.

12 ¦ In 1999, an additional site just upstream of the GE facility near Taconic Valley

13 Trucking Co. and a site at Threemile Pond State Wildlife Management Area in

14 Sheffield were added.

15 ¦ Eggs and young to be monitored and pippers and nestlings were collected as

16 appropriate. Collected tree swallows were euthanized, and stomach contents were

17 removed and pooled for analysis separate from the carcasses.

18 ¦ Approximately 322 selected tissue samples were analyzed for PCBs (total and

19 congeners), and percent lipids. Additionally, aliphatic hydrocarbons, trace elements,

20 and semivolatile constituents were analyzed in a subset of approximately 10% of the

21 tissue samples. Dioxin and furan constituents were analyzed in approximately 40%

22 of the tissue samples.

23 C.8.3.15 Earthworm and Terrestrial Utter Invertebrate Sampling

24 Soil invertebrates form a pathway by which soil contamination may be passed on to receptors

25 such as the short-tailed shrew (Blarina brevicauda), American robin (Turdus migratorius), and

26 American woodcock (Scolopax minor). These species feed on invertebrates, particularly

27 earthworms, for a major portion of their diet. For the purpose of this study, soil invertebrates

28 were divided into two groups based on their availability to receptors and their degree of exposure

29 to contaminated soils:

30 ¦ Invertebrates living in the soil, represented by earthworms.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g

-------
1

¦ Invertebrates living primarily on the soil surface in the leaf litter, as represented by
litter invertebrates.

The primary objective of the earthworm study was to collect earthworms for toxicological
analysis. In addition, results of tissue analysis and co-located soil samples were intended to be
used to determine the relationship between earthworm tissue concentrations and corresponding
soil concentrations. Earthworms were collected in Reach 5 from the soil surface to 6 inches
below the surface from three sites, which were co-located with small mammal collection sites
(Sites 13, 14, and 15) along the Housatonic River floodplain. Two sites (Sites 13 and 14)
occurred in transitional floodplain forest communities and one (Site 15) in a black ash-red
maple-tamarack calcareous seepage swamp community.

Three species of earthworms were collected during earthworm sampling: Aporrectodea longa,
Aporrectodea trapezoids, and Eisenoides carolinensis. A. longa and A. trapezoids were collected
from Site 13. A. trapezoids and E. carolinensis were collected from Sites 14 and 15. It is also
likely that some Lumbricus species were present; however, these species dwell deeper than 15
cm below the surface (McKeegan personal communication 2000). Morphometric data collected
from specimens included total individual weight and total composite weight. A total of 10
composite tissue samples from each of the three locations were analyzed for PCBs (total and
Aroclors), PCB congeners, percent lipid content, and 10% of the tissue samples were analyzed
for Appendix IX (40 CFR 264) pesticides and dioxin/furan parameters.

Terrestrial litter invertebrates were collected in conjunction with the earthworm collection.
Invertebrates were collected from the leaf litter and beneath decaying woody debris. They were
identified to Order, and a total mass for each Order per site was recorded. Slugs and snails made
up the greatest percentage of the total mass. Sow bugs (Isopoda) were the most abundant
invertebrates, but due to their smaller size, did not comprise the greatest mass. Beetles, spiders,
harvestman spiders ("daddy long-legs"), centipedes, and millipedes were all common. Earwigs,
caterpillars, and one cicada were also collected. Eight composite terrestrial litter invertebrate
samples from each of the locations were analyzed for PCBs (total and Aroclors), PCB congeners,
Appendix IX (40 CFR 264) pesticides, and percent lipid content.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g

-------
1

C.8.3.16 Small Mammal Sampling

The primary objective of this study was to collect small mammals for tissue analysis for PCBs,
and to provide data on the bioaccumulation of these materials in the food web. Analytical results
will be used in the ecological risk assessment and in the food chain model to estimate potential
risks to higher consumers, such as large mammals or raptors that feed upon small mammals.
Small mammals were captured in several areas of floodplain forest habitat within the study area
with varying PCB concentrations (i.e., <1 mg/kg, from 1 to 30 mg/kg, and >30 mg/kg).

C.8.3.16.1 Floodplain Samples

Soil sampling was conducted at the 13 small mammal trapping locations within the floodplain.
Results from the initial screening performed under the Preliminary Work Plan (WESTON 1998)
were reviewed and used to determine specific trapping areas for small mammals.

Approximately 10 soil samples were collected from 0 to 15 cm at each of the 13 locations,
resulting in 157 samples analyzed for PCBs (total and Aroclors). Approximately 10% of the
floodplain samples were analyzed for TOC and grain size parameters.

C.8.3.16.2 Tissue Samples

Small mammal trapping was conducted in September 1998 and from August 1999 to
September 1999. Three sites were chosen as trapping locations in 1998 (Sites IB, 3, and 8) and
in 1999 (Sites 13, 14, and 15). Where small mammal runways were apparent, traps were placed
on the runways to increase capture efficiency. Morphometric data collected on each individual
captured included species, sex (if possible), age, total length, total weight, tail length, hind foot
length, and ear length. During 1999, 116 small mammals were collected, including:

¦ White-footed mouse (Peromyscus leucopus)

¦ Shorttail shrew (Blarina brevicaudata)

¦ Meadow jumping mouse (Zapus hudsonius)

¦ Meadow vole (Microtuspennsylvanicus)

¦ Redback vole (Clethrionomys gapperi)

¦ Masked shrew (Sorex cinereus)

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 34

-------
1 A summary of the 1999 trapping survey results is presented in Table C.8-5. Small mammal

2 whole body tissue from a sample of specimens collected in 1999 was analyzed for PCBs (total,

3 Aroclors, and congeners), percent moisture, percent lipids, organochlorine pesticides, and

4 dioxins/furans (15%).

5 Table C.8-5

7 1999 Trapping Summary Per Location

Site 13

Site 14

Site 15

Common Name

Male

Female

Unknown

Male

Female

Male

Female

Total

White-footed mouse

Shorttail shrew

Meadow jumping
mouse

Meadow vole

Redback vole

Masked shrew

9 C.8.3.17 Macrophytes, Filamentous Algae, Periphyton, and Plankton/Detritus

10 Study

11 The objective of sampling macrophytes, filamentous algae, periphyton, and plankton/detritus

12 was to obtain information on biomass per unit area (standing crop) during a period when

13 significant biomass was present in the Housatonic River study area and to determine contaminant

14 concentrations in these groups for use in the modeling study. Sampling for each of these four

15 community types was conducted in each of the major habitat types in the study area (shallow

16 stream, deep river, backwater, and pond) when the community type was observed to be present.

17 C.8.3.17.1 Macrophyte Sampling

18 ¦ Three samples of each macrophyte community were collected from each sampling

19 area and analyzed individually for biomass; a composite sample was collected and

20 analyzed for tissue residue. Field observations determined that macrophyte samples

21 were not present in one area selected for sampling.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 35

-------
1

¦ All residue samples were analyzed for PCBs (total, Aroclors, and congeners). A
subset of the samples was analyzed for dioxins/furans and organochlorine pesticides.
Biomass samples were analyzed for wet weight, dry weight, TOC, and ash-free dry
matter.

¦ Within each area sampled, the distribution of macrophyte communities was estimated
to allow for a determination of total biomass (standing crop). Macrophyte voucher
samples were collected for identification.

C.8.3.17.2 Filamentous Algae

¦ Filamentous algae sample locations were selected following qualitative surveys of the
filamentous algae distribution in the study area. Three samples of filamentous algae
were collected from each area and analyzed individually for biomass. A single
composite sample was analyzed for tissue residue. Field observations determined that
filamentous algae were not present in three areas selected for sampling.

¦ All samples were analyzed for PCBs (total, Aroclors, and congeners). A subset of
samples was analyzed for dioxins/furans and organochlorine pesticides. Biomass
samples were analyzed for chlorophyll a and phaeophytin, wet and dry weight, TOC,
and ash-free dry matter.

¦ Within each area sampled, the distribution of filamentous algae was estimated,
allowing for a determination of total biomass (standing crop). Voucher samples were
collected and preserved for taxonomic identification.

C.8.3.17.3 Periphyton

¦ Cobble and gravel riffle, soft bottom, and aquatic macrophyte bed locations
containing periphyton communities were selected for sampling. Three samples of
periphyton communities were collected from each study area and analyzed
individually for biomass; a composite sample was analyzed for tissue residue. Field
observations determined that periphyton (from macrophytes) was not present in one
area selected for sampling.

¦ All samples were analyzed for PCBs (total, Aroclors, and congeners). A subset of
samples was analyzed for dioxins/furans and organochlorine pesticides. Periphyton
biomass samples collected from cobble riffles and macrophytes were analyzed for
chlorophyll a, phaeophytin, dry matter, ash-free dry matter, and TOC. Periphyton
samples collected from soft bottom and gravel substrates were analyzed for
chlorophyll a and phaeophytin.

¦ Within each area sampled, the distribution of periphyton was estimated to allow for a
determination of total biomass (standing crop). Voucher samples were collected and
preserved for taxonomic identification.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 36

-------
1 C.8.3.17.4 Plankton/Detritus

2 ¦ Three samples of plankton (phytoplankton and zooplankton) and detritus from each

3 reach were collected and analyzed for biomass; composite phytoplankton and

4 zooplankton samples from each reach were analyzed for tissue residue.

5 ¦ All samples were analyzed for PCBs (total, Aroclors, and congeners); a subset of

6 samples was analyzed for dioxins/furans and organochlorine pesticides. Plankton

7 biomass samples were analyzed for chlorophyll a and phaeophytin, wet and dry

8 weight, TOC, and ash-free dry matter.

9 ¦ Detritus biomass samples were analyzed for total organic matter (TOM) and

10 dissolved organic matter (DOM).

11 C.8.3.18 Rare Plants and Natural Communities Survey

12 The rare plant and natural community survey included the following activities:

13 ¦ Compilation of information concerning known and historic distributions of rare

14 species and communities known or suspected to occur in the study area.

15 ¦ Taxonomic identification of rare occurrences—Rare plant species to be identified to

16 subspecific level based on morphological, phenological, habitat, and distributional

17 information. Collection of voucher specimens and photographs, if appropriate, to

18 allow confirmation of identification.

19 ¦ Natural community and population data—Each rare occurrence (plant or community)

20 to include collection of descriptive information (e.g., associated plant species,

21 population area, canopy height, areal cover of species, canopy tree age and diameter,

22 and substrate).

23 ¦ Collection of tree cores, enumeration of growth rings, and samples saved for

24 verification if needed.

25 Seven communities of state conservation concern were identified in the study area. A total of 20

26 state-listed species were documented from 37 sites.

27 C.8.3.19 Dragonfly Survey

28 The objective of the dragonfly survey was to determine the species that occur or may occur

29 within the study area, with a special emphasis on rare species.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g

-------
1

Dragonfly surveys were to consist of exuvia (shed exoskeleton of nymph stage) collection along
the riverbanks. Opportunistic aerial netting of adults was also to be conducted during exuvia
collections and other field surveys. The planned surveys are described below:

4 ¦ Surveys were conducted on foot in the shallow upstream reaches and by canoe in the

5 deeper downstream reaches.

6 ¦ Two observers walked or floated slowly along the shore to collect exuvia from

7 vegetation, rocks, logs, and exposed substrates. A total of 12 areas were surveyed

8 during five site visits for a total of 651 identifications.

9 ¦ Exuvia were placed in round paperboard containers and sent to a contracted

10 laboratory for identification.

11 ¦ Surveys were conducted over a 2-day period and repeated five times from

12 approximately mid-May to August.

13 ¦ Adult dragonflies were netted, killed in a killing jar, and then mounted as reference

14 specimens. These specimens were sent to a contract laboratory for verification. A

15 total of 69 adult tissue samples were identified.

16 C.8.3.20 Avian Field Survey

17 The objective of the avian field survey was to identify the species of birds that occur in the study

18 area.

19 The survey procedures are described below:

20 ¦ Playback point counts were used to survey raptors within the study area and in two

21 reference areas.

22 ¦ Transects were established along the river or other water body (for reference area)

23 and adjacent roads, with point counts being taken at 300-m intervals.

24 ¦ Approximately 10 minutes was spent at each point, with calls being broadcast, at

25 various angles, for 10 seconds followed by 30 seconds of silence for each call.

26 ¦ All raptors observed were identified and recorded along with the type of observation.

27 Raptor surveys were conducted between one-half hour before sunrise to sunset,

28 except owl surveys, which were conducted one-half hour after sunset to sunrise.

29 ¦ Transects were visited two to three times during breeding season, at least once during

30 mating season, and once during the nesting-fledgling period.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 38 7/10/2003

-------
1 Table C.8-6 summarizes the completed avian field survey.

2 Table C.8-6

4 Summary of Avian Surveys

Avian Group

No. of Areas

No. of Survey Stations

No. of Visits

Species Observed

Marsh Birds

Hawks

Owls

Forest Birds

6 No samples were proposed or collected for chemical analysis. Information recorded at each

7 survey site included location, start and end times, observer, date, visit number, wind speed, cloud

8 cover, precipitation, responses per species, and all other wildlife sightings.

9 C.8.3.21 River Otter, Mink, and Bat Surveys

10 The objectives of these surveys are described as follows:

11 "To determine if mink (Mustela vison) and river otter (Lutra canadensis) are present in

12 the study area.

13 ¦ To determine which species of bats are present in the study area and what habitats

14 they are using for feeding, and, potentially, roosting.

15 River Otter and Mink Surveys:

16 ¦ Mammal snow track counts were conducted in various habitat types.

17 ¦ Several 500-m-long (1,600-ft-long) transects were established so that each habitat

18 type (forested and shrub swamp, emergent marsh, forested upland, and agricultural

19 field) was represented.

20 ¦ Transects in Reaches 5 and 6 and reference areas were walked a minimum of two

21 times after fresh snowfall. However, the number of Year 2000 snow tracking visits

22 for mink and otter depended upon snow conditions. Lack of fresh snowfall and early

23 thaw prevented the proposed number of snow tracking visits. Additionally, access to

24 some reference areas was blocked by heavy snowfall.

25 ¦ Scent stations were used to detect the presence or absence of mink and otter.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 39

-------
1 ¦ Transects were set up parallel to the river. Each transect was 600-m (2,000-ft) long

2 and contained 10 scent stations at 60-m (200 ft) intervals.

3 ¦ Fine sand was placed around each scent post in an approximate 0.5-m (1.6-ft) radius

4 to facilitate track observation.

5 ¦ Transects were visited for 3 days following setup, weather permitting.

6 Table C.8-7 summarizes the completed mink and otter survey.

7 Table C.8-7

9 Survey Summary - Mink and Otter

Method

No. of Transects

No. of Survey Stations

No. of Visits

1999 Snow Tracking

N/S

2-4

2000 Snow Tracking and Scent Post

150

1-2

1999 Scent Post

3-4

1998 Scent Post

11 Bat Survey:

12 ¦ Bat species were surveyed using echolocation.

13 ¦ Three 1-km (0.6-mile) transects were set up parallel to the river.

15 Surveys were conducted starting 15 minutes after sunset and performed for 120 minutes. Table

16 C.8-8 summarizes the bat survey.

17 Table C.8-8

19 Survey Summary - Bats

Method

Location

No. of Calls Recorded

No. of Species Recorded

Transect 1

363

Echolocation

Transect 2

805

Transect 3

639

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g

-------
1 C.8.3.22 Dietary Exposure of Mink

2 This study was designed to evaluate whether farm-raised mink fed diets containing PCB-

3 contaminated fish from the Housatonic River would exhibit impaired reproductive performance

4 and/or offspring (kit) growth and survival. Results of the study would provide information and

5 data for the Ecological Risk Assessment.

6 ¦ Fish were collected from the Housatonic River in areas of mink habitat. The fish

7 were ground and blended into a homogeneous mixture. Three composite samples

8 were collected and analyzed for organochlorine pesticides, tPCBs, PCB congeners,

9 percent lipids, and dioxins/furans.

10 ¦ The mink toxicity tests used six dietary treatments, one of which would be a control

11 diet containing uncontaminated ocean fish. The remaining five diets contained

12 mixtures of ocean fish and homogenized fish from the Housatonic River.

13 ¦ Three random grab samples from each dietary treatment (for a total of 18) were

14 collected for analysis of tPCBs, PCB congeners, percent lipids, and dioxins/furans; an

15 additional sample from each dietary treatment was collected for nutrient analysis.

16 ¦ Liver tissue was collected from 12 adult females, six mink kits at 6 weeks of age, and

17 an additional six kits at 6 months of age from each dietary treatment. The liver

18 samples were analyzed for tPCBs, PCB congeners, percent lipids, and dioxins/furans.

19 A reduction in samples occurred during the study due to the death of five adult

20 females during the trial period; therefore, these liver samples were not retained for

21 analysis.

22 ¦ Measurement endpoints for the mink toxicity study include body weight, length of

23 gestation, reproductive success, survival, histopathology, biochemical analyses

24 (including cytochrome P-450 levels), and organ weights.

25 C.8.3.23 Crayfish Tissue Sampling

26 The primary objective of the crayfish study was to collect crayfish from areas of the Housatonic

27 River representing a range of sediment PCB concentrations for tissue residue analysis and to

28 provide data on the bioaccumulation of PCBs, dioxins/furans, and organochlorine pesticides in

29 the aquatic food web for fate and effects and exposure models.

30 Sampling was conducted during September and October 1999 using seine nets, hand nets, and

31 baited crayfish traps. A total of six locations were sampled, including four sites along the main

32 stem of the Housatonic River and two reference sites (the West Branch of the Housatonic River

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 41

-------
1

upstream of the confluence, and along the East Branch of the Housatonic River upstream of
Center Pond in Dalton). Morphometric data recorded for captured individuals included species,
sex, weight, total length, and carapace length. The virile crayfish (iOrconectes virilis) was the
only crayfish species captured and identified during the surveys in 1999. A total of 153 crayfish
were collected for tissue analysis: 90 from Reach 5 and 63 from the two reference areas. A
subset of 60 crayfish samples was selected for tissue analysis for PCBs (total,
congeners/homologs), organochlorine pesticides, percent moisture, and percent lipids.
Approximately 18% of these samples were analyzed for dioxin/furan parameters.

C.8.3.24 Reptile and Amphibian Survey

The objective of the amphibian and reptile surveys was to determine species present, abundance,
and habitat use in the Housatonic River and floodplain. The reptile and amphibian survey is
described below:

¦ Vernal pool locations were mapped and the habitat characterized.

¦ Pools were visited and visual and acoustic surveys were conducted to document the
presence of reptiles and amphibians.

¦ Aquatic funnel trapping was conducted during the period when larval amphibians
were present, and all individuals captured were identified and measured.

If incidental mortality occurred during sampling, these individuals were to be collected for
analysis of PCBs (total, Aroclors, congeners, and homologs), lipids, and moisture. However, no
incidental mortality occurred during the survey.

C.8.3.25 Waterfowl Collection and Tissue Sampling

Waterfowl, including wood ducks (Aix sponsa) and mallards (Anas platyrhynchos), had been
observed using Woods Pond and upstream floodplain wetlands for breeding, brood rearing, and
feeding. Waterfowl hunting was also a common activity along this portion of the Housatonic
River. For these reasons, Woods Pond and its backwaters were selected, along with a reference
area, as wood duck and mallard collection sites in support of both human health and ecological
risk assessments.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g ^

-------
1 During August 1999, the Massachusetts Division of Fisheries and Wildlife (MDFW) captured

2 waterfowl in Woods Pond in the evening by using an airboat, headlights, and hand nets. To

3 supplement those ducks captured by MDFW, trapping was conducted in Woods Pond and

4 adjacent backwaters. Two floating box traps and one walk-in trap were used to capture

5 waterfowl in backwaters near Woods Pond and from Threemile Pond State Wildlife

6 Management Area, a reference area located in Sheffield, MA, in August and September 1998.

7 Morphometric data collected from specimens included age, sex, wing chord length, and total

8 weight. A total of 20 wood ducks (11 males, nine females) and five mallards (four males, one

9 female) were collected from the Housatonic River, and 20 wood ducks (12 males, eight females)

10 were collected from Threemile Pond State Wildlife Management Area. Liver and breast tissues

11 were analyzed for PCBs (total, Aroclors, congeners/homologs), dioxins/furans, organochlorine

12 pesticides, percent lipids, and percent moisture.

13 C.8.3.26 June 2000 Flood Sampling

14 Photographic documentation of the flood event that began on the evening of 6 June 2000 was

15 collected at various locations along the East Branch of the Housatonic River and the surrounding

16 floodplain from Newell Street to Woods Pond. Visual inspections of the material deposited by

17 the floodwaters were conducted from Lyman Street Bridge to Woods Pond and eight locations

18 were chosen for sampling of recently deposited sediments as scrapings from soil and vegetation.

19 Seven samples recovered from soil scrapings were analyzed for PCBs (total and Aroclors), TOC,

20 and grain size parameters. Vegetation samples were analyzed for PCBs (total and Aroclors)

21 only.

22 C.8.3.27 Discretionary Sampling

23 An iterative approach to the overall sampling program was used whenever possible to improve

24 the effectiveness of the field investigations. A review of newly collected data from recently

25 completed studies was used to design focused discretionary field sampling activities to fill in

26 identified data gaps and better fulfill data quality objectives.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 43

-------
1 Discretionary sampling activities were initiated to provide further characterization of spatial and

2 temporal variability in the data to support and complete the human health and ecological risk

3 assessments and river system modeling activities.

4 In Reach 5, an additional 428 sediment samples and 327 floodplain samples, including duplicate

5 samples, were collected and analyzed for PCBs (total and Aroclors), TOC, and sediment grain

6 size parameters. Including duplicate samples, a total of seven floodplain samples were collected

7 in Reach 6 and analyzed for PCBs (total and Aroclors) and approximately 10% of the samples

8 were analyzed for TOC and grain size parameters. In addition, nearly 2% of the floodplain

9 samples were analyzed for PCB congeners.

10 C.8.4 Water Quality Sampling and Modeling Studies

11 Water quality sampling was conducted primarily to support the modeling study. Specific

12 activities undertaken to support the modeling data needs include surface water sampling and

13 storm event sampling, measurement of channel geometry cross sections, and flow monitoring.

14 Water sampling programs are summarized in the following sections. Each summary includes a

15 brief description of the program proposed in the SIWP for sample collection and analysis.

16 C.8.4.1 Surface Water Sampling—Monthly

17 Surface water samples were collected monthly at 17 locations along the Housatonic River and

18 tributaries from August 1998 through September 1999 (15 months, due to an overlap in sampling

19 at the end of July and beginning of August 1998). A total of 14 of the locations were on the East

20 Branch or main stem of the Housatonic River. Two locations were on tributaries to the river, and

21 the remaining location was on the West Branch of the Housatonic River near the confluence with

22 the East Branch.

23 The following parameters were analyzed:

24 ¦ Total suspended solids

25 ¦ Total dissolved solids

26 ¦ Filtered and unfiltered PCBs (total, Aroclors, and congeners)

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g ^

-------
1

¦ Biochemical oxygen demand (5-day)

¦ Phosphorus (ortho- and total -P)

¦ Appendix IX parameters (filtered and unfiltered metals)

¦ Calcium

¦ Magnesium

¦ Alkalinity

¦ Hardness

¦ Chlorophyll a

¦ Total Kj el dahl nitrogen

¦ Ammonia nitrogen

¦ Nitrite nitrogen

¦ Nitrate nitrogen

¦ Total organic carbon (TOC) and dissolved organic carbon (DOC) (particulate organic
carbon by difference)

¦ Cyanide

¦ Sulfide

During each monthly sampling activity, all locations were also analyzed for field parameters
including hydrogen ion concentration (pH), temperature, dissolved oxygen, turbidity, and
specific conductance using field instruments.

From February 1999 to September 1999, high-volume water sampling was initiated for analysis
of dissolved and particulate PCB congener analysis.

C.8.4.2 Supplemental Surface Water Study

The supplemental surface water study was proposed to measure PCBs, organic carbon (OC), and
related parameters in surface water in Reaches 5 and 6, and the West Branch of the Housatonic
River. Samples were collected from four locations:

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 45

-------
1 ¦ Pomeroy Avenue Bridge (Reach 5)

2 ¦ West Branch of the Housatonic River near the confluence

3 ¦ New Lenox Bridge (Reach 5)

4 ¦ Woods Pond Footbridge (Reach 6)

6 Three sampling events at the four locations were conducted to evaluate the Housatonic River at

7 low-, medium-, and high-flow conditions. Prior to the three sampling events, a single sampling

8 event was conducted with three replicate samples at the Pomeroy Avenue Bridge location as a

9 trial of the sampling, processing, and analytical protocols (trial event).

10 During each event, the following parameters were measured: dissolved and particulate PCB

11 congeners; total, dissolved, and particulate OC; total suspended solids (TSS); and chlorophyll a.

12 Staff gage readings for each of the stations and at the USGS Coltsville gage during the time of

13 each sampling were also recorded at the time of sampling.

14 C.8.4.3 Water Column Profile(s) for Woods Pond

15 Water quality measurements were collected to evaluate the degree of eutrophication in Woods

16 Pond. The study included a water column profile of the deep basin at the eastern side of Woods

17 Pond and six locations within and upstream of Woods Pond for the following parameters: pH,

18 temperature, dissolved oxygen, specific conductivity, and turbidity.

19 C.8.4.4 Storm flow Monitoring and Sampling

20 Water samples were collected from the Housatonic River and selected tributaries under

21 conditions when water quality and suspended sediment transport were influenced by storm -

22 induced flows. Sampling was conducted periodically from August 1999 to October 1999, along

23 the East and West Branches of the Housatonic River, and tributaries of the East Branch to

24 provide suspended solids, PCB, and water quality data for use in the modeling study. Surface

25 water samples and suspended sediment samples were collected from three primary locations and

26 five secondary locations. Water quality samples were collected hourly throughout each event,

27 but only representative samples from the baseline, rising limb, plateau, and falling limb of the

28 hydrograph at each station were submitted for analysis.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g

-------
1

C.8.4.4.1

Primary Locations

The primary sampling locations were the following:

¦ Pomeroy Avenue (ST000004)—Located at the intersection of Pomeroy and
Appleton Avenues, on residential property. In addition to water quality and water
chemistry analysis, depth-integrated sampling and velocity measurements were
completed from the Pomeroy Avenue Bridge adjacent to the entrance to Fred Garner
Park.

¦ New Lenox Road (ST000007)—Located at the Decker Canoe Launch, on the south
side of New Lenox Road. Depth-integrated sampling and velocity measurements
were completed from the New Lenox Road Bridge.

¦ Woods Pond (ST000009)—Located at the footbridge across the outlet of Woods
Pond. Depth-integrated sampling and velocity measurements were completed from
this location.

Water samples from the primary locations were analyzed for:

¦ Ammonia nitrogen

¦ Nitrite nitrogen

¦ Nitrate nitrogen

¦ Total Kj el dahl nitrogen

¦ Organic phosphorus

¦ Ortho-phosphorus

¦ Total phosphorus

¦ Chlorophyll a

¦ Biochemical oxygen demand (5-day)

¦ Chemical oxygen demand

¦ Total organic carbon

¦ Dissolved organic carbon

¦ Particulate organic carbon

¦ Total suspended solids

¦ Polychlorinated biphenyls (PCBs) (total, Aroclors, and congeners)

¦ Dissolved PCBs (total, Aroclors, and congeners)

¦ Alkalinity

¦ Hardness

¦ Turbidity (field measurement)

¦ Temperature (field measurement)

¦ pH (field measurement)

¦ Dissolved oxygen (field measurement)

¦ Specific conductance (field measurement)

Water samples from the secondary locations were analyzed for the same parameters as the
primary locations, with the exception of PCBs, alkalinity, and hardness. Optional PCB analyses

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC O AH 7/10/2003

-------
1 were conducted at selected secondary locations based on the initial results. Additional

2 parameters were to be analyzed if needed for the Housatonic River modeling. Additional

3 volumes of water at the secondary locations were sampled for laser or sieve/hydrometer analysis

4 of suspended solids.

5 In addition to the analytical parameters described here, water quality measurements were

6 collected and recorded at regular intervals from both primary and secondary locations. These

7 measurements included turbidity, temperature, pH, dissolved oxygen, specific conductivity, and

8 staff gage height. At primary locations, velocity measurements were collected from the left,

9 middle, and right portions of the channel at depths of 20% and 80% of the total depth below the

10 water surface.

11 Suspended sediments were collected from the water column and collected in filter bags with 5-

12 micron screen size. Sediments were analyzed into four grain size categories:

13 ¦ 5 to 10 |im

14 ¦ 10 to 62 |im

15 ¦ >62 to 250 |im

16 ¦ >250 |im

18 An aliquot from each grain size fraction was analyzed for tPCBs and TOC. A limited number of

19 samples were tested for 1,2,4-trichlorobenzene from the second, third, and fourth fraction (>62 to

20 250 |im, 10 to 62 |im, and 5 to 10 |im, respectively).

21 C.8.4.4.2 Secondary Locations

22 The secondary sampling locations were the following:

23 ¦ Hubbard Avenue (ST000002)—Located on the City Tire parking lot, on the north

24 side of Hubbard Avenue, near Wal-Mart. There is no staff gage at this location,

25 however, the USGS Coltsville gaging station is adjacent to the Hubbard Avenue

26 station.

27 ¦ Unkamet Brook (ST000003)—Located at the outflow pipe off Merrill Road (north

28 side) in the Conrail train yard behind the Pittsfield School Bus Depot.

29 ¦ West Branch (ST000005)—Located approximately 100 yards upstream of the

30 confluence, in Fred Garner Park, off Pomeroy Avenue.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g 7/10/2003

-------
1 ¦ Sackett Brook (ST000006)—Located in the Canoe Meadows Audubon preserve,

2 accessed through Tween Brook Farm, East New Lenox Road.

3 ¦ Roaring Brook (ST000008)—Located approximately V2 mile south of intersection of

4 New Lenox Road and October Mountain Road (Roaring Brook Road).

5 C.8.4.5 Flow Monitoring

6 Flow monitoring was completed at numerous locations along the East and West Branches of the

7 Housatonic River and in selected tributaries to characterize flows at different stages of river

8 height. Table C.8-9 summarizes the locations and flow measurements collected.

9 Table C.8-9

11 Measurement Summary - Flow Monitoring

Location

Dawes Ave.
Bridge

Pomeroy Ave.
Bridge

West Branch

Holmes Rd.
Bridge

New Lenox Rd.
Bridge

Woods Pond
Bridge

Lenoxdale
Bridge

Unkamet Brook

Sackett Brook

Roaring Brook

Goodrich Pond

Conducted

Measurements

13 In September 2001, pressure transducers were installed at Pomeroy Avenue Bridge, West

14 Branch, Electric Power Research Institute (EPRI), New Lenox Road Bridge, and the Woods

15 Pond footbridge to provide additional data on the surface water elevation and temperature.

16 C.8.4.6 River Channel Geometry Measurements

17 Floodplain and channel cross sections were surveyed at locations between the GE facility and

18 Woods Pond Dam to provide the channel geometry needed to support the modeling study.

19 ¦ Channel measurements included water depth, sediment depth, and distance to the top

20 of bank at each transect.

21 "A total of 286 cross sections were completed in Reach 5 to support advanced

22 modeling approaches.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC q g ^g

-------
1 C.8.4.7 Vertical Definition Cores

2 Subsurface materials were sampled in the river in Reaches 5 and 6. These samples, known as the

3 vertical definition cores (VDCs), had three objectives:

4 "To determine the vertical extent of PCB contamination.

5 "To identify subsurface lithologic features that may have an impact on hydrology or

6 contaminant migration.

7 "To look for evidence of an upper, "active" layer of sediment, which is subject to

8 resuspension in certain flow regimes.

9 Sampling was completed at 26 VDC locations in the river. Each VDC location comprised two

10 sample locations on opposite sides of the river, each midway between the thalweg (deepest part

11 of the channel) and the bank. The VDC sampling program is described as follows:

12 ¦ Samples were collected from four 6-inch intervals: the interval at the surface of the

13 core (0 to 0.5 ft below ground surface [bgs]); 3 to 3.5 ft bgs; the interval halfway

14 between 3.5 ft bgs and the bottom of the core; and the interval at the bottom of the

15 core. In addition, where unique lithologic units were identified, additional samples of

16 such layers were considered for sampling.

17 Supplemental floodplain samples were collected at 26 locations in the floodplain adjacent to the

18 VDC in-river locations. Each supplemental floodplain location comprised two sample locations,

19 one on each bank of the river. The sampling program is described as follows:

20 ¦ PCB/TOC samples to be collected from each core. Any fine units between 2.5 ft and

21 the bottom of the core were sampled, as was the deepest material from each core.

22 ¦ Two cores from each reach of the river (Reaches 5A, 5B, and 5C for a total of six

23 cores) in the study area were dated by means of a lead-210 method so that

24 depositional rates could be calculated.

25 Samples that appeared appropriate for carbon-14 dating were selected at the discretion of the

26 samplers. Table C.8-10 provides a summary of the number of analytical results by parameter for

27 the VDC sampling program.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 50

-------
1

Table C.8-10

Analysis Summary - VDCs and Supplemental Floodplain Sampling

Media

Total PCB

Aroclors

TOC

Grain Size

Bulk Density

Lead-210

Radiocarbon

Sediment

102

Riverbank

Floodplain

C.8.4.8 Sediment Microscopy

A total of 10 riverbank sediment samples located between the confluence of the East and West
Branches of the Housatonic River and New Lenox Road (Reach 5) were visually evaluated under
a binocular microscope to estimate the mineralogical composition of the sediments. A portion of
each of the 10 samples was sent to a laboratory for x-ray diffraction and for scanning electron
microscopy analysis for mineralogical determination.

C.8.4.9 Toe Pins

Based on the results of an initial survey to locate areas of bank erosion in Reach 5, five locations
were selected for the installation of toe pins. In October 2001, four pins of half-inch rebar, each
5 ft (1.5 m) in length, were driven into the bank to a depth of 4 ft (1.2 m) at each location,
leaving a 1 -ft (30-cm) length of each bar exposed. The pins were installed during low- to
medium-flow conditions at approximately the water surface and each pin was marked to indicate
the amount exposed. The toe pins were re-measured in May 2001, August 2001, April 2002, and
June 2002.

C.8.4.10 River Channel Resurveys

Nine transects were selected for periodic survey to evaluate changes in channel morphology over
time and after significant high-flow events. Transects were surveyed in September 2001, April
2002, and June 2002.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 5 1 7/10/2003

-------
1 C.8.4.11 Additional Velocity Measurements

2 Acoustic Doppler Current Profiler (ADCP) velocity data were collected in several areas of

3 Reaches 5 and 6. These areas included the headwaters of Woods Pond, Woods Pond, New

4 Lenox Road, portions of the Test Reach (so-called ADCP bend), and Pomeroy Avenue. The

5 ADCP work had two primary objectives:

6 "To collect high-frequency velocity data for use in calibration of the hydrodynamic

7 model and to support development of stage/discharge rating curves.

8 "To collect detailed bathymetry information to evaluate potential changes in the

9 sediment bed over time.

10 C.8.4.12 Additional Stormflow Monitoring (Supplemental Modeling Study)

11 The primary objective of the additional stormflow monitoring was to evaluate selected

12 characteristics of the riverbed and suspended sediment loads in the Housatonic River during

13 high-flow events. Representative samples of TSS and bedload sediments were proposed along

14 with velocity measurements. Up to four events were proposed with sampling to be conducted at

15 three locations: the Pomeroy Avenue Bridge, adjacent to EPRI, and at the New Lenox Road

16 Bridge. For each event, nine bedload sediment samples were proposed for PCB, TOC, and grain

17 size analysis, and up to 15 surface water samples were proposed for PCB, TSS, and TOC

18 analysis. At the time of preparation of this Ecological Risk Assessment report, one sampling

19 event has been completed, with samples from Pomeroy Avenue Bridge (only) submitted for

20 tPCB, TOC, grain size parameters (sediment only), and TSS (surface water only) analysis.

21 C.8.4.13 Sediment Flume Studies

22 A Sediment Flume (Sedflume) study was conducted to evaluate the erosion potential of bottom

23 sediments as a function of shear stress. Sediment was analyzed for TOC, bulk density, and grain

24 size parameters in support of the study. A total of 24 total cores (17 from Reach 5 and 7 from

25 Reach 6, including duplicate and reconstituted cores) were collected.

26 C.8.4.14 Meandering and Bank Erosion Study

27 The objectives of the meander survey were the following:

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 52 7/10/2003

-------
1

¦ To identify long-term changes in the course of the river and provide estimates of the
extent of riverbank erosion by reach.

¦ To estimate short-term changes in the volume of riverbank soil loss per reach after a
bank-full flow event (1.5-year flood event).

¦ To map areas that are currently eroding or accreting and determine the percent of
linear riverbank eroding for each reach.

¦ To characterize sediment grain size for a sample of eroding banks.

8 Stereo-pairs of aerial photos from 1972 and 1952 were used to create a georeferenced river line

9 for each of these time periods, and these river lines were overlaid with the 1990 and 2000 river

10 lines in AutoCAD. The comparison of these river lines was used to estimate the amount of long-

11 term erosion.

12 Field visits were conducted to map areas of ongoing erosion and accretion in Reach 5. Site-

13 specific erosion and accretion characteristics (e.g., extent of erosion/accretion area, soil type,

14 vegetation) were recorded at each location. A total of 15 sites that covered the range of potential

15 low to high erosion rates were selected in Reaches 5A and 5B. Transects at these locations were

16 surveyed twice, with the second survey following a bank-full flood event. The second terrain

17 surface survey was overlaid onto the initial terrain surface to estimate the volume of soil loss at

18 each site.

19 A total of 29 soil samples (one to three per site) were collected and analyzed for sediment grain

20 size parameters.

21 C.8.4.15 Vegetative Stem Counts

22 A vegetative stem count survey was conducted in the floodplain of the PSA to provide data on

23 vegetation density that potentially could be incorporated into the hydrodynamic model to

24 simulate attenuation of flow velocity during flood events.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC

C.8-53

7/10/2003

-------
1 C.8.5 References

2 Burton, G.A., Jr. 2001. Assessment of In Situ Stressors and Sediment Toxicity in the Lower

3 Housatonic River, Revised Draft Report. Institute of Environmental Quality, Wright State

4 University, Dayton, OH. 2 November 2001.

5 McKeegan, C., Ohio State University. 18 August 2000. Personal Communication (email) to

6 Arthur Haines, Woodlot Alternatives, Inc. Re: Earthworms Identified.

7 Quinney, T.E., and C.D. Ankley. 1985. Prey size selection by tree swallows. Auk 102:245-250.

8 WESTON (Roy F. Weston, Inc.). 1998. Preliminary Work Plan for Engineering Evaluation and

9 Cost Analysis (EE/CA) and Remedial Investigation (RI) Work Plan for OU 2, Housatonic River.

10 Prepared for U.S. Army Corps of Engineers and U.S. Environmental Protection Agency.

11 WESTON (Roy F. Weston, Inc.). 2000. Supplemental Investigation Work Plan for the Lower

12 Housatonic River. Prepared for U.S. Army Corps of Engineers and U.S. Environmental

13 Protection Agency. 22 February 2000. DCN GEP2-020900-AAME.

14 WESTON (Roy F. Weston, Inc.). 2001. Quality Assurance Project Plan, Vol. I - Text, Vol. II -

15 Appendix A, Vol. IIA - Appendix A, cont'd., Vol. IV - Appendices E and F. Prepared for U.S.

16 Army Corps of Engineers and U.S. Environmental Protection Agency. DCN GE-021601-

17 AAHM.

18 WESTON (Weston Solutions, Inc.). 2002. Rest of River Site Investigation Data Report. Prepared

19 for U.S. Army Corps of Engineers and U.S. Environmental Protection Agency. DCN GE-

20 080202-ABDK.

O:\20123001.096\ERA_PB\ERA_APC8_PB.DOC C 8 54

-------
APPENDIX C.9
SUMMARY OF ANALYTICAL METHODS

MK01 |O:\20123001.096\ERA_PB\ERA_APC9_PB.DOC

-------
1

APPENDIX C.9

SUMMARY OF ANALYTICAL METHODS

4 C.9.1 Introduction

5 This appendix provides an overview of the analytical methods and associated quality

6 assurance/quality control (QA/QC) procedures used to generate the numerical data upon which

7 this ecological risk assessment is based. Substantial detail on analytical methods is provided in

8 the Quality Assurance Project Plan (WESTON 2001a), Supplemental Investigation Work Plan

9 (WESTON 2000), Field Sampling Plan (WESTON 2001b), and in the appropriate EPA and

10 ASTM documents that describe the standards for the cited procedures. Primary emphasis in this

11 summary has been placed upon describing the analyses for PCBs, which are the primary COC

12 evaluated in the risk assessment.

13 C.9.2 Analytical Overview

14 A number of different analytical methods and laboratories were used to develop data on PCB

15 concentrations in the Housatonic River. Different methods were used to serve different data

16 quality objectives, which were designed to provide a balance between extensive coverage of the

17 area, achieving particular analytical detection limits, measuring specific analytes (e.g., PCBs as

18 Aroclors versus as congeners), and cost. In addition, the measurement of PCBs in different

19 media dictated the use of different analytical methods. These analyses, each of which will be

20 discussed in a subsequent section, included the following:

21 Water

22 ¦ PCBs as Aroclors (Severn-Trent Laboratories, Modified EPA Method 8082).

23 ¦ PCBs as congeners (Alta Analytical Laboratories, Modified EPA Method 1668).

24 ¦ Northeast Analytical (Green Bay Mass Balance Method).

26 Soil/Sediment

¦ PCBs as Aroclors, field screening laboratory (On-Site Laboratories, Modified EPA
Method 8082).

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC

C.9-1

7/10/03

-------
1

¦ PCBs as Aroclors, fixed laboratory (Severn Trent Laboratories, Modified EPA
Method 8082).

¦ PCBs as congeners (Pacific Analytical, Inc., Modified EPA Method 1668;
Geochemical and Environmental Research Group, Modified EPA Method 8082;
Northeast Analytical, Inc.; Green Bay Mass Balance Method [Swackhamer]).

6 Tissue

¦ PCBs as congeners (Geochemical and Environmental Research Group, Modified EPA
Method 8082).

9 C.9.3 Analysis of Water for PCBs

10 C.9.3.1 PCBs as Aroclors

11 The majority (approximately 500) of the water samples collected in support of this investigation

12 were analyzed for PCBs as Aroclors using quantitation techniques that avoided double-counting

13 of congener peaks common to two or more Aroclors. The Aroclor values were summed to

14 provide a determination of tPCBs. Samples were provided to the laboratory in both filtered and

15 unfiltered form, thereby allowing determination of both dissolved and total (dissolved plus

16 adsorbed to particulates) concentrations of analytes.

17 Details of the analytical method are provided in Appendices A-24 and A-37 to the QAPP

18 (WESTON 2001a). This method involved the extraction of approximately 1 L of raw sample

19 with methylene chloride. The resultant extract was injected into a gas chromatograph with

20 electron capture detector (GC/ECD) calibrated using a five-point calibration curve for Aroclor

21 1260, the most common Aroclor identified in the study area. Individual Aroclors were

22 quantitated using three to five major peaks per Aroclor and an individual response factor for each

23 peak, averaging the three to five determinations. Method detection limits and related

24 information for this analytical method are presented in Table C.9-1.

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC

C.9-2

7/10/03

-------
Table C.9-1

PCB Aroclor Reporting Limits (SW-846 8082)

Analytical Parameter

CAS Number

Soil/Sediment
and NAPLa
Reporting Limit

(Mg/kg)

Waterb
Reporting Limit

(Mg/L)

SW-846 8082 (SOPs A-24, A-37, A-48, A-49, A-50, A-73, A-74, A-75, and A-79)

PCB-Aroclor 1016

12674-11-2

0.014

PCB - Aroclor 1221

11104-28-2

0.014

PCB - Aroclor 1232

11141-16-5

0.014

PCB - Aroclor 1242

53469-21-9

0.014

PCB - Aroclor 1248

12672-29-6

0.014

PCB - Aroclor 1254

11097-69-1

0.014

PCB - Aroclor 1260

11096-82-5

0.014

SW-846 Modified 8082 (FLD MTHD) (SOP A-37)

PCB - Aroclor 1248

12672-29-6

500

PCB - Aroclor 1254

11097-69-1

500

PCB - Aroclor 1260

11096-82-5

500

aNAPL reporting limits will reflect these levels whenever achievable.

bAroclor reporting limits are 0.5 |ig/L for field blanks associated with soil/sediment sample.

MK01 |O:\20123001.096\ERA_PB\ERA_APC9_PB.DOC O ^ 7/10/2003

-------
1 C.9.3.2 PCBs as Congeners

2 Approximately 175 water samples were also analyzed for congeners by Alta Analytical

3 Laboratory (Sacramento, CA) using high-resolution mass spectrometry (HRMS). This method

4 provides determinations for 52 individual PCB congeners (including the most toxicologically

5 significant congeners), each of the 10 PCB homolog groups, each of 7 target Aroclors, and

6 tPCBs. Details of this analytical method are provided in Appendix A-38 to the QAPP (WESTON

7 2001a).

8 In the GC/MS method, stable isotopically labeled analogs of PCB target congeners were spiked

9 into the approximately 1-L sample and either vacuum filtered through a solid-phase extraction

10 (SPE) disk with subsequent methylene chloride extraction or liquid-liquid extracted directly with

11 methylene chloride. Extracts were concentrated to near dryness and injected into the GC

12 followed by HRMS with selected ion monitoring (SIM). Individual PCB congeners were

13 identified by comparing the GC retention time and ion abundance ratio with corresponding

14 standards. For PCB congeners with labeled analogs, the concentration of each compound was

15 determined using the isotope dilution technique. Homologs, and congeners without labeled

16 analogs, were determined using the internal standard technique. Method detection limits and

17 related information are presented in Table C.9-2.

18 In addition to the monthly and storm sampling programs, a supplemental surface water sampling

19 study was conducted in 2001-2002 specifically to evaluate partitioning behavior. The study was

20 conducted in cooperation with GE, who contracted the chemistry analyses to Northeast

21 Analytical, Inc. (NEA); the samples were analyzed for PCB congeners using the Green Bay

22 Mass Balance Method.

23 C.9.4 Analysis of Soil and Sediment for PCBs

24 C.9.4.1 PCBs as Aroclors

25 The majority (approximately 12,000) of the soil and sediment samples collected in conjunction

26 with this project were analyzed for PCBs as Aroclors using quantitation techniques that avoided

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 4

-------
Table C.9-2

PCB Congener/Homolog Reporting Limits [HRGC/HRMS] (Modified EPA 1668)

(SOP A-38)

Congener
Number/Homolog
Group

Soil/Sediment
Reporting Limits
(US/kg)

Water
Reporting Limits (ng/L)

PCB-1

0.05

0.50

PCB-3

0.05

0.50

PCB-8

0.05

0.50

PCB-15

0.05

0.50

PCB-18

0.05

0.50

PCB-28

0.05

0.50

PCB-3 7

0.05

0.50

PCB-44

0.05

0.50

PCB-49

0.05

0.50

PCB-52

0.05

0.50

PCB-66

0.05

0.50

PCB-70

0.05

0.50

PCB-74

0.05

0.50

PCB-77

0.05

0.50

PCB-81

0.05

0.50

PCB-87/115

0.05

0.50

PCB-90/101

0.05

0.50

PCB-99

0.05

0.50

PCB-110

0.05

0.50

PCB-119

0.05

0.50

PCB-118

0.05

0.50

PCB-123

0.05

0.50

PCB-105

0.05

0.50

PCB-114

0.05

0.50

PCB-126

0.05

0.50

PCB-151

0.05

0.50

PCB-128/167

0.05

0.50

PCB-138/158

0.05

0.50

PCB-149

0.05

0.50

MK01 |O:\20123001.096\ERA_PB\ERA_APC9_PB.DOC n Q Z 7/10/2003

-------
Table C.9-2

PCB Congener/Homolog Reporting Limits [HRGC/HRMS] (Modified EPA 1668)

(SOP A-38)

(Continued)

Congener
Number/Homolog
Group

Soil/Sediment
Reporting Limits
(US/kg)

Water
Reporting Limits (ng/L)

PCB-153/168

0.05

0.50

PCB-156

0.05

0.50

PCB-157

0.05

0.50

PCB-169

0.05

0.50

PCB-170

0.05

0.50

PCB-177

0.05

0.50

PCB-180

0.05

0.50

PCB-183

0.05

0.50

PCB-184

0.05

0.50

PCB-187

0.05

0.50

PCB-189

0.05

0.50

PCB-201

0.05

0.50

PCB-202

0.05

0.50

PCB-194

0.05

0.50

PCB-195

0.05

0.50

PCB-206

0.05

0.50

PCB-207

0.05

0.50

PCB-209

0.05

0.50

Total monochlorobiphenyl

0.05

0.50

Total dichlorobiphenyl

0.05

0.50

Total trichlorobiphenyl

0.05

0.50

Total tetrachlorobiphenyl

0.05

0.50

Total pentachlorobiphenyl

0.05

0.50

Total hexachlorobiphenyl

0.05

0.50

Total heptachlorobiphenyl

0.05

0.50

Total octachlorobiphenyl

0.05

0.50

Total nonachlorobiphenyl

0.05

0.50

Total decachlorobiphenyl

0.05

0.50

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 6 7/10/03

-------
1

double-counting of congener peaks common to two or more Aroclors. The Aroclor values were
summed to provide a determination of tPCBs. Most samples were analyzed at a field laboratory
maintained by On-Site Laboratories at the Pittsfield field office. Approximately 10% of the
samples were split, and a second aliquot was also analyzed by Severn Trent Laboratories
(Burlington, VT, and Chicago, IL). Criteria were established to define if the results from the
field laboratory and fixed laboratory were considered "comparable" - when results were
considered to be comparable, additional criteria were established to determine which result was
entered into the database. When results fell outside the comparability criteria, the fixed
laboratory result was entered into the project database. These criteria and their application are
discussed in detail in the QAPP (WESTON 2001a). This project quality assurance criterion was
met.

Details of the analytical method for soils and sediments by Aroclor are provided in Appendices
A-24 and A-37 of the QAPP (WESTON 2001a). In outline, the method involved the extraction
of approximately 2 g to 30 g of raw sample with hexane:acetone or methylene chloride:acetone.
The resultant extract was injected into a gas chromatograph with electron capture detector
(GC/ECD) calibrated using a five-point calibration curve for Aroclor 1260, the most common
Aroclor identified in the study area. Individual Aroclors were quantitated using three to five
major peaks per Aroclor and an individual response factor for each peak, averaging the three to
five determinations. Method detection limits and related information for this analytical method
are presented in Table C.9-1.

C.9.4.2 PCBs as Congeners

Approximately 500 soil/sediment samples were also analyzed for PCBs as congeners by Pacific
Analytical, Inc. (Carlsbad, CA) (Pacific) using GC with low-resolution mass spectrometry
(LRMS). The analytical SOP used by Pacific followed details contained in EPA Methods 680,
1668, and 8082. A 10-g aliquot of sample was spiked with internal standards and extracted using
methylene chloride:acetone and sonication. Extracts were processed through appropriate
cleanup and concentration steps; the sample extract was analyzed via HRGC/LRMS. Five
calibration standards were used to define the calibration curve for the instrument.

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 7

-------
1 Quantitation for congeners with labeled analogs was conducted using the isotope dilution

2 method, and congeners with labeled analogs were determined using the internal standard method.

3 The procedure targeted 52 congeners, including the 12 World Health Organization (WHO)

4 congeners listed in Van den Berg et al. (1998). Values for PCB homologs and tPCBs were

5 calculated as the sum of the concentrations of the appropriate individual congeners. Complete

6 details of the soil/sediment PCB congener analytical methodology are included in Appendix

7 A-47 of the QAPP (WESTON 2001a). Method detection limits and related information for this

8 method are presented in Table C.9-3.

9 In 2001 and 2002, approximately 50 additional samples of soil and sediment were collected in

10 Reaches 5 and 6. These samples were analyzed at the Geochemical and Environmental Research

11 Group (GERG) for planar polychlorinated biphenyls (PL-PCBs) using isotope dilution high-

12 resolution gas chromatography/high-resolution mass spectrometry (HRGC/HRMS).

13 In 2001, a study of PCB partitioning in sediments and overlying water was conducted in

14 cooperation with GE. Sediment samples were collected from 50 locations representing a range

15 of PCB concentrations and physical characteristics. The samples were partitioned into solid and

16 porewater, and these two phases were analyzed for PCB congeners by Northeast Analytical, Inc.

17 (NEA) using the Green Bay Mass Balance Method under contract to GE.

18 C.9.5 Analysis of Biological Tissues for PCBs

19 All analyses of biological tissues collected by EPA were conducted by GERG at Texas A&M

20 University (College Station, TX). Some tissue samples collected by an investigator as a

21 component of a particular study were analyzed through other laboratories, as specified in the

22 study plans. Approximately 2,000 tissue samples of various types were analyzed for PCBs as

23 congeners, homologs, Aroclors, and tPCBs via GC/ECD. Sample aliquots were extracted with

24 methylene chloride using a Tissuemizer. The extracts were then concentrated and purified with

25 various chromatographic techniques, including silica gel chromatography, prior to instrumental

26 analysis.

27 Instrument calibration was conducted using four calibration standard mixtures interspersed with

28 actual samples during the analysis. Qualitative identification of target compounds was based on

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 8

-------
Table C.9-3

PCB Congener/Homolog Reporting Limits (Modified EPA 1668) (SOP A-47)

HRGC/LRMS

Congener
Number/Homolog Group

Soil/Sediment and
NAPL Reporting
Limits (ju.g/kg)

Water
Reporting Limits (ng/L)

Large Volume Water
Reporting Limits

(Mg/L)

PCB-1

0.016667

0.5

0.000042

PCB-3

0.016667

0.5

0.000042

PCB-8

0.016667

0.5

0.000042

PCB-15

0.016667

0.5

0.000042

PCB-18

0.016667

0.5

0.000042

PCB-28

0.016667

0.5

0.000042

PCB-37

0.016667

0.5

0.000042

PCB-44

0.033333

1.0

0.000083

PCB-49

0.033333

1.0

0.000083

PCB-52

0.033333

1.0

0.000083

PCB-66

0.033333

1.0

0.000083

PCB-70/74

0.033333

1.0

0.000083

PCB-77

0.033333

1.0

0.000083

PCB-81

0.033333

1.0

0.000083

PCB-87/119

0.100000

3.0

0.000250

PCB-90/101

0.100000

3.0

0.000250

PCB-99

0.100000

3.0

0.000250

PCB-110/115

0.100000

3.0

0.000250

PCB-158

0.133333

4.0

0.000333

PCB-119

0.100000

3.0

0.000250

PCB-118

0.100000

3.0

0.000250

PCB-123

0.100000

3.0

0.000250

PCB-105

0.100000

3.0

0.000250

PCB-114

0.100000

3.0

0.000250

PCB-126

0.100000

3.0

0.000250

PCB-151

0.133333

4.0

0.000333

PCB-128

0.133333

4.0

0.000333

PCB-138

0.133333

4.0

0.000333

PCB-149

0.133333

4.0

0.000333

MK01 |O:\20123001.096\ERA_PB\ERA_APC9_PB.DOC O O 7/10/2003

-------
Table C.9-3

PCB Congener/Homolog Reporting Limits (Modified EPA 1668) (SOP A-47)

HRGC/LRMS
(Continued)

Congener
Number/Homolog Group

Soil/Sediment and
NAPL Reporting
Limits (ju.g/kg)

Water
Reporting Limits (ng/L)

Large Volume Water
Reporting Limits

(Mg/L)

PCB-153/168

0.133333

4.0

0.000333

PCB-156/157

0.133333

4.0

0.000333

PCB-167

0.133333

4.0

0.000333

PCB-169

0.133333

4.0

0.000333

PCB-170

0.166667

5.0

0.000417

PCB-177

0.166667

5.0

0.000417

PCB-180

0.166667

5.0

0.000417

PCB-183

0.166667

5.0

0.000417

PCB-184

0.166667

5.0

0.000417

PCB-187

0.166667

5.0

0.000417

PCB-189

0.166667

5.0

0.000417

PCB-194

0.200000

6.0

0.000500

PCB-195

0.200000

6.0

0.000500

PCB-201

0.200000

6.0

0.000500

PCB-202

0.200000

6.0

0.000500

PCB-206

0.333333

10.0

0.000833

PCB-207

0.333333

10.0

0.000833

PCB-209

0.333333

10.0

0.000833

Total monochlorobiphenyl

0.016667

0.5

0.000042

Total dichlorobiphenyl

0.016667

0.5

0.000042

Total trichlorobiphenyl

0.016667

0.5

0.000042

Total tetrachlorobiphenyl

0.333333

1.0

0.000083

Total pentachlorobiphenyl

0.100000

3.0

0.000250

Total hexachlorobiphenyl

0.133333

4.0

0.000333

Total heptachlorobiphenyl

0.166667

5.0

0.000417

Total octachlorobiphenyl

0.200000

6.0

0.000500

Total nonachlorobiphenyl

0.333333

10.0

0.000833

Total decachlorobiphenyl

0.333333

10.0

0.000833

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 10

-------
a comparison of the retention times of the target compounds in calibration standards with the
same target compounds in the sample extract. This procedure provides for quantification of
approximately 120 PCB congeners, some of which coeluted as unresolved doublet or triplet
peaks. Appendices C and D to the QAPP (WESTON 1999) provide full details of procedures
relative to the tissue analyses, including SOPs for laboratory procedures. Method detection
limits and related information for this method are presented in Table C.9-4.

C.9.6 Other Chemistry Analyses

In addition to the PCB analyses discussed in Section C.9.2, a subset of samples was also
examined for other target analytes. Most of the PCB analyses also provide data on other
chlorinated hydrocarbons, particularly pesticides. These results were delivered along with the
PCB results, and therefore, chlorinated pesticide data are also available for the majority of the
samples analyzed.

Approximately 10% of the soil and sediment samples collected for PCB analysis were also
analyzed for a modified list of Appendix IX constituents (40 CFR 264), including semivolatile
organic compounds (SVOCs), dioxins, furans, and inorganics. The complete list of Appendix IX
constituents analyzed for is included in both the Supplemental Investigation Work Plan
(WESTON 2000) and the QAPP (WESTON 2001a). Approximately 2% of the samples were
analyzed for a modified list of Appendix IX organophosphate pesticides and herbicides. Water
samples were analyzed for a variety of conventional and hazardous constituents in addition to
PCBs; the complete list of analytes is presented in Appendix C.8. Methods and analytical
details, including method detection limits, for the procedures used in the analysis of these
additional constituents will not be presented herein, but are available in the QAPP.

C.9.7 Supplemental Analyses

In addition to the analyses for chemical constituents discussed in the preceding sections, all
sediment samples and approximately 10% of the soil (i.e., riverbank and floodplain) samples
were analyzed for total organic carbon (TOC) and grain-size distribution.

MK01 |O:\20123001.096\ERA_PB\ERA_APC9_PB.DOC

C.9-11

-------
Table C.9-4

PCB Congener/Homolog Reporting Limits (SOP-9304)
(Modified SW-846 Method 8082)

Congener Number/Homolog
Group

Tissue

Reporting Limits (ppb)**

PCB-1

0.01

PCB-7/9

0.01

PCB-8*/5

0.01

PCB-15

0.01

PCB-16/32

0.01

PCB-18*/17

0.01

PCB-22/51

0.01

PCB-24/27

0.01

PCB-25

0.01

PCB-26

0.01

PCB-28*

0.01

PCB-29*

0.01

PCB-30

0.01

PCB-31

0.01

PCB-33/20

0.01

PCB-3 9

0.01

PCB-40

0.01

PCB-41/64

0.01

PCB-42/59/37

0.01

PCB-44*

0.01

PCB-45

0.01

PCB-46

0.01

PCB-47/75

0.01

PCB-48

0.01

PCB-49

0.01

PCB-52*

0.01

PCB-53

0.01

PCB-60/56

0.01

PCB-63

0.01

MK01 |O:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 12 7/10/2003

-------
Table C.9-4

PCB Congener/Homolog Reporting Limits (SOP-9304)
(Modified SW-846 Method 8082)
(Continued)

Congener Number/Homolog
Group

Tissue

Reporting Limits (ppb)**

PCB-66*

0.01

PCB-67

0.01

PCB-69

0.01

PCB-70

0.01

PCB-72

0.01

PCB-74/61

0.01

PCB-77*

0.01

PCB-81*

0.01

PCB-82

0.01

PCB-83

0.01

PCB-84

0.01

PCB-85

0.01

PCB-87*/115

0.01

PCB-91/55

0.01

PCB-92

0.01

PCB-95/80

0.01

PCB-97

0.01

PCB-99

0.01

PCB-101*/90

0.01

PCB-105*

0.01

PCB-107

0.01

PCB-110*/77

0.01

PCB-114

0.01

PCB-118*

0.01

PCB-119

0.01

PCB-126*

0.01

PCB-128*

0.01

PCB-129

0.01

PCB-130

0.01

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 13

-------
Table C.9-4

PCB Congener/Homolog Reporting Limits (SOP-9304)
(Modified SW-846 Method 8082)
(Continued)

Congener Number/Homolog
Group

Tissue

Reporting Limits (ppb)**

PCB-135

0.01

PCB-136

0.01

PCB-138*/160

0.01

PCB-141/179

0.01

PCB-146

0.01

PCB-149/123

0.01

PCB-151

0.01

PCB-153*/132

0.01

PCB-156

0.01

PCB-158

0.01

PCB-166

0.01

PCB-167

0.01

PCB-169*

0.01

PCB-170*/190

0.01

PCB-171/202

0.01

PCB-172

0.01

PCB-174

0.01

PCB-175

0.01

PCB-176/137

0.01

PCB-177

0.01

PCB-178

0.01

PCB-180*

0.01

PCB-183

0.01

PCB-185

0.01

PCB-187*

0.01

PCB-189

0.01

PCB-191

0.01

PCB-193

0.01

PCB-194

0.01

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 14

-------
Table C.9-4

PCB Congener/Homolog Reporting Limits (SOP-9304)
(Modified SW-846 Method 8082)
(Continued)

Congener Number/Homolog
Group

Tissue

Reporting Limits (ppb)**

PCB-195*/208

0.01

PCB-197

0.01

PCB-199

0.01

PCB-200

0.01

PCB-201*/157/173

0.01

PCB-203/196

0.01

PCB-205

0.01

PCB-206*

0.01

PCB-207

0.01

PCB-209

0.01

Total monochlorobiphenyl

1.0

Total dichlorobiphenyl

1.0

Total trichlorobiphenyl

1.0

Total pentachlorobiphenyl

1.0

Total hexachlorobiphenyl

1.0

Total heptachlorobiphenyl

1.0

Total octachlorobiphenyl

1.0

Total nonachlorobipenyl

1.0

* Target analytes included in the calibration mixtures. Indicates co-
eluting congeners on a DB-5 column. The order of co-eluting
congeners is given according to their relative contribution in
common Aroclor mixtures.

**Tissue reporting limits are based on a 10-gram initial sample
weight.

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 15

-------
1 The majority of the TOC analyses were conducted at the On-Site field laboratory using the Lloyd

2 Kahn Method; the same method was followed at Severn-Trent Laboratories for the confirmation

3 samples (see Section C.9.4.1). The Lloyd Kahn Method uses pyrolysis to convert the organic

4 carbon to carbon dioxide, which is then quantified by non-dispersive infrared detection. The

5 pyrolysis is preceded by an acid digestion step to remove inorganic carbonates and bicarbonates.

6 Detection limits and related information for this method are presented in the QAPP, Appendix

7 A-16.

8 Sediment grain-size analysis was conducted at GZA GeoEnvironmental, Inc. (Newton, MA)

9 following ASTM Method D422. Particle size classes greater than 75 |j,m (No. 200) were

10 determined gravimetrically following dry sieving. Finer fractions were determined via

11 hydrometer. Analytical details for this method as applied are included in Appendix A-35 of the

12 QAPP (WESTON 2001a).

13 C.9.8 References

14 Swackhamer, D.L. 1987. "Quality Assurance Plan, Green Bay Mass Balance Study, 1. PCBs and

15 Dieldrin." U.S. EPA Great Lakes National Program Office.

16 Van den Berg, M., L. Birnbaum, A.T.C. Bosveld, B. Brunstrom, P. Cook, M. Freely, J.P. Giesy,

17 A. Hanberg, R. Hasegawa, S.W. Kennedy, T. Kubiak, J.C. Larsen, F.X.R van Leeuwen, A.K.

18 Djien Liem, C. Nolt, R.E. Peterson, L. Poellinger, S. Safe, D. Schrenk, D. Tillitt, M. Tysklind,

19 M. Younes, F. Waern, and T. Zacharewski. 1998. Toxic equivalency factors (TEFs) for PCBs,

20 PCDDs, PCDFs for humans and wildlife. Environmental Health Perspectives 106(12):775-792.

21 WESTON (Roy F. Weston, Inc.). 1999. Quality Assurance Project Plan, Volume III, Appendix

22 C (27 October 1999, GEP2-060499-AAIY) and Appendix D (7 January 1999, DCN GEP2-

23 123098-AAET).

24 WESTON (Roy F. Weston, Inc.). 2000. Supplemental Investigation Work Plan for the Lower

25 Housatonic River. Prepared for U.S. Army Corps of Engineers and U.S. Environmental

26 Protection Agency. 22 February 2000. DCN GEP2-020900-AAME.

27 WESTON (Roy F. Weston, Inc.). 2001a. Quality Assurance Project Plan, Vol. I - Text, Vol. II -

28 Appendix A, Vol. IIA - Appendix A, cont'd., Vol. IV - Appendices E and F. Prepared for U.S.

29 Army Corps of Engineers and U.S. Environmental Protection Agency. DCN GE-021601-

30 AAHM.

31 WESTON (Roy F. Weston, Inc.). 2001b. Final Field Sampling Plan. Prepared for U.S. Army

32 Corps of Engineers and U.S. Environmental Protection Agency. DCN GE-053001-AAMA.

MK010:\20123001.096\ERA_PB\ERA_APC9_PB.DOC C 9 16

-------
APPENDIX C.10

APPROACH FOR CALCULATING
TOXIC EQUIVALENCE (TEQ)

MK01 |O:\20123001,096\ERA_PB\ERA_APC10_PB.DOC

-------
1

APPENDIX C.10
APPROACH FOR CALCULATING TOXIC EQUIVALENCE (TEQ)

C.10.1 Introduction

A number of methods for calculating toxic equivalence (TEQ) for polychlorinated hydrocarbons
(PCHs) have been proposed (e.g., Van den Berg et al. 1998; Safe 1994; Kennedy et al. 1996).
For this ERA, the toxic equivalency factors (TEFs) developed for the World Health Organization
(WHO) by Van den Berg et al. (1998) have been adopted. The WHO TEFs are the most recent
estimate of 2,3,7,8-TCDD equivalency and are based on current scientific research (Dyke and
Stratford 2002). They have been accepted and applied in numerous jurisdictions worldwide
(Dyke and Stratford 2002). These TEF values were developed for compounds that show a
structural relationship to PCDDs and PCDFs, bind to the aryl hydrocarbon (Ah) receptor, elicit
an Ah-receptor-mediated biochemical and toxic response, and are persistent and accumulate in
the food chain. The WHO TEFs for deriving TEQ for mammals, fish, and birds as predators are
listed in Table C.10-1.

Two issues that have to be addressed in calculating a TEQ value are congener concentrations
reported below the method detection limit (MDL) (i.e., non-detects), and congeners that cannot
be resolved due to co-elution during analysis. Each of these issues, and how it was addressed in
this ERA, is discussed in the following sections.

C.10.2 Congener Non-Detects

Congeners detected at or below the MDL were included in the TEQ calculations by investigating
three options: first, setting the value for the congener equal to zero (0), then setting it to half the
MDL, and, finally, setting it equal to the MDL, as described in Appendix C.2. A comparison of
the three separate results of this bounding analysis provides a description of the uncertainty
surrounding the TEQ value due to one or more congeners being detected at or below the sample
detection limit. This approach is also useful for determining the relative influence of individual
non-detected congeners on the estimated TEQ value. For example, use of the

MK01 |O:\20123001,096\ERA_PB\ERA_APC10_PB.DOC

C.10-1

-------
Table C.10-1

Toxic Equivalency Factors from Van den Berg et al. 1998 for Mammals, Fish, and

Birds as Predators

No.

Congener

Mammals

Fish

Birds

TEF

PCB-77

0.0001

0.05

PCB-81

0.0001

0.0005

0.1

PCB-126

0.1

0.005

0.1

PCB-169

0.01

0.00005

0.001

PCB-105

0.0001

<0.000005*

0.0001

PCB-114

0.0005

<0.000005*

0.0001

PCB-118

0.0001

<0.000005*

0.00001

PCB-123

0.0001

<0.000005*

0.00001

PCB-156

0.0005

<0.000005*

0.0001

PCB-157

0.0005

<0.000005*

0.0001

PCB-167

0.00001

<0.000005*

0.00001

PCB-189

0.0001

<0.000005*

0.00001

1,2,3,4,6,7,8-HpCDD

0.01

0.001

<0.001*

1,2,3,4,6,7,8-HpCDF

0.01

1,2,3,4,7,8,9-HpCDF

0.01

1,2,3,4,7,8-HxCDD

0.1

0.5

0.05

1,2,3,4,7,8-HxCDF

0.1

1,2,3,6,7,8-HxCDD

0.1

0.01

1,2,3,6,7,8-HxCDF

0.1

1,2,3,7,8,9-HxCDD

0.1

0.01

0.1

1,2,3,7,8,9-HxCDF

0.1

1,2,3,7,8-PeCDD

1,2,3,7,8-PeCDF

0.05

0.1

2,3,4,6,7,8-HxCDF

0.1

2,3,4,7,8-PeCDF

0.5

2,3,7,8-TCDF

0.1

0.05

2,3,7,8-TCDD

OCDD

0.0001

<0.0001*

0.0001

OCDF

0.0001

<0.001*

0.0001

* Values that are less than (<) should be considered to be the upper limit in TEQ calculations.

MK01|O:\20123001.096\ERA_PB\ERA_APC10_PB.DOC C 10 2 7/10/2003

-------
1

MDL for non-detected congeners generates a conservative TEQ value that may result in over-
estimates in any subsequent risk evaluations.

Table C.10-2 shows that several congeners (i.e., 1,2,3,4,7,8-HxCDD, 1,2,3,7,8-PeCDD,
2,3,4,7,8-PeCDF) reported at or below the MDL potentially contribute substantially to the TEQ
value; and the difference in the TEQ value, depending upon the treatment of the non-detects, is
significant. This type of uncertainty and other uncertainties identified using the bounding
analysis are further discussed in the appendices when TEQ calculations are performed.

C.10.3 Congener Co-Elution

Concentrations for all 29 congeners (PCBs plus dioxin/furans) in Table C.10-1 must be available
as separate values in order to derive a complete TEQ value using the Van den Berg et al. (1998)
approach. This is not always the case because most, but not all, samples were analyzed for both
PCB and dioxin and furan congeners.

Additionally, even when all congeners are analyzed for and reported above detection limits,
some congeners co-elute in the analysis and, therefore, are not quantified separately. In the case
of the congeners required for TEQ determination, the analytical protocol used for the tissue
analysis resulted in PCB-157 and PCB-123 being reported as part of a triplet (PCB-201/157/173)
and doublet (PCB-149/123), respectively. Assuming that these co-eluting peaks consist entirely
of the TEQ congener could lead to an over-estimation of the TEQ concentration, and therefore,
an alternate method of dealing with co-eluting congeners was explored. A description of the
approach used to generate the TEQ when PCB-201/157/173 and/or PCB-149/123 occur in a
tissue sample is presented for each tissue type below.

C.10.3.1 Fish Tissue

The PCB-149/123 doublet and the PCB-201/157/173 triplet were reported in fish tissue samples
analyzed by GERG throughout the PSA. Data on PCB congener concentrations in largemouth
bass from the PSA in which these individual congeners were resolved were available from
samples analyzed by USGS as part of the mink feeding study. From these data, the relative
proportion of each of the congeners that make up the doublet (PCB-149/123) and triplet (PCB-
201/157/173) was determined using Equation C.10-1 and Equation C.10-2.

MK01|O:\20123001.096\ERA_PB\ERA_APC10_PB.DOC p ia ^ 7/10/2003

-------
Table C.10-2

Example Calculation Showing the Potential for Over-Estimation of the TEQ Value from Assuming that Non-
Detected Values are Equal to Zero, Half the Sample Detection Limit, and the Full Sample Detection Limit

Congener

Concentration

(mg/kg)

Result Flag

TEF

cTEF (SDL = 0)

cTEF (ND = 1/2 SDL)

cTEF (ND = SDL)

1,2,3,4,6,7,8-HpCDD

0.0000342

0.001

1.71E-08

3.42E-08

1,2,3,4,6,7,8-HpCDF

0.0000308

0.01

0.000000154

0.000000308

1,2,3,4,7,8,9-HpCDF

0.0000342

0.01

0.000000171

0.000000342

1,2,3,4,7,8-HxCDD

0.0000342

0.5

0.00000855

0.0000171

1,2,3,4,7,8-HxCDF

0.0000342

0.1

0.00000171

0.00000342

1,2,3,6,7,8-HxCDD

0.0000342

0.01

0.000000171

0.000000342

1,2,3,6,7,8-HxCDF

0.0000342

0.1

0.00000171

0.00000342

1,2,3,7,8,9-HxCDD

0.0000342

0.01

0.000000171

0.000000342

1,2,3,7,8,9-HxCDF

0.0000342

0.1

0.00000171

0.00000342

1,2,3,7,8-PeCDD

0.0000342

0.0000171

0.0000342

1,2,3,7,8-PeCDF

0.0000342

0.05

0.000000855

0.00000171

2,3,4,6,7,8-HxCDF

0.0000342

0.1

0.00000171

0.00000342

2,3,4,7,8-PeCDF

0.0000342

0.5

0.00000855

0.0000171

2,3,7,8-TCDD

0.0000068

0.0000034

0.0000068

2,3,7,8-TCDF

0.0000074

0.05

0.000000185

0.00000037

OCDD

0.0000852

0.0001

8.52E-09

OCDF

0.0000685

0.0001

3.425E-09

6.85E-09

PCB-105

0.005643

0.000005

2.8215E-08

PCB-114

0.0000613

0.000005

1.5325E-10

3.065E-10

PCB-118

0.032745

0.000005

1.63725E-07

MK01|O:\20123001.096\ERA_PB\ERA_APC10_PB.DOC C* "\ C\ A 7/10/2003

-------
Table C.10-2

Example Calculation Showing the Potential for Over-Estimation of the TEQ Value
from Assuming that Non-Detected Values are Equal to Zero, Half the Sample
Detection Limit, and the Full Sample Detection Limit

(Continued)

Congener

Concentration

(mg/kg)

Result Flag

TEF

cTEF (SDL = 0)

cTEF (ND = 1/2 SDL)

cTEF (ND = SDL)

PCB-123

0.162041

0.000005

8.10205E-07

PCB-126

0.000558

0.005

0.00000279

PCB-156

0.014904

0.000005

7.452E-08

PCB-157

0.006829

0.000005

3.4145E-08

PCB-167

0.016693

0.000005

8.3465E-08

PCB-169

0.0000613

0.00005

1.5325E-09

3.065E-09

PCB-189

0.004427

0.000005

2.2135E-08

PCB-77

0.000229

0.0001

2.29E-08

PCB-81

0.0000613

0.0005

1.5325E-08

3.065E-08

TEQ

4.038E-06

5.02224E-05

9.6407E-05

TEFs are from Van den Berg et al. 1998, Mammalian TEFs.
Note: Result Flags: U = non-detect; J = estimated value.
SDL = Sample Detection Limit.

MK01 |O:\20123001.096\ERA_PB\ERA_APC10_PB.DOC P1 1 fi ^ 7/10/2003

-------
1

PCB- 123ff

.PC.Z? - 123r * (PC-g-123/149g)
PC5 - 123r +PCB-\49t

(Eq. 1)

PCB-157H =

PCB - 157r *(PCB — 201/157/173g)

(Eq. 2)

PCB - 157r +PCB - 20lr +PCB - 173r

where:

PCB-\23h = Concentration of PCB-123 in a Housatonic River fish sample
PCB-\51h = Concentration of PCB-157 in a Housatonic River fish sample
PCB-\23T = Concentration of PCB-123 in the largemouth bass analyzed by USGS
PCB-157T = Concentration of PCB-157 in the largemouth bass analyzed by USGS
PCB-\2?>I\A9H = Concentration of PCB-123/149 doublet in fish tissue data
/>C5-201/157/173ff = Concentration of PCB-201/157/173 in fish tissue data.

This analysis showed that PCB-123 was, on average, 0.3% of PCB-149/123 and that PCB-157
was 19.5% of the triplet PCB-201/157/173. These proportions were then applied to the fish
tissue data in the Housatonic River database for calculation of a TEQ. Table C.10-3 is an
example of the application of this approach for one fish sample.

In addition to the inherent variability in data generated by different laboratories, there are a
number of additional sources of uncertainty in this approach to treating the co-eluted congeners.
The data used to develop the congener ratios in the co-eluted groups were developed using only
largemouth bass and could vary for other fish species, or due to differences in spatial or temporal
factors associated with sample collection.

C.10.3.2 Mammals, Birds, Amphibians, Invertebrates, and Vegetation

There are no data currently available for mammals, birds, amphibians, invertebrates, and
vegetation that can be used to derive ratios for the congeners of interest versus the co-eluters.
The applicability of the ratios developed using largemouth bass samples to these other taxonomic
groups is not known. A factor that must be addressed, particularly for mammals and birds, is the

MK01|O:\20123001.096\ERA_PB\ERA_APC10_PB.DOC p ia / 7/10/2003

-------
Table C.10-3

TEQ Congener Data Isolated for One Unique Fish Sample (Field Sample ID# H3-TW11GF03-0-8c20) and TEQ (Mammalian)

Calculation Results

Caption

Result

Units

Result Flag

TEF
(Mammalian)

Co-Elution
Ratio

cTEF (non-detect = 0)

cTEF (non-detect =
1/2 SDL)

cTEF (non-detect = SDL)

1,2,3,4,6,7,8-HpCDD

0.0000028

mg/kg

0.001

2.8E-09

1,2,3,4,6,7,8-HpCDF

0.0006509

mg/kg

0.01

0.000006509

1,2,3,4,7,8,9-HpCDF

3.13E-06

mg/kg

0.01

1.565E-08

3.13E-08

1,2,3,4,7,8-HxCDD

3.13E-06

mg/kg

0.5

7.825E-07

0.000001565

1,2,3,4,7,8-HxCDF

3.13E-06

mg/kg

0.1

1.565E-07

0.000000313

1,2,3,6,7,8-HxCDD

3.13E-06

mg/kg

0.01

1.565E-08

3.13E-08

1,2,3,6,7,8-HxCDF

3.13E-06

mg/kg

0.1

1.565E-07

0.000000313

1,2,3,7,8,9-HxCDD

3.13E-06

mg/kg

0.01

1.565E-08

3.13E-08

1,2,3,7,8,9-HxCDF

3.13E-06

mg/kg

0.1

1.565E-07

0.000000313

1,2,3,7,8-PeCDD

3.13E-06

mg/kg

0.000001565

0.00000313

1,2,3,7,8-PeCDF

0.0007208

mg/kg

0.05

0.00003604

2,3,4,6,7,8-HxCDF

3.13E-06

mg/kg

0.1

1.565E-07

0.000000313

2,3,4,7,8-PeCDF

0.0000414

mg/kg

0.5

0.0000207

2,3,7,8-TCDD

0.000002

mg/kg

0.000002

2,3,7,8-TCDF

0.0000126

mg/kg

0.05

0.00000063

OCDD

0.0000109

mg/kg

0.0001

1.09E-09

OCDF

6.27E-06

mg/kg

0.0001

3.135E-10

6.27E-10

PCB-105

0.463743

mg/kg

0.000005

2.31872E-06

PCB-114

0.0002

mg/kg

0.000005

5E-10

0.000000001

PCB-118

1.728937

mg/kg

0.000005

8.64469E-06

PCB-126

0.003347

mg/kg

0.005

0.000016735

MK01|O:\20123001.096\ERA_PB\ERA_APC10_PB.DOC C* '\ (\ 1 7/10/2003

-------
Table C.10-3

TEQ Congener Data Isolated for One Unique Fish Sample (Field Sample ID# H3-TW11GF03-0-8c20) and TEQ (Mammalian)

Calculation Results
(Continued)

Caption

Result

Units

Result Flag

TEF
(Mammalian)

Co-Elution
Ratio

cTEF (non-detect = 0)

cTEF (non-detect =
1/2 SDL)

cTEF (non-detect = SDL)

PCB-149/123

8.269354

mg/kg

0.000005

0.003

1.2404E-07

PCB-156

0.422

mg/kg

0.000005

0.00000211

PCB-167

0.212862

mg/kg

0.000005

1.06431E-06

PCB-169

0.001258

mg/kg

0.00005

6.29E-08

PCB-189

0.029064

mg/kg

0.000005

1.4532E-07

PCB-201/157/173

0.329227

mg/kg

0.000005

0.195

3.20996E-07

PCB-77

0.005447

mg/kg

0.0001

5.447E-07

PCB-81

0.000862

mg/kg

0.0005

0.000000431

TEQ

9.84E-05

0.000101

0.000104

Note: Result Flags: U = undetected; J = estimated; cTEF = calculated TEF value.
SDL = Sample Detection Limit.

MK01 |O:\20123001.096\ERA_PB\ERA_APC10_PB.DOC C 10 8 7/10/2003

-------
1 rate at which different species metabolize the congeners. Boon et al. (1997) demonstrated that

2 for different fish-eating mammals (e.g., otter, dolphin, seals), there were substantial differences

3 in ability to metabolize PCB congeners. These results indicate that species-specific metabolism

4 can be an important factor even for closely related species.

5 A bounding analysis was performed to account for the co-elution of the two congeners in non-

6 fish tissue. The lower bound was determined by assuming that the contribution of PCB-123 and

7 PCB-157 to the doublet and triplet, respectively, was zero. The upper bound was determined by

8 assuming that the full value of the doublet and triplet corresponded to PCB-123 and PCB-157,

9 respectively. The bounding analysis allows a description of the uncertainty introduced by

10 including the doublet and triplet in the TEQ calculation.

11 C.10.4 References

12 Boon, J.P., J. Van der Meer, C.R. Allchin, R.J. Law, J. Klungsoyr, P.E.G. Leonards, H. Spliid,

13 E. Storr-Hansen, C. McKenzie, and D.E Wells. 1997. Concentration-dependent changes of PCB

14 patterns in fish-eating mammals: Structural evidence for induction of cytochrome P450. Arch.

15 Environ. Contam. Toxicol. 33:298-311.

16 Dyke, P.H. and J. Stratford. 2002. Changes to the TEF schemes can have significant impacts on

17 regulation and management of PCDD/F and PCB. Chemosphere 47:103-116.

18 Kennedy, S.W., A. Lorenzen, and R.J. Norstrom. 1996. Chicken embryo hepatocyte bioassay for

19 measuring cytochrome P4501A-based 2,3,7,8-tetrachlorodibenzo-p-dioxin equivalent

20 concentrations in environmental samples. Environmental Science and Technology 30:706-715.

21 Safe, S.H. 1994. Polychlorinated biphenyls (PCBs): Environmental impact, biochemical and

22 toxic responses, and implications for risk assessment. Critical Reviews in Toxicology

23 24(2):87-149.

24 Van den Berg, M., L. Birnbaum, A.T.C. Bosveld, B. Brunstrom, P. Cook, M. Freely, J.P. Giesy,

25 A. Hanberg, R. Hasegawa, S.W. Kennedy, T. Kubiak, J.C. Larsen, F.X.R van Leeuwen,

26 A.K. Djien Liem, C. Nolt, R.E. Peterson, L. Poellinger, S. Safe, D. Schrenk, D. Tillitt,

27 M. Tysklind, M. Younes, F. Waern, and T. Zacharewski. 1998. Toxic equivalency factors (TEFs)

28 for PCBs, PCDDs, PCDFs for humans and wildlife. Environmental Health Perspectives

29 106(12):775-792.

MK01 |O:\20123001,096\ERA_PB\ERA_APC10_PB.DOC

C.10-9

-------
APPENDIX C.11
SUMMARY OF ANALYTICAL VARIABILITY

MK01 |O:\20123001,096\ERA_PB\ERA_APC11_PB.DOC

-------
1

SUMMARY OF ANALYTICAL VARIABILITY

APPENDIX C.11

5 C.11.1 Introduction

6 Measurements of contaminant concentrations in environmental media are recognized as being

7 subject to variability, or "error" of varying degrees of magnitude. The sources of such variability

8 are numerous and include factors such as inherent small-scale spatial or temporal variability in

9 the parameter and/or medium of interest; variability introduced by sampling, sample storage, or

10 sample shipment; variability due to analysis by different instruments and/or laboratories; and

11 inherent variability and stochasticity in analytical methods. This analytical variability is an

12 additional source of uncertainty that affects any end use of the data. The magnitude and sources

13 of analytical variability in the tPCB data developed for the GE/Housatonic River Project were

14 investigated via a comparison of quality control (QC) sample results; this document presents a

15 summary of the findings.

16 C.11.2 Quality Control Sample Types

17 Three separate types of QC samples were available for determination of variability: duplicate

18 samples, comparability samples, and split samples. Each sample type provided an integrated

19 measurement of several different sources of variability, as follows:

20 ¦ Duplicate Samples: Duplicate samples were two (occasionally more) aliquots of the

21 same environmental sample, collected in a manner to minimize, to the extent

22 possible, any sources of variability prior to collection of the aliquots (e.g., small-scale

23 spatial variability, sampling methods, etc.). In this study, duplicate samples were

24 collected from the same sample matrix following homogenization of the sample as

25 collected. Duplicate samples were intended to quantify a base level of variability due

26 to inherent stochasticity in the analytical methodology, but also to integrate any

27 variability introduced by incomplete homogenization of the sample matrix and

28 differential handling (if any) in transport and/or in the laboratory. Duplicate samples

29 are typically included in QC programs.

30 ¦ Comparability Samples: Comparability samples were, in effect, duplicate samples

31 that were analyzed by two different types of laboratories and were a special

MK01 |O:\20123001,096\ERA_PB\ERA_APC11_PB.DOC

C.ll-1

7/11/2003

-------
1

component of the GE/Housatonic River Project. To allow for the analysis of as many
samples as possible from the study area within the available budget, most analyses
were conducted in a field laboratory. Comparability samples were collected
following the same procedures as described for duplicate samples, but then one
aliquot was shipped to a standard fixed laboratory for analysis using the same
methodology followed in the field laboratory. Comparability samples were collected
for the soil/sediment matrix only because this was the only matrix analyzed by the
field laboratory. This procedure provided an ongoing QC check on the field
laboratory and also provided a means of evaluating potential variability added by the
use of a different laboratory and different transportation and storage.

¦ Split Samples: As is typical in projects involving regulatory agencies and potentially
responsible parties (PRPs), aliquots of some samples were provided to GE for their
use in confirming the analytical results obtained by the EPA laboratories. Split
samples were collected in the same manner as duplicates, with one aliquot provided
to GE for their analysis. In the case of the GE/Housatonic River Project, split
samples integrate all of the sources of variability described above for comparability
samples and, in addition, introduce variability derived from a different analytical
methodology.

C.11.3 Results

The results of the analysis of sources and magnitude of variability are discussed below by sample
type and media. Throughout this discussion, the magnitude of variability is described in terms of
relative percent difference (RPD), which is the difference between two measurements divided by
the average of the two measurements, expressed as a percentage.

C.11.3.1 Duplicate Samples

Mean RPD for duplicate samples is presented by medium in Table C. 11-1. The RPDs range
from approximately 29% for biota samples to approximately 60% for surface water samples.
Because homogenization issues are presumably not a factor for water samples, the higher
variability in this matrix may be due to many of the samples being close to the detection limit for
PCBs.

C. 11.3.2 Comparability Samples

Mean RPD for comparability samples is presented in Table C.ll-2. As noted above,
comparability samples were only available for the soil/sediment matrix. To further investigate

MK01 |O:\20123001,096\ERA_PB\ERA_APC11_PB.DOC

C.ll-2

-------
1

potential differences between the field and fixed laboratory results, the comparability samples
were also separated and examined by concentration range and geomorphological terrain type to
determine if any systematic differences due to these two factors were evident. In addition, a chi-
square goodness-of-fit test was used to test for a consistent bias between the two laboratories.
The null hypothesis tested was that the higher of the two paired results should occur equally
between the two laboratories.

The mean RPD for soil/sediment comparability samples was approximately 60%, which is nearly
double the RPD for duplicate samples analyzed by the same laboratory. This additional variance
component likely reflects the additional variability introduced by the second laboratory. In
addition, a statistically significant bias was found, with the fixed laboratory (STL) generally
reporting higher concentrations than the field laboratory (OS). This bias was persistent across
nearly all concentrations and across all geomorphological terrain types.

C. 11.3.3 Split Samples

Mean RPDs for samples split between EPA and GE are presented in Tables C.ll-3 through
C.ll-5 for surface water, soil/sediment, and fish tissue, respectively. RPDs for surface water
samples increased to 87% (vs. approximately 35% for the duplicate samples). This added
variability reflects the integrated effect of different laboratories and different methods. RPDs for
soil/sediment increased only slightly over the comparability samples, indicating that the added
variance component due to a change in analytical method is minor overall, although greater
differences are evident in some of the concentration and geomorphological groupings. As was
the case for duplicate samples, the fish tissue matrix again had the lowest RPD, at approximately
42%.

An additional component of the evaluation of split sample results was to test for consistent bias
using a procedure identical to that followed for the comparability samples, but in this case
comparing the GE split sample results to the EPA results. For the water and tissue matrices there
was no statistical indication of a bias. However, for the soil/sediment matrix, there was a highly
significant difference overall that persisted through most concentration and geomorphological
groupings. GE results were consistently higher than those reported by EPA. The cause of this
observed difference was not determined.

MK01|O:\20123001.096\ERA_PB\ERA_APC11_PB.DOC n 7/11/2003

-------
Table C.11-1
Variability in Duplicate Samples, All Matrices

Medium

Total Samples

Total Detects

Mean RPD

Surface Water

60.05%

Soils and Sediments

611

428

34.71%

Biota Tissue

28.67%

MK01|0:\20123001.096\ERA_PB\ERA_APC11TBLS_PB.DOC C* \ \ A 7/10/2003

-------
Table C.11-2

Variability in Comparability Samples, Soil/Sediment Matrix

Sample Grouping

Sample Size

Percent
STL>OS

Percent
OS>STL

Mean
RPD

Chi-square

All data with both results "detect"

779

63.03

36.84

59.96%

53.49

<0.0001

Grouped by Concentration:

STL result < 7 ppm

255

49.41

50.59

65.10%

0.035

STL result > 7 ppm but <25 ppm

248

67.74

31.85

49.79%

32.07

<0.0001

STL result > 25 ppm

276

71.38

28.62

64.36%

50.45

<0.0001

Grouped by Terrain Type:

Bars and terraces

140

60.71

39.29

55.87%

6.43

0.01

Floodplain

211

59.24

40.76

49.61%

7.21

0.007

Main channel

208

64.42

35.58

75.38%

17.31

<0.0001

Banks

128

68.75

30.47

48.88%

18.91

<0.0001

SCOX

71.70

28.3

73.88%

9.98

0.002

MK01|O:\20123001.096\ERA_PB\ERA_APC11TBLS_PB.DOC 1 1 ^ 7/10/2003

-------
Table C.11-3
Variability in Split Samples, Water Matrix

GE>EPA
(%)

EPA>GE
(%)

Mean RPD

Chi-square

Surface Water Splits

190

Surface Water Splits - Detects

54.67

44.00

87.02%

0.865

MK01|0:\20123001.096\ERA_PB\ERA_APC11TBLS_PB.DOC p i i / 7/10/2003

-------
Table C.11-4
Variability in Split Samples, Soil/Sediment Matrix

Sample Grouping

GE>EPA
(%)

EPA>GE
(%)

Mean
RPD

Chi-square

Total N of Soil and Sediment Splits

805

Both results >DL

562

71.53

28.47

64.29%

312.30

<0.0001

Grouped by Terrain Type:

Bars and terraces

64.15

35.85

53.98%

106.98

<0.0001

Main channel

182

68.68

31.32

77.36%

10.95

<0.001

Floodplains

73.97

26.03

39.60%

76.93

<0.0001

Pond sediment

80.77

19.23

67.44%

3.53

0.06

Riverbanks

139

73.38

26.62

65.10%

0.43

Grouped by River Reach:

Reach 4

106

74.53

25.47

85.81%

17.49

<0.0001

Reach 5

179

75.98

24.02

54.90%

4.08

<0.043

Reach 6

82.05

17.95

71.13%

401.28

<0.0001

MK01|O:\20123001.096\ERA_PB\ERA_APC11TBLS_PB.DOC n 1 7/10/2003

-------
Table C.11-5
Variability in Split Samples, Tissue Matrix

GE>EPA
(%)

EPA>GE
(%)

Mean RPD

Chi-square

Fish Tissue

41.48%

0.889

MK01|0:\20123001.096\ERA_PB\ERA_APC11TBLS_PB.DOC

C.ll-8

-------
APPENDIX D

ASSESSMENT ENDPOINT—COMMUNITY STRUCTURE, SURVIVAL,
GROWTH, AND REPRODUCTION OF BENTHIC INVERTEBRATES

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

-------
1

TABLE OF CONTENTS

2 Section Page

3 D. ASSESSMENT ENDPOINT—COMMUNITY STRUCTURE, SURVIVAL,

4 GROWTH, AND REPRODUCTION OF BENTHIC INVERTEBRATES D-l

5 D. 1 INTRODUCTION D-l

6 D.l.l Overview of Approach for the Benthic Risk Assessment D-2

7 D. 1.1.1 Conceptual Model D-2

8 D. 1.1.2 Exposure Assessment D-4

9 D. 1.1.3 Effects Assessment D-4

10 D.l.l.4 Risk Characterization D-5

11 D. 1.2 Organization D-6

12 D.2 EXPOSURE ASSESSMENT D-7

13 D.2.1 Introduction D-7

14 D.2.1.1 Selection of COCs for Benthic Invertebrates D-7

15 D.2.1.2 Site Identifiers D-9

16 D.2.1.3 Types of Exposure Data D-10

17 D.2.2 Habitat Characterization D-12

18 D.2.2.1 Physical Substrate D-14

19 D.2.2.2 Benthic Community Types D-14

20 D.2.3 Statistical Approach and Methods D-15

21 D.2.4 Assessment of Sediment Chemistry D-16

22 D.2.4.1 Overview D-16

23 D.2.4.2 Sources of Sediment Data D-16

24 D.2.4.3 Distributions and Concentrations of Physical Parameters D-20

25 D.2.4.4 Distribution and Concentrations of PCBs D-21

26 D.2.4.5 Distribution and Concentration of Other COCs D-26

27 D.2.4.6 Discussion of Results D-27

28 D.2.5 Tissue Chemistry Assessment D-29

29 D.2.5.1 Overview D-29

30 D.2.5.2 Tissue Data Sources D-29

31 D.2.5.3 Distribution and Concentrations of PCBs D-31

32 D.2.5.4 Distribution and Concentration of Other COCs D-33

33 D.2.6 Surface Water Chemistry Assessment D-34

34 D.2.6.1 Overview D-34

35 D.2.6.2 Data Sources D-34

36 D.2.6.3 Distribution and Concentrations of COCs D-35

37 D.3 EFFECTS ASSESSMENT I)-37

38 D.3.1 Sediment Toxicity D-37

39 D.3.1.1 Methods I)-3 8

40 D.3.1.2 Relevance of Selected Endpoints D-39

41 D.3.1.3 Particle Size Sensitivities of Test Organisms D-40

42 D.3.1.4 Data Evaluation Approach D-40

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC • • • 7/10/2003

-------
1

TABLE OF CONTENTS
(Continued)

D.3.1.5 Results D-41

D.3.1.6 Interpretation of Test Results D-45

D.3.1.7 Conclusions D-46

D.3.2 Concentration-Response Analysis -Toxicity Endpoints D-47

D.3.2.1 Approach 1: Calculation of Individual Toxicity Test

Endpoints D-48

D.3.2.2 Approach 2: General Linear Model of Concentration-

Response D-54

D.3.2.3 Relationships of Effects with Other COCs D-57

D.3.3 Toxicity Identification Evaluations D-59

D.3.4 Tissue Effects Thresholds D-60

D.3.4.1 Review of Literature D-60

D.3.4.2 Estimation of a Tissue Effects Threshold D-61

D.3.5 Sediment Quality Values D-61

D.3.5.1 Application of SQVs D-62

D.3.5.2 SQVs Selected for Application to the Housatonic River D-62

D.3.6 Water Quality Benchmarks D-64

D.3.7 Benthic Macroinvertebrate Community Evaluation D-64

D.3.7.1 Identification of Benthic Community Metrics D-65

D.3.7.2 Average Rank Plots D-66

D.3.7.3 Multi-Dimensional Scaling (All Stations) D-67

D.3.7.4 Multi-Dimensional Scaling (By Habitat Type) D-69

D.3.7.5 Analysis of Variance (ANOVA) for Summary Metrics D-69

D.3.7.6 Modified Hilsenhoff Biotic Index D-71

D.3.7.7 Analysis of Invertebrate Biomass D-73

D.3.7.8 Potential Role of Habitat in Benthic Community

Observations D-74

D.3.7.9 Conclusions D-76

D.3.8 Concentration-Response Analysis - Benthic Community Assemblages.... D-76

D.3.8.1 Total PCBs D-76

D.3.8.2 Other COCs D-78

D.4 RISK CHARACTERIZATION D-79

D.4.1 Field Surveys D-79

D.4.2 Comparison of Field-Measured Exposures to Effects Levels or

Benchmarks D-80

D.4.2.1 Sediment Chemistry D-82

D.4.2.2 Water Chemistry D-85

D.4.2.3 Tissue Chemistry D-86

D.4.3 Site-Specific Toxicity Study Results D-87

D.4.3.1 Sediment Toxicity Tests D-87

D.4.3.2 Toxicity Identification Evaluation D-88

D.4.4 Integrated Station-by-Station Triad Assessment D-90

D.4.4.1 Sediment Toxicity Endpoints D-90

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC • 7/10/2003

-------
TABLE OF CONTENTS
(Continued)

1 D.4.4.2 Benthic Community Endpoints D-91

2 D.4.4.3 PCB Chemistry Endpoints D-91

3 D.4.4.4 Summary D-93

4 D.4.5 Weight-of-Evidence Procedure for Assessing Risk from PCBs in the

5 Housatonic River PSA D-93

6 D.4.5.1 Weighting of Measurement Endpoints D-94

7 D.4.5.2 Magnitude of Responses in Measurement Endpoints D-99

8 D.4.5.3 Concurrence Among Measurement Endpoints D-100

9 D.4.5.4 Summary D-100

10 D.4.6 Sources of Uncertainty D-101

11 D.4.6.1 Exposure Assessment D-101

12 D.4.6.2 Effects Assessment D-102

13 D.4.6.3 Risk Characterization D-104

14 D.4.7 Extrapolation to Other Species D-105

15 D.4.8 Downstream Assessment D-105

16 D.4.9 Conclusions D-107

17 D.5 REFERENCES I)-109

19 TABLES

20 FIGURES

22 LIST OF ATTACHMENTS

24 ATTACHMENT D.l

26 ATTACHMENT D.2

28 ATTACHMENT D.3

30 ATTACHMENT D.4

32 ATTACHMENT D.5

34 ATTACHMENT D.6

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC 7/10/2003

DESCRIPTIONS OF STUDIES CONSIDERED IN
DETERMINATION OF PCB TISSUE BENCHMARKS

RATIONALE USED IN DEVELOPMENT OF SEDIMENT
QUALITY VALUES FOR PCBs

DISCUSSION OF METRICS ADOPTED FOR THE ASSESSMENT
OF BENTHIC COMMUNITY ASSEMBLAGES

GRAIN SIZE SENSITIVITY OF HOUSATONIC RIVER TOXICITY
TEST ORGANISMS

ALTERNATIVE CONCENTRATION-RESPONSE ASSESSMENT
FOR PCB AND TOXICITY TEST ENDPOINTS

RAW DATA FOR MULTIVARIATE ASSESSMENT OF BENTHIC
COMMUNITY METRICS

-------
1

LIST OF ACRONYMS

ANOVA

analysis of variance

BEAST

benthic assessment of sediment

BSAF

biota-sediment accumulation factor

C/C

coarse-grained contaminated location

C/R

coarse-grained reference location

coc

contaminants of concern

dry weight

effect concentration

EDTA

ethylenediaminetetra-acetic acid

EPA

U.S. Environmental Protection Agency

EPT

Mayflies (Ephemeroptera), Stoneflies (Plecoptera), and Caddisflies (Tricoptera)

ERA

ecological risk assessment

F/C

fine-grained contaminated location

F/R

fine-grained reference location

General Electric Company

HBI

Hilsenhoff Biotic Index

hazard quotient

inhibition concentration

LC20

lethal concentration causing 20% mortality

LC50

lethal concentration causing 50% mortality

LOAEL

lowest observed adverse effect level

MATC

maximum acceptable threshold concentration

MDS

multidimensional scaling

MEC

mid-range effect concentration

MHBI

Modified Hilsenhoff Biotic Index

NOEC

no observed effect concentration

NOAEL

no observed adverse effect level

OMEE

Ontario Ministry of the Environment and Energy

PAH

polycyclic aromatic hydrocarbon

PCA

principal components analysis

PCB

polychlorinated biphenyl

PSA

Primary Study Area

RIVM

Rijksinstituut voor Volksgezondheid en Milieu

RPP

relative performance proportion

SCOX

side channel/oxbow

MK01|O:\20123001.096\ERA PB\ERA PB APD.DOC • 7/10/2003

I _ _ V1

-------
1

LIST OF ACRONYMS
(Continued)

SIWP

Supplemental Investigation Work Plan

SPE

solid phase extraction

SQV

sediment quality value

wet weight

TEF

toxic equivalency factor

TEQ

toxic equivalence

TIE

toxicity identification evaluation

TOC

total organic carbon

tPCB

total polychlorinated biphenyls

TSK

Trimmed Spearman-Karber

USACE

U.S. Army Corps of Engineers

WOE

weight-of-evidence

WSU

Wright State University

WWTP

wastewater treatment plant

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC •• 7/10/2003

-------
1

APPENDIX D

ASSESSMENT ENDPOINT—COMMUNITY STRUCTURE, SURVIVAL,
GROWTH, AND REPRODUCTION OF BENTHIC INVERTEBRATES

D.1 INTRODUCTION

The purpose of this appendix is to characterize and quantify the current and potential risks posed
to benthic invertebrates exposed to contaminants of concern (COCs) in the Housatonic River,
focusing on total PCBs (tPCBs) and other COCs originating from the General Electric (GE)
facility in Pittsfield, MA. The river is located in western Massachusetts and Connecticut,
discharging into Long Island Sound, with the GE facility located near the headwaters of the
watershed. The Primary Study Area (PSA) includes the river and the 10-year floodplain from
the confluence of the East and West Branches of the Housatonic River downstream of the GE
facility to Woods Pond (Figure 1.1-2).

A Pre-ERA was conducted to narrow the scope of the ecological risk assessment (ERA) by
identifying contaminants, other than tPCBs, that pose potential risks to aquatic biota in the PSA
(Appendix B). A three-tiered deterministic approach was used to determine the COPCs. In
general, a hazard quotient (environmental media concentration/threshold effect benchmark)
greater than 1 for aquatic organisms in the Housatonic River resulted in the COPC being
screened through to the next tier of the assessment. The ERA further screened COPCs for
specific relevance to benthic invertebrates occupying the main channel of the Housatonic River.
This additional screening included comparison of invertebrate tissue chemistry to conservative
screening benchmarks. In summary, COPC groups that screened through to the detailed risk
assessment for benthic invertebrates were tPCBs, several metals, several polycyclic aromatic
hydrocarbons (PAHs), and dibenzofuran. Total PCBs detected in Housatonic River media
samples most closely resemble the commercial PCB mixture Aroclor 1260, with a lesser
contribution of Aroclor 1254-derived congeners.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-l

-------
1

D.1.1 Overview of Approach for the Benthic Risk Assessment

A step-wise approach was used to assess the risks of COPCs to benthic invertebrates in the
Housatonic River watershed. The four main steps in this process included:

1. Derivation of a conceptual model (Figure D.l-1), including selection of
representative taxa and establishment of measurement and assessment endpoints.

2. Assessment of exposure of benthic invertebrates to COPCs (Figure D.l-2). The
first step in the exposure assessment was to refine the list of COPCs to focus on the
contaminants of concern (COC) for detailed evaluation in the ERA.

3. Assessment of the effects of COCs on benthic invertebrates (Figure D.l-3).

4. Characterization of risks to benthic invertebrates (Figure D. 1-4).

In addition, a summary of the studies conducted in conjunction with the ecological risk
assessment for benthic invertebrates and their linkage to the ERA is presented in Figure D.l-5.

D. 1.1.1 Conceptual Model

Total PCBs, dioxins, and furans are persistent and hydrophobic and lipophilic. Therefore,
organic carbon pools (both living and non-living) are the primary uptake vectors for benthic
invertebrates. Less hydrophobic COPCs, such as low molecular weight PAHs and metals, are
not as strongly associated with organic pools, and exhibit more complex partitioning behavior.
Essential metals, such as copper, are also homeostatically regulated. In summary, the COCs
identified for benthic invertebrates exhibit both direct (i.e., contact with contaminated source
media) and indirect (i.e., food web bioaccumulation/biomagnification) pathways.

The conceptual model presented in Figure D.l-1 illustrates the exposure pathways for benthic
invertebrates in the PSA. The benthic invertebrate portion of the ERA considered organisms that
reside in, or are in direct contact with, Housatonic River sediment. For sediment invertebrates,
the dominant abiotic exposure media were sediment (solid phase and/or porewater) and surface
water; the former was assigned greater weight in the analysis because of greater ecological
relevance. Contaminant concentrations in sediment represent a more reliable indicator of benthic
invertebrate exposure than those in surface water because benthic organisms reside within or on
top of sediment, which serves as a sink for most COCs. Although some invertebrate species

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-2

-------
1

respire overlying water, aqueous exposure for many COCs (particularly hydrophobic substances
such as PCBs) is a less important pathway for uptake than direct contact with the sediment bed
and/or dietary uptake. Concentrations of COCs in tissues of benthic invertebrates were also
considered (Figure D.l-1). The tissue data provide an organism-based measure of
bioavailability, and provide an additional line of evidence to consider with the conventional
Sediment Quality Triad approach (matched measurements of sediment chemistry, toxicity tests,
and invertebrate communities). Tissue concentrations reflect the net COC uptake from sediment,
food, overlying water, and porewater, and therefore integrate all primary exposure pathways of
interest.

The problem formulation (see Section 2) identified species used in toxicity tests as surrogates for
the Housatonic River freshwater benthic community (i.e., Chironomus tentans, Hyalella azteca,
Lumbriculus variegatus, Daphnia magna, Ceriodaphnia dubia). Both the status of sensitive taxa
and the overall community composition are considered indicators of overall health and
productivity of the benthic community. The benthic invertebrate ERA emphasizes community-
level responses. For example, greater importance was placed on the overall abundance, richness,
biomass, and diversity of invertebrates than was placed on the presence or absence of individual
species. Where individual taxa were investigated to address the assessment endpoint (e.g.,
individual species toxicity tests), they were considered to be representative of the overall
community. This approach is commonly applied for benthic ERAs.

The assessment endpoints for this ERA are benthic invertebrate community structure, survival,
growth, and reproduction. Measurement endpoints were used to evaluate the assessment
endpoint and were integrated using the Sediment Quality Triad approach (Chapman 1996).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-3

-------
1

Measurement Endpoints for Benthic Invertebrates

¦ Determine, based on field studies, the extent to which reductions in benthic
community abundance, biomass, species richness, and other community metrics
have occurred, including species-specific indications of adverse effects.
Determine if these changes can be related to exposure to PCBs or other COCs
in the sediment of the river.

¦ Determine, based on laboratory toxicity studies, the extent to which the

concentrations of PCBs and other COCs in the sediment may cause impairment
of the benthic community or reductions in the survival, growth, or reproduction of
benthic taxa.

9
10

¦ Determine, based on in situ and laboratory toxicity studies performed for this
ERA, the extent to which the exposure to PCBs and other COCs in the river
sediment may result in adverse impacts to survival, growth, and/or reproduction
of the benthic community.

¦ Determine, based on in situ studies and literature values, the extent to which the
concentrations of PCBs or other COCs bioaccumulated in the tissues of the
benthic organisms will cause effects to the survival, growth, or reproduction of
one or more benthic taxa.

20 D.1.1.2 Exposure Assessment

21 The exposure assessment estimates the exposure of benthic invertebrates to tPCBs and other

22 COCs in the Housatonic River PSA (Figure D.l-2). Because the conceptual model for

23 invertebrate exposures (Figure D.l-1) was relatively simple, a complex exposure model was not

24 required. First, the COPCs retained in the Pre-ERA were further screened specifically for the

25 benthic invertebrate endpoint, resulting in the COCs that would be evaluated in the exposure

26 assessment. Exposures were assessed as either the COC concentrations in abiotic site media

27 (i.e., sediment, water), or as the tissue body burdens that represent integrated exposure from all

28 sources.

29 Because habitat factors are known to affect the composition of benthic communities, the

30 exposure assessment also partitioned the PSA into areas of similar macro-habitat. Accordingly,

31 the stations were grouped on the basis of physical substrate and broad biological characteristics.

32 D.1.1.3 Effects Assessment

33 The effects assessment for benthic invertebrates (Figure D.l-3) emphasized the site-specific

34 biological investigations undertaken at the 13 benthic sampling stations because these studies

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-4

7/10/2003

-------
1

provided direct indications on the bioavailability, toxicity, and effects of site COCs. Both the
toxicity assessments (laboratory toxicity, in situ toxicity, and toxicity identification evaluation
[TIE]) and the community evaluations (benthic macroinvertebrate community composition) were
compared to appropriate field references to determine whether the exposed sites on the
Housatonic River exhibited biological impairment. In addition, the concentration-response
relationships for these studies were evaluated to derive site-specific benchmarks for COCs in
sediment.

In developing concentration-response relationships for the sediment toxicity endpoints, two
variations of exposure analysis were applied. Concentration-response modeling was conducted
using the single PCB value that matched the effects data most closely in space and time (i.e.,
"most synoptic" chemistry data). A separate assessment was also conducted in which additional
nearby samples were included to integrate the small-scale variations in PCB concentrations seen
in the sediment chemistry results; in this case, the median was used as the central tendency
measure due to non-normality of chemistry distributions. Both quantitative approaches yielded
similar findings.

An overview of the literature on the effects of tPCBs and other COCs to survival, growth, and
reproduction of benthic invertebrates is also included in the effects assessment. Acceptability
criteria were used to evaluate the literature to determine studies that may be used to develop an
effects metric. Studies were selected and used to derive appropriate effects metrics for tissue,
sediment, and water. In recognition of the uncertainty inherent in threshold effects concentrations
for these media, ranges of benchmarks were derived instead of relying on point estimates for
literature effects thresholds.

D.1.1.4 Risk Characterization

The risk characterization evaluates the likelihood that adverse effects may occur as a result of
invertebrate exposure to tPCBs and/or other COCs (Figure D. 1-4). Three lines of evidence were
available to characterize risks to invertebrates:

¦ Field Surveys (Macroinvertebrate Community Structure) - The risk
characterization considered the magnitude of biological alteration at exposed sites as
compared to reference locations with comparable macro-habitat.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw r 7/10/2003

-------
1 ¦ Linked Exposure and Literature Effects - The risk characterization derived ranges

2 of hazard quotients (HQs) by relating observed COC concentrations in tissue,

3 sediment, and/or water to corresponding effects benchmarks from the literature.

4 ¦ Site-Specific Toxicity Investigations - The risk characterization considered the

5 magnitude of response observed at exposure sites relative to reference locations and

6 negative control treatments. Furthermore, information on cause-effect linkages was

7 obtained from the TIE treatments.

8 A weight-of-evidence approach was used to combine the results from each line of evidence.

9 This included a station-by-station assessment of each sampling location, as well as an overall

10 weight-of-evidence assessment for the assessment endpoint using the approach described in

11 Menzie et al. (1996). The section concludes with a discussion of sources of uncertainty in the

12 assessment of risks, extrapolates the findings to sediment downstream of the PSA, and provides

13 the conclusions of the risk characterization.

14 D.1.2 Organization

15 This appendix is organized as follows:

16 ¦ Section D.2 (Exposure Assessment) describes the quantification of exposures, both

17 for specific benthic sampling locations and for the PSA as a whole.

18 ¦ Section D.3 (Effects Assessment) describes the effects to benthic invertebrates

19 exposed to site COCs, as indicated by the toxicity tests and benthic community

20 evaluations, and summarizes the ranges of benchmarks derived from the literature.

21 ¦ Section D.4 (Risk Characterization) discusses the three lines of evidence, provides a

22 discussion of the sources of uncertainty regarding risk estimates, and the conclusions

23 regarding risk for benthic invertebrates in the Housatonic River.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-6

-------
1

D.2 EXPOSURE ASSESSMENT

The purpose of this section is to determine and evaluate exposure of benthic invertebrates to
COCs, using data that can be appropriately linked to effects information. The exposure
assessment for benthic invertebrates also incorporates consideration of the influences of habitat
and sediment substrate, and assesses the degree to which exposure data can be appropriately
linked to biological effects measurements.

D.2.1 Introduction

D.2.1.1 Selection of COCs for Benthic Invertebrates

The contaminants initially considered in the benthic invertebrate exposure assessment were
identified in the Pre-ERA (Appendix B). The Pre-ERA included screening on a reach-by-reach
basis and subdivision of COPCs by major hydrological/geomorphological category. For the
benthic invertebrate assessment, only samples from the "main channel and aggrading bars"
sediment types were considered to be relevant because benthic locations were specifically
selected to represent main channel habitat. COPCs that screened through to the detailed risk
assessment for benthic invertebrates were total tPCBs, several metals, several polycyclic
aromatic hydrocarbons, and dibenzofuran.

PCBs were screened in as sediment COPCs in all reaches. PAHs were retained throughout the
PSA, although the number of individual PAH compounds screened in was greater for Reaches
5A and 5C, relative to the other reaches. Dibenzofuran was retained for Reach 5A only. Metals
were retained in Reaches 5C and 6 only.

Several additional COPCs (mainly pesticides) were determined in the Pre-ERA to be below
detection limits in sediment, but had detection limits that exceeded screening benchmarks
(Appendix B). These included aldrin, gamma-BHC, chlordane, dieldrin, endosulfan, endrin,
heptachlor epoxide, methoxychlor, benzyl alcohol, and hexachlorobenzene. To assess the
potential importance of these pesticides with respect to invertebrate toxicity, a semi-quantitative
screening procedure was conducted to build on the Pre-ERA screening. This additional screen
considered the frequency of detections in available tissue chemistry data for these contaminants

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw q 7/10/2003

-------
1

(which were not evaluated in the Pre-ERA), combined with a preliminary conservative
ecotoxicity screen using conservative tissue concentrations.

An examination of the detection limits for invertebrate tissues indicated that concentrations for
most of the pesticides of concern (except dieldrin, endosulfan, and hexachlorobenzene) were
below detection limits. These low concentrations were observed despite the highly lipophilic
behavior of these contaminants. To ensure that this finding was not an artifact of high detection
limits, and to assess the potential impact of the more frequently detected pesticides, a comparison
was made to a database of tissue concentration effects data compiled by Jarvinen and Ankley
(1998). Tissue concentration effects data were available for many pesticides, and screening of
maximum observed tissue concentrations showed that even the contaminants detected in
invertebrate tissues were below concentrations shown to be of ecological concern. For example,
the lowest reported aquatic toxicity threshold for dieldrin (no observed effect concentration
[NOEC] for growth) was 0.26 mg/kg tissue, nearly 100 times the maximum observed
concentration. The maximum observed endosulfan concentration (0.06 mg/kg) was below all
invertebrate effect and no-effect thresholds. Aldrin and chlordane were each detected in only 1
out of 21 tissue samples, and all tissue concentrations were below 0.005 mg/kg. A few pesticides
(heptachlor epoxide, methoxychlor, benzyl alcohol, and hexachlorobenzene) could not be
definitively ruled out due to lack of sediment quality values (SQVs) or tissue effects information;
however, these chemicals were detected at concentrations of 0.1 mg/kg or lower. The
concentrations of these pesticides were typically low relative to benchmarks for other pesticides.
Furthermore, some pesticides detections in tissue samples may be attributable to laboratory
interference. Based on these considerations, the entire suite of organochlorine pesticides listed
above was eliminated from further consideration in the invertebrate portion of the ERA.

Surface water COPCs identified in the Pre-ERA (Appendix B) included dioxins/furans, PCBs,
and silver. Therefore, the water chemistry screening did not result in any additional
contaminants that were not already retained as sediment COCs for invertebrates.

The final list of COCs carried forward into the risk assessment for benthic invertebrates is
provided below.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-8

-------
1

COCs for Benthic Invertebrates

¦ Chlorinated organic compounds - tPCBs, dioxins/furans.

¦ Metals - antimony, barium, cadmium, chromium, copper, lead, mercury, silver,
tin.

¦ Semivolatile organic compounds - dibenzofuran.

¦ PAHs - numerous individual PAH compounds, including low- and high-molecular
weight PAHs.

D.2.1.2 Site Identifiers

To match exposure data with effects-based measures, many of the data considered were derived
from sampling conducted in association with the 13 benthic community stations and/or the 7
sediment toxicity stations. For the purposes of this report, the 13 benthic macroinvertebrate
sampling stations are referenced using the "simplified IDs" presented in Figure D.2-1 rather than
the field sampling ID. This labeling approach identifies the benthic stations in the PSA
sequentially (Stations 1 to 9), moving downstream from the confluence of the East and West
Branches. Upstream reference stations are labeled A1 to A3, and the reference location at
Threemile Pond (an uncontaminated reference location within the Housatonic River watershed
but at a higher elevation than the river) is labeled as Station R4. Table D.2-1 presents a cross-
reference of these sample IDs to the original field sample designations. Station A3 (West
Branch) may be affected by PCB contamination because of its location downstream of a known
source other than the GE facility. Although the nomenclature "reference station" is used for
Station A3, it likely represents a location with reduced (but not insignificant) exposure, relative
to the PSA.

Summary of IDs for 13 Benthic Invertebrate Sampling Locations

¦ Upstream Reference Locations: A1, A2, A3 (arranged north to south).

¦ Exposed Locations on Housatonic River: 1 to 9 (arranged north to south).

¦ Downstream (watershed) Reference Location: R4 (Threemile Pond).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-9

-------
1

D.2.1.3 Types of Exposure Data

The approach used to characterize exposure to benthic invertebrates was based on evaluation of
numerous data sources, including both sediment and water column COC concentrations and
invertebrate tissue COC concentrations. Some of the data were collected specifically for the
benthic risk assessment, whereas others were collected for other risk assessment and/or modeling
purposes. This section summarizes general technical issues that affected the selection of
exposure data and subsequent interpretation.

Many of the data applicable to benthic endpoints relate to the Sediment Quality Triad. The
Sediment Quality Triad approach to assessment of sediment quality is based on synoptic
measurement of sediment chemistry, site-specific sediment toxicity, and benthic invertebrate
community structure (Long and Chapman 1985; Chapman 1996). The term "synoptic" refers to
data for two or more variables that are collected at comparable times and locations, allowing for
quantitative comparisons between the variables. It is also a relative term, meaning that there are
varying degrees to which data sets are synoptic.

Use of the Triad approach requires selection of exposure data to achieve maximum
correspondence between exposure and effects endpoints, while also recognizing the need to
address spatial and temporal variability in the data, and inherent limitations in field sampling.
The advantages of maximizing synoptic data collection must be balanced against the
considerations of maximizing sample size and characterizing variability (i.e., spatial and
temporal replication). In the Housatonic River project, data vary in how synoptic they are over
both time and space with associated effects data. Because of the high degree of variability in
PCB concentrations in sediment observed over small spatial and temporal scales and from
analytical sources (see Appendix C.ll), data were used that were not fully synoptic with (but
that have an acceptable correspondence to) associated effects data.

Accordingly, the benthic invertebrate ERA adopted the following criteria in the development of
the exposure data set for comparison to effects endpoints:

¦ Sediment data were considered acceptably synoptic if collected within a radius of 5
meters from the original location used to define the sampling location. In most cases,
the data were tightly clustered within this radius (i.e., within 2 meters or less) due to

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-10

-------
1

efforts to target the same station on multiple occasions while designing the benthic
program.

¦ Exposure data were related to biological effects endpoints only if collected between
March and October of 1999. This time period contains the dates of the site-specific
effects measurements (toxicity tests, benthic community structure evaluations, and
toxicity identification evaluations). The summer of 1999 was consistently a relatively
low-flow season, which minimized the potential for confounding by major storm
events and the associated redistribution of sediment contaminants.

¦ For sediment, samples were considered only if they were collected from within the
top 6 inches (15 cm) of sediment, and included at least the top 2 inches (5 cm).

¦ In merging data sets from multiple studies, exposure data were combined (using a
median of individual data) for all points collected on the same calendar day at the
same location. This was done to avoid bias resulting from a higher number of
samples or replicates on one day, which would have potentially obscured temporal
variability.

¦ The median was used as the preferred measure of central tendency for benthic
invertebrate exposures. Whereas other receptors (e.g., wildlife) integrate their
exposures over foraging areas (justifying the use of means in exposure models), most
benthic invertebrates remain relatively localized over the exposure periods considered
in this ERA. For example, during toxicity tests, organisms were constrained by the
test vessels or in situ enclosures. Therefore, the parameter of interest is the "most
likely" concentration encountered by the organism during the exposure. To derive this
50th percentile concentration, which is equally likely to under- or over-estimate
exposures, the surrounding concentration data were summarized using medians.
Analysis of the individual concentration data indicated lognormal distributions; the
median is an appropriate central tendency measure for lognormally distributed data.

The results of the exposure data processing approach described above form the basis for the
concentration-response modeling presented in Appendix D and Section 3. In addition, a
confirmatory evaluation was also conducted in which the single "most synoptic" chemistry value
was paired with each toxicity endpoint (Attachment D.5). For example, in the confirmatory
evaluation, the effects data from the 48-hour toxicity tests were paired with the sediment
chemistry data (single composite sample) collected on the day of termination of the 48-hour
tests. In most cases, the two approaches yielded similar results, showing that the interpretations
in the risk assessment were not an artifact of the data processing methods. Attachment D.5
presents summary statistics for concentration-response analysis of individual effect endpoints as
well as for linear modeling using combined endpoints.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-ll

-------
1

In addition to the three standard components of the Sediment Quality Triad, additional lines of
evidence were incorporated in the study design to further refine exposure and effects
characterization. These included the contaminant concentrations in benthic invertebrate tissue
and the toxicity identification evaluation (TIE). The former was directly relevant to the exposure
assessment because these data provide a more refined indication of exposure to benthos at the
organism level (Krantzberg et al. 2000). Measurements of contaminants in tissues of resident
benthic fauna provide evidence of bioavailability and a means of determining whether the
contaminants accumulated are responsible for any observed effects. TIE results were not
considered in the exposure assessment, but are instead evaluated in the effects assessment and
risk characterization sections.

The Triad sampling design adopted for the benthic ecological risk assessment (i.e., modest
number of sampling stations, with highly detailed exposure and effects analyses at each station,
including explicit consideration of variability) was not tailored to identification of specific
localized zones of potential impacts in all portions of the river. Instead, the broader purpose of
the study was to address whether PCBs or other COCs are responsible for impairment of
Housatonic River benthic invertebrate communities. The evaluation of potential risks in reaches
of the river not sampled for biology/toxicity required extrapolation of concentration-response
relationships from the toxicity testing location. Such extrapolation is common in ecological risk
assessments, particularly those for sites of this large scale.

D.2.2 Habitat Characterization

To provide the foundation for the risk characterization, the preliminary results of physical and
ecological investigations were used to define appropriate "clustering" of benthic sampling
locations for the exposure assessment. In this case, clustering refers to the grouping of stations
of similar properties (e.g., sediment types), to discriminate between substrate-related responses
and those attributable to contaminants. Limiting statistical contrasts to similar substrate types
helped minimize the confounding factors of particle size and TOC, such as reduced abundance
and diversity in sandy sediment relative to fine-grained sediment (Ward 1992, Minshall 1984,
Allan 1995).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-12

-------
1

Rationale for "Clustering" of Benthic Sampling Locations in ERA

Clustering was relevant for the exposure assessment for three primary reasons:

¦ To determine the appropriate reference stations for making statistical
comparisons to exposed stations.

¦ To provide a means of separating physical and ecological "regimes" in a manner
consistent with both the exposure and effects assessment.

¦ To provide a tool to assess if there are other influences that have not been well
characterized.

10 For a heterogeneous system such as the Housatonic River, habitat considerations are likely to be

11 a major factor in determining the nature of benthic communities, with chemical stressors

12 secondary. Organic carbon in sediment serves as a food source for sediment infauna, and

13 particle size distributions are closely related to invertebrate life-histories (burial behavior and

14 feeding). Therefore, these substrate characteristics (i.e., total organic carbon [TOC] and grain

15 size) are considered important factors in the exposure assessment. Separating the effects of

16 substrate and habitat, to the extent possible, helps when evaluating the potential risk from

17 contaminants.

18 To minimize the potential effects of such differences in sediment characteristics along the PSA,

19 the benthic invertebrate community sampling was restricted to softer sediment. However, the

20 general change in sediment type in the PSA from coarse sands (upstream) to silts and clays

21 (downstream) resulted in some unavoidable differences in the nature of the sediment sampled.

22 Two approaches were used to evaluate the Triad stations in terms of their gross habitat

23 characteristics:

24 ¦ Physical substrate variables were evaluated within and among benthic sampling

25 locations downstream of identified contaminant sources to identify significant break

26 points in physical habitat features (e.g., substrate type).

27 ¦ Aquatic habitat was evaluated to identify broad biological regimes within the benthic

28 sampling locations. Parameters such as flow rate, velocity, depth, canopy cover,

29 stream width, water temperature, sedimentation rate, pool variability, and other

30 factors can also affect the distribution of benthos (Plafkin et al. 1989).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-13

7/10/2003

-------
1

Based on these approaches, a clear demarcation in substrate and habitat type was observed
between Stations 5 and 6, which is coincident with the transition in river regime from Reach 5A
to Reach 5B, and also with the location of the Pittsfield wastewater treatment plant (WWTP)
outfall. This shift in habitat characteristics (e.g., particle size distributions and organic carbon
content of sediment) was used to identify appropriate statistical contrasts in the ERA.

D.2.2.1 Physical Substrate

The sediment characteristics at the study locations can be broadly categorized as either "coarser
grained" (all stations upstream of the Pittsfield wastewater treatment plant [WWTP]) or "finer
grained" (all remaining stations). The patterns in substrate composition (for benthic community
samples) are summarized in Figure D.2-2, which depicts the trends in organic carbon and percent
fines with distance downstream. The graph shows that the downstream stations contained a high
percentage (mean of 60%) of fines (i.e., silt and clay fractions). This contrasts with the upstream
stations for which sediment with greater than 5% fines were rare. A similar pattern was
observed for total organic carbon. Mean TOC was less than 1% at all stations upstream of the
WWTP, except Station A3. Mean TOC ranged from 2 to 8% at the five downstream stations.

D.2.2.2 Benthic Community Types

Preliminary analyses of the community composition data revealed broad patterns relevant to the
assessment of habitat influences. Specifically, the composition of benthic communities changed
from numerical dominance by dipterans (true flies) in upstream portions of the study area to
dominance by gastropods and bivalves in the lower portions. These changes are consistent with
expectations based on the life histories and habitat preferences of the species present in the river.
Multivariate statistical analyses (cluster analyses and non-metric multidimensional scaling) of
community composition supported this finding (Figures D.2-3 and D.2-4), and revealed a
marked distinction between the assemblages found in the "coarser-grained" upstream
contaminated locations (Stations 1-5), and the "finer-grained" downstream contaminated
locations (Stations 6-9; Station R4). Both analyses also indicated a difference between coarser-
grained contaminated sediment and coarser reference sediment that could not be attributed to
obvious habitat influences.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-14

-------
1 In summary, preliminary analyses of biological and physical components of habitat indicate a

2 pronounced habitat change near the WWTP due to changes in the hydrologic regime in the river.

3 The broad changes in community composition were directly proportional to alterations in

4 physical substrate, in particular TOC and percent fines. Accordingly, the stations were grouped

5 as follows (see Figure D.2-1):

Grouping of Benthic Sampling Stations Based on Habitat

Benthic sampling stations in Figure D.2-1 were assigned to one of the following

categories:

¦ "Coarser" Reference Locations (C/R) - Low total organic carbon (TOC) (typically
less than 1%), sandy sediment found either upstream of influence from the GE
facility or on the West Branch. Three locations (A1, A2, A3).

¦ "Coarser" Contaminated Locations (C/C) - Low TOC (typically less than 1%),
sandy sediment found between the confluence and the Pittsfield WWTP. Five
locations (1, 2, 3, 4, 5).

¦ "Finer" Reference Locations (F/R) - High TOC, silty sediment found outside the
PSA at Threemile Pond (Location R4).

¦ "Finer" Contaminated Locations (F/C) - High TOC (typically a few percent or
greater), silty sediment found downstream of the Pittsfield WWTP. Four
locations (6, 7, 8, 9).

20 D.2.3 Statistical Approach and Methods

21 Three approaches were used to quantitatively estimate exposure metrics in the benthic

22 invertebrates exposure assessment:

23 ¦ Comparison to Reference Approach - The station groupings described above (C/C,

24 C/R, F/C, F/R) were used in statistical contrasts for hypothesis testing, using an

25 analysis of variance (ANOVA) design. Differences in exposure, independent of

26 habitat factors, were isolated by minimizing habitat differences within each treatment

27 and by performing contaminated-to-reference comparisons for treatments with similar

28 habitats.

29 ¦ Gradient Approach - This approach was used to evaluate the interactions and

30 correspondence between exposure metrics. Regression analyses were used to identify

31 the strength of the quantitative association between variables, such as TOC and

32 percent fines, or tPCB concentration and TOC. In some cases, the regression

33 approach was combined with the reference approach by performing separate

34 regressions for each of the four station types identified.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-15

-------
1 ¦ Spatial Distributions - For variables that were independent of habitat type or that

2 indicated weak associations with habitat variables, simple plots of central tendency

3 and variability were used to display spatial trends over the PSA. The stations were

4 sorted from north to south to correspond to changes in exposure variables with

5 distance from the primary source of contamination (i.e., GE facility).

6 D.2.4 Assessment of Sediment Chemistry

7 D.2.4.1 Overview

8 Sediment chemistry is often used as a measure of exposure to benthic invertebrates for several

9 reasons:

10 ¦ Sediment serves as a sink for numerous contaminants, particularly those that partition

11 strongly to the organic carbon pool in the sediment bed.

12 ¦ The short life span and relative lack of mobility of most benthic invertebrates means

13 that sediment concentrations measured in a given area are a good indication of the

14 contaminant concentrations to which local benthos are exposed.

15 ¦ Databases of effects concentrations exist for many COCs, enabling comparison of

16 sediment chemistry to effects-based benchmarks.

17 This section discusses both COCs and ancillary parameters that were considered in the

18 evaluation of sediment chemistry data. These other parameters (TOC, grain size) were

19 considered for two primary reasons:

20 ¦ They are important in defining physical habitat conditions that influence biological

21 communities.

22 ¦ They can influence contaminant partitioning and/or bioavailability.

23 D.2.4.2 Sources of Sediment Data

24 The following data sources are described in terms of their relevance/applicability to the benthic

25 invertebrate assessment endpoints. Relevance was defined as a function of their compatibility

26 with effects data, degree of replication in time/space at each station, number of samples

27 collected, types of analytes measured, and other factors.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-16

-------
1

D.2.4.2.1 Benthic Community Grab Samples

Exposure data were collected by EPA in 1999 during collection of the benthic community
structure samples. At each of 13 benthic sampling locations within the study area, 12 replicate
grabs were collected for benthic taxonomic enumeration. This resulted in a total of 156
individual sediment grabs, of which 108 (9 stations x 12 replicates) represent contaminated sites
between the confluence and Woods Pond, and 48 (4 stations x 12 replicates) represent reference
conditions not affected by the primary source of PCB contamination. Small core samples were
extracted from each grab for analysis of PCBs (Aroclors and tPCBs), grain size distribution, and
TOC. Other COCs were not evaluated in these samples due to insufficient sample volume
obtained from the mini-core extraction. The 12 replicates were collected over a relatively small
area (i.e., within a 2-meter radius of the "center" of the station), providing representation of
micro-scale spatial variability in analyte concentrations at the time of sampling. Because the
core samples consist of a small subsample of sediment from each sediment grab (to minimize
disturbance of the biota in the remainder of the grab), the mini-cores are not precisely synoptic
with the benthic community replicates. However, the benthic community chemistry data provide
a good indication of both central tendency and chemistry variability at each of the 13 stations.

D.2.4.2.2 Toxicity Testing Samples

Toxicity testing of Housatonic River sediment conducted by Wright State University (WSU) in
1999 yielded additional sediment chemistry data that can be linked to benthic invertebrate effects
data (EVS 2003). A total of seven stations were sampled for in situ and laboratory toxicity
analyses; six of these stations were co-located (within a few meters) with one of the group of 13
benthic community locations listed above. Table D.2-1 indicates the correspondence between
WSU site IDs and the IDs for other Triad data.

For five of the seven toxicity stations, a complete Sediment Quality Triad evaluation was
conducted, including laboratory toxicity and in situ toxicity, in addition to chemistry and benthic
community structure. A modification to this design was made for the remaining two stations
(Stations 5 and 8A) due to the proximity of Stations 8 and 8A. Stations 8 and 8A were situated
approximately 12 meters apart, with Station 8A positioned along the center of the channel (i.e.,
thalweg) and Station 8 located just to the north adjacent to a point bar. These two stations were

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-17

-------
1

both evaluated for laboratory toxicity. However, Station 8A was not evaluated for in situ
toxicity or benthic community structure, due to its proximity to Station 8. Instead, a different
upstream station (Station 5) was included in the in situ toxicity testing program to represent a
moderately contaminated station in the PSA.

The sediment chemistry data collected during the toxicity study did not include small-scale
spatial replication as was performed for the assessment of benthic community structure. Instead,
five separate sediment grabs were composited at each location to provide an estimate of the
average sediment contamination, and a single chemistry value was obtained from the composite.
Multiple measurements were made during the test, and separate measurements were made for
laboratory and in situ testing. A broader suite of analytes was measured in conjunction with this
program, which included:

¦ Laboratory toxicity samples - Composite samples (mixture of five cores) were
collected; samples were submitted for measurement of chemistry and laboratory
toxicity. Chemical analyses included tPCBs and TOC.

¦ In situ toxicity samples - Composite samples were taken in a similar fashion to the
laboratory samples; however, samples were also taken at different time periods (i.e.,
end of 48-hour, 7-day, and 10-day exposure periods). Analysis for tPCBs was
conducted for all sediment. The 7-day sediment samples also were evaluated for PCB
congeners, PAHs, pesticides, metals, chlorinated benzenes, and TOC.

D.2.4.2.3 Other Housatonic River Sediment Data

The EPA database contains thousands of entries of sediment chemistry parameters from a variety
of sources, including data collected by GE. Although most of these data are not precisely
synoptic with effects measurements, many of the data are applicable to the ERA endpoints,
provided the data are properly screened for spatial and temporal relevance. Inclusion of these
data better characterized spatial and temporal heterogeneity in sediment data that could not be
addressed using a more limited data set. Therefore, sediment quality data collected during 1999
in proximity to the benthic and toxicity stations were considered for inclusion in the exposure
assessment; the criteria for screening these data are outlined above in Section D.2.1.3.

Examples of data not collected for the toxicity and benthic community studies, but considered
relevant for comparison to the toxicological endpoints, include:

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-18

-------
1 ¦ April-May 1999 Toxicity Screening Sampling - Several locations were sampled

2 several weeks prior to initiation of the toxicity testing and analyzed for tPCB

3 concentrations in sediment; these results were used in the final selection of stations

4 for testing. Many locations were tested within meters of the toxicity stations.

5 ¦ April 1999 Discrete River Sampling - Several locations were sampled in proximity

6 to the toxicity stations 2 months before toxicity tests were conducted; analyses

7 included tPCBs, PCB Aroclors, and TOC.

8 ¦ May 1999 and July 1999 Mussel Exposure Sampling - Several additional samples

9 were collected in the area of the toxicity stations in conjunction with the discontinued

10 caged mussel study and were tested for tPCBs, PCB congeners, inorganics, Appendix

11 IX semivolatiles, grain size, herbicides, pesticides, and other parameters.

12 ¦ September 1999 Porewater TIE Sampling - WSU resampled most of the toxicity

13 study locations and conducted screening tests. Chemistry data collected included

14 sediment PCBs, metals, dioxins/furans, Appendix IX semivolatiles, and TOC.

15 ¦ October 1999 Discrete River Sampling - Several locations were sampled in

16 proximity to the benthic stations a few months after toxicity tests were conducted;

17 analyses included tPCBs, PCB congeners, grain size, and TOC.

18 More recent sampling was performed in the vicinity of some of the benthic stations, including

19 April 2002 discrete river sampling and September 2001 porewater study sampling. However,

20 these data were not used in the development of concentration-response relationships because of

21 the elapsed time (and event-based hydrodynamics) between effects measurements and exposure

22 data.

23 Sediment chemistry data not meeting the criteria for direct comparison to effects data were still

24 used in the exposure assessment, although in a more limited way. For the purpose of this benthic

25 invertebrate ERA, the term "generic" was used to refer to data not specifically comparable with

26 effects measurements made in the 1999 site-specific studies. These data were collected for other

27 ERA or modeling objectives. These sediment data were used for:

28 ¦ Screening against sediment quality values.

29 ¦ Describing general exposure patterns throughout the PSA.

30 ¦ Assisting in the extrapolation of concentration-response relationships derived using

31 the more synoptic data categories described above.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-19

-------
1

D.2.4.3 Distributions and Concentrations of Physical Parameters

D.2.4.3.1 Grain Size

Figure D.2-5 compares the sediment grain size distributions in each of the four station groupings
(C/C, C/R, F/C, F/R). The reference sites for each habitat classification were representative in
terms of particle size distributions. Analysis of the median particle sizes (D50; log-transformed
to address non-normality) also yielded similar findings. The central tendency estimates of D50 in
the coarse sediment (both contaminated and reference locations) were approximately 1.0 mm,
whereas the central tendency estimates for the fine sediment groups (both contaminated and
reference locations) were well below 0.1 mm.

Figure D.2-6 provides histograms of the silt and clay content of sediment samples (i.e., percent
fines). The marked difference in particle size distributions among location types was apparent
from these graphs. While the distribution of percent fines was somewhat wider in F/C sediment
relative to that found in F/R sediment, the central tendency estimates are similar (60 to 65%
fines). In contrast, the vast majority of "coarse" stations exhibited less than 5% fines.

D.2.4.3.2 Organic Carbon

TOC concentrations in the benthic grabs indicate a similar pattern to that observed for particle
sizes (Figure D.2-7). The distribution of TOC in coarse sediment (C/C and C/R) was skewed
because of the large number of stations with low organic carbon content (i.e., less than 1%). In
fine-grained sediment, TOC concentrations were not skewed and exhibited approximately the
same range in both contaminated and reference samples. The difference in mean TOC in the
benthic grabs between F/C and F/R locations was small and was not considered to be
ecologically significant.

To ensure that the TOC concentrations in benthic grabs were representative of the broader data
set, TOC data considered representative of the sediment toxicity studies were plotted on a
logarithmic scale (Figure D.2-8). Although variability in the TOC values was observed at each
station, the general conclusions were the same as for the benthic grab data. The C/C and C/R
stations had median TOC concentrations of approximately 1%, whereas the F/C stations had

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n OH 7/10/2003

-------
1

TOC in the 1 to 10% range. Because of the representativeness of the benthic grab data and
because of the large and consistent sample sizes, these data were used in the statistical analyses.

D.2.4.3.3 Correlation Between TOC and Grain Size

Figure D.2-2 suggests a strong relationship between TOC and percent fines in the benthic
invertebrate grabs. The logarithmic regression between these variables was highly significant (p
< 0.001), with percent fines explaining approximately 80% of the variance in TOC (R2 = 0.80).
However, as shown in Figure D.2-9, the strength of the relationship was not consistent across
sediment types. The coarse-grained contaminated sites exhibited no significant relationship
between TOC and percent fines (p = 0.57), whereas the coarse-grained reference sites and fine-
grained contaminated sites had strongly significant regressions (R2 = 0.73 and 0.81,
respectively).

D.2.4.4 Distribution and Concentrations of PCBs
D.2.4.4.1 Benthic Grabs

Sediment samples collected synoptically using core subsamples from each of the 156 benthic
grabs (13 stations x 12 grabs) were analyzed for tPCB and Aroclors, at a detection limit of
approximately 0.5 mg/kg dw. PCBs, quantitated as Aroclor 1260, were detected in all replicates
in 8 of the 9 contaminated stations. At the remaining contaminated station, Aroclor 1260 was
detected in 11 of 12 replicates. No PCBs were detected in sediment at any of the reference
locations.

Individual replicate concentrations of tPCB for each benthic sampling station are presented on a
logarithmic scale in Figure D.2-10. Both means and medians are presented in Table D.2-2.
Because of the lognormal distribution of these data, the median was the more appropriate
measure of central tendency for discussion of these data. The data indicate highly elevated tPCB
concentrations in the C/C sites, with median values of approximately 5 to 25 mg/kg. In the F/C
sites, median PCB concentrations were generally lower (1 to 5 mg/kg), with the exception of
Station 8, which exhibited concentrations similar to the coarse-grained stations. A pooled
variance t-test was conducted to evaluate whether PCB concentrations were higher in the coarse-

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw ^7/10/2003

-------
1

grained contaminated sites relative to the fine-grained contaminated sites. In spite of the high
variability observed at all contaminated stations (Figure D.2-11), the difference was large enough
to be statistically significant (p < 0.001).

There was considerable variability in tPCB concentrations between replicates at some stations,
and an intra-station range in tPCB concentrations of 1 to 2 orders of magnitude was common
(Figure D.2-10). This indicates that small-scale variability in PCB concentrations was
significant, and is an important consideration in the risk assessment. Appendix C.ll provides a
discussion of the variability observed due to analytical considerations.

To provide a measure of potential bioavailability, tPCB concentrations were also normalized on
the basis of total organic carbon (TOC). These TOC-normalized data are summarized in Table
D.2-2. TOC-normalized PCB concentrations in the C/C stations were substantially elevated
relative to reference stations and F/C.

Regression analysis was used to test whether tPCB concentrations were significantly correlated
with TOC and/or percent fines, on both a river-wide and habitat-specific basis. The regression of
PCB concentrations on percent fines yielded a statistically significant negative relationship (i.e.,
PCBs decreased as percent fines increased; p = 0.011). However, this relationship did not hold
when the data were separated into the four station groupings; p-values for the individual
regressions ranged from 0.18 to 0.29. The coarse and fine sediment each form separate clusters
within which there was no apparent relationship between PCB concentration and percent fines.
Similar findings were observed for regression of tPCB against TOC.

The TOC-normalized data imply that the bioavailable concentrations of PCBs may be greater in
the C/C samples relative to the F/C samples. However, the poor correlation between PCB
concentrations and organic carbon in the upstream portions of the PSA complicates this
interpretation. The lack of association between PCBs and organic carbon is somewhat unusual
given the tendency for PCBs to partition to organic carbon pools in the environment (equilibrium
partitioning). PCB partitioning in the upstream portions of the PSA does not appear to follow
conventional equilibrium partitioning behavior. It is possible that the distributions could be
partly explained on the basis of proximity to the source of PCB loadings, or to the existence of a
separate PCB phase not strongly associated with TOC. Because chemical and biological

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n OO 7/10/2003

-------
1

(bioaccumulation) data (see Section D.2.5) do not indicate that PCB bioavailability increases in
the low TOC sediment, the benthic ERA focuses on non-normalized PCB concentrations.

D.2.4.4.2 Toxicity Testing Samples

Concentrations of tPCBs were measured in Housatonic River sediment in conjunction with
laboratory and in situ toxicity and bioaccumulation tests conducted between May and July of
1999 (EVS 2003). Locations (with one exception) were a subset of those used for the benthic
macroinvertebrate sampling; two of the four benthic reference locations (A1 and A3) and four of
the nine contaminated benthic locations (4, 5, 7, 8) were included in the toxicity testing program.
Location 5 (just upstream of the WWTP) was used for the in situ experiments only, and sediment
from Location 8A was used only in the laboratory exposures:

Station ID

Benthic Grabs
Collected?

In situ Toxicity
Tests Conducted?

Laboratory
Toxicity Tests
Conducted?

TIE Analyses
Conducted?

Concentrations of tPCBs were measured in sediment samples collected from each of the six
stations for laboratory toxicity testing. Sediment samples were also collected from each station
at the end of each of the three in situ experiments (i.e., after 48-h, 7-d, and 10-d). Sediment
samples collected after the 7-d in situ experiment were also analyzed for PCB congeners and

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n 7/10/2003

-------
1

other analytes. The tPCB concentrations for these various sampling events are presented in
Table D.2-3.

Table D.2-3 and Figure D.2-12 show that tPCB concentrations were quite variable within
stations across the four toxicity sampling events. This was consistent with the findings of the
chemistry analyses conducted in conjunction with the benthic community grabs, and otherwise
observed in the sediment data collected for the project. At the four contaminated stations, tPCB
concentrations varied by up to 30-fold within a given station over the four measurements.
Furthermore, the apparent spatial pattern of tPCB concentrations in the data from the
toxicity/bioaccumulation samples was different from that described for the benthic grab samples.
There was an increase in tPCB concentrations at downstream contaminated stations that
contrasted with the decrease in tPCB concentrations observed in the benthic grab samples.

D.2.4.4.3 Other Data Relevant to the Sediment Quality Triad Analysis

These results suggest that PCB exposure data from the toxicity testing data sets should not be
extrapolated to benthic community effects endpoints, and also indicate that the variability in the
chemistry data associated with the toxicity program must be considered when deriving
concentration-response relationships. Because sediment chemistry samples were not replicated
in individual toxicity sampling events, data from all relevant sampling events were therefore
included in the development of concentration-response relationships for toxicity endpoints.

Figure D.2-13 shows the tPCB values from all relevant sampling events conducted in 1999 in
proximity to the sediment toxicity stations. These data include the benthic community grab
samples and the four toxicity testing grab samples described above. In addition, data collected
just before or after toxicity testing were included, resulting in an additional six sampling events.
These additional studies help to quantify the small-scale variability over space and time. Figure
D.2-13 shows that the spatial pattern in the toxicity testing tPCB data was generally consistent
with that of the other sampling investigations conducted in 1999. The range of tPCB
concentrations was at least one order of magnitude at all sampling locations. The graph also
shows that the sediment tPCB concentrations at benthic community Stations 7 and 8 were the
lowest of all sampling investigation results. These differences may have been due to small-scale
variability; however, because the results are based on 12 replicates, there is likely a true

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n O/l 7/10/2003

-------
1

difference in the tPCB concentrations for each of these events. Therefore, concentration-
response analysis for benthic community metrics considered only the 12 replicate values at each
site. The concentration-response analysis for toxicity studies utilized the median value from all
applicable sampling events at each location.

D.2.4.4.4 PSA Sediment Characterization

Many data collected by EPA and GE throughout the PSA were not directly related to studies
establishing biological effects endpoints. However, these data are still applicable to the ERA, in
that they can be used to define broad COC exposure patterns. These exposures can be combined
with concentration-response relationships derived from the biological investigations and used to
estimate risk.

Figure D.2-14 depicts the spatial distribution of tPCB concentrations within the PSA. The data
indicate that median PCB concentrations are highest in the upstream reaches of the PSA and
decrease with distance from the GE facility. The median concentrations are lowest just
downstream of the WWTP, but increase moving farther downstream to Woods Pond. This
spatial distribution was different from that evident in the toxicity samples (Figure D.2-13). This
should not be surprising, however, because the toxicity stations were chosen specifically to
represent a range of PCB concentrations when designing the study rather than to represent
broader reach-wide exposure patterns.

For areas downstream of Woods Pond, tPCB data were evaluated separately in the Pre-ERA.
Figure D.2-15 summarizes the tPCB data between Woods Pond and the Connecticut border.
Figure D.2-15 also presents the median sediment tPCB concentration within each river segment.
Overall, the tPCB concentrations are much lower than in the PSA (by approximately an order of
magnitude) and are lowest in Connecticut. However, localized elevated concentrations are
observed in several river segments, with a few concentrations above 30 mg/kg within 2 of the 10
river segments.

Sediment tPCB concentrations downstream of the Connecticut border were generally below
1 mg/kg, reflecting the general trend of decreasing concentration with distance downstream.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-25

-------
1

D.2.4.5 Distribution and Concentration of Other COCs

Although only PCBs were measured in the core subsamples collected in conjunction with the
benthic community assessment, there are data available on other COCs in Housatonic River
sediment from other aspects of the investigation. The most relevant sediment chemistry data for
other COCs are:

¦ Sediment Toxicity Testing (EVS 2003) - In addition to PCBs, measurements of
PAHs, metals, and Appendix IX semivolatiles and pesticides were made for samples
from 7-d in situ toxicity testing (June 1999), and for most laboratory toxicity testing
samples (May 1999). Additional sediment chemistry data were obtained from
collections in September 1999, in conjunction with the porewater TIE investigation.
The latter sampling effort was conducted in two phases; the first round (denoted
"AW") was conducted during September 2-9, 1999, and the second round (denoted
"AD") was conducted during September 15-20, 1999.

¦ Supplemental Investigation Work Plan (WESTON 2000) - Other COCs were
measured at a subset of the Sediment Quality Triad stations. For example, samples
were obtained at toxicity testing stations in July 1999 in conjunction with the
discontinued caged mussel exposure study. These samples were analyzed for a
variety of COCs including PCBs, PAHs, metals, dioxins/furans, and Appendix IX
semivolatiles and pesticides.

Figures D.2-16 to D.2-28 portray the spatial distributions of other COCs from the above two
studies. Principal findings are:

¦ Dioxin/furan concentrations (Figures D.2-16 and D.2-17) are elevated at downstream
fine-grained locations relative to upstream and/or reference locations. Total dioxin
and total furan concentrations are below 1 ng/kg at reference stations, but median
concentrations are close to 10 ng/kg at F/C stations (Stations 7, 8, and 8A). In
addition, the moderate dioxin/furan contamination observed at A3 provides evidence
that A3 does not represent a pristine or true background condition.

¦ Dibenzofuran concentrations (Figure D.2-18) do not indicate a pronounced spatial
pattern throughout the PSA; most concentrations were in the 0.1 to 1.0 mg/kg range.
In the Pre-ERA, dibenzofuran was identified as a COPC only for Reach 5A. The
sediment benchmark used in the Pre-ERA was 0.42 mg/kg. The HQs calculated
using this value were low; in the Pre-ERA, only two of 15 measured dibenzofuran
concentrations in Reach 5A exceeded this benchmark, and never by a factor of 10 or
greater. In the sediment data set considered for the benthic invertebrates, all but two
individual concentrations were within a factor of 2 of the screening benchmark.
There was no significant trend in dibenzofuran concentrations with distance
downstream in the PSA, and the majority of PSA stations had dibenzofuran
concentrations within the range observed at background stations. The

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n OA 7/10/2003

-------
1

correspondence between dibenzofuran concentrations in sediment and toxicity
endpoints in the site-specific toxicity testing was very weak compared to other COCs.
Because of the relatively minor differences between contaminated and reference sites,
the low HQs, and limited spatial extent identified in the Pre-ERA, this contaminant
was not considered further in the ERA.

¦ Total PAH concentrations were highly variable (Figure D.2-19), both spatially and
between sampling events. In all samples, PAHs tended to be dominated by higher-
molecular weight PAH compounds. Station 7 had the highest individual
measurement, exceeding 200 mg/kg total PAH in the EVS (2003) 7-day in situ
sampling program. The median concentrations were greatest near the urbanized areas
of the Housatonic River watershed, and were lowest at A1 and at the Woods Pond
headwaters. Similar to dioxins/furans, Station A3 exhibited elevated concentrations
of PAHs, with total PAHs ranging from approximately 20 to 60 mg/kg. The broad
spatial pattern of PAH concentrations in the toxicity locations was opposite the
pattern observed for PCBs.

¦ Inorganic concentrations (Figures D.2-20 to D.2-28) were typically lower at upstream
sites relative to the fine-grained sediment found downstream. Inorganic constituents
exhibited generally similar concentration patterns. Antimony, barium, chromium,
copper, and lead all exhibited gradual increases downstream, with an order of
magnitude difference between the stations with the highest and lowest concentration.
Cadmium, mercury, silver, and tin exhibited a steeper gradient, with a two order-of-
magnitude difference between upstream and downstream stations. Inorganic
concentrations were significantly correlated (p<0.05) with total organic carbon
concentrations.

D. 2.4.6 Discussion of Results

The various data sources indicate that there was considerable variability in sediment PCB and
other COC concentrations across a variety of spatial and temporal scales. Because the benthic
invertebrate risk characterization used a weight-of-evidence approach to evaluate potential
chemical-induced effects, it is useful to evaluate these sources of variation and their implications
for linkages to effects-based endpoints.

D.2.4.6.1 Small-Scale Spatial Variability

The PCB concentrations in samples taken from the benthic macroinvertebrate sampling program
(Figure D.2-10) indicate high variability, even at a spatial scale of a few meters or less. The
lognormal distributions evident in these grabs indicated that PCBs were not homogeneously

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC ~T\ 07 7/10/2003

-------
1

distributed in the sediment, but tended to be concentrated in localized areas at each station.
Small-scale variability and analytical variability (Appendix C.ll) combined to create uncertainty
in the PCB exposure concentrations for benthic invertebrates.

4 D.2.4.6.2 Small-Scale Temporal Variability

5 It is difficult to separate the effects of small-scale spatial and temporal variability because it is

6 impossible to sample exactly the same sediment on two separate occasions. Therefore, any

7 assessment of temporal changes will be confounded by the small-scale spatial variation described

8 above. The toxicity testing samples taken at the same site several days apart (Figure D.2-12)

9 often indicated order-of-magnitude differences in tPCB concentrations. However, the overall

10 spatial trends in PCB concentrations were reasonably consistent for each of the four sampling

11 events associated with the toxicity study. It is likely that the observed variability in PCB

12 concentrations over these closely spaced sampling events was due to small-scale spatial

13 variability rather than broader changes in PCB distribution over time. This is supported by

14 analyses of sediment and water data collected over a period of years at the same locations

15 showing little or no temporal patterns (BBL and QEA 2003).

16 In summary, the following assumptions were made to guide the exposure assessment and to

17 address variability, while also taking into account the benefits of sample size (and statistical

18 power):

19 "For benthic community endpoints, only data from the 12 benthic grab samples were

20 considered. The synoptic collection of these chemistry data with the biota samples

21 and the large sample size allows for a robust assessment of concentration-response for

22 PCBs. Because the 12 samples were collected from a highly localized area

23 (approximately 2 meters maximum diameter), it is best to treat these chemistry data

24 as replicates for use in estimating the central tendency and variation exposures.

25 Pairing of exposure data with effects data for each individual replicate may not be

26 appropriate because of the dynamic nature of the sediment microenvironment at this

27 spatial scale. Sampled organisms were exposed to sediment areas larger than the

28 surface area of the micro-core tube used to extract sediment for chemical analysis.

29 "For sediment toxicity endpoints, selection of data from a single sampling event could

30 potentially lead to bias because of the lack of replication and the high variability

31 observed. Combining data from sampling events that are approximately (but not

32 exactly) synoptic with the effects data more accurately portrays the COC

33 concentrations at each station.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC D 28 7/10/2003

-------
1 The validity of this approach was tested by comparing each effects measurement with the single

2 most synoptic chemistry measurement. Comparability of the overall results of the two

3 approaches indicated that the concentration-response analysis was not biased by the methods

4 used to select the exposure data. For brevity, the statistics for the concentration-response

5 analyses conducted with the "single most synoptic chemistry" have not been presented in this

6 appendix.

7 D.2.5 Tissue Chemistry Assessment

8 D.2.5.1 Overview

9 Use of exposure media concentrations as dose surrogates has several limitations (Landrum et al.

10 1992), including:

11 ¦ The bioavailable and toxicologically active fraction of the total chemical

12 concentration may not be known.

13 ¦ Multiple uptake routes of chemicals (e.g., dermal absorption, dietary uptake) are not

14 considered.

15 ¦ Intermittent or variable exposure cannot be readily assessed (particularly relevant for

16 small-scale heterogeneity observed in Housatonic River sediment).

17 ¦ Factors influencing exposure (i.e., bioaccumulation kinetics, organism avoidance) are

18 not considered.

19 By associating the effects endpoint with the tissue concentration of the contaminant causing the

20 effect, many of the complicating factors listed above are reduced or eliminated. The basic

21 principle in the evaluation of tissue concentrations of COCs is that there is a relationship

22 between chemical concentration in tissue at the site of toxic action and the toxic response of

23 interest (Eaton and Klaassen 1995).

24 D.2.5.2 Tissue Data Sources

25 There are two major sources of data on PCB tissue concentrations in benthic invertebrates in the

26 Housatonic River:

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-29

-------
1 ¦ Analysis of composite samples of predators and shredders, respectively, conducted in

2 1999 as part of EPA site characterizations.

3 ¦ Data from a 7-day in situ bioaccumulation study (EVS 2003) conducted with the

4 oligochaete worm Lumbriculus variegatus.

5 D.2.5.2.1 Predator/Shredder Tissue Chemistry Analysis

6 Composite samples of benthic invertebrates were collected from 11 of the 12 stations (2

7 reference and 9 exposure stations - no sample was analyzed from Station A2 due to insufficient

8 volume of material) for tissue analysis for selected contaminants (Table D.2-1). The station

9 locations corresponded to those used for the benthic community analysis, 6 of which were also

10 used for in situ toxicity testing.

11 At each station, benthic invertebrates were collected using a D-net (i.e., kick sampler) from an

12 area within approximately 10 m of the grab sampling location. The benthic organisms collected

13 by the D-net were separated into two functional feeding groups, predators and shredders, to yield

14 two composite tissue samples for each station. These two groups were selected to evaluate

15 whether there are consistent differences in concentrations from different benthic feeding guilds.

16 There was no shredder sample analyzed for Station 3, and no predator samples were analyzed for

17 Stations 2 and 6, resulting in a total of 21 benthic tissue samples. The missing samples were due

18 to tissue volume limitations for chemical analyses (i.e., insufficient tissue volume was sampled

19 to perform both analyses). Each tissue sample was analyzed for a number of organic

20 contaminants, including PCBs (as congeners, as Aroclors, and as tPCBs), dioxin/furans, and

21 pesticides. In addition, the lipid content of each sample was measured.

22 D.2.5.2.2 In Situ Bioaccumulation Study

23 In situ bioaccumulation tests (7-day exposure) were conducted at 6 of the 13 benthic

24 macroinvertebrate sampling stations using the oligochaete worm Lumbriculus variegatus (EVS

25 2003). Following this exposure, the tPCB concentration and lipid content of the surviving

26 organisms from each station were analyzed, yielding one tissue sample per station.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-30

-------
1

D.2.5.3 Distribution and Concentrations of PCBs

The concentrations in predators and shredders were compared to determine whether PCB
concentrations differed significantly between these feeding groups. Both a two-sample t-test and
a paired t-test indicated that there was no significant difference between functional feeding types
(p > 0.5), regardless of whether equal variances were assumed or whether data were log-
transformed. The statistical power of these tests was low due to the small sample size.
Therefore, the results were compared with other data collected in the Housatonic River. Benthic
invertebrate tissue data are available from the Connecticut portion of the Housatonic River.
Observations from 1978 to 1998 at West Cornwall, CT (Academy of Natural Sciences of
Philadelphia 1999) found no significant differences between PCB concentrations in filter feeders
and those in predatory invertebrates. Collectively, these observations confirm that PCB
concentrations in benthic invertebrates do not vary consistently based on feeding strategy.
Nevertheless, in the following discussions, predator and shredder tissues are treated separately to
preserve the ability to observe differences in tissue contaminant concentrations.

Detailed statistical analyses of PCB and other COC concentrations (i.e., contrasts between
contaminated and reference stations, or among sediment substrate types) were not conducted due
to the small sample sizes and associated low statistical power. Instead, a graphical approach was
employed, whereby the PCB tissue burdens were examined over the study area.

Figure D.2-29 presents the distribution of tPCB concentrations by sampling location and tissue
type. With the exception of the anomalously elevated value for shredders at Station A3, all
reference samples had tPCB tissue concentrations below 1.0 mg/kg. In contrast, contaminated
locations had elevated concentrations ranging from 2 to 48 mg/kg. There was no apparent spatial
trend that could be attributed to either distance from PCB source or to sediment substrate type.
For example, tissue PCB concentrations at contaminated stations were lowest at Stations 2 and 3,
but much higher at Stations 4 and 5, even though all of these locations are similar coarse-grained
habitat upstream of the WWTP.

It is unknown whether the A3 shredder concentration greater than 30 mg/kg represents a
significant localized contamination source in the West Branch of the Housatonic River. A
duplicate analysis of the A3 shredder sample confirmed the validity of the result. Localized areas

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw ^ i 7/10/2003

-------
1

of PCB contamination in this stretch of the Housatonic River are possible due to the proximity of
Dorothy Amos Park. However, the low predator PCB concentration and the low sediment PCB
concentrations observed at this site suggest that contamination is unlikely to extend over a large
spatial scale.

Figures D.2-30 and D.2-31 present Aroclor compositions as identified by the laboratory in
shredders and predators across the PSA. The shredder tissue samples at contaminated locations
exhibited a high and relatively constant percentage of Aroclor 1260 (approximately 50 to 70%),
with the majority of the remaining PCB resembling Aroclor 1254. A similar composition was
observed in predator tissue samples, although the percentage reported as Aroclor 1260 was
somewhat more variable (45 to 100%). Reference locations exhibited different and more
variable patterns of Aroclor compositions. Upstream reference locations (Al, A3) tended to
exhibit a higher percentage reported as Aroclor 1242, whereas the Threemile Pond reference
(R4) exhibited a pattern similar to some of the PSA.

Although the tissue data confirm that PCBs in sediment of the PSA are bioavailable, the spatial
patterns in tissue chemistry are not consistent with the patterns in sediment collected from nearby
sites during the benthic macroinvertebrate sampling. This difference cannot be explained on the
basis of bioavailability related to TOC or particle size differences because Stations 1 through 5
exhibit low TOC and relatively coarse particle size. Lipid normalization of PCB concentrations
in invertebrates also did not explain the observed differences. Therefore, it appears that PCBs in
the upstream portions of the PSA are less bioavailable than in downstream areas. This pattern
has also been observed in the fish tissue data and therefore appears not to be a sampling or
analytical artifact.

The duration of the in situ Lumbriculus bioaccumulation test (7 days) may not have been
sufficient for tissue PCB concentrations to reach equilibrium with the surrounding sediment;
therefore, tissue contaminant concentrations should be interpreted with caution. For many
hydrophobic substances, a month or longer is required to reach quasi-equilibrium with
surrounding sediment in bioaccumulation tests. For example, McLeese et al. (1980) indicated
that the polychaete worm Nereis virens did not attain equilibrium concentrations during 32 days
exposure. Brunson et al. (1998), in discussing the results of laboratory and field

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

-------
1

bioaccumulation studies of Lumbriculus variegatus, stated that "laboratory exposures longer than
28 days may be required to reach equilibrium for super-hydrophobic chemicals" (Ankley et al.
1992). PCB concentrations measured in the toxicity tests conducted in the PSA were lower than
those in the field samples, supporting the conclusion that the 7-day tests may not have reached
equilibrium concentrations. For example, wet weight PCB concentrations in Lumbriculus at
contaminated locations (Stations 4, 5, 7, 8) ranged from 1.5 to 4.5 mg/kg (Figure D.2-32),
whereas the field samples ranged from 5 to 45 mg/kg at the same locations. The lower tissue
concentrations and biota-sediment accumulation factors (BSAFs) observed for Lumbriculus
suggest that the effects endpoint of 7-day laboratory Lumbriculus survival may not be a
conservative indication of field responses.

BSAFs were calculated for the Lumbriculus bioaccumulation data by normalizing both the
measured 7-day synoptic sediment PCB concentration to sediment TOC and the tissue PCB
concentrations to lipid content. The lipid contents observed in the Lumbriculus tissue samples
were consistently around 1%. BSAFs were not calculable for two samples because of the lack of
detected TOC in sediment. The remaining four samples yielded BSAF values ranging from 0.8
to 6.1 g-lipid/g-OC. These values are consistent with equilibrium partitioning theory for PCBs,
which yields a BSAF of approximately 2 (Parkerton 1993; McFarland 1994). Uncertainty
inherent in the measurement of tPCBs, organic carbon, and lipids potentially reduces both
accuracy and precision in the BSAF estimates.

D.2.5.4 Distribution and Concentration of Other COCs

Concentrations of some other COCs, such as Appendix IX pesticides, are also available in the
tissue chemistry data set. However, because pesticides were screened out of the benthic ERA
(based on detection limit considerations and conservative tissue concentration screening), these
contaminants were not considered further. Although PAHs and metals were screened in as
COCs, the tissue analyses did not include these parameters due to the lack of sufficient sample
volume.

Dioxin and furan concentrations were measured in a subset of the benthic invertebrate kick-net
samples collected in 1999. At nine stations (i.e., all benthic sampling stations except A2, 5, 6,
and 8), a predator tissue and/or a shredder tissue sample was analyzed for a suite of 17 individual

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC T\ 7/10/2003

-------
1 dioxin and furan congeners. Of the data points collected (n = 187), less than 10% were above

2 respective detection limits, which ranged from 4.7 to 2,500 ng/kg. None of the concentrations of

3 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) were measured above detection limits. The

4 most frequently detected congener was 1,2,3,4,6,7,8-heptachlorodibenzofuran, with four of nine

5 concentrations above detection limits (maximum detected value = 51 ng/kg). Because of the low

6 detection frequency, no meaningful quantitative analyses could be performed with these tissue

7 data, either as individual congeners or as total dioxins/furans.

8 D.2.6 Surface Water Chemistry Assessment

9 D.2.6.1 Overview

10 Surface water chemistry data have limited application to the benthic invertebrate risk assessment

11 due to the uncertainty in extrapolating from water chemistry to effects in sediment-dwelling

12 biota. Although epifauna and infauna are exposed to overlying water, either through direct

13 exposures or via cycling of water through constructed tubes, correlating these exposures with

14 effects is uncertain. The variability in water column concentrations over time adds additional

15 uncertainty. For these reasons, sediment chemistry and porewater chemistry are more commonly

16 used as abiotic exposure metrics for benthic invertebrates.

17 Comparison of water chemistry data to toxicity thresholds for invertebrates and fish was

18 conducted in the Pre-ERA screening. Surface water COCs retained were limited to PCBs and

19 dioxins/furans (Table 2.4-3 in Section 2).

20 D.2.6.2 Data Sources

21 Because of the uncertainty in relating water column chemistry to effects in benthos, the only data

22 considered relevant were those collected in conjunction with effects measurements. Unfiltered

23 overlying site water was collected in conjunction with the in situ toxicity testing (EVS 2003).

24 Samples collected in conjunction with the 7-day exposures were evaluated for tPCBs, PCB

25 congeners, PAHs, pesticides, and metals. Only tPCBs were measured for the 48-hour and 10-

26 day exposures.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-34

-------
1 Water samples were not collected as part of the benthic community sampling. Therefore, the

2 water chemistry data considered for the benthic invertebrate exposure assessment were limited to

3 the EVS (2003) results.

4 D.2.6.3 Distribution and Concentrations of COCs

5 Figure D.2-33 presents the tPCB concentrations in surface water for samples collected during the

6 WSU 7-day in situ toxicity testing. The tPCB values were calculated both as the sum of the

7 homologs and the sum of the congeners. For both calculations, only detected values were

8 included. Total PCB concentrations were nearly 2 orders of magnitude higher at contaminated

9 stations relative to reference stations. In all samples, the sum of congeners was about 70% of the

10 tPCB concentrations calculated by summing the homologs.

11 Figure D.2-34 presents tPCB measurements (as the sum of homologs) for surface water samples

12 collected in conjunction with all three in situ toxicity tests (48-hour, 7-day, 10-day). These

13 results indicate little variability in the surface water tPCB concentrations throughout the duration

14 of the in situ toxicity testing. Station 7 consistently had the highest PCB concentrations, with the

15 remaining contaminated sites having concentrations half as large.

16 The only other COCs considered in surface water were dioxins/furans. No dibenzo-p-dioxins

17 were detected in the 7-day in situ water column chemistry samples; however, dibenzofurans were

18 detected (tetra-, penta-, and hexa- compounds only). Figure D.2-35 illustrates that the spatial

19 patterns in furans concentrations are similar to those of PCBs, with contaminated sites having

20 significantly elevated furans concentrations. Furans were analyzed for, but not detected, at

21 Station Al.

22 Water quality conditions during the laboratory and in situ sediment toxicity tests were generally

23 within acceptable ranges. Although occasional dissolved oxygen and temperature measurements

24 were recorded outside of specified tolerances, these excursions were rare and would have had no

25 adverse impact on the test organisms. Total ammonia was monitored in the overlying water

26 during sediment toxicity testing, and concentrations were not high enough to be of concern for

27 any treatments. Some high ammonia concentrations were detected in some of the porewater

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-35

-------
1 used in the TIE portion of the study, but the TIE results overall were not indicative of ammonia

2 toxicity.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-36

-------
1

D.3 EFFECTS ASSESSMENT

The following effects assessment for benthic invertebrates presents information related to
toxicological and biological responses to COCs in aquatic media of the PSA. These responses
were predicted (using sediment or tissue benchmarks) or observed (using laboratory and field
investigations of invertebrate toxicity and benthic community structure). For some measurement
endpoints, the effects assessment considered individual COCs; in other cases the entire
contaminant mixture to which organisms are exposed is considered.

The assessment of potential environmental and biological impacts of contaminated sediment has
traditionally relied upon two approaches:

¦ The comparison of sediment and tissue chemistry data to sediment quality values
(SQVs) and tissue benchmarks.

¦ Laboratory and/or field sediment toxicity tests (and/or benthic community
evaluations).

Both approaches were used in this ERA; however, emphasis was placed on the latter due to
greater site-specificity. The purpose of SQVs is to establish sediment benchmarks for
comparison with site sediment chemistry to gauge the potential toxicity of a contaminant to
aquatic biota. Site-specific effects benchmarks were also derived, against which exposure data
could be compared.

D.3.1 Sediment Toxicity

Wright State University (WSU) conducted site-specific toxicity testing of Housatonic River
sediment (EVS 2003). Seven stations were sampled for in situ and laboratory toxicity analyses.
Of these, 6 were within the group of 13 benthic sampling locations where the benthic
macroinvertebrate sampling was conducted (Figure D.2-1).

This section summarizes the results of the toxicity tests, assesses the magnitude of toxicological
impact at each station, and defines the range of thresholds considered for the development of
HQs. Sediment PCB concentrations in each toxicity treatment are presented for information
purposes; both the single "most synoptic" PCB chemistry concentration (used in Attachment D.5

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC D 3 7

-------
1 for concentration-response modeling) and the median of all relevant data (used in this appendix)

2 are presented.

3 The sediment toxicity tests were conducted over a range of tPCB concentrations, with station

4 locations selected based on previous chemical characterization data from the Housatonic River.

5 The objective of the study was to investigate biological responses over a gradient in tPCB

6 concentrations, in order to develop a concentration-response relationship, rather than to focus on

7 biological responses in specific areas of the river. For this reason, the pattern of tPCB

8 concentrations in the sediment toxicity studies (i.e., increase in concentrations with distance

9 downstream) does not match the general trend across the PSA (see Figure D.2-14).

10 D.3.1.1 Methods

11 Test protocols, study methods, and other detailed documentation for the Housatonic River

12 sediment toxicity testing program are presented in EVS (2003) and in the SIWP (WESTON

13 2000). A brief summary is provided below.

14 D.3.1.1.1 Laboratory Toxicity

15 Laboratory testing was conducted on two freshwater species using long-term (chronic) testing

16 protocols endorsed by EPA (EPA 2000):

17 ¦ Chronic 42-d bulk sediment test using a freshwater amphipod (Hyalella azteca).

18 Duration and endpoints were 28-d, 35-d, and 42-d for survival; 28-d and 42-d for

19 growth (dry weight); 35-d and 42-d for reproduction.

20 ¦ Chronic 43-d sediment test using a freshwater midge (Chironomus tentans). Duration

21 and endpoints were 20-d for survival and growth, and 23-d to 43-d for mortality and

22 emergence.

23 D.3.1.1.2 In Situ Toxicity

24 There are no formalized standard methods for toxicity testing of aquatic organisms in the field.

25 Burton (1996) summarized the literature on the available methods, which are designed to

26 minimize the uncertainty of extrapolation of responses from laboratory conditions to field

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC D 3 8 7/10/2003

-------
1

conditions. Four test organisms were evaluated in short-term tests to complement the laboratory
testing program:

¦ 48-h toxicity test using a freshwater cladoceran (Daphnia magna, 48-h old).
Endpoint: survival.

¦ 48-h and 10-d toxicity test using a freshwater midge (Chironomus tentans, 8 to 12
days post-hatch). Endpoint: survival.

¦ 48-h and 7-d toxicity test using a freshwater oligochaete worm (Lumbriculus
variegatus, multiple ages). Endpoint: survival and tissue bioaccumulation.

¦ 48-h and 10-d toxicity test using a freshwater amphipod (Hyalella azleca, 7 to 14
days old). Endpoint: survival.

The in situ tests were performed during low-flow periods and were conducted using two
exposure configurations. First, organism exposures were limited to the water column only, via
placement of the chamber in the top tray of the in situ basket. Second, interaction with both
sediment and overlying water was simulated via placement of the chamber against the sediment
surface (with one window facing the water column). This sampling design enabled
discrimination between water-borne and sediment-associated toxicity.

D.3.1.2 Relevance of Selected Endpolnts

The organisms selected are both environmentally relevant to the Housatonic River (e.g.,
chironomid and oligochaete species were present in the river sediment) and have a large
toxicological database demonstrating their relative sensitivity to the COCs. The selected species
inhabit sediment during the life stages tested, remain relatively immobile, and have a high
potential for exposure. The test organisms selected are tolerant to a broad range of sediment
physico-chemical characteristics (EPA 2000). This was a particularly important consideration in
the determination of toxicity to the benthos of the Housatonic River, given the changes in
distribution of sediment total organic carbon (TOC) and grain size upstream and downstream of
the Pittsfield WWTP (i.e., relatively coarse grained, low-TOC substrate above the WWTP and
finer-grained, higher-TOC substrate below the WWTP).

In general, Lumbriculus spp. and Chironomus spp. have been shown to be less sensitive to both
single chemicals and complex mixtures than the other organisms used in the WSU study, such as

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-39

-------
1 D. magna and H. azteca (EPA 2000). These findings are in agreement with the pollution

2 sensitivity values used in the calculation of a modified Hilsenhoff Biotic Index, for which

3 Lumbriculus spp. and Chironomus spp. are assigned pollution tolerance values of 8 and 10,

4 respectively (indicating high tolerances to pollution) (Bode 1988).

5 D.3.1.3 Particle Size Sensitivities of Test Organisms

6 Sediment grain size is a significant variable that can impact benthic organisms, both in the field

7 and in some laboratory toxicity studies. Because the field reference sediment (Stations A1 and

8 A3) represented coarse-grained sediment, it was important to assess the potential for

9 confounding effects of particle sizes in toxicity test treatments with fine-grained sediment. A

10 literature review was conducted to document the sensitivity of the test organisms to changes in

11 particle size distributions (Attachment D.4). The review indicated that the indicator species

12 chosen for toxicity testing are quite tolerant of a broad range of particle sizes (hence their

13 selection as measurement endpoints). Therefore, comparison of all contaminated sediment

14 treatments to the reference sediment A1 and A3 was appropriate.

15 D.3.1.4 Data Evaluation Approach

16 Pairwise comparisons between treatment responses and negative control responses (i.e.,

17 hypothesis testing) were conducted using ToxCalc v.5.0.23 software. In addition, exposed

18 treatments were separately compared to each reference sediment. An a = 0.05 level of

19 significance was applied in the determination of statistically significant differences.

20 In evaluating the pairwise comparisons, Shapiro-Wilk's Test for normality and an F-Test for

21 equality of variances were first applied. When distributions were normal and homoscedastic

22 (equal variances), a Homoscedastic t-test was applied. When distributions were normal and

23 heteroscedastic (unequal variances), a Heteroscedastic t-test was applied. When distributions

24 were non-normal, appropriate transformations (e.g., arcsine square root transformation) were

25 attempted prior to rerunning the analysis. If the data remained non-normal, data were not

26 transformed and, instead, tested using non-parametric Steel's Many-One Rank test or

27 Wilcoxon's Rank Sum test.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-40

-------
1

Although assessment of statistical significance was informative, there are a number of reasons
why statistical significance should not be the only method used to evaluate toxicity test results.
For example, the level of statistical significance is greatly impacted by the magnitude of
variability in the negative control treatments. Testing that exhibits low variability in the negative
control results in small effect sizes being deemed "significant" even if these differences are not
ecologically significant. Conversely, high variability could mask ecologically relevant toxicity.
An example of the latter is the results of the 48-h and 10-d testing of Hyalella, which yielded
approximately 50% reductions in survival without statistical significance being achieved. For
this reason, this section emphasizes comparisons based on effect sizes for various endpoints that
are deemed "ecologically relevant." Effect sizes of 20% and 50% were selected as indicators of
moderate and major toxic effects; the use of the EC2o approach is consistent with regulatory
guidance from other jurisdictions (e.g., BC MELP 1997).

Two general approaches to the integrated assessment of toxicity were employed:

¦ A station-by-station assessment that considered the toxicity endpoints for each station
(relative to negative controls), without considering the comparative "exposure" for
stations in different locations (i.e., exposed sites versus reference sites). This
assessment allowed for the identification of any localized areas of toxicity observed
within the broader study design.

¦ A comparative integrated toxicity assessment (i.e., assessment of relative risks) was
conducted. This assessment considered the endpoints in relation to reference station
responses (i.e., results from Stations A1 and A3). The assessment of relative risks is
considered to carry greater importance in the overall assessment of risk because the
approach considers the local physicochemical factors in Housatonic River sediment
that may mediate toxicity.

D.3.1.5 Results

Table D.3-1 presents the results of statistical tests of significance for most toxicity test endpoints
(in situ water-only exposure results are excluded; few responses were observed in these tests). In
some cases, test statistics could not be calculated due to 100% mortality observed in a treatment.
The following discussion focuses on the qualitative findings and comparisons; Table D.3-1
provides details of specific statistical comparisons. Toxicological responses for each test type
and treatment are also presented graphically (Figure D.3-1 to Figure D.3-11). PCB
concentrations are presented in two ways, based on the two approaches discussed in the exposure

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-41

-------
1

assessment (Section D.2.4.6.2): (1) the values in bold represent the median of all spatially and
temporally relevant PCB concentrations at each station; and (2) the values in italics represent the
single PCB concentration measurement taken closest to the effects endpoint (i.e., most synoptic
concentration).

D.3.1.5.1 Laboratory Hyalella Test

Hyalella toxicity test results are shown in Figures D.3-1, D.3-2, and D.3-3 (for survival, growth,
and reproduction, respectively). The survival and reproduction endpoints exhibited strong
concordance, whereas variations in growth rates were minor among treatments/controls and
showed no obvious pattern. No statistically significant differences were observed for
comparisons of either 28-d or 42-d dry weight to the reference sediment at A1 or A3, or to the
negative (Trout Farm) control sediment.

For survival and reproduction, the negative control performance was similar to the treatments
from both reference locations (Al, A3). Station 4 exhibited a modest but statistically significant
reduction in survival and reproduction (p < 0.05) relative to the negative control, for all three
exposure durations (Table D.3-1). Significant differences in survival were also observed
between Station 4 and reference sediment A3, for all exposure durations.

In contrast, the remaining "exposed" treatments (i.e., the three most downstream locations) had
larger reductions in both survival and reproduction. Survival and mean young reductions were
statistically significant relative to the negative control and to reference sediment at Al and A3, at
all three exposure durations (Table D.3-1). Station 7, corresponding to the highest PCB
concentration, exhibited 100% mortality by Day 28, and therefore sublethal endpoints for this
treatment could not be derived.

Both survival and growth measurements were made at multiple time periods throughout the 42-
day test. However, the patterns of statistical differences observed across treatments generally
remained the same for these endpoints over the duration of the test (Figures D.3-1 and D.3-2).
The patterns of toxicity were similar for the reproductive endpoints, whether the mean
reproduction metric was standardized to number of females (i.e., young produced per female) or
left unstandardized (i.e., total number of young).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-42

-------
1

D.3.1.5.2 Laboratory Chironomid Test

Chironomus toxicity test results are shown in Figures D.3-4 and D.3-5 (for survival/emergence
and growth, respectively). For this test, three separate controls were conducted (a-cellulose [C],
florissant [F], and Trout Farm sediment [T]); all yielded similar results. Trout Farm survival and
emergence were slightly higher than the other two controls, but Trout Farm growth was the
lowest of the three treatments. Statistical comparisons were made to the T-control treatment,
because Trout Farm sediment was used in other test endpoint comparisons, and because
differences among controls for the chironomid test were not large. As with the Hyalella test, the
two reference stations (Al, A3) yielded test performance similar to that of controls. Survival,
dry weight, and emergence did not differ significantly (p > 0.05) between references and
controls. However, ash-free dry weight was significantly reduced at reference location Al.

All four "exposed" locations tested had large and statistically significant reductions in all
Chironomus test endpoints (Table D.3-1; Figures D.3-4 and D.3-5). Both Stations 7 and 8A
exhibited 100% mortality by Day 20, such that sublethal endpoints could not be evaluated for
these treatments. Stations 4 and 8 exhibited mean survival less than 10%, and the surviving
animals exhibited near zero growth.

D.3.1.5.3 In Situ Hyalella Test

Hyalella were tested in situ over two exposure regimes (water column only, and sediment plus
water column) and two different time periods.

For the 48-h exposure (Figure D.3-6), survival in four of six treatments (including reference
sediment at Al and A3) was similar to the negative control, for both water column and sediment
exposures. However, two treatments (Stations 5 and 7) yielded significant reductions in survival
during the 48-h sediment exposures. Despite the high median PCB concentration measured at
Station 8 (77.2 mg/kg), no reduction in survival was observed in either water column or sediment
in situ exposures at this station.

Survival of Hyalella during the 10-d in situ exposures (Figure D.3-7) was generally comparable
to the 48-h exposures, except that significant reductions in survival were observed at the highest
median PCB concentration (77.2 mg/kg) after 10 days. Part of the apparent difference in

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n /I ^ 7/10/2003

-------
1

observed toxicity for the high PCB exposures in sediment may be attributable to inadequate
exposure time in the 48-h treatment. In the longer-term exposure, there was near complete
(>95%) mortality after 10 days in the two treatments with PCB concentrations greater than 50
mg/kg (Stations 7 and 8). Modest (<40%) reductions in survival were also observed for water-
only exposures at the three most downstream stations tested, but these were not statistically
significant.

D.3.1.5.4 In Situ Chironomid Test

No significant differences in mean chironomid survival were observed following short-term (48-
h) exposures to sediment and/or the water column in the in situ exposures (Figure D.3-8).
However, as with the Hyalella tests, toxicity was clearly evident following longer sediment
exposure durations (10-d) (Figure D.3-9), particularly at stations with high PCB concentrations.
The two stations with tPCB concentrations of over 50 mg/kg yielded 80-90% mortality after 10
days exposure to sediment.

D.3.1.5.5 In Situ Daphnid Test

In spite of the short test duration, Daphnia magna exposed to PSA sediment for 48 hours in situ
yielded significant reductions in survival (Figure D.3-10). The stations with the highest PCB
concentrations (7 and 8; both >50 mg/kg PCB) had 100% mortality. There was an approximate
50%) response at Station 5 (median of 8.3 mg/kg), but this was not statistically significant due to
high variability. No significant responses were observed for water-only exposures. The pattern
of response was similar to that observed for 10-d Hyalella exposures (Figure D.3-7), providing
strong concordance in toxic responses.

D.3.1.5.6 In Situ Lumbriculus Test

In both water and sediment exposures, no adverse effects on survival were observed for
Lumbriculus (Figure D.3-11). The exposure duration for this test was brief (survival assessed at
48 hours). This species is also known to be resistant to contaminant-induced mortality (hence its
use as a preferred bioaccumulation organism).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-44

-------
1

D.3.1.6 Interpretation of Test Results

The station-by-station assessment is presented in Tables D.3-2 and D.3-3. Table D.3-2 provides
the summary of in situ toxicity endpoints, and Table D.3-3 summarizes the laboratory studies.
These tables provide a means of evaluating the concordance among toxicity test endpoints. The
degree of confidence in the assessment of potential for ecological harm is in large part a function
of the degree of concordance observed among endpoints. Each endpoint was rated as exhibiting
negligible, moderate, strong, or very strong evidence for toxicological effects to freshwater
organisms, based on the effect magnitude observed relative to the negative control(s). PCB
concentrations in sediment and in oligochaete tissues are also provided as an indicator of gross
exposure at each station.

For the in situ exposures (Table D.3-2), there was negligible toxicity at both reference locations
(Al, A3) and the most upstream "exposed" location with the lowest PCB concentration (Station
4). However, toxicity was evident in multiple tests for the remaining three exposed locations
(Stations 5, 7, 8), with the magnitude of response generally greatest at the two most downstream
locations (with the highest PCB concentrations). Although modest effects were observed in
some water-column exposures, the most pronounced effects were observed in the sediment-
exposure treatments. Due to the multiple severe responses observed in the in situ sediment
exposures (Daphnia, Hyalella, and Chironomus), Stations 7 and 8 were assigned a toxicity rating
of "high." Station 5 was assigned a rating of "moderate," because the toxic responses, although
frequent, were not as large in magnitude.

For the laboratory exposures (Table D.3-3), the overall frequency of toxic responses was greater
than for in situ exposures. This finding was consistent with the longer test duration and
measurement of more sensitive (sublethal) endpoints used in the laboratory tests. Unlike the in
situ exposures, there were some indications of reduced endpoint performance (relative to
negative control) in both reference locations (Al, A3). These responses consisted mainly of
marginal reductions in Hyalella reproduction and chironomid endpoints. The endpoints also
exhibited large variability, so the source of the observed 20% reductions is unclear. However,
the magnitudes of responses were much greater at all four exposed locations. Significant
chironomid toxicity was observed at all four exposed locations, and Hyalella at these locations

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-45

-------
1

either died by Day 28 or exhibited strong reproductive effects. These findings support a
conclusion of a "high" toxicity rating for all four exposed locations.

A comparative approach was also applied to help distinguish background responses from those
observed at exposed locations. In this assessment, comparisons were made not to the negative
control, but to the two reference locations (Al, A3), to account for site-specific background
responses in Housatonic River sediment. Despite the modest toxicity observed in the laboratory
toxicity endpoints for Al and A3, the comparative assessment (Table D.3-4) still indicated a
moderate to strong incremental toxicity associated with contaminated PSA sediment. The three
most downstream stations (7, 8, and 8A) had "high" ratings due to the consistency and severity
of toxicity observed for numerous endpoints.

D.3.1.7 Conclusions

The assessment supports a conclusion of impact of contaminated sediment in the PSA to
Housatonic River benthos. The probability of impact was judged to be high, based on the
following:

¦ The large number of endpoints indicating toxicity.

¦ The severity of the responses observed.

¦ The responses in the in situ exposures (that reflect site-specific conditions).

¦ The observations of toxicity in multiple species even for short-term exposures and for
lethal endpoints.

Based on toxicity testing results, EVS (2003) concluded that PCBs are a major stressor to aquatic
biota in the Housatonic River PSA. Examination of the PCB concentrations relevant to the
toxicity tests (as shown in Figures D.3-1 to D.3-11) support this conclusion. Sediment PCB
concentrations associated with the largest toxicity effects were also well above sediment quality
values for PCBs (discussed in Section D.3.5). There were a few observations that were not
consistent with the hypothesis of PCB concentrations causing the observed toxicity. The high
variability in measured PCB concentrations over small spatial and temporal scales may explain
some of these atypical findings. Relationship of PCBs to toxicity endpoints and other measures
of ecological response are discussed in detail in the Risk Characterization (Section D.4).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n /I A 7/10/2003

-------
1

D.3.2 Concentration-Response Analysis -Toxicity Endpoints

A statistical assessment was conducted to quantify the observed relationship between toxicity
endpoints and COC concentrations measured concurrent with the biological tests. The objectives
of the concentration-response assessment were to:

¦ Determine whether a concentration-response relationship existed for some or all
toxicity endpoints.

¦ Where concentration-response relationships were observed, determine whether the
data supported a threshold COC concentration for biological effects in invertebrates.
Such thresholds are required for extrapolation of effects to portions of the river
without biological data.

¦ Evaluate the degree of concordance among various toxicological endpoints with
respect to concentration-response relationships and/or effects thresholds.

This evaluation was based on two approaches to processing the sediment chemistry data.

¦ First, using the assumption that exposures are best represented by combining multiple
chemistry sampling events to effects measurements.

¦ Second, assuming that the individual sediment concentration measured at the
termination of a given test (i.e., most synoptic individual concentration) provided the
best estimate of the average exposures to organisms during that test. The output from
this exercise is presented in Attachment D.5.

The use of these two approaches acknowledges that there is some uncertainty with respect to
concentration-response derivations in the benthic invertebrate ERA, primarily because of small-
scale variability in sediment COC concentrations. Performing the evaluation using these
assumptions does not bias the assessment, and strong concentration-response relationships were
still apparent, despite the uncertainty introduced by spatial or temporal variability in sediment
PCB concentrations.

The assessment focused on the relationship between PCBs and toxicity endpoints. The results of
several measurement endpoints for benthos (particularly the suite of toxicity tests) indicated a
high probability that PCBs were a causal agent for toxicity to benthic invertebrates within the
Housatonic River PSA. Other COCs were assessed in a semi-quantitative manner and are
presented at the end of the section.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-47

-------
1

The following methods were used to evaluate PCB concentration-response relationships for site-
specific toxicity data:

¦ Each toxicity endpoint was investigated individually using conventional descriptive
statistics that related degree of effect to PCB concentrations. The results of each
series of laboratory or in situ toxicity tests were used to calculate metrics for each test
(e.g., LC50, IC20, NOAEL, LOAEL), based on PCB concentrations measured
concurrent with the tests. This provided numerical estimates of tPCB concentrations
likely to be associated with specific effects. Although no observed adverse effect
level (NOAEL) and lowest observed adverse effect level (LOAEL) values were
calculated and presented herein, the risk characterization focused on 20% and 50%
effect levels, because statistical significance may not represent ecological significance
or effect magnitude.

¦ The toxicological endpoints were integrated using a general linear modeling approach
to identify similarities and differences in concentration-response relationships across
species and endpoints.

D.3.2.1 Approach 1: Calculation of Individual Toxicity Test Endpoints

Point estimates were calculated for each toxicity test, including LC20 and LC50 values for
survival endpoints, and IC2o and IC50 values for sublethal response endpoints (e.g., growth,
reproductive success). Effect sizes of 20% and 50% were selected to be consistent with the
evaluations made in the station-by-station toxicity assessment (Section D.4.4), and to represent
"moderate" and "major" indications of response. LC values represent "lethal concentrations" at
which a specified mortality rate occurs (i.e., 20% or 50% lethality). EC values and IC values
represent "effect concentrations" or "inhibition concentrations" at which a specified level of
sublethal effect occurs (e.g., 20% or 50% reduction in reproduction or growth). NOAEL and
LOAEL values were also determined. All endpoints were calculated using (1) the sediment PCB
concentrations measured in conjunction with each testing event, and (2) using median chemistry
from the vicinity of the station.

Test endpoints were calculated using the ToxCalc program (Version 5.0.23; Tidepool Scientific
1994). LC50 and LC20 values were calculated using the Probit method whenever possible. In
some instances, Chi-square values exceeded the critical value, indicating that the data were not
well suited to the Probit method. Therefore, LC50 values were also calculated using the Trimmed
Spearman-Karber (TSK) method (LC20 values cannot be calculated by TSK). These TSK LC50

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "n /I Q 7/10/2003

-------
1 values have been reported for comparative purposes to assess the sensitivity of the findings to the

2 interpolation method. IC50 and IC20 values were calculated using the linear interpolation method.

3 The 95% confidence limits were also determined for each point estimate, whenever possible.

4 After testing for data normality and homogeneity of variance, NOAEL and LOAEL values were

5 determined using appropriate parametric tests (i.e., ANOVA with Dunnett's or Bonferroni's I-

6 tests) or non-parametric tests (i.e., Steel's Many-One Rank test). Hypothesis tests (to determine

7 NOAEL and LOAEL values) were performed using a probability of a = 0.05.

8 For each data set, test endpoints were calculated three ways:

9 ¦ Comparison to Negative Control - Data from all six Housatonic River stations,

10 including the two upstream reference and four "contaminated" stations, were included

11 for comparisons to the negative (clean) control.

12 ¦ Comparison to Reference Station A1 (Dalton Reference) - Data from the four

13 contaminated Housatonic River sediment were compared to Station Al. Laboratory

14 controls and the other potentially impacted reference station were excluded.

15 ¦ Comparison to Reference Station A3 (West Branch near Confluence) - Data

16 from the four contaminated Housatonic River sediment were compared to Station A3.

17 Laboratory controls and the other reference station were excluded. Although the

18 statistical comparisons to A3 are somewhat redundant with the comparisons to Al,

19 the few observations of elevated PCB chemistry data at A3 (sediment and tissue)

20 warranted separate comparisons.

21 D.3.2.1.1 Comparison to Negative Laboratory Controls

22 Metrics from the 48-h and 10-d in situ toxicity tests and the laboratory toxicity tests are

23 summarized in Tables D.3-5 to D.3-7. Figures D.3-12 and D.3-13 illustrate the relationships

24 between the point estimates (LC20/IC20 and LC50/IC50) calculated for each toxicity test endpoint,

25 based on comparisons to the control treatment. In Figure D.3-12, these values are arranged in

26 order of increasing LC50/IC50 value, whereas Figure D.3-13 presents the same point estimates in

27 order of increasing test duration. Overall, the thresholds for occurrence of adverse effects are

28 observed at a range of concentrations. This variability is not unexpected given the range of

29 species, endpoints, and test durations. Nevertheless, the majority of endpoints exhibited 50%

30 effect levels in the range of 1 to 30 mg/kg tPCBs.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-49

-------
1

Analysis of the impact of test duration on magnitude of response is complicated by differences in
the sensitivity of biological endpoints. However, with the exception of the H. azteca dry weight
endpoints, there was a tendency for the sensitivity to increase with increasing test duration
(Figure D.3-13). For H. azteca and C. tentans, which were tested for multiple durations and
endpoints, responses were greater for 10-d and chronic exposures compared to the 48-h
exposures.

Interpretation of statistical comparisons to negative laboratory controls may have some
limitations because of the differences in test performance between negative controls and
reference sediment. There were several instances for which responses were observed in
sediment at one or both of the reference locations, resulting in calculation of very low effects
thresholds based on comparison to negative controls. Because differences in responses between
negative control and reference sediment cannot be explained on the basis of PCB concentrations
(with the exception of A3), it is appropriate to make comparisons between contaminated stations
and reference locations.

D.3.2.1.2 Comparison to Reference Station A1

Metrics from the 48-h and 10-d in situ toxicity tests are summarized in Table D.3-5; survival was
the only endpoint measured. Overall, there were few major differences between the negative
controls and the reference sediment for the 48-h and 10-d metrics; therefore, the LC20 and LC50
results were similar to those presented above for comparisons to negative controls.

The LC50 values for all in situ endpoints except 48-h C. tentans and 48-h L. variegatus fell
within the range of 6 to 30 mg/kg tPCBs. This range is similar to that observed based on
comparisons to the negative control (as well as to the other reference station, A3). Furthermore,
there was consistency in results for different statistical calculation methods (i.e., calculations
using both the Probit model and Trimmed Spearman-Karber models). These findings indicate a
high probability (and magnitude) of PCB toxicological responses within this concentration range.

The lack of mortality in the 48-h C. tentans and 48-h L. variegatus tests may be attributable to
either lack of organism sensitivity, such as the relative pollution tolerance of L. variegatus, or to

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-50

-------
1

short test duration. C. tentans exhibited significant toxicity when test duration was increased
beyond 48-h in both in situ and laboratory exposures, which supports the latter explanation.

Metrics from the H. azteca life cycle tests are summarized in Table D.3-6, and metrics from the

C. tentans life cycle tests are summarized in Table D.3-7. The lowest PCB concentration
reported for the contaminated sediment was 4.56 mg/kg tPCBs, so any point estimates lower than
that value required extrapolation using ToxCalc. These extrapolated values should be treated
with caution but have been included here to provide information about relative sensitivity.
Similar to the negative control comparisons, statistics for the H. azteca survival metrics were
generally consistent from Day 28 through 42. Point estimates for the 43-d C. tentans life cycle
test were confounded by the significant mortality in the treatments from all four "contaminated"
stations, resulting in mean 20-d survival ranging from zero to 7.5%. Therefore, the NOAELs
were reported as <4.56 mg/kg PCB and the unbounded LOAELs were 4.56 mg/kg PCB.

Overall, in spite of a few differences between negative control and A1 performance, the PCB
concentrations resulting in toxicity fell within comparable ranges for most endpoints (i.e., 3 to 25
mg/kg tPCBs for an effect magnitude of 50%) (Figure D.3-14). On the whole, the findings
suggest that, all other factors being equal, tPCB concentrations of 30 mg/kg or higher are
expected to cause significant effects to nearly all of the biological endpoints studied in this
program. Effects at lower concentrations (i.e., 3 to 10 mg/kg tPCBs) are also expected for
sensitive species and endpoints.

D.3.2.1.3 Comparison to Reference Station A3

Comparisons to Station A3 yielded responses that were qualitatively similar to those obtained for
the comparisons to both Station A1 and the negative control. These results are summarized in
Tables D.3-5 to D.3-7, and illustrated in Figure D.3-15. The concentrations associated with an
effect size of 50% fell mainly within the range of 3 to 30 mg/kg tPCBs. More than half of the
LC50/EC50 values were below 10 mg/kg tPCBs.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-51

-------
1

D.3.2.1.4 Derivation of a Site-Specific MATC for tPCB in Sediment

Overall, although there are small differences in the toxicity threshold values calculated using
different statistical methods (i.e., choice of extrapolation model or choice of reference sediment),
the data indicate an increase in the frequency and magnitude of adverse biological responses with
increasing sediment tPCBs concentration. The following ranges of tPCBs concentrations and
associated responses were developed, based on comparisons of contaminated station responses to
reference stations:

¦ <3 mg/kg - Some sensitive endpoints exhibited apparent responses, but the
magnitude of responses was not large. These subtle responses were difficult to
evaluate due to statistical power limitations, caused in part by the high variability in
some treatments.

¦ 3 to 10 mg/kg - Numerous endpoints indicated ecologically significant responses,
with many LC50/EC50 values falling in this range. Statistically significant responses
were observed in most Hyalella and Chironomus life-cycle endpoints at 4.56 mg/kg.

¦ 10 to 30 mg/kg - Nearly all toxicity endpoints indicated large (>50%) responses
relative to reference stations. The only endpoints that did not exhibit large responses
in this concentration range were either growth endpoints or were short-term (48-h)
tests and/or tests using tolerant species.

¦ >30 mg/kg - The concentration-response analyses indicated that most survival and
reproduction endpoints exhibited very large effects at these concentrations, with
complete mortality of some species.

This analysis is empirically based, and only tests for correspondence of tPCBs concentrations
with biological effects. Though not definitive proof of dose/response causality, the results can be
evaluated together with the SQVs and TIE findings (Section D.3.3), thus increasing the
confidence in the conclusion that PCBs are causing adverse effects in the Housatonic River and
that the magnitude of impact increases with the PCB concentration.

This section emphasized concentration-response using the "median" sediment PCB exposure
data at each station. An alternative analysis, using only the "most synoptic" exposure data is
presented in Attachment D.5. Generally, the two approaches were in agreement (i.e., most
endpoints within a factor of 2); the "median" analysis yielded effects thresholds that were
slightly lower than the "most synoptic" method.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-52

-------
1 Threshold effects concentrations were calculated using the individual endpoint data, in order to

2 allow derivation of site-specific hazard quotients in the Risk Characterization, and to serve as

3 MATCs for downstream risk extrapolations. To calculate threshold effects concentrations, the

4 average of values from the six most sensitive endpoints was calculated for both 50% effects and

5 20% effects levels. This approach was based on the rationale that thresholds should consider

6 multiple sensitive endpoints, but should not be based on the single most sensitive endpoint. The

7 50% effects level corresponds to major impacts, for which there is a high degree of confidence in

8 a significant biological impact. The 20% effects levels correspond to lower but potentially

9 biologically significant effect sizes.

10 Calculations were performed for comparisons to negative control sediment and also to field

11 reference sediment. In general, comparisons to field references were preferred for derivation of

12 sediment MATC values, since field references account for physicochemical factors that may

13 mediate sediment toxicity.

26 Using the "median" exposure data, the 50% effect level for sensitive toxicity endpoints is

27 approximately 3 mg/kg tPCB. The analysis conducted using "most synoptic" exposure data only

28 (Attachment D.5) indicated that the 50% effect level for sensitive toxicity endpoints is

29 approximately 6-7 mg/kg tPCB, and that the 20% effect level for sensitive toxicity endpoints is

30 approximately 3 mg/kg tPCB. Based on these results, 3 mg/kg tPCB was selected as the site-

31 specific threshold for sediment tPCB.

Summary of 50% and 20% Effects Levels

¦ Comparison to Negative Control - The mean of the lowest six 50% effects levels
was 1.3 mg/kg tPCB. The mean of the lowest six 50% effects levels was 0.1
mg/kg tPCB.

¦ Comparison to Reference A1 - The mean of the lowest six 50% effects levels
was 3.5 mg/kg tPCB. The mean of the lowest six 50% effects levels was 0.9
mg/kg tPCB.

¦ Comparison to Reference A3 - The mean of the lowest six 50% effects levels
was 3.3 mg/kg tPCB. The mean of the lowest six 50% effects levels was 0.9
mg/kg tPCB.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-53

-------
1

D.3.2.2 Approach 2: General Linear Model of Concentration-Response

The assessment of individual endpoints is sensitive to test variability, which can mask broader
trends in toxicity of PCBs. Therefore, a supplemental approach was applied that combined the
toxicity results from various endpoints to identify the overall trend(s) in concentration-response
observed. The endpoints for all toxicity tests were standardized so that the response variables
were equivalent (i.e., responses represented proportion of their control mean response). This
transformation of all endpoints to the relative performance proportion (RPP) values standardized
results from different toxicological endpoints to similar ranges and facilitated the search for a
single unified model among all endpoints. The negative control selected for the concentration-
response modeling was the Trout Farm reference sediment, and the field reference responses
were included as low exposure treatments in the model.

The linear model was run on the following subset of indicator toxicity endpoints:

¦ 10-d Chironomus survival (in situ, sediment exposures)

¦ 10-d Hyalella survival (in situ, sediment exposures)

¦ 20-d Chironomus survival (laboratory exposures)

¦ 43-d Chironomus emergence (laboratory exposures)

¦ 48-h Daphnia survival (in situ, sediment exposures)

¦ 42-d Hyalella survival (laboratory exposures)

¦ 42-d Hyalella growth (laboratory exposures)

¦ 42-d Hyalella young-per-female (laboratory exposures)

The analysis was restricted to these eight indicator endpoints primarily to minimize redundancy
and multicolinearity in the endpoints. For example, the 42-d Hyalella test included
measurements at 28-d and 35-d that correlated strongly with 42-d measures; therefore, only the
42-d measures were considered. Chironomus growth was also eliminated because a large
number of replicates exhibited no value due to 100% mortality in these replicates, which
precluded growth endpoint determination. Furthermore, some endpoints were not included
because of an obvious lack of observed effects (e.g., no responses in 48-h Lumbriculus survival
endpoint). Similarly, water column exposure endpoints for in situ tests were not included
because the toxicity endpoints were more pronounced in the sediment exposure treatments. In
addition, consideration was given to ensuring that the eight endpoints chosen for statistical
analysis represented a range of species, test durations, and endpoint types.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-54

-------
1

Treatment of data prior to application of the model included the following:

¦ The sediment PCB data were logio-transformed to correct for non-normality observed
in the chemistry distributions.

¦ A unit scalar was added to the sediment chemistry concentrations prior to logio-
transformation to accommodate zero results.

¦ One RPP value for Hyallella young per female (42-d test) was 1.8, indicating that this
replicate performed much better than negative controls. To avoid giving this value
undue weight in the analysis, this value was excluded as an outlier.

An evaluation of the logio-transformed PCB (median) concentrations vs. RPP data indicated that
a segmented (i.e., hockey stick) regression model was reasonable for each endpoint, although
inflection points and change in response varied among groups of endpoints. Separate hockey
stick models were pursued to accommodate the apparent differences in inflection points, slopes,
and intercepts between the following groups:

¦ Acute endpoints (10-d Hyalella survival, 10-d Chironomus survival, 48-hour
Daphnia survival).

¦ Chronic endpoints excluding 42-d Hyalella growth (20-d Chironomus survival, 43-
day Chironomus emergence, 42-d Hyalella survival, and 42-d Hyalella young per
female).

¦ 42-day Hyalella growth.

These results are summarized below, and are shown in Figure D.3-16. All parameter estimates
derived were found to be significantly different from zero (V = 0.05). Statistical summaries
from the models are provided in Table D.3-8. This section emphasized concentration-response
using the "median" sediment PCB exposure data at each station. An alternative analysis, using
only the "most synoptic" exposure data is presented in Attachment D.5.

D.3.2.2.1 Acute Endpoints

The acute endpoints (i.e., 10-d Hyalella survival, 10-d chironomid survival, 48-h Daphnia
survival) exhibited similar concentration-response relationships, and all appeared to have an
inflection point below 10 mg/kg tPCBs. Up to the inflection point, RPP values were consistently
high (i.e., close to one).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-55

-------
1

D.3.2.2.2 Chronic Endpoints

In contrast to the acute endpoints, the chronic endpoints (i.e., 20-d chironomid survival, 43-day
chironomid emergence, 42-d Hyalella survival, and 42-d Hyalella reproduction) had a lower
inflection point near 0.3 mg/kg tPCBs. Furthermore, the RPP at low PCB concentrations was
about 0.6, reflecting the differences observed between the negative control and the reference
station responses.

The data for 42-d Hyalella survival generally had stronger responses (higher RPP values) than
the other chronic endpoints. For this reason, the hockey stick model for the set of chronic
endpoints had one inflection point but separate slopes and intercepts for the two groups. Figure
D.3-16 shows the 42-d Hyalella survival endpoint separated from the other endpoints.

D.3.2.2.3 Hyalella Growth

The 42-d Hyalella growth endpoint exhibited an inflection point similar to the acute endpoints.
However, the magnitude of response was very low, even at the highest PCB concentrations,
indicating that this was not a sensitive endpoint.

D.3.2.2.4 Summary

Overall, the linear modeling indicated that seven of eight toxicity endpoints evaluated correlate
significantly with log-transformed PCB concentration. Differences between acute endpoints and
chronic endpoints were observed; these are likely related to the greater sensitivity of chronic
endpoints in toxicity tests. Similar to Section D.3.2.1, the modeling procedure enabled the
identification of threshold tPCB concentrations in sediment. The inflection points are indicative
of a no-effect level for adverse responses. Figure D.3-16 indicates that the point at which RPP is
reduced by 50% relative to the background stations corresponds to logio values of 0.3, 0.4, and
1.2 for chronic endpoints, Hyalella survival, and acute endpoints, respectively. These values
equate to 2.5, 3.2, and 15.8 mg/kg tPCBs in sediment. These results are in agreement with the
summary of individual test endpoints provided in Section D.3.2.1. In summary, sediment tPCBs
concentrations above 3 mg/kg indicate significant adverse effects for sensitive (chronic)
endpoints, and tPCB concentrations in the 10 to 30 mg/kg tPCBs range elicit acute mortality to
multiple organisms.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC x-v r/ 7/10/2003

-------
1

As a check on the robustness of the method, the linear modeling procedure was repeated using
the individual chemistry data most synoptic with each individual effects measure. Attachment
D.5 presents the results of the linear modeling using only the "most synoptic" exposure data.
Results were qualitatively similar to the above; the analysis indicated that the threshold for tPCB
effects is likely greater than 1 mg/kg but less than 10 mg/kg. There is some uncertainty within
this concentration range due to variations in exposure and effects data, with the frequency of
adverse effects increasing toward the upper end of this range.

D.3.2.3 Relationships of Effects with Other COCs

The data for other COCs were also evaluated qualitatively to assess whether the concentrations
of these contaminants were likely to have confounded the results of the PCB concentration-
response presented above. In the in situ toxicity tests, sediment concentrations of secondary
COCs such as PAHs and metals were only measured with the 7-d treatments. A HQ analysis
comparing the COC concentrations to effects thresholds is provided in Section D.4.2; therefore
this section emphasizes trends in concentration rather than absolute values.

As indicated in Section D.3.1, the in situ toxicity tests for Stations 7 and 8 indicated major signs
of toxicity (i.e., greater than 50% inhibition of endpoints). Station 5 had moderate signs of
toxicity for multiple acute endpoints, suggesting an intermediate level of response. The
remaining stations (Al, A3, and 4) did not exhibit overt toxicity in the acute in situ exposures. If
other COCs exerted a major effect on toxicity endpoints, a large gradient in COC concentrations
would be expected, with the highest COC concentrations at Stations 7 and 8. Assessments of the
trends for the other COCs were provided in the exposure assessment (Section D.2.4.5). A
comparison of these patterns follows:

¦ PAHs - The highest individual bulk sediment PAH concentration (over 200 mg/kg)
was observed at Station 7, in conjunction with the 7-d in situ testing. The magnitude
of this value is indicative of potential incremental toxicity due to PAHs. However,
the remaining PAH data show a trend of reduced PAH concentrations with distance
downstream (Figure D.2-19). This trend is the reverse of the observed trend for toxic
responses. For example, pronounced indications of toxicity were observed for Station
8, which had relatively low concentrations of PAHs and other SVOCs. Therefore,
with the possible exception of Station 7, there is no evidence that PAHs provided a
major contribution to the observed pattern of sediment toxicity.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-57

-------
1

¦ Dioxins/Furans - These analytes exhibited a spatial trend similar to the trend in
toxicity (Figures D.2-16 and D.2-17). This is likely due to co-occurrence between
PCBs and dioxins/furans in environmental samples. Although the magnitudes of
benchmark exceedances are much lower for dioxins/furans relative to PCBs, the
potential contribution of these chemicals to toxicity cannot be ruled out.

¦ Metals - Concentrations of COC metals were highly correlated with each other, such
that the spatial trends are fairly consistent for the eight metals. Metals generally
exhibited a pattern of increasing concentration with distance downstream, which
matched the pattern of toxic responses (Figures D.2-20 to D.2-28). Most of the
increases in metal concentrations were about an order of magnitude, relative to
background, but some metals (cadmium, mercury, silver) exhibited two order of
magnitude increases at the downstream stations with high toxicity. Although this
would appear to suggest a potential confounding effect of sediment metals
concentrations, there are several lines of evidence that indicate a low risk potential for
metals.

The trends in metals concentrations match the sediment organic carbon and
particle size distributions. This is evident from a comparison of Figure D.2-2 to
Figures D.2-20 to D.2-28; the highest metals concentrations are consistently
observed at the downstream Stations 7, 8 and 8A. The metals associated with
high TOC and high percent fines in downstream areas were likely less
bioavailable due to their environmental partitioning characteristics. Metals bind
to acid volatile sulfides in organic carbon-rich sediment, making them less
available to biota.

The magnitudes of exceedance of environmental quality criteria were low for
these metals, such that increases above background do not necessarily equate
with risk.

The high toxicity observed at Station 7 was not supported by elevated metals
chemistry in the 7-d in situ chemistry data that are most synoptic with the
observed responses. Instead, the concentrations of detected metal COCs
measured at Station 7 during that study were within the range of metal
concentrations observed at the Reference Stations A1 and A3 (or were non-
detect).

- Finally, the TIE (Sections D.3.3 and D.4.3.2) did not suggest that metals were the
cause of the observed toxicity, even though the TIE treatments had the highest
bulk sediment chemistry for metals.

On the basis of the information presented above, other COCs (other than tPCBs) do not explain
the patterns of toxicity observed in the in situ toxicity tests. One possible exception is for
dioxins and furans, which correlate strongly with tPCBs.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-58

-------
1

D.3.3 Toxicity Identification Evaluations

Toxicity identification evaluations (TIEs) were conducted to broadly define the
physical/chemical characteristics of the contaminants causing observed toxic responses (EVS
2003). TIE procedures involve physical and chemical manipulation of samples. By comparing
the results of the manipulated sample to the untreated sample with respect to toxicity,
information regarding the cause of toxicity can be obtained. For instance, toxicity due to metals
such as copper, cadmium, or zinc is identified by a reduction in toxicity as a result of addition of
the chelating agent EDTA. Organic contaminant-driven toxicity is generally indicated by a
reduction in toxicity following solid-phase extraction of organics from solution. Generally, TIEs
follow an iterative process in which the results of previous manipulations are used to direct
subsequent tests. Detailed discussion of methods and results of the TIEs are provided in EVS
(2003); further discussion and interpretation of TIE results are also provided in Section D.4.3.2.
Therefore, only a brief summary is provided below.

TIEs were conducted using porewater from Housatonic River sediment to which Ceriodaphnia
dubia was exposed for 48 hours. The TIE approach showed that acute toxicity was significantly
reduced when non-polar organic compounds were removed from the porewater at the two most
contaminated sites (using pH-adjusted solid-phase extraction). Other treatments indicated mild
evidence for potential effects of other COCs, such as metals (i.e., filterable compounds);
however, the strongest TIE evidence pointed toward non-polar organic compounds. Based on a
number of lines of evidence, the general conclusion of these test results was that PCBs are the
most likely candidate as the cause of the observed toxic responses. Elevated concentrations of
PAHs (also non-polar organics) in TIE treatments mean that the implication of PCBs is not
definitive. However, Figure D.2-19 indicates that the pattern of sediment PAH concentrations is
inconsistent with PAHs being responsible for the broad pattern of toxicity observed (i.e., median
PAH concentrations decreased with distance downstream, whereas the toxic responses
increased). Sediment dioxin/furan concentrations showed a similar spatial pattern to PCBs and
cannot be conclusively ruled out as a contributor to toxicity. A more comprehensive Phase II or
Phase III TIE would be required to make a definitive conclusion. However, the indications of
PCB toxicity in the Phase I TIE are consistent with the large exceedances of sediment quality
values and water quality guidelines for PCBs observed in site media.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-59

-------
1

D.3.4 Tissue Effects Thresholds

D.3.4.1 Review of Literature

Data were compiled on PCB tissue concentrations associated with lethal or sublethal effects in
aquatic invertebrates. The purpose was to estimate threshold tissue concentrations where
adverse effects might occur in Housatonic River benthos. Because the thresholds were derived
from literature studies (i.e., not conducted specifically for the Housatonic project, or for the exact
contaminants and conditions found in the river), they were considered to be less relevant than the
site-specific toxicity studies. Nevertheless, the tissue effects assessment was expected to be
more reliable than the comparison to sediment quality thresholds (Section D.3.5) because the
former endpoint addresses the site-specific bioavailability of PCBs. The tissue contaminant
concentration assessment focused on PCBs because other sediment COCs were not measured in
the invertebrate tissue samples collected near the Triad sampling locations.

Tissue concentration effects data were gathered primarily from two databases (1) the United
States Army Corps of Engineers (USACE) on-line Environmental Residue Effects Database
(ERED; www.wes.army.mil/el/ered), and (2) Jarvinen and Ankley (1999). Literature searches
were also performed to identify additional data sources. Only studies that provided concurrent
measurement of biological effects and whole-body tissue concentrations (reported on a wet
weight or dry weight basis) were considered for estimating invertebrate PCB effects thresholds.
Studies that reported PCB tissue concentrations (with or without sediment/water exposure data)
without associated effects measurements were excluded, as were those uptake studies with
radiolabeled compounds that did not report tissue concentrations on the basis of weight.

The review focused on data for Aroclors 1260 and 1254, in addition to tPCBs. Because tissue
effects data were limited for these PCB metrics, both freshwater and marine/estuarine
invertebrate species were considered in the review. Although some tissue effects data were
available for fish, there were no invertebrate data specifically for Aroclor 1260. Five studies of
Aroclor 1254, two of Clophen A50 (both commercial PCB mixtures containing approximately
54% chlorine by weight), and three of tPCBs were identified, along with a number of studies that
investigated other Aroclors or individual PCB congeners. In several cases, multiple species were
evaluated in a study, which increased the number of endpoints available for derivation of a

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw r- rx 7/10/2003

-------
1

threshold. These studies (summarized in Attachment D.l) were reviewed and evaluated for
relevance and data quality. Results from these studies are presented in Table D.3-9.

D.3.4.2 Estimation of a Tissue Effects Threshold

To estimate PCB effects thresholds, the tissue effects data for Aroclor 1254 and tPCB were
combined and then ranked in order of increasing tissue concentration to illustrate studies where
effects did and did not occur (Figure D.3-17). The figure includes only the subset of studies
from Table D.3-9 deemed applicable to threshold derivation (rationales for exclusion are
provided below). All tissue benchmark data in Table D.3-9 and Figure D.3-17 were reported in
units of mg/kg wet weight (ww), unless otherwise noted. Where the original study reported
tissue data on a dry weight basis, a conversion factor of 0.2 was applied; this method was also
applied in the USACE ERED and Jarvinen and Ankley (1999) databases.

Ten studies of a total of 11 freshwater species and seven estuarine/marine species were included
in Figure D.3-17. There were a total of 30 no effect measurements (ranging from 0.04 to 425
mg/kg ww) and nine effect measurements (ranging from 3.9 to 425 mg/kg ww). The majority of
data applied to effects on mortality; however, there were also data for growth, development,
behavior, physiological, and cellular effects endpoints.

Figure D.3-17 shows the distribution of no effect and effect tissue concentrations. No adverse
effects were observed at tissue concentrations of 3 mg/kg ww or less, whereas above 10 mg/kg
ww the frequency of occurrence of adverse effects increased significantly. Nearly half of the
studies above 10 mg/kg exhibited significant effects, and these effects included acute mortality.
There were three instances of adverse effects occurring between 3 and 10 mg/kg. Based on this
distribution, it appears that adverse effects are unlikely to occur at tissue concentrations at or
below 3 mg/kg, that they are likely to occur to sensitive organisms above 10 mg/kg, and that
there is some uncertainty about whether they will occur at concentrations between these points.

D.3.5 Sediment Quality Values

Numerous sediment quality benchmarks have been developed, using various derivation
procedures. A brief discussion of sediment quality values (SQVs) is presented below.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-61

-------
1 D.3.5.1 Application of SQVs

2 SQVs provide a range of contaminant-specific benchmarks that, when applied to site-specific

3 data, result in a bounded estimate of potential sediment toxicity.

4 The limitations associated with the derivation of the SQVs must be considered in their

5 application. These limitations include:

6 ¦ Inconsistent Degree of Conservatism - Depending upon the environmental

7 protection goals under which the SQVs were determined, some SQVs may not be

8 protective when applied to a given site, while others may be overly conservative.

9 Generally, generic criteria are inherently conservative, resulting in a high frequency

10 of "false positives" for marginal criteria exceedances (Chapman and Mann 1999).

11 ¦ Inability to Predict Bioaccumulation and Biomagnification Potential - SQVs are

12 typically based upon acute or chronic toxicity to aquatic biota. They are not

13 developed for assessing risks associated with the long-term bioaccumulation or

14 biomagnification potential of chemicals like PCBs (Chapman and Mann 1999).

15 ¦ Contaminant Mixtures - SQVs are most functional and predictive when used to

16 evaluate the potential adverse effects of single contaminants. However, a complex

17 mixture of contaminants is present at many contaminated sites. The confounding

18 effects of the various contaminants can affect the ability of the SQVs to accurately

19 predict toxicity for a single contaminant.

20 SQVs were used in the benthic invertebrate ERA as an additional line of evidence, rather than as

21 a conclusive statement, regarding the toxicity of COCs in Housatonic River sediment, in

22 recognition of the limitations described above.

23 D.3.5.2 SQVs Selected for Application to the Housatonic River

24 The benchmark selection process used in the benthic invertebrate risk assessment was similar to

25 that applied in the Pre-ERA screening, although additional sources were considered to provide

26 information on benchmark variability across jurisdictions.

27 Sources of North American benchmarks included (but were not limited to):

28 ¦ EPA Ecotox benchmarks (EPA 1996).

29 ¦ EPA national water quality criteria (EPA 1999).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-62

-------
1

Canadian environmental quality guidelines (CCME 1999).

Environment Canada sediment quality guidelines (Environment Canada 1995).

New Jersey Department of Environmental Protection guidelines (NJDEP 1998).

New York State sediment screening guidelines (NYSDEC 1999).

Washington State water quality standards (WDOE 1997).

Ontario guidelines for the protection and management of aquatic sediment quality
(OMEE 1996).

Consensus-based sediment quality guidelines (Long et al. 1995; MacDonald et al.
2000a, 2000b).

10 Other benchmark sources used for SQV derivation are presented in Table D.3-10.

11 A recent global review of derivation protocols used to determine environmental quality

12 guidelines for the protection of aquatic life, with a focus on metals/metalloids was conducted

13 (EVS 2002). The report concluded that incorporation of issues unique to metals in the derivation

14 of numerical guideline values is a new area of scientific investigation; however, two non-North

15 American jurisdictions were noted as leaders in this field:

16 ¦ Australia (ANZECC and ARMCANZ 2000) - Provides a clear framework for

17 addressing metal-related issues, even if exact technical methodologies need additional

18 development.

19 ¦ Netherlands (de Bruijn et al. 1999) - The "added risk approach" is an innovative

20 system to address both naturally elevated background concentrations and the

21 relatively high bioavailability of metals under laboratory testing conditions.

22 Therefore, the international benchmarks from Australia and the Netherlands were added to the

23 list of relevant benchmarks for screening metals. The Dutch values (de Bruijn et al. 1999) were

24 adjusted using the modification formulae for organic carbon content and clay content of

25 sediment. Two values were calculated for metals in sediment in this manner; one considered

26 parameters typical for stations upstream of the WWTP (Reach 5A), and the other considered

27 parameters representative of finer-grained sediment in Reaches 5B, 5C, and 6 (Woods Pond).

28 The lists of benchmarks, with references, are provided in Table D.3-10. Definitions of the

29 benchmark types are found in Table D.3-11. Note that the benthic invertebrate ERA list of SQV

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw ^ 7/10/2003

-------
1 citations is broader than that applied in the Pre-ERA (Appendix B); this is because the ERA

2 considers the range of available benchmarks and conservatism level, rather than being simply a

3 contaminant screening step. Furthermore, no single SQV was selected since the risk

4 characterization (Section D.4) emphasized the range and median of all available benchmarks.

5 Sediment benchmarks could not be found for dibenzofuran or tin; these substances are evaluated

6 qualitatively in the risk characterization. Because PCBs are the COC of greatest concern within

7 the Housatonic River sediment, the rationale used in the development of PCB SQVs is discussed

8 in detail in Attachment D.2.

9 D.3.6 Water Quality Benchmarks

10 Table D.3-12 presents water quality criteria considered in the risk characterization. Only PCBs

11 are included because there are no ecologically relevant water quality criteria for dioxins/furans.

12 Silver, the only other water column COC, was retained in only a few river

13 reach/geomorphological category combinations, and therefore will be considered qualitatively.

14 Other metals (e.g., copper, lead) not retained as water column COCs were evaluated against

15 water chemistry data from the EVS (2003) in situ toxicity studies. No exceedances of water

16 quality criteria (compiled using similar data sources to those for PCBs in Table D.3-12) were

17 observed, further justifying their exclusion from consideration as water column COCs in the risk

18 assessment.

19 D.3.7 Benthic Macroinvertebrate Community Evaluation

20 Assessment of benthic community structure is always complicated by the fact that species

21 composition is an expression of both habitat conditions and stressor-related effects. This

22 community evaluation controlled for macro-habitat, and compared exposed stations with

23 appropriately matched reference locations.

24 Statistical approaches applied in the risk characterization for benthic invertebrates included:

25 ¦ Comparison of benthic assemblages between contaminated locations within the PSA

26 and reference locations (e.g., Housatonic River upstream of the GE facility, or off of

27 the main stem). Tools used to make these comparisons included: (1) average rank

28 plots, combining relevant summary metrics in a non-parametric multivariate

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw s-a 7/10/2003

-------
1

approach; (2) multidimensional scaling plots, using the same summary metrics in a
parametric multivariate approach; and (3) univariate tests using key summary metrics.

¦ Analysis of the relationship between sediment COC concentrations and benthic
community structure indices, using a regression/correlation approach. This required
partitioning of the data set into broad habitat types to help reduce the confounding
effect of habitat type on benthic assemblages. Because the regression/correlation
approach represents an integration of exposure and effects assessments, these
analyses were deferred to the risk characterization (Section D.4).

D.3.7.1 Identification of Benthic Community Metrics

The scientific literature contains many benthic community metrics that potentially can be used as
indicators of environmental disturbance; however, not all of these metrics are necessarily
applicable to the Housatonic River.

A list of candidate metrics was prepared based on their biological relevance for the PSA and
their suitability for use with the available benthic community data (Table D.3-13). Metrics
selected for statistical analyses were those judged to be most indicative of environmental
perturbations (or lack thereof). A secondary objective was to avoid excessive redundancy in
metrics, which would introduce colinearity in the multivariate analyses. The relevance of
individual metrics was determined using a combination of best professional judgment and
application in the scientific literature.

Based on the rationale provided in Attachment D.3, the following six benthic community metrics
were included in multivariate statistical analyses:

¦ Organism abundance (number of animals per replicate or station).

¦ Taxonomic richness (number of unique taxa per replicate or station).

¦ Relative abundance of "EPT" taxa (Ephemeroptera, Plecoptera, Trichoptera) (i.e.,
mayflies, stoneflies, and caddisflies, respectively).

¦ Relative abundance of tolerant dipterans.

¦ Relative abundance of tolerant oligochaetes.

¦ Relative abundance of tolerant gastropods.

The use of the relative abundance metrics recognizes that organism tolerance to organic
pollutants can vary both within and among taxonomic groups (Figure D.3-18). In addition to
multivariate assessments of these six metrics, univariate assessments were also conducted for
some benthic metrics, including abundance, richness, and Modified Hilsenhoff Biotic Index

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC x-v s-c 7/10/2003

-------
1

(MHBI) scores (Sections D.3.7.5 and D.3.7.6). MHBI scores were not included in the
multivariate assessment because of the high degree of overlap with the pollution tolerance
measures; instead, these scores were separately evaluated in Section D.3.7.6.

Raw data for these six metrics, which were used for multivariate analyses, are presented in
Attachment D.6.

Appendix A. 13 of the SIWP (WESTON 2000) presents a number of potential metrics that would
be considered for the ERA. Of these, only the diversity (Shannon-Wiener H') and evenness
(Pielou's J') indices were not selected as metrics for the multivariate analysis. The reasons for
their exclusion are:

¦ Difficulty in ascribing ecological relevance - Diversity and evenness values alone
cannot be easily related to a "desirable" or "undesirable" state of a community.

¦ These indices are duplicate measures of information - The evenness index measures
the extent to which a community is dominated by a few resilient species. This was
already addressed in a more rigorous manner using the selected metrics for relative
abundances of tolerant taxa. Diversity is strongly correlated with evenness and with
taxa richness, and the latter is a more intuitive and simple metric.

¦ Such indices can mask underlying distributions (by collapsing information into a
single value that oversimplifies the data)

D.3.7.2 Average Rank Plots

Average rank plots were used as a means of integrating the results of multiple benthic metrics to
provide a weighted measure of potential impact. The replicate sample results for each metric
were ranked from 1 to 155, with 1 representing the most favorable response.

Figure D.3-19 displays the replicate average ranks for all stations, using a boxplot presentation
format. The box for each station is the interquartile range (1st through 3rd quartiles), and the
whiskers are the entire range. The median of the average ranks is shown as a black dot within
the box. Figure D.3-19 indicates spatial differences in average ranks, along with variability at
each station. There was a tendency for larger rank values at the five coarse-grained upstream
stations, relative to downstream and/or reference stations.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-66

-------
1

Although the ranking analyses provided indications of stations that suggest greater or lesser
disturbance, additional statistics were used to assess whether the observed patterns of ranks were
significantly different than what would be expected by chance. Therefore, as a means of
determining the relative magnitudes of the average ranks, a randomization distribution was
formed, under the null hypothesis that there was no trend in ranks among stations. The
randomization distribution was formed by assuming that there was no order or trend to the
station replicates, such that any ranking was equally likely. If this null hypothesis were true, it
would be expected that most of the station medians displayed in Figure D.3-19 would lie
between the 2.5th and the 97.5th percentiles of the appropriate random median distribution. These
percentiles are displayed as solid lines in the figure. The dashed lines represent the 2.5th and the
97.5th percentiles of the random replicate average results (i.e., before the median of 12 random
replicates was taken).

The results in Figure D.3-19 show that the median ranks at Stations 3, 4, and 5 are significantly
higher than all reference stations, indicating degraded conditions for the six metrics evaluated.
Although the median ranks for Stations 1 and 2 were higher than at the coarse-grained reference
sites, these differences were not statistically significant.

The spatial pattern in responses in the ranking plots is different from that observed in the
sediment toxicity program. This difference can at least partly be ascribed to differences in the
PCB concentrations measured in conjunction with these programs. The benthic community
sediment PCB concentrations at downstream stations were lower than those observed in the
toxicity study. For this reason, caution must be exercised when linking the benthic community
and toxicological impacts at these downstream stations.

D.3.7.3 Multi-Dimensional Scaling (All Stations)

The ranking procedure discussed above does not consider the magnitude of the differences
between locations. Multidimensional scaling (MDS) represents the dissimilarities of a group of
objects in a two-dimensional space, and is comparable to a scatter plot of the first two principal
components in a principal components analysis (PCA). Stations that group together on an MDS
plot are "more similar" with respect to the metrics evaluated than are stations more distant from
one another. When PCA or MDS is used alone, one can determine differences among stations,

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw 7/10/2003

-------
1

but it is often difficult to interpret whether these differences are necessarily indicative of
degraded conditions. Used together, MDS and the average ranking procedure yield better
insight. The following four conditions describe the possible outcomes:

¦ If a station has a higher than expected average rank (i.e., consistently worse than
other stations with respect to summary benthic metrics), and the MDS plot shows that
it is distant from reference stations, the station is clearly degraded.

¦ If the station has a normal average rank and is near the reference stations in the MDS
plot, there is no evidence of adverse effects.

¦ If the station has a high average rank, but is located near the reference stations in the
MDS plot, the result is equivocal. The best explanation is that although the station
indicates adverse effects on the basis of the chosen metrics, the magnitude of the
difference is small in comparison to the overall station variability.

¦ If the station has a normal average rank but is far from the reference stations in the
MDS plot, the best interpretation is that the station is different from the reference
stations, but is not consistently degraded relative to the other stations.

MDS was performed on the median result of the 12 replicates at each station. MDS was also
conducted using arithmetic means in place of medians. The results for the means are not
presented here because the findings were similar, and the use of a median was considered to
most closely match the approach used in the rank analysis. The results are displayed in Figure
D.3-20. The shading of the ovals on the figure is not statistically based, but indicates subjective
groupings of stations based on their MDS scores and physical location in the study area. In the
MDS plot, all coarse-grained contaminated (C/C) stations (Stations 1 through 5) were set apart
from the remaining stations. Concordance with the rank analysis provided a strong weight of
evidence for impairment at Stations 3 through 5. Stations 1 and 2 indicated benthic communities
that were different but not consistently degraded relative to the coarse-grained reference sites.

The fine-grained stations did not indicate benthic community alteration using the MDS approach.
All four fine-grained contaminated (F/C) locations were located near the fine-grained reference
(F/R) location in the MDS plot. This finding was in concordance with the relative ranking
assessment.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-68

-------
1

D.3.7.4 Multi-Dimensional Scaling (By Habitat Type)

Separate MDS evaluations were conducted with coarse- and fine-grained sediment to determine
if habitat differentiation significantly affected the outcome. To compare coarse-grained stations
and fine-grained stations to their respective reference stations, all replicate values were included
in the MDS calculation, rather than analysis of medians only. This was necessary to have
adequate samples for spatial representation. This added some "noise" to the MDS plot, but also
showed the range of results within a particular station, thus characterizing within-station
variability. Raw data for the MDS plots are provided in Attachment D.6.

The results for fine-grained stations are plotted in Figure D.3-21. The centroid (i.e., mean of x
and y coordinates) for each station was plotted with rays extending to individual replicates for
each station. Although there was only one F/R station (Station R4), the MDS scores for this
station were similar to the other fine-grained stations, indicating a lack of obvious impairment at
the F/C stations.

The results for the coarse-grained stations are displayed in Figure D.3-22. The figure is "zoomed
in" such that the ends of some rays were truncated, for ease of presentation. There was not a
large separation among the five contaminated stations, with Stations 3 and 4 indistinguishable
from one another. All five of the C/C stations were somewhat separate from the C/R stations,
although these differences are less obvious due to the large variability in individual replicate
values. As with the relative ranking analysis, Station 5 indicated the largest deviation from the
reference condition.

D.3.7.5 Analysis of Variance (ANOVA) for Summary Metrics

Univariate methods were also applied for the benthic metrics that are highly integrative in nature,
as described in the SIWP (WESTON 2000). One-way analyses of variances (ANOVAs) were
performed for the total abundance and total taxa (richness) metrics. As an additional refinement,
separate analyses were performed on coarse and fine-grained data to minimize habitat as a
confounding factor. The results of the ANOVAs are presented below, by metric and substrate
type. In all cases, the null hypothesis tested was that contaminated locations on the Housatonic
River were equal to comparable reference locations for the given metric. For statistical purposes,

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-69

-------
1 log transformations were required for fine-grained stations. Log transformations were not

2 adequate to approximate normality for coarse-grained stations, so a rankit transformation was

3 applied in these cases, providing a non-parametric version of the test.

4 D.3.7.5.1 Results for Total Abundance

5 For fine-grained stations, the log-transformed ANOVA indicated that Stations 6 and 9 had

6 significantly higher total abundance compared to reference Station R4 (p < 0.05). The effect size

7 was not large, as both stations exhibited mean abundance within a factor of 2 of the reference.

8 There was no indication of impairment in fine-grained sediment relative to reference based on

9 these results.

10 For coarse-grained stations, a rankit-transformed ANOVA using only the reference stations

11 indicated no significant differences among these stations (p = 0.33); therefore, the reference

12 stations were pooled for statistical comparison with contaminated stations. The rankit-

13 transformed ANOVA with all coarse-grained stations (combined) yielded statistical significance

14 for the "contaminated versus reference" contrast. The 95% simultaneous confidence intervals

15 for individual comparisons to reference stations indicated that all five contaminated stations had

16 significantly lower total abundances. This finding is in concordance with the MDS evaluation

17 presented above.

18 D.3.7.5.2 Results for Taxa Richness

19 The mean number of unique taxa in fine-grained sediment ranged from 13.1 (Station 7) to 26.9

20 (Station 9). The log-transformed ANOVA indicated that these two stations were statistically

21 different from the reference sediment R4. Given the small effect size and the lack of consistency

22 in the direction of the "effect," this result did not indicate evidence of ecological effects in the

23 fine-grained sediment.

24 In coarse-grained sediment, the rankit-transformed ANOVA indicated significant differences

25 among the three reference stations (p = 0.014); therefore, they were not pooled and separate

26 ANOVAs were run for each reference station. When comparisons were made to Reference

27 Stations A1 and A3, all five contaminated stations yielded statistically lower taxa richness; this

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-70

-------
1

finding was consistent with the multivariate MDS finding described in Section D.3.7.3.
However, when comparisons were made to Reference Station A2, only Stations 3, 4, and 5
yielded statistically lower taxa richness; this finding was consistent with the rank analysis
finding described in Section D.3.7.2. Overall, whereas the ANOVA results indicated an overall
difference between contaminated and reference stations, the differences were somewhat more
pronounced for Stations 3 through 5, compared to Stations 1 and 2.

D.3.7.6 Modified Hilsenhoff Biotic Index
D.3.7.6.1 Methods and Limitations

A number of combined indices exist for gauging the degree of perturbation in aquatic
communities, including the Hilsenhoff (HBI) and modified Hilsenhoff Biotic Index (MHBI), the
Lenat's Index, the Biotic Condition Index, and the Florida Index. The MHBI is essentially the
same as the HBI, but with greater taxonomic resolution. These indices were developed to
summarize information on the pollution tolerance of a macroinvertebrate community. Wildhaber
and Schmitt (1996), in their evaluation of the relationship between sediment contamination and
biotic indices, found that the HBI was significantly correlated with laboratory-measured
sediment toxicity. The MHBI integrates information on the relative abundance of various taxa
combined with estimates of their pollution tolerance (see Attachment D.3). High weighted
pollution tolerance values (i.e., biotic indices) indicate potential ecological effects. The
traditional interpretations of MHBI values for evaluation of water quality are summarized in the
following table:

Biotic Index

Water Quality

Degree of Organic Pollution

0.00-3.50

Excellent

No apparent organic pollution

3.51-4.50

Very Good

Possible slight organic pollution

4.51-5.50

Good

Some organic pollution

5.51-6.50

Fair

Fairly significant organic pollution

6.51-7.50

Fairly poor

Significant organic pollution

7.51-8.50

Poor

Very significant organic pollution

8.51-10.00

Very Poor

Severe organic pollution

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-71

-------
1

An important caveat to the application of the MHBI is that the index was developed as a measure
of organic pollution, primarily nutrient enrichment. The index, therefore, is not necessarily
sensitive to pollution by toxic substances such as chlorinated organic compounds or other COCs.
The majority of applications of the MHBI in the literature are to evaluate changes in the more
traditional organic water quality parameters, rather than for specific chemical pollutants such as
PCBs. Intuitively, a prerequisite for a useful indicator system is the demonstration of specific
benthic invertebrate sensitivities to the stressors of interest. In the case of the Housatonic River
sediment COCs, no existing benthic invertebrate pollution sensitivity classification system is
sufficiently comprehensive to allow these organisms to be used as indicators of habitat
contamination (Wogram and Liess 2001). Nevertheless, to provide a "combined index" type
assessment for the Housatonic River, the MHBI was applied, consistent with the SIWP
(WESTON 2000).

D.3.7.6.2 Results

The mean MHBI scores for all stations are shown in Figure D.3-23. There was a narrow range
of biotic index scores at all stations (from 6.0 to 7.6). Most observations in this range are
classified as "fair to fairly poor" water quality. The mean biotic index score for the three coarse-
grained reference stations was 6.43, while that of the coarse-grained contaminated stations was
6.49; these values were not significantly different. The single fine-grained reference station had
a biotic index score of 7.48; the mean score of the four fine-grained contaminated stations was
6.81. Using the MHBI, there was no indication that the degree of pollution or disturbance at the
contaminated stations was greater than at the reference stations. The Pittsfield WWTP is located
between Stations 5 and 6. It appears that the WWTP is not having an impact on most stations
below the plant, but may have some impact on Station 6 immediately downstream.

In summary, there was no compelling evidence of incremental habitat degradation due to PCBs
at any of the contaminated stations using the MHBI metric. However, the appropriateness of this
metric was questionable for the study area because the MHBI was not developed to address
effects of PCBs, and because the reference locations indicated a high "background" proportion of
pollution tolerant taxa.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-72

-------
1

D.3.7.7 Analysis of Invertebrate Biomass

Measures of invertebrate biomass were not included in the multivariate assessment because
biomass measures are strongly correlated with abundance measures, and review of the literature
for freshwater biometrics did not indicate that biomass measures have been broadly applied for
evaluating environmental effects. However, because biomass does provide a measure of the
amount of tissue available as a food source to higher trophic level organisms, invertebrate
biomass data were evaluated in the ERA.

Figure D.3-24 illustrates the mean biomass per replicate for each of the benthic community
stations, subdivided by habitat type. The most obvious finding was that the biomass was
significantly greater in fine-grained sediment relative to coarse-grained sediment. The lowest
biomass station in fine substrate (R4) contained more biomass than the highest-biomass coarse-
grained station (A3).

In coarse-grained sediment (Figure D.3-24), dipterans (true flies) comprised the greatest
proportion of biomass, in both reference and contaminated locations. Oligochaetes and bivalves
also comprised a substantial proportion of biomass at both contaminated and reference locations,
although the reference stations were more variable with respect to the relative mass of these two
taxa. Reference stations differed somewhat from contaminated stations in that odonates (e.g.,
dragonflies) comprised more mass in the reference locations. For total biomass, the results were
similar to the total abundance analysis in that contaminated stations exhibited significantly less
biomass relative to reference locations. Only Reference Station A2 had total biomass
comparable to contaminated stations. Stations 3 through 5 had lower biomass than Stations 1
and 2, which is in concordance with the results of the total abundance analysis.

In fine-grained sediment, Station 9 had a very high biomass primarily because of large bivalves
in the grabs. With bivalves removed, the F/C stations exhibited similar biomass, and all
contaminated stations had biomass greater than the Reference Station R4. However Figure D.3-
24 also indicates that the variability in taxonomic composition among stations was high,
suggesting that micro-habitat or other factors may play an important role in shaping the fine-
sediment communities.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-73

-------
1

Overall, the biomass assessment yielded similar findings to the abundance assessment in that
lower biomass was evident in coarse-grained contaminated stations compared to reference
stations. No impairment was evident in fine-grained sediment; however, the very high
variability in taxonomic distributions among fine-grained locations suggests that habitat
variations may confound the ability to detect perturbations due to chemical factors in the benthic
community biomass.

D.3.7.8 Potential Role of Habitat in Benthic Community Observations

The role of habitat variation in shaping benthic communities is a potential confounding factor in
the evaluation of risks from PCBs and other COCs. The inability to establish direct relationships
between contaminant concentrations and benthic community composition may be the result of
unaccounted for habitat differences among sampling stations (Canfield et al. 1994). Differences
in in-stream flow velocity or percent cover of submerged macrophytes may contribute to
differences in benthic community structure (such as changes in abundance of predators, grazers,
or dinger taxa). Reynoldson et al. (1995) note the value of multivariate approaches that account
for habitat differences by examining deviations from expected community composition.

To address the role of habitat, available ecological characterization data were used to assess
whether the division of benthic community stations into coarse- and fine-grained stations
provided an adequate refinement of the analysis based on habitat. The evaluation included
analysis of macrophyte maps, review of photos taken at each station, and review of the habitat
classifications along the riverbanks adjacent to where the samples were collected.

Table D.3-14 summarizes the in-river and bank vegetation observed in the vicinity of the benthic
grab locations. For the coarse-grained contaminated stations, algal films were observed covering
a majority of the substrate at all locations, with only a modest degree of variation among stations.
Mapped algae was predominantly green algae along with some brown algae, both forming a non-
filamentous film on the substrate. Downstream stations exhibited a transition in vegetation status
coincident with the change in river hydrology and substrate type. Specifically, all fine-grained
stations had macrophyte beds along the riverbanks, typically containing the species
Myriophyllum spicatum, Elodea canadensis, and Potamogeton sp. The characterizations were
qualitatively similar for all fine-grained stations, although the density of macrophyte growth

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw 7/10/2003

-------
1

tended to increase with proximity to Woods Pond. Table D.3-14 also contains descriptions of the
surrounding (and sometimes overhanging) vegetation in proximity to the sampling stations. The
nearshore vegetation regime exhibits a pronounced shift between the upstream and downstream
stations. Near all coarse-grained stations, the surrounding vegetation is dominated by
transitional floodplain forest. The fine-grained stations were situated adjacent to wet meadows,
shrub swamps, and shallow emergent marshes, with greater variability among stations relative to
coarse-grained stations.

Less habitat information was available for the reference locations relative to the nine
contaminated stations within the PSA. Reference Stations Al, A2, and R4 were outside the
boundary of the available natural community GIS maps. For reference stations without GIS
habitat information, slightly older (1995) aerial photographs were evaluated. These were
combined with visual observations made in the field and assessment of hydrology and substrate
type, to make qualitative assessments of habitat (Table D.3-14).

Overall, for both coarse- and fine-grained sediment, there were no major differences between
contaminated stations and reference stations. When accounting for grain size reference(s), the
confounding effects of habitat variation were expected to be low, especially for the coarse-
grained stations. The lack of significant macrophyte development at Stations Al, A2, and A3 is
similar to all five coarse-grained stations. All coarse-grained stations also had transitional
floodplain forest near the location of sampling. Although the upstream reference stations were
located close to some shrub wetlands and residential areas, relative to the coarse-grained
contaminated stations, there is no indication of a pronounced habitat shift that would explain the
qualitative differences observed in benthic communities.

In summary, the benthic community data reflect the major habitat transition that occurs
coincident with the shift in hydrological regime near the WWTP. Although the field sampling
crew endeavored to sample "similar" substrates throughout the program, the qualitative
differences between upstream and downstream conditions within the PSA made controlling for
habitat impossible via field sampling alone. The evaluation of ecological data collected in the
PSA indicated that there was a pronounced shift in substrate, vegetative status, and hydrological
regime between Stations 5 and 6, primarily due to the backwater effect from Woods Pond Dam.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-75

-------
1

For this reason, the data set was partitioned to control for these differences, and statistical
comparisons were made only to stations with comparable gross habitat conditions. This served
to remove some of the confounding effect of substrate type.

D.3.7.9 Conclusions

Overall, the benthic macroinvertebrate community evaluation indicated a high degree of
variability, both within and among locations. Despite within-station variability, some significant
locational differences were observed that were consistent across the metrics considered.
Specifically, for most metrics, the coarse-grained contaminated stations exhibited impaired
benthic communities relative to the three coarse-grained reference stations; impairment was most
pronounced at Stations 3 through 5. No strong or consistent differences in benthic assemblages
were observed among the fine-grained stations. The four fine-grained contaminated stations
exhibited assemblages that were variable but qualitatively similar to Reference Station R4.

D.3.8 Concentration-Response Analysis - Benthic Community Assemblages

As described in the exposure assessment, the concentration-response assessment for benthic
community assemblages was conducted using only the PCB data collected synoptic with the
benthic community grab sampling. The replication at each station (i.e., characterization of
micro-variation by using 12 replicates), combined with the synoptic collection of these data,
justifies this approach.

D. 3.8.1 Total PCBs

Figure D.3-25 illustrates the relationship between taxa richness and sediment tPCB
concentrations for coarse- and fine-grained sediment. Figure D.3-26 shows a similar plot with
total abundance as the measurement endpoint. The significant differences between C/C and C/R
locations are readily apparent in the figures, providing confirmation of the significant difference
identified in the ANOVA analyses. However, in the coarse-grained sediment at contaminated
locations, there was no apparent relationship between tPCB concentrations and these summary
benthic metrics. For example, the higher median sediment PCB concentration at Station 2 was
not associated with a community condition that is more degraded than other C/C stations. In fine-

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw q^ 7/10/2003

-------
1

grained sediment, there was also no apparent relationship between sediment PCB concentrations
and richness or abundance.

From Figure D.3-25 it is apparent that alterations in the benthic community occur in coarse-
grained sediment at concentrations as low as 5 mg/kg tPCB. An assessment of whether
alterations would occur at concentrations below 5 mg/kg could not be conducted because the
median tPCB concentrations exceeded this value at all 5 coarse-grained contaminated locations.
In fine-grained sediment, adverse responses were not observed at median tPCB concentrations
up to 13 mg/kg (Station 8). This could be explained by several factors, including bioavailability
differences for PCBs in higher TOC sediment, natural background variation in habitat that
obfuscated contaminant effects, and uncertainty in measures of both exposures and effects.
Overall, the concentration-response analyses indicate that adverse effects may occur at tPCB
concentrations close to 5 mg/kg. This value has moderate to high uncertainty, but is close to the
threshold of 3 mg/kg obtained from the concentration-response analyses for toxicity endpoints,
and is also near the upper bound of sediment quality values.

Figures D.3-27 and D.3-28 display scatterplots of the replicate data in coarse-grained locations
used to further evaluate the relationships described above. A clear difference between the three
reference locations (with tPCBs concentrations below 1 mg/kg) and the downstream locations
(tPCBs concentrations ranging up to a few hundred mg/kg) was evident, despite the variability in
the data. However, no trend in benthic summary metrics was evident in the coarse-grained
contaminated sediment.

In summary, while the comparison-to-reference approach yields significant differences for
coarse-grained contaminated stations, these differences were not supported by a linear
relationship with PCB concentrations over a wide range of PCB concentrations in coarse-grained
sediment. It is possible that the micro-scale variation in PCB sediment chemistry confounded
the ability to determine a relationship between PCB chemistry and benthic abundance and/or
richness. Because the micro-core used to subsample the benthic grabs was very small (a required
procedure, due to the need to maintain an intact grab sample for benthic community assessment),
an individual replicate PCB measurement in the grab may not have been representative of the

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-77

-------
1 PCB exposures to the animals collected in the entire grab. If this were the case, broader trends in

2 PCB effects would be masked by small-scale variability.

3 D.3.8.2 Other COCs

4 Because the mini-cores used to extract sediment from each of the benthic community grabs had

5 limited volume, concentrations of secondary COCs could not be assessed synoptic with the

6 benthic community assemblage data. Therefore, a qualitative assessment of concentration-

7 response was conducted for benthic community endpoints using other (non-synoptic) chemistry

8 data from Housatonic River sediment.

9 The largest benthic community responses were observed at Stations 1 through 5, which are all

10 located in Reach 5A. If other COCs were driving the observed responses, an increase in the

11 mean concentrations of COCs in 5A (particularly Station 5), relative to downstream areas, would

12 be expected. However, the available data for metals in main channel surface sediment suggest

13 that the reverse is the case. Specifically, all metal COCs for benthic invertebrates (antimony,

14 barium, cadmium, chromium, copper, lead, mercury, selenium, silver, and tin) had a trend

15 toward increasing concentrations with distance downstream, with a typical increase in mean

16 sediment concentration of an order of magnitude or more.

17 PAH results at the benthic sampling locations (Figure D.2-19) indicated a weak pattern of

18 decreasing PAH concentration with distance downstream. Although Stations 4 and 5 exhibited

19 among the highest PAH chemistry, the median concentrations were no higher than at Station A3.

20 Therefore, PAH chemistry cannot explain the reduced abundance and richness observed at the

21 coarse-grained contaminated locations relative to A3.

22 In summary, no pattern in other COC concentrations was observed that would explain the

23 impaired benthic communities observed in coarse-grained contaminated sediment.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-78

-------
1

D.4 RISK CHARACTERIZATION

The risk characterization for benthic invertebrates integrates the exposure assessment (Section
D.2) and effects assessment (Section D.3) to evaluate risks to the benthos. The overall objectives
of the risk characterization were to evaluate the extent to which benthic communities are
adversely impacted by exposure to COCs, and to identify threshold effects concentrations
relevant to the site. The threshold effects concentrations were used to extrapolate the benthic
risk assessment findings to areas both within and downstream of the PSA that have not been
evaluated with detailed biological studies.

Three lines of evidence were used to develop the risk characterization in the Housatonic River
benthic invertebrate risk assessment (Figure D.l-4):

¦ Field Surveys (i.e., benthic community structure) - For these endpoints, care was
exercised to discriminate, to the extent possible, between responses related to COCs
and those related to other factors, such as substrate or habitat type.

¦ Comparison of Field-Measured Exposures to Effects Levels or Benchmarks -

For these endpoints, the risk characterization integrated exposure and effects by
relating the two terms quantitatively (e.g., hazard quotient method for chemistry data
and derivation of concentration-response relationships for toxicity data).

¦ Site-Specific Toxicity Study Results - These endpoints (e.g., in situ and laboratory
toxicity tests, toxicity identification evaluations) directly evaluated biological
responses to COCs.

These three lines of evidence were independent, allowing for a robust weight-of evidence
assessment of the potential for risk using the approach of Menzie et al. (1996) in Section D.4.5.

All three lines of evidence suggested some degree of harm to benthic invertebrates in the
Housatonic River. In addition, for line of evidence, there were indications that PCBs are the
primary COC responsible for the observed patterns of responses.

D.4.1 Field Surveys

The benthic invertebrate community study (Section D.3.7) directly assessed the assemblages of
organisms found at several locations in the PSA, and related these assemblages to concentrations
of COCs and other stressors. After controlling for broad habitat factors (sediment particle size

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw 7/10/2003

-------
1

distributions and organic carbon content), significant differences between coarse-grained
contaminated sites and coarse-grained references were observed. These differences were not
observed in fine-grained sediment, however.

4 There are several possible explanations for the lack of community responses observed in the

5 downstream fine-grained sediment within the PSA, including:

6 ¦ Microhabitat variation - Unlike the coarse-grained sediment, the fine-grained

7 portions of the PSA exhibited considerable inter-station differences in invertebrate

8 communities. These variations may have masked any subtle impacts due to PCBs.

9 ¦ Lower sediment chemistry - The concentrations of tPCB in the benthic community

10 sampling program were lower than for other sampling efforts associated with effects

11 endpoints (e.g., toxicity studies). As shown in Figure D.2-10, the median sediment

12 tPCB concentration was generally in the 1-10 mg/kg range in the fine-grained

13 sediment collected synoptic with the benthic community grabs. Because these

14 concentrations are close to the site-specific toxicity threshold of 3 mg/kg derived

15 from sediment toxicity endpoints, large alterations in community structure would not

16 necessarily be observed at these levels. Although some biological alteration may be

17 occurring at this concentration range, the statistical power for detecting these

18 differences is very low given the other sources of variability in the study.

19 ¦ Reduced bioavailability of tPCB - As shown in Figure D.2-8, some of the

20 downstream stations exhibited high organic carbon contents, which may act to

21 sequester hydrophobic PCBs. Although the fine-grained sediment were clearly toxic

22 at the higher exposure concentrations in the sediment toxicity tests, the high TOC

23 may have been sufficient to reduce effects in the benthic community grab samples

24 that had tPCB concentrations close to the 3 mg/kg threshold.

25 Overall, due to a relatively narrow range of exposure concentrations and high natural variability,

26 the benthic community study was not suited to the identification of low-level environmental

27 perturbations in fine-grained sediment. Responses in coarse-grained sediment were evident, and

28 were consistent across a number of biologically relevant effects metrics (e.g., abundance,

29 taxonomic richness, multivariate community structure).

30 D.4.2 Comparison of Field-Measured Exposures to Effects Levels or Benchmarks

31 For chemistry data (water, sediment, and invertebrate tissue), hazard quotients (HQs) were used

32 to quantify the degree to which chemistry measurements exceeded environmental benchmarks

33 deemed protective of the assessment endpoint. In theory, adverse ecological responses are

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC D 80 7/10/2003

-------
1

possible if any HQ exceeds 1 (i.e., if exposure exceeds the protective threshold). However, most
ecological benchmarks were developed for screening and ranking purposes; inherent
conservatism in many benchmarks typically yields a high percentage of false positives.
Furthermore, the degree of conservatism in environmental benchmarks varies greatly within and
among jurisdictions. Benchmarks for a single contaminant-media combination often span
several orders of magnitude. This uncertainty translates to the resulting HQs.

To address the uncertainty described above, the HQ assessment used in the benthic ERA
considered multiple benchmarks and calculated a range of HQs. Benchmarks for many
jurisdictions were used, emphasizing federally promulgated criteria and benchmarks, as well as
state or provincial criteria that were based on commonly used ecological toxicity databases. The
full annotated lists of benchmarks, with references, are provided in Tables D.3-10 and D.3-12,
for sediment and water, respectively. Definitions of the various benchmarks are found in Table
D.3-11.

SQVs derived from the literature are generally conservative and have high associated
uncertainty; hazard quotients greater than 1 based on literature SQVs must be interpreted in this
context. However, HQs based on site-specific effects thresholds are more reliable indicators of
potential effects. This section discusses both types of HQs; however, only literature-based HQs
were derived for COCs other than PCBs.

For each contaminant and medium, the full range of HQs was considered. Furthermore, to depict
the "central tendency" of the benchmarks, HQs were also calculated using the median value of
all applicable benchmarks. Because the benchmarks range in terms of the level of protection
afforded (e.g., low-effect level versus severe-effect level) and also vary across jurisdiction, the
median value represents an intermediate level of protection. The extremes of the HQ distribution
are called "upper-bound" and "lower-bound" HQs in the following subsections.

The HQ assessment is similar to the contaminant screening process conducted in the Pre-ERA,
with two important differences:

¦ These analyses are based on a wider range of screening benchmarks.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-81

-------
1

¦ This section includes the calculation of HQs for samples of specific relevance to the
benthic invertebrate risk assessment, including samples collected at Triad sampling
stations.

D.4.2.1 Sediment Chemistry

HQs were derived using two sets of data. First, HQs were derived for sediment chemistry data
collected in 1999 in the vicinity of the Sediment Quality Triad stations. Second, HQs were
calculated for the average exposures observed across the entire PSA, divided into river
subreaches. The reason for two separate analyses is that the data collected at the Triad stations
are of use in assessing the likelihood that COCs contributed to observed toxicity in station-
specific analyses. The broader data set was used to interpret the larger scale contamination in the
study area and to assist in extrapolation of responses to the study area as a whole. For both sets
of data, the median was used as the measure of central tendency. The HQs derived do not depict
the variation of COC concentrations at each site, but instead portray the variability in SQVs.

D.4.2.1.1 Total PCBs

Figure D.4-1 shows the ranges of HQs for the PCB measurements made at the Triad stations in
1999. Eleven sampling events were conducted within the time period (March to October 1999)
deemed to be relevant to the effects data. The bars for each station indicate that the range of
benchmarks (and hence HQs) is more than two orders of magnitude. The median HQs for the
contaminated stations are all greater than one, usually by a large amount. Due to the
conservatism inherent in SQVs, the HQ values close to 1 (such as those observed at Station A3)
are unlikely to pose large risk to benthic invertebrate communities. However, the large median
HQs (30 to 500) observed at exposed locations imply a greater risk.

Because generic SQVs can be highly uncertain (as indicated by the wide ranges of HQs derived)
and can be overly conservative, it is informative to compare the SQVs with the site-specific
thresholds for toxicity observed in the Housatonic River. The median literature benchmark of
0.5 mg/kg is slightly lower than the concentration at which ecologically significant adverse
effects were observed in toxicity tests with sensitive endpoints. The maximum literature
benchmark (8.1 mg/kg) corresponds to a concentration at which significant adverse effects were
observed in the majority of species, including acute mortality. Therefore, the median HQs in

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC q ^ 7/10/2003

-------
1

Figure D.4-1 represent a reasonable indication of site-specific risk for sensitive endpoints, and
the upper-bound HQs are indicative of risk of more severe effects. The lower-bound HQs
indicate a level of risk consistent with the most conservative literature benchmarks, but which
does not appear to be occurring in the PSA.

Figure D.4-1 also depicts HQs derived using the site-specific effects threshold (MATC) of 3
mg/kg. From a comparison of the two types of HQs, it is apparent that site-specific thresholds
for toxicity observed in the Housatonic River fall within the range of values found in the
literature, but fall toward the higher end of SQVs (and therefore the lower end of HQs). All
contaminated stations yielded HQ values greater than one.

Figure D.2-14 shows the reach-wide distributions of tPCBs throughout the PSA. The median
tPCB value in Reach 5A throughout its length is approximately 10 mg/kg. Based on both the
site-specific analysis of toxicity concentration-response and the assessment of literature
benchmarks, these concentrations yield HQs greater than 1. These HQs exceed 1 even when
very liberal assumptions are made in the HQ derivation, such as adopting the highest (i.e., least
conservative) sediment quality value, or requiring that numerous site-specific toxicity endpoints
yield 50% responses to be deemed ecologically significant. Therefore, the HQ analysis indicates
a very high probability of risk in Reach 5A. HQs of similar magnitude result from the analysis
of Woods Pond (Reach 6) sediment. Reaches 5B, 5C, and 5D yield lower HQs, but median
concentrations in these reaches exceed both the site-specific toxicity threshold of 3 mg/kg and
the literature-derived median SQV of 0.5 mg/kg tPCBs. The variability of sediment PCB
concentrations in all reaches means that there are localized areas of sediment with PCB
concentrations below the threshold values, but the majority of the PSA contains sediment tPCBs
concentrations for which environmental impairment is predicted.

D.4.2.1.2 Other COCs

HQs were also determined for other COCs that have SQVs. Dioxin and furans were not assessed
using an HQ approach because the WHO TEF system for toxicity (Van den Berg et al. 1998)
does not include a specific TEF model for invertebrates. Dioxin-like toxicity was not considered
relevant to invertebrates because of the lack of an Ah-receptor-mediated mechanism of toxicity

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-83

-------
1

for freshwater invertebrates. Dibenzofuran was not assessed using HQs because of the lack of
reliable SQVs for this contaminant.

3 D.4.2.1.3 Triad Station HQs

4 Figures D.4-2(a) to D.4-2(i) show the HQs for sediment chemistry data collected near Triad

5 stations in 1999. As with PCBs, the HQs close to 1 are unlikely to yield major adverse responses

6 due to the conservatism in the benchmarks. Accordingly, the HQ analysis for metals indicated

7 low risks, in concordance with the results of concentration-response analyses and the TIE:

8 ¦ Median antimony HQs were below 1, and maximum antimony HQs barely exceeded

9 one at downstream stations.

10 ¦ Barium HQs barely exceeded 1, and only at downstream stations.

11 ¦ Median cadmium HQs exceeded 1 only at Station 7, and maximum HQs were 10 or

12 less even at the most contaminated sites.

13 ¦ Median chromium concentrations barely exceeded 1 at downstream locations, and

14 maximum HQs were below 10.

15 ¦ Maximum copper and lead HQs were 10 or less, even at the most contaminated

16 stations.

17 ¦ Mercury and silver exhibited median HQs between 1 and 10 at most downstream

18 locations.

19 The HQs for total PAHs (Figure D.4-2(i)) also indicated low risk at Triad stations from these

20 compounds, with median HQs below 3 at all stations. However, the wide range of PAH SQVs

21 resulted in higher HQs (i.e., greater than 10) if lower-bound SQVs are applied.

22 D.4.2.1.4 Evaluation Throughout the PSA

23 To provide an indication of the relative risks posed by other COCs in broader areas of the

24 Housatonic River, HQs were calculated for the main channel and side channel/oxbow (SCOX)

25 sediment throughout the PSA (Figure D.4-3). Woods Pond sediment was excluded from these

26 graphs because the contaminant concentrations were similar to Reach 5C, and because Woods

27 Pond substrate was assigned different substrate coding (i.e., data were assigned

28 geomorphological codes other than "main channel" or "SCOX" in the data base). HQ

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC "T\ Q/l 7/10/2003

-------
1

calculations were based on median chemical concentrations from each habitat, as obtained from
a database extract conducted in May 2002. Antimony, barium, and cadmium HQs (Figure D.4-
3a-c) were close to or below 1 for both substrate types and all reaches, indicating low risks,
similar in magnitude to those calculated for the toxicity stations. Chromium, copper, lead, and
mercury HQs (Figure D.4-3d-g) were slightly lower than for the toxicity stations, with median
HQs very close to 1 for most downstream stations. Silver HQs in the reach-wide analysis were
significantly lower than for the Triad station assessment, with median HQs below 1 for all
stations. Because the toxicity data exhibited no indications of metals-induced toxicity and the
PSA reach-wide metals concentrations were equal to or lower than the Triad station data, metals
are unlikely to have significant toxicological impacts within the PSA.

The HQs for total PAHs exhibited significant differences compared to the HQs from the Triad
station data. Specifically, the HQs were lower using the broader PSA data, with median HQs
below 1 for all reaches and both substrate types. This result suggests that the PAH
concentrations in toxicity studies were conservative representations of PAH exposures (i.e.,
overestimates of general conditions). Because intermediate concentrations of PAHs in the in situ
toxicity studies yielded no significant toxicity, it is apparent that the lower concentrations of
PAHs observed over most of the PSA are unlikely to exert toxicological impacts.

Overall, the HQ assessment for sediment indicated that the chemical hazard for tPCBs was much
higher than for other COCs. The median HQ for tPCBs was often 100 to 1,000, compared to
other COCs that rarely exceeded an HQ of 10. Use of a site-specific benchmark for tPCB
reduced the magnitude of HQs for tPCB; however, the HQ values were still greater than 1 over
large areas. This finding is in agreement with the TIE conclusions, which implicated PCBs
and/or other non-polar organics as the dominant causative agents in toxicity tests. When HQs
based on site-specific tPCB effects information are considered, risks are moderate to high for
most sediment found within the PSA.

D.4.2.2 Water Chemistry

Water chemistry data were considered to be relatively unimportant, because sediment serves as
the primary sink for most COCs and because the water-only exposures in the toxicity study
yielded low toxicity relative to the sediment exposures. Most analytes, with the exception of

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC D 8 5 7/10/2003

-------
1

PCBs, PAHs, and metals, were not detected in water column samples taken in conjunction with
in situ toxicity tests. The Pre-ERA identified a small number of COPCs in water, including
PCBs, dioxins/furans, and silver. Measurements of other contaminants in the in situ water
samples taken synoptically with the toxicity testing, either yielded non-detected results, or
concentrations below screening benchmarks. Dioxins/furans and silver were not considered
quantitatively because many analytes were not detected or because marginal HQs were identified
in the Pre-ERA screening. Accordingly, HQs are presented for PCBs only (Figure D.4-4).
These HQs were calculated by comparing the PCB water column data derived from the toxicity
study (EVS 2003) to water quality criteria for PCBs. The median HQs for both reference
stations (Al, A3) were less than 1 in all three sampling events. In contrast, the PCB
concentrations at contaminated locations exhibited median HQs that were elevated and fairly
consistent among stations and across monitoring events (i.e., median HQ of approximately 10).
The maximum HQs, using worst-case PCB benchmarks, were approximately 100. Overall, the
results indicate a moderately high hazard based on PCB chemistry in the water column, with
negligible risk from other water column contaminants.

D.4.2.3 Tissue Chemistry

HQs were derived for tissue PCB burdens in benthic invertebrates sampled near the Triad
stations. Two sets of HQs were derived representing different levels of conservatism (Figure
D.4-5). One set of HQs was based on comparison of observed tissue concentrations to an effects
benchmark of 3 mg/kg tPCBs, which represents the lowest concentration at which significant
adverse effects were found in the literature. Nearly all HQs derived in this manner were greater
than 1, and three HQs were greater than 10. The tissue values were variable across stations and
did not exhibit a clear trend with distance from the confluence for either predators or shredders.
The second method compared observed tissue concentrations to 10 mg/kg tPCBs, a
concentration that the literature review suggested would cause impacts to numerous species.
Even with this less conservative benchmark, most HQs still exceeded 1.

Cumulatively, these HQs suggest a high probability of adverse effects to benthic invertebrates
based on the tissue accumulations of PCBs. Although the assessment is somewhat uncertain, due
to the inclusion of marine species and non-resident organisms in the analysis, the degree of

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-86

-------
1 conservatism is not particularly high, because many of the significant effects concentrations were

2 for mortality rather than sublethal endpoints.

3 D.4.3 Site-Specific Toxicity Study Results

4 D.4.3.1 Sediment Toxicity Tests

5 Both the in situ and laboratory toxicity tests (Section D.3.1) exhibited significant adverse effects

6 in both coarse- and fine-grained sediment, relative to both negative controls and field reference

7 sediment. The only toxicity test endpoints that did not yield significant adverse responses at the

8 highest tPCB concentrations were: (a) limited exposure pathways, such as water-only in situ

9 exposures; (b) short test durations; and/or (c) tolerant test species, such as freshwater

10 oligochaetes used for bioaccumulation. The large number of endpoints indicating significant

11 toxicity (even for some acute lethal endpoints), and the high magnitude of response at the highest

12 PCB concentrations (100% mortality in some treatments), indicate a significant potential for

13 environmental harm. The evaluation of concentration-response (Section D.3.2) and the TIE

14 study (Section D.3.3) both indicated that non-polar organics (principally PCBs) were likely the

15 dominant toxicological agents in the toxicity tests.

16 Section D.3.2.1.3 presents the derivation of a site-specific MATC for tPCB in sediment, which is

17 summarized here. To calculate threshold effects concentrations, the average of values from the

18 six most sensitive endpoints was calculated for both 50% effects and 20% effects levels. This

19 approach was based on the rationale that thresholds should consider multiple sensitive endpoints,

20 but should not be based on the single most sensitive endpoint. The 50% effects level corresponds

21 to major impacts, for which there is a high degree of confidence in a significant biological

22 impact. The 20% effects levels correspond to lower but potentially biologically significant effect

23 sizes.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-87

-------
1

13 Based on this evaluation, and considering the alternative statistical evaluation presented in

14 Attachment D.5, 3 mg/kg tPCB was selected as the site-specific threshold (MATC) for sediment

15 tPCB.

16 D.4.3.2 Toxicity Identification Evaluation

17 The benthic ERA endpoints discussed in the preceding sections provide several indications of

18 biological impairment associated with concentrations of tPCBs in sediment. For example, the

19 toxicity observed in the in situ and laboratory bioassays was generally correlated with the

20 magnitude of sediment tPCBs contamination (Section D.3.1 and D.3.2). However, the

21 relationships were based on empirical associations rather than determinations of conclusive

22 cause and effect. To help address this limitation, toxicity identification evaluations (TIEs) were

23 conducted with the purpose of narrowing the range of potential stressors that could explain the

24 biological responses observed.

25 TIE procedures involve physical and chemical manipulation of samples. By comparing the

26 toxicity of the manipulated sample to the untreated sample, information regarding the cause of

27 toxicity can be obtained. For example, toxicity due to metals such as copper, cadmium, or zinc

28 is indicated by a reduction in toxicity following the addition of the chelating agent EDTA.

29 Organic contaminant-driven toxicity is generally indicated by a reduction in toxicity following

30 solid phase extraction (SPE) of organics from solution. TIEs are an iterative process in which

31 the results of previous manipulations are used to direct subsequent tests.

Summary of 50% and 20% Effects Levels

¦ Comparison to Negative Control - The mean of the lowest six 50% effects levels
was 1.3 mg/kg tPCB. The mean of the lowest six 50% effects levels was 0.1
mg/kg tPCB.

¦ Comparison to Reference A1 - The mean of the lowest six 50% effects levels
was 3.5 mg/kg tPCB. The mean of the lowest six 50% effects levels was 0.9
mg/kg tPCB.

¦ Comparison to Reference A3 - The mean of the lowest six 50% effects levels
was 3.3 mg/kg tPCB. The mean of the lowest six 50% effects levels was 0.9
mg/kg tPCB.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-88

-------
1

EVS (2003) provides detailed methods and discussion of the TIE testing of Housatonic River
sediment; therefore, only a synopsis is provided here. Ceriodaphnia dubia was selected as the
test species because acute effects were observed in this species following short-term exposures to
multiple test sediment. Several TIE treatments were initiated from August 15-17, 1999,
including baseline tests, oxidant reduction addition tests, EDTA chelation addition, pH-adjusted
filtration, pH-adjusted aeration, and pH-adjusted Ci8 SPE. None of the individual treatments
provided a definitive identification of toxic agent; however, integration of the results of various
treatments provides strong indications of the class of toxic agents. Key findings included:

¦ Significant Reduction in Toxicity in the pH-Adjusted/Filtration Treatments -

Higher survival in the filtration test was attributed to organic colloids in the samples
being filtered out and/or pH-mediated toxicity alteration of organic compounds (EVS
2003). Filterable compounds can include non-polar organics, such as PAHs, PCBs,
and some metals.

¦ Significant Reduction in Toxicity in the pH-Adjusted Ci8 SPE Treatments - The

results of these manipulations indicated that the filtration reduced the toxicity of the
original samples. Therefore, this test implicated non-polar organics, pesticides,
and/or some metals.

¦ EDTA Treatments - These treatments did not result in a reduction of toxicity, and in
fact, toxicity increased overall in this test, with observation of increased toxicity as
EDTA addition concentrations increased. This provides evidence against metals as
the dominant causal agent.

¦ Sediment and Porewater Chemistry - PCB concentrations in TIE treatments were
observed to be well above upper-bound sediment quality guidelines and water quality
criteria applied to porewater. Conversely, PAH chemistry in these TIE treatments
(EVS 2003) were below applicable criteria (i.e., Swartz [1999] consensus-based
sediment quality guidelines for total PAHs). Furthermore, the two samples
demonstrated to be most toxic in the initial 24-h screening toxicity test had the
highest sediment tPCBs concentrations, in both the in situ and laboratory
measurements.

In summary, the TIE treatments indicated that acute toxicity was significantly reduced in
treatments that reduced organic compounds and some metals in porewater of the two most
contaminated stations. Because the EDTA chelation treatment failed to reduce toxicity, the TIE
results appear to implicate non-polar organics as the dominant causative agent. More
comprehensive Phase II and Phase III TIE would be required to more precisely identify the
specific compounds responsible for altered biological responses. However, the presence of

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-89

-------
1

PCBs well above ecotoxicological thresholds and the absence of other contaminants with similar
magnitudes of exceedance of criteria confirm that PCBs may be responsible for the observed
responses. Other non-polar organic contaminants, such as PAHs, also exhibited elevated bulk
sediment chemistry relative to screening benchmarks in other environment samples (e.g., Station
7 in the in situ toxicity testing program). However, the relatively low PAH chemistry observed
for the two TIE samples indicated that PAHs were unlikely contributors to the toxicity observed
in the TIE treatments. Dioxins and furans are also non-polar organic compounds that could not
be eliminated as potential causal agents in the TIE. Dioxins and furans in sediment were
strongly correlated with PCB concentrations.

D.4.4 Integrated Station-by-Station Triad Assessment

Potential impacts of contaminated sediment to local ecological resources at each contaminated
station were assessed using a graphical approach that considered multiple lines of evidence
(Figure D.4-6). Multiple measurement endpoints were used, and the results of each were
integrated into a single conclusion regarding potential ecological impacts. For the purposes of
evaluating each measurement endpoint, results were categorized/simplified based on ecological
decision criteria. The categorizations facilitated the interpretation of the results for each leg of
the Sediment Quality Triad, on a station-by-station basis. The decision criteria used in the
categorizations also provided the rationale for the values used to rate the magnitude of ecological
harm in the formal weight-of-evidence assessment (Section D.4.5).

Each measurement endpoint was assigned a rating of high, medium, or low impact. Where
applicable, indications of potential for harm were standardized to appropriate background
conditions (e.g., toxicity endpoints were compared to reference Stations A1 and A3 rather than
negative control sediment). The decision criteria used to make the evaluations are summarized
briefly below. Some degree of professional judgment was required in determining some of the
ratings.

D.4.4.1 Sediment Toxicity Endpoints

Ratings were assigned for five stations at which toxicity tests were conducted. A response of a
toxicity test endpoint by a factor less than 20% relative to the reference sediment at A1 and A3

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw 7/10/2003

-------
1

was considered to be a negligible effect and not ecologically relevant. Effect magnitudes of 20
to 50% relative to reference locations were deemed moderate indications of significant biological
effects, and effect magnitudes greater than 50% were deemed to represent indications of
significant biological harm.

D.4.4.2 Benthic Community End points

To evaluate the average rank plots, the tails of the randomization distribution were considered.
Stations for which median ranks were in the upper tail (97.5th percentile or higher) of the
randomization distribution were considered indicative of major effects. Stations for which the
medians differed considerably, but which were between the 2.5th and 97.5th percentiles of the
random median distribution, were considered indicative of "moderate" effects. Stations with
medians comparable to reference locations were considered indicative of "negligible" effects.

To rate the multi-dimensional scaling (MDS) scores, the degree of separation in the MDS plot
was interpreted. The separation between Station 5 and the other locations (including coarse-
grained reference stations) was considered sufficiently large to be considered a major effect,
whereas the smaller degree of separation for Stations 1-4 was considered moderate. All other
stations clustered closely with the matched reference stations, indicating negligible effects.

The remaining benthic community endpoints were evaluated using ANOVAs with pairwise
differences comparing stations to appropriate controls. Stations for which statistically significant
differences (p < 0.05) were consistently observed based on comparisons to all reference
sediment, and for which the effect size was at least two-fold, were assigned a rating of major
effect. A rating of moderate effect was assigned to those stations that were significantly different
from at least one control, but for which significance did not hold for all reference stations, or for
which the effect size was less than a factor of 2. Responses for all other stations were rated as
negligible.

D.4.4.3 PCB Chemistry Endpoints

The median sediment chemistry data were used in the ratings, consistent with the rationale
presented in Section D.2.1.3. Because the site-specific effect benchmarks were derived using

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-91

-------
1

median PCB concentrations in the derivation of exposure-response relationships, this provides an
"apples-to-apples" comparison consistent with other components of the ERA. Sediment PCB
concentrations greater than the consensus-based Extreme Effect Concentration (EEC; 1.7 mg/kg
tPCBs) of MacDonald et al. (2000b) were considered major indications of hazard. This
concentration is comparable, considering analytical variability, to the site-specific threshold of 3
mg/kg tPCBs identified using the results of the sediment toxicity tests.

Use of a median concentration in the ratings acknowledges that, due to small-scale and analytical
variability, a portion of the sediment at a specific location may exceed the effects threshold while
still maintaining an overall conclusion of acceptable risk. However, the use of median
concentrations to evaluate acceptability of sediment risks is not intended to apply to broad scale
areas of the Housatonic River. Within a river reach that has a median PCB concentration below 3
mg/kg, there may be localized deposits of sediment that exceed the threshold. Provided localized
areas are of sufficient size to be ecologically important and also managed as a practical unit,
these areas should not be grouped with other areas with lower PCB concentrations.

The highest benchmark available for prediction of low- to mid-range effects (i.e., LEL, ER-L,
ER-M, TEC, MEC) was chosen as the threshold for indication of moderate potential risk. The
value used was the consensus-based Mid-Range Effect Concentration (MEC; 0.4 mg/kg) of
MacDonald et al. (2000b).

For PCB tissue concentrations in invertebrates, the literature-derived tissue effects
concentrations were applied to evaluate the stations. Tissues with less than 3 mg/kg tPCBs were
considered indicative of negligible to low risk, values from 3 to 10 mg/kg were considered
indicative of moderate risk, and values above 10 mg/kg were considered indicators of major risk.
The thresholds of 3 and 10 mg/kg were developed considering the frequency of adverse effects
observed in the literature studies. Also, the majority of experiments yielded significant effects
above 10 mg/kg tPCBs, whereas none of the available studies yielded toxic responses below 3
mg/kg tPCBs.

Chemistry data for analytes other than tPCBs were not considered in the weight-of-evidence
evaluation presented above. The very high HQs for PCBs relative to other COCs, combined

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-92

-------
1 with the results of the TIE, provide a strong indication that PCBs are the dominant toxicant of

2 concern. Therefore, the weight-of-evidence focused on the evidence related to PCBs.

3 D.4.4.4 Summary

4 The ratings in Figure D.4-6 indicate evidence for ecological disruption for all three components

5 of the Sediment Quality Triad. For each component, there are multiple indications of major risk,

6 and at multiple stations. The overall assessment yielded a rating of high overall risk for all

7 stations except Stations 6 and 9, for which no toxicity testing was conducted. Although there

8 was a high degree of overall concordance, one area of discrepancy was in the benthic community

9 endpoints for fine-grained stations. The strong toxicological responses at these stations were not

10 associated with strong indications of benthic community alterations. The differences in PCB

11 chemistry associated with these endpoints (i.e., higher tPCBs concentrations observed in the

12 toxicity samples) may explain this apparent difference.

13 The weight-of-evidence for risk to benthos is discussed further in the following section, in which

14 a systematic means of weighting various lines of evidence is followed.

15 D.4.5 Weight-of-Evidence Procedure for Assessing Risk from PCBs in the

16 Housatonic River PSA

17 A formal weight-of-evidence process was applied to determine whether PCBs pose a significant

18 risk to the Housatonic River benthos. The three-phase approach of Menzie et al. (1996) and the

19 Massachusetts Weight-of-Evidence Workgroup was applied for this purpose, in which weight-

20 of-evidence was reflected in the following three characteristics: (a) the weight assigned to each

21 measurement endpoint, (b) the magnitude of response observed in the measurement endpoint,

22 and (c) the concurrence among outcomes of the multiple measurement endpoints. The Menzie et

23 al. (1996) weight-of-evidence framework was integrated with the Sediment Quality Triad

24 concept, which incorporates measurement endpoints for benthic invertebrates using chemistry,

25 toxicity, and macroinvertebrate community structure. The application of a weight-of-evidence

26 approach has the benefits of providing a transparent means of synthesizing the numerous

27 endpoints and using a systematic means of evaluating the uncertainty for each endpoint.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-93

-------
1

The weight-of-evidence procedure assessed whether tPCBs were responsible for risk to the
benthic communities in the PSA. For this reason, potential confounding effects of other factors,
such as habitat-induced alterations or stresses due to other COCs, were considered explicitly in
the evaluation of measurement endpoints. Endpoints that could not discriminate between these
factors (e.g., raw chemistry, MHBI metric) were assigned lower weight in the overall evaluation
relative to those that implicated specific stressors or contaminant classes (e.g., TIE studies).

D.4.5.1 Weighting of Measurement Endpoints

Menzie et al. (1996) described the weighting factors for each of the attributes of measurement
endpoints. These weighting factors reflect three major attribute categories— presumed strength
of association between assessment and measurement endpoints, data quality, and study design
and execution.

The measurement endpoints for the benthic ERA were organized by the three components of the
Triad, and each Triad component was further subdivided into three categories reflecting the
major studies applied in the ERA program. Although it is possible to perform separate weight-
of-evidence evaluations for each assessment endpoint in the benthic ERA, the individual
measurement endpoints were often applicable to many or all of the assessment endpoints, and
assessment of each endpoint would result in an extremely complicated discussion. Therefore, a
single weight-of-evidence evaluation was applied to benthic invertebrate endpoints
simultaneously, reflecting a comprehensive assessment endpoint related to protection of benthic
invertebrates from mortality, community alteration, or adverse sublethal responses. The weight-
of-evidence assessment was directed toward discriminating PCB-induced impacts from other
potential stressors, such as other COCs or habitat factors.

The weighting factors and evaluation of the measurement endpoints are summarized in Table
D.4-1. A brief discussion of each attribute is provided in the following sections:

D.4.5.1.1 Relationship Between Measurement and Assessment Endpoint

¦ Degree of Association - This attribute represents the degree of biological association
expected between measurement and assessment endpoints, in terms of known
biological processes and level of ecological organization.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw qa 7/10/2003

-------
1

The values were lowest for chemistry endpoints, because there are a number of
factors that mediate the toxicity of a given concentration of an environmental
contaminant.

Toxicity endpoints received moderate or moderate/high values for this attribute,
because although toxicity tests provide a more direct measure of adverse effects,
there is uncertainty associated with the extrapolation to a local population and/or
community response.

- Benthic community endpoints received low/moderate values, because although
the benthic community abundance and composition directly reflect the
assessment endpoint, measures of standing stock are not an integration of
conditions or productivity on a larger scale.

¦ Stressor/Response - This attribute represents the ability to correlate effects (i.e.,
susceptibility as well as magnitude of response) with the degree of exposure.

Chemistry endpoints received relatively low values for this attribute, because
there is a level of uncertainty associated with SQVs and other criteria with the
definition of the causal agent. Although benchmarks are useful as screening
tools, particularly to predict circumstances where no effect would be anticipated
and to discriminate between the relative impact of different contaminants, they
are not highly predictive of magnitude of effect on a site-specific basis.

Toxicity endpoints received high values for this attribute, because toxicity tests
were developed with the specific purpose of identifying toxicant responses, and
laboratory toxicity to PCBs has been demonstrated in spiked sediment tests. TIEs
are able to associate responses with specific classes of toxicants, and their ability
to quantitatively relate stressors to response corresponds to the level of TIE
performed.

- Benthic community endpoints received low/moderate values, because these
measurement endpoints may respond to a broad range of stressors other than the
COCs (including physical and habitat variables). These confounding factors
make it more difficult to interpret results related to PCBs alone.

¦ Utility of Measure - This attribute refers to the ability to judge measurement
endpoint results against well-accepted criteria, standards, or objective performance-
based measures. The measure must be applicable, reasonably certain, and have a
sound scientific basis.

Chemistry endpoints received low to moderate designations for this attribute.
Although benchmarks and criteria have the endorsement of regulatory bodies,
there is often considerable variation in the procedures used in deriving these
values. Because criteria are derived to be conservative (i.e., protective), they may
yield false positives.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-95

-------
1

Toxicity endpoints received high scores for this attribute, because toxicity test
protocols have broad acceptance, and include consideration of statistical issues
such as statistical power and methods for calculation of toxicity endpoints (e.g.,
IC2o values).

- Benthic community endpoints received a fairly low ranking for this attribute,
because the benthic metrics have not been developed specifically for PCB-related
impacts. Even though many of the individual metrics, such as pollution-tolerance
measures and composition measures, have been studied for other pollutants, few
have resulted in criteria, standards, or objective performance-based measures.

D.4.5.1.2 Data Quality

¦ Data quality—This attribute refers to the extent to which data are collected in

conformance with standard operating procedures (SOPs) and meet data quality

objectives (DQOs). All site-specific evaluations received high values for this
attribute.

D.4.5.1.3 Study Design

¦ Site Specificity - This attribute refers to the extent to which data, media, species,
environmental conditions, and habitat types used in the study design reflect the site of
interest (Menzie et al. 1996).

The chemistry endpoints have low to moderate site specificity, because although
the criteria and/or benchmarks were derived from studies conducted outside the
Housatonic River watershed, many of these benchmarks were derived using data
collected from many sites. Another limitation is that effects data and/or
benchmarks were not always available for the specific PCB composition found in
the study area (i.e., Aroclor 1260), requiring extrapolation from slightly different
PCB mixtures. The lack of literature-based effects data for some endpoints
sometimes required inclusion of effects data for species not resident in the
Housatonic River, such as the use of marine species in developing critical tissue
residues for invertebrates.

The toxicity endpoints have high values for this attribute, because the selected
species are appropriate and sensitive surrogates for the freshwater organisms
found in the river, and because the results of these tests reflect site-specific
bioavailability. The in situ tests received the highest ratings because they
simulate the conditions in the field, including water quality, micro-habitat, and
other factors that cannot be simulated in the laboratory.

- Benthic community endpoints were assigned high ratings for this attribute,
because the grab samples reflect true field conditions and community responses
at the time of sampling.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-96

-------
1

¦ Sensitivity to Detecting Change - This attribute refers to the ability of the endpoint
to detect variation in the stressor.

Chemistry endpoints were assigned low to moderate values as indicated by the
wide ranges in benchmark regulatory values available.

- Moderate to high ratings were assigned to toxicity endpoints because of the
replication included in these tests, which increased statistical power. In some
cases, the toxicity tests yielded uncertain estimates of 20% response levels;
however, the 50% response levels were generally reliable and statistically
significant, and strong concentration-response relationships were observed for
most endpoints to help bound the toxic threshold concentrations. Many of the
conditions that could vary in the field can be controlled and/or measured in the
laboratory.

- Low/moderate ratings for this attribute were assigned to benthic community
structure endpoints. Discriminating between PCB effects and effects related to
micro-habitat or other factors present in the field was not always possible. This
issue was partly accounted for by explicitly considering substrate differences in
the study design and later in the classification of stations when performing
statistical analyses, and also by considering semi-quantitative information on
other habitat parameters, such as vegetative status. Sensitivity was also reduced
in the concentration-response assessment by the high variability of PCB
concentrations in samples collected synoptically with biological endpoints (i.e.,
variability reduced the statistical power).

¦ Spatial Representativeness - This attribute relates to the degree of compatibility
between the locations of samples, distribution of stressors, and locations of receptors.

The chemistry endpoints have high ratings for this attribute because there was an
intensive sampling study covering a range of habitats and substrate, particularly
for PCB concentrations in sediment.

The toxicity endpoints have moderate to moderate/high values for this attribute.
A range of habitat types was evaluated, resulting in high quality data at each
toxicity station; however, there were limitations on the number of stations that
could be tested. However, when designing the study, the stations were carefully
selected to provide a gradient in the concentration of PCBs. The in situ tests
evaluated water-only exposures separately from sediment exposures, which
provided additional information on the relevant exposure pathways. For the TIE,
only the two most contaminated stations were investigated because these yielded
pronounced acute toxicity in the preliminary Ceriodaphnia tests, providing
limited spatial representation resulting in a lower value.

- For benthic community structure endpoints, moderate to high values were
assigned to this attribute. With 13 stations, the program sampled multiple

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-97

-------
1

stations from coarse- and fine-grained habitats, as well as reference sediment of
similar substrate type.

¦ Temporal Representativeness - This attribute relates to the temporal compatibility
between data collection and the period during which effects of concern or variability
would occur.

The chemistry data, with the exception of invertebrate tissue (collected in a single
sampling event), reflect measurements over a wide temporal range, and can be
considered representative of potential effects manifested throughout the year;
therefore, moderate/high to high values were assigned.

The toxicity, invertebrate tissue collections, and benthic invertebrate
enumerations represent data collected from a fairly narrow window of time (i.e.,
summer 1999). However, the timing of laboratory toxicity testing is not a
sensitive consideration, and invertebrate studies were conducted during a
representative period of benthic productivity; therefore, a moderate value was
assigned.

In situ toxicity tests were planned for a high flow event; however, the weather
conditions did not allow for the high-flow study to be conducted as planned.

¦ Quantitativeness - This attribute relates the degree to which the magnitude of the
response of the measurement endpoint to the stressor can be quantified.

- For chemistry endpoints, the values assigned are moderate; HQs are quantitative
in the sense that HQ values can be calculated numerically. However, the
ecological effects benchmarks are not absolute values; therefore, the results,
while quantitative, may not have a linear correlation to the biological response.

In general, the toxicity endpoints reflect more quantitative measures; studies were
designed to produce a dose/response relationship that can be tested for statistical
significance. However, in the study implementation, although broad sediment
concentration gradients were maintained, a high degree of small-scale spatial and
temporal variability was observed, as had been observed in other site data sets.
Therefore, a value of moderate to high was assigned.

The benthic community endpoints were assigned moderate to high values for this
attribute, for the same reasons. In addition, MHBI scores may be more reflective
of general water quality condition and habitat, rather than PCB toxicity.

¦ Standard Method - This attribute refers to the extent to which the study followed
recognized scientific protocols.

The chemistry and benthic community endpoints had high values assigned for
this attribute, because sampling and analysis were conducted in conformance
with published protocols and standard methods.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-98

-------
1

- For toxicity testing, the laboratory toxicity tests were conducted following draft
EPA/ASTM testing protocols, which were subsequently finalized by EPA in
March 2000; therefore, this measure was assigned a high rating. In situ toxicity
tests and TIEs required adaptation of standard guidance and procedures for these
tests to the site and sample conditions (because no specific detailed protocols
were available); therefore, a value of moderate to high was assigned.

The values for each endpoint are provided at the bottom of Table D.4-1. The chemistry
endpoints yielded the lowest overall values because of lower site-specificity and some
uncertainties in the biological association between the measurement endpoints and the
assessment endpoint(s). Of the chemistry endpoints, tissue chemistry analysis and associated
comparison to benchmarks was valued highest. This is due to the more direct biological linkage
to effects expected when exposures are considered at the organism tissue concentration. The
tissue chemistry metric incorporates bioavailability from the exposure media to the animal,
which is closer to the site of toxic action.

The toxicity testing endpoints yielded the highest overall values, because of the high degree of
biological relevance of the tests. The benthic community structure endpoints had intermediate
values. Although these endpoints were site-specific, collected at a time when effects would be
expected, and were measures of the community structure component of the assessment endpoint,
the potential for the confounding effects of other factors in the direct attribution of the response
to the stressor reduced the utility of these endpoints to some degree.

D.4.5.2 Magnitude of Responses in Measurement Endpoints

The magnitude of the response in the measurement endpoint is considered together with the
measurement endpoint weight in judging the overall weight-of-evidence (Menzie et al. 1996).
This requires assessing the strength of evidence that ecological harm has occurred, as well as an
indication of the magnitude of response, if present. The weighting scores, evidence of harm, and
magnitudes of responses were combined in a matrix format and are presented in Table D.4-2.
Although professional judgment was, by necessity, applied in the designation of high, medium,
and low magnitudes of impact, the decision criteria used to rate the individual endpoints that
formed the basis for the designations in Table D.4-2 were the same as those applied for the
station-by-station assessment in Section D.4.4.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-99

-------
1

D.4.5.3 Concurrence Among Measurement Endpoints

A graphical method was used for displaying concurrence among measurement endpoints (Table
D.4-3). The method entailed plotting the nine symbols representing the toxicity (T), benthic
community (B), and chemistry (C) endpoints in a matrix, with the weight of the measurement
endpoint and the degree of response as axes. These graphics indicate that the majority of
endpoints suggest some risk for benthic communities in both coarse- and fine-grained sediment.
The plots also indicate that several of the endpoints suggest a high degree of risk with a
relatively high weight (e.g., toxicity endpoints).

The conclusion from interpretation of Table D.4-3 is that there is a moderate to high likelihood
of harm to much of the benthic community indicated by the weight-of-evidence evaluation, with
the exception of Measurement Endpoint B-3 (MHBI index of pollution tolerance) and benthic
community measures for fine-grained sediment.

D.4.5.4 Summary

Overall, the benthic ERA indicates significant risk to aquatic invertebrates based on a weight-of-
evidence evaluation of multiple Sediment Quality Triad endpoints. Furthermore, the available
data suggest that PCBs are the primary chemical stressor responsible for such impairment. The
confidence in the conclusion is moderate to high, based upon the concordance in predictions of
risk from multiple measurement endpoints. Screening of sediment, water column, and tissue
chemistry to ecologically based benchmarks indicated a high probability of effects attributable to
PCBs, with lesser degree of hazard from other COCs.

Compelling evidence for ecological risk comes from the sediment toxicity tests, which not only
indicated significant toxicological effects in multiple appropriate indicator species and endpoints,
but also indicated a correlation between the level of effect and sediment PCB concentration.
This correlation was consistent with the TIE results, which implicated non-polar organics as the
dominant toxicants in Housatonic River sediment. The evidence of effects to benthic community
structure was not as compelling because significant alteration relative to reference conditions
was observed, primarily in the coarse-grained sediment upstream of the WWTP.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-100

-------
1

D.4.6 Sources of Uncertainty

The assessment of risks to benthic invertebrates contains uncertainties. Each source of
uncertainty can influence the estimates of risk; therefore, it is important to describe and, when
possible, specify the magnitude and direction of such uncertainties. The sources of uncertainty
associated with the assessment of risks of tPCBs and other COCs to benthic invertebrates are
described below.

D.4.6.1 Exposure Assessment

The greatest uncertainty in the benthic invertebrate exposure assessment was the potential for
small-scale variability in exposure concentrations to complicate the development of
concentration-response relationships. The actual field variability combines with analytical
variability (Appendix C.l 1) and creates incertitude in the exposure concentrations that are paired
with effects data. For studies that had replicate measurements of PCBs at a given station over a
short period (e.g., benthic macroinvertebrate sampling), the spatial variability can be quantified
and considered explicitly in the derivation of concentration-response relationships. Where
spatial replication was not available, characterization of variability requires the incorporation of
additional data sets that are not fully synoptic with the effects data. The main approach to
concentration-response assessment applied in this ERA assumes that uncertainty arising from use
of data that are not 100% synoptic (resulting from such incorporation of broader data sets) was
more than offset by the increase in sample size and characterization of variability. Note,
however, that the qualitative conclusions of the study were very similar irrespective of the choice
of exposure data.

There were some unusual singular data points that were carefully assessed in the exposure
assessment:

¦ One replicate in one benthic community grab indicated a TOC value of 57%. Because
this value contrasted greatly with the low TOC values in the remaining 11 replicates
(mean = 0.7%), this value was replaced with the mean of the other replicates for
statistical analysis purposes.

¦ The observation of elevated PCB concentrations in the A3 shredder tissue sample
(i.e., 33.8 mg/kg, while the associated predator tissue sample was 0.35 mg/kg)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw -> rx-> 7/10/2003

-------
1

indicated that this station may not be an ideal reference station, and/or it was not
possible to exclude the possibility that there was another issue with the sample, such
as sample handling. PCBs were not detected in sediment collected at this location.

¦ The lipid concentrations in benthic invertebrate tissues were highly variable. A
number of lipid measurements were only 0.1%, which seems unusually low,
particularly given that other samples collected from the same location yielded lipid
contents as high as 11.7%. The mean lipid contents observed in biota (approximately
1%) appear reasonable, however.

¦ The mass of invertebrate tissue samples was small. As a result, the analytical
laboratory was operating near the lower end of their ability to detect PCBs and
quantify the results.

¦ The Lumbriculus bioaccumulation tests yielded lower PCB concentrations in
invertebrates relative to field samples of predators and shredders. It is not known
whether this difference was due to lack of equilibrium achieved in the
bioaccumulation tests, interspecies differences in PCB uptake, or small-scale
spatial/temporal variability in exposures to these organisms. Laboratory studies from
the literature suggest that the exposure duration was not adequate for equilibrium to
be achieved.

¦ The lack of replication placed limitations on the interpretation of the tissue chemistry
data. Because of the small-scale variability in sediment PCB concentrations, there is
uncertainty in the derivation of site-specific BSAFs.

In summary, the variability in exposure concentrations within and among studies, and the
differences in spatial trends across some studies require careful characterization of exposures that
are appropriately matched to effects data, particularly for sediment concentrations. The approach
taken was to bound the assessment using two different techniques. The first technique combined
data sets that are reasonably (but not always fully) synoptic and used the central tendency of
these data (i.e., median) to provide an estimate of the COC concentrations at the time of effects
measurements. The second technique used only the study data that were most synoptic with the
effects data. Evaluation of the data using this bounding approach provided a more robust
assessment.

D.4.6.2 Effects Assessment

A few pesticides (heptachlor epoxide, methoxychlor, benzyl alcohol, hexachlorobenzene) could
not be definitively ruled out as COCs due to lack of available SQVs or tissue effects information.
However, it is highly unlikely that these contaminants were present at concentrations that would

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-102

-------
1 exert effects. Given the low risks identified for other pesticides for which screening values were

2 available, the low tissue concentrations of these pesticides (0.1 mg/kg or lower), and the

3 possibility that pesticides detections in tissue samples may be attributable to laboratory

4 interference artifacts, the uncertainty associated with these data gaps is expected to be very low.

5 The effects benchmarks derived from the literature carry a high degree of uncertainty because of

6 the need to extrapolate across sites and species. This uncertainty was explicitly addressed in the

7 weight-of-evidence evaluation. The evaluation of site-specific studies indicates that the most

8 conservative SQVs for tPCB are at concentrations that do not appear to be associated with

9 effects in the Housatonic River, but that the upper-bound SQVs (i.e., Extreme Effect or Severe

10 Effect Levels) are in agreement with levels at which biological responses are observed at the site.

11 The effects benchmarks for tissue and sediment were based on toxicity information for Aroclor

12 1254 and Clophen A50 mixtures. Because these formulations may contain different ratios of

13 congeners relative to the Housatonic River site media some uncertainty is introduced. Aroclor

14 1260 was included in the literature review; however, the lack of effects data for this Aroclor adds

15 uncertainty to the evaluation of tissue-based effects for tPCB.

16 Individual toxicity effects endpoints carry some uncertainty because individual taxa have

17 specific tolerances to both chemical and background environmental factors. The strength of the

18 Sediment Quality Triad approach comes from the multiple lines of evidence (lethal and sublethal

19 test endpoints with different exposure durations) from multiple test species. The concurrence of

20 findings from different taxa substantially reduced this uncertainty.

21 There is some uncertainty with respect to the conclusions of the TIE study. While the conclusion

22 was reached in the study (EVS 2003) that PCBs are the most likely causal agent responsible for

23 the observed patterns of responses, the specific TIE treatments conducted were not able to

24 conclusively eliminate other non-polar organic compounds as causal agents. For this reason,

25 other lines of evidence, such as correlation between COC chemistry and effect levels, were also

26 used to support the argument that PCBs are the dominant toxicant in the Housatonic River.

27 There is uncertainty with respect to the confounding role of micro-habitat for benthic

28 communities. Although the study design controlled for habitat (physical and biological) to the

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-103

-------
1

extent possible, variations in micro-habitat factors may have obscured alterations due to
chemical stressors. Downstream of the WWTP, there was high natural variability in the
composition of benthic assemblages; therefore, the statistical power to detect environmental
perturbations was weak.

There is uncertainty in the relationship between benthic community endpoints and PCB
concentrations. For example, while the ANOVA analysis identified a significant difference
between C/C and C/R stations, a significant relationship between PCB and benthic community
effects was not observed at a finer spatial resolution. Micro-scale habitat influences, a different
mode of toxicity in site macrobenthos versus toxicity test organisms, bioavailability issues, or
PCB sediment concentration heterogeneity may be responsible for the non-correlated responses
at the individual station level.

D.4.6.3 Risk Characterization

The interpretation of responses in the station-by-station Triad evaluation was complicated by the
fact that PCB exposure levels were not consistent across different studies used to evaluate effects
endpoints. The patterns of PCB concentrations observed in the benthic community study,
sediment toxicity study, and benthic invertebrate tissue sampling were not always consistent.
For example, the PCB concentrations measured at Stations 7 and 8 during the benthic
community sampling, were lower than most other PCB measurements made at those locations.
This complicated the integrated station-by-station assessment presented in Section D.4.4.

Calculations of site-specific effects thresholds used in the risk characterization (e.g., sediment
MATC of 3 mg/kg) have the following uncertainties: (a) uncertainty due to application of dose-
response models required for interpolation; (b) uncertainty regarding the choice of exposure data
that is synoptic with effects information; (c) uncertainty due to natural variability in exposure
data and effects data. These uncertainties were addressed in the ERA by conducting multiple
assessments (e.g., applying different statistical models and exposure assumptions). The general
concordance of the findings using many different data processing assumptions provides
confidence that the derived thresholds are not based on spurious statistical outcomes.
Furthermore, the indications from each major line of evidence are that tPCB concentrations
above 3 mg/kg can result in adverse effects to benthic invertebrates.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw -> 7/10/2003

-------
1

D.4.7 Extrapolation to Other Species

The benthic invertebrate ERA included the entire benthic community; benthic community
composition analysis was a measurement endpoint considered in the weight-of-evidence
assessment. Individual species were also used in toxicity tests as surrogates for the Housatonic
River freshwater benthic community. Both the status of sensitive taxa and community
composition are considered indicators of overall health and productivity of the benthic
community. As a result, no formal extrapolation to other species was required. The toxicity test
species and endpoints encompass a range of toxicological sensitivities, ranging from sensitive
(e.g., Hyalella chronic reproduction) to tolerant (e.g., Lumbriculus survival); similar variation in
sensitivity can be expected in the field.

D.4.8 Downstream Assessment

Because of the more limited amount and spatial coverage of data on contaminant concentrations
downstream of the PSA, the more rigorous approach followed in assessing ecological risks in the
PSA was not appropriate or possible. A preliminary estimate of potential ecological risks was
developed using mapping (GIS) techniques and threshold concentrations that would indicate
potential risk to benthic invertebrates.

Maximum acceptable threshold concentrations (MATCs) for tPCBs in relevant media were
developed based primarily on the detailed risk assessment performed for the PSA. MATC
values were compared to available medium-specific data for areas downstream of Woods Pond
to Long Island Sound, and areas of exceedances, indicating potential risk, were plotted on maps
of the river.

For benthic invertebrates, the media of interest were river sediment and benthic invertebrate
tissue:

¦ The sediment MATC of 3 mg/kg tPCB was used as a measure of the potential for
adverse affects to benthic invertebrates downstream of Woods Pond. This
concentration was developed in the risk assessment for the PSA using multiple lines
of evidence (e.g., benthic community studies, in situ and laboratory toxicity testing,
bioaccumulation testing, Sediment Quality Triad) and was selected as the
concentration at which some sensitive endpoints exhibited apparent responses, but the

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-105

-------
1

magnitude of responses was not large. Above a concentration of 3 mg/kg tPCB,
numerous endpoints indicated ecologically significant responses, with many
LC50/EC50 values falling in this range.

¦ The tissue MATC of 3 mg/kg tPCB was used as a conservative measure of the
potential for adverse affects to benthic invertebrates downstream of Woods Pond.
This concentration was developed considering the frequency of adverse effects
observed in the literature studies; none of the available studies yielded toxic responses
below 3 mg/kg tPCB, but approximately 40% of the studies above 3 mg/kg yielded
significant adverse responses (Figure D.3-17).

The 3 mg/kg sediment tPCB threshold concentration was compared to available recent surficial
sediment data downstream of Woods Pond and the results plotted (Figure D.4-7) to indicate
samples above and below the threshold value. In addition, inverse distance weighting was used
to interpolate sediment concentrations between discrete sampling points, and the potential for
risk to benthic invertebrates is indicated in Figure D.4-7 by the shading of the appropriate
sections of the river channel.

In general, potential risks to benthic invertebrates occur in limited areas downstream of Woods
Pond to Rising Pond where pockets of sediment contaminated with higher concentrations of
PCBs appear to have accumulated. Below Rising Pond through the remainder of Massachusetts
and Connecticut, sediment does not contain concentrations of PCBs that are sufficiently elevated
to represent a potential risk to benthic invertebrates.

The tissue MATC can be compared against historical benthic invertebrate tissue data collected at
Cornwall, CT. Table D.4-4 presents data for 1994 onward, which reflect current conditions and
exclude the 1992 calendar year that reflects a short-term episodic release of PCB-contaminated
sediment from Rising Pond. Since 1994, the benthic invertebrate tissue concentrations for three
species (caddisfly, dobsonfly, and stonefly) have remained fairly constant. Although the
dobsonfly larvae marginally exceeded the 3 mg/kg tissue MATC (3.03 mg/kg in 1994 and 3.94
mg/kg in 1998), the tissue concentrations for the other two species have remained below the
MATC. Therefore, the estimated risks for this downstream portion of the Housatonic River are
considered to be low, matching the sediment MATC assessment.

The use of both sediment and tissue-based MATCs is expected to provide a sufficient level of
protection to downstream habitats. Whereas the sediment MATC is most appropriate for

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-106

-------
1

depositional habitats downstream of Woods Pond, the tissue-based MATC also provides
protection for invertebrate exposures that occur in free-flowing habitats (i.e., erosional riffle
habitats) for which invertebrate exposures may occur via contact with PCBs in particulate
organic matter.

D.4.9 Conclusions

The weight-of-evidence assessment of benthic invertebrate endpoints indicates a high risk of
ecologically significant effects at the PCB concentrations observed at the Sediment Quality Triad
stations. The available data suggest that PCBs are the primary chemical stressor responsible for
such impairment. Compelling evidence for ecological risk comes from the sediment toxicity
tests, which not only indicated significant toxicological effects in multiple appropriate indicator
species and endpoints, but also indicated a correlation between the level of effect and sediment
PCB concentration. This correlation was consistent with the TIE results, which implicated non-
polar organics as the dominant toxicants in Housatonic River sediment. The evidence of effects
to benthic community structure was not as compelling, because significant alteration relative to
reference conditions was not observed in the fine-grained sediment downstream of the WWTP.

The magnitude of risk to benthic invertebrates in the Housatonic River varies spatially, primarily
as a function of sediment tPCB concentration. Extrapolation of risk estimates to the broader
reaches of the PSA and to downstream reaches requires use of concentration-response
relationships derived from the site-specific studies. The toxicity studies indicated that
ecologically significant effects were observed at sediment tPCB concentrations of 3 mg/kg or
higher, and that effects were large in magnitude (i.e., 50% responses in most test species) at 10
mg/kg tPCBs. The benthic community data indicated that five of six stations with median tPCBs
above 5 mg/kg also had significant benthic community alteration. These values are in general
agreement with the higher end of the SQVs for tPCBs identified in a literature review (i.e., 1 to
10 mg/kg). Although the site-specific studies suggest that tPCB sediment quality values may be
overly conservative for direct application for the Housatonic River, they support the use of a
threshold value of 3 mg/kg tPCBs to protect against adverse effects to benthic invertebrates.

Figure D.2-14 depicts the spatial distribution of tPCBs concentrations in the PSA, with medians
and interquartiles (25th and 75th percentiles) for each reach and subreach. The median

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC -pw -> rsq 7/10/2003

-------
1

concentration within Reach 5A is 10 mg/kg tPCBs, which indicates unacceptable risk for the
majority of sediment sampled within this reach. In the downstream reaches in the PSA, there
may be reduced risk; however, the tPCBs data indicate that a high percentage of samples exceed
the site-specific thresholds described above. Furthermore, because the quarter-mile subreaches
shown in Figure D.2-14 represent very large areas of sediment, with variations in substrate
within each subreach, it is likely that even areas with median concentrations below 3 mg/kg will
contain significant contiguous areas of sediment that exceed the effects threshold. Downstream
of Woods Pond, risks are reduced relative to the PSA, and are negligible to low downstream of
Rising Pond.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC D 108 7/10/2003

-------
1 D.5 REFERENCES

2 Academy of Natural Sciences of Philadelphia, Patrick Center for Environmental Research. 1999.

3 PCB Concentrations in Fishes and Benthic Insects from the Housatonic River, Connecticut, in

4 1984-1998. November 15, 1999. Report No. 99-10F.

5 Allan, J.D. 1995. Stream Ecology: Structure and Function of Running Waters. Kluwer Academic

6 Publishers, Dordrecht, Netherlands. 388 pp.

7 Ankley, G.T., K. Lodge, D.J. Call, M.D. Balcer, L.T. Brooke, P.M. Cook, R.G. Kreis, Jr., A.R.

8 Carlson, R.D. Johnson, G.T. Neimi, R.A. Hoke, C.W. West, J.P. Giesy, P.D. Jones, Z.C. Fuying.

9 1992. Integrated assessments of contaminated sediments in the Lower Fox River and Green Bay,

10 Wisconsin. Ecotoxicol. Environ. Safety 23: 46-63.

11 ANZECC and ARMCANZ. 2000. Australia and New Zealand Guidelines for Fresh and Marine

12 Water Quality. Australia and New Zealand Environment and Conservation Council/Agriculture

13 and Resource Council of Australia and New Zealand, Canberra ACT, Australia.

14 BBL (Blasland, Bouck & Lee, Inc.) and QEA (Quantitative Environmental Analysis, LLC).

15 2003. Housatonic River - Rest of River RCRA Facility Investigation Report. Prepared for

16 General Electric Company. January 2003.

17 BC MELP (British Columbia Ministry of Environment, Lands and Parks). 1997. Checklist for

18 the Tier 1 Ecological Risk Assessment of Contaminated Sites in British Columbia. Submitted to

19 the British Columbia Ministry of Environment, Lands and Parks, Environmental Protection

20 Department, Industrial Waste and Hazardous Contaminants Branch, Victoria, B.C., Canada.

21 Submitted by Ecological Risk Assessment Guidance Team. Contract 1070-20/IWHC7-012.

22 BC MELP (British Columbia Ministry of Environment, Lands and Parks). 1999. Criteria for

23 Managing Contaminated Sediment in British Columbia. 46 pp.

24 BC MWLAP (British Columbia Ministry of Water, Land and Air Protection). 2001. A

25 Compendium of Working Water Quality Guidelines for British Columbia. 1998 edition, updated

26 23 August 2001. British Columbia Ministry of Water, Land and Air Protection, Water

27 Management Branch, Environment and Resource Management Department, Victoria, BC.

28 Boese, B.L., M. Winsor, H. Lee II, S. Echols, J. Pelletier, and R. Randall. 1995. PCB congeners

29 and hexachlorobenzene biota sediment accumulation factors for Macoma nasuta exposed to

30 sediments with different total organic carbon contents. Environmental Toxicology and Chemistry

31 14:303-310.

32 Bode, R.W. 1988. Quality assurance work plan for biological monitoring in New York State. NY

33 Department of Environmental Conservation. 15pp.

34 Borgmann, U., W.P. Norwood, and K.M. Ralph. 1990. Chronic toxicity and bioaccumulation of

35 2,5,2',5'- and 3,4,3',4'-tetrachlorobiphenyl and Aroclor® 1242 in the amphipod Hyalella azteca.

36 Archives of Environmental Contamination and Toxicology 19:558-564.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-109

-------
1 Brunson, E.L., T.J. Canfield, F.J. Dwyer, C.G. Ingersoll, and N.E. Kemble. 1998. Assessing the

2 bioaccumulation of contaminants from sediments of the Upper Mississippi River using field-

3 collected oligochaetes and laboratory-exposed Lumbriculus variegatus. Arch. Environ. Contam.

4 Toxicol. 35: 191-201.

5 Burton, G.A., Jr., C. Hickey, T. DeWitt, D. Morrison, D. Roper, and M. Nipper. 1996. In situ

6 toxicity testing: Teasing out the environmental stressors. SETAC News 16(5):20-22.

7 Canfield, T.J., N.E. Kemble, W.G. Brumbaugh, M.C. Swift, G.A. Burton, F.J. Dwyer, C.G.

8 Ingersoll, and J. Fairchild. 1994. Use of benthic invertebrate community structure and the

9 sediment quality triad to evaluate metal-contaminated sediment in the Upper Clark Fork River,

10 Montana. Environmental Toxicology and Chemistry 13:1999-2012.

11 CCME (Canadian Council of Ministers of the Environment). 1999. Canadian tissue residue

12 guidelines for the protection of wildlife consumers of aquatic biota: polychlorinated biphenyls

13 (PCBs). In: Canadian Environmental Quality Guidelines, 1999. Canadian Council of Ministers

14 of the Environment, Winnipeg.

15 Chapman, RM. 1996. Presentation and interpretation of sediment quality triad data.

16 Ecotoxicology 5: 327-339.

17 Chapman, P.M. and G.S. Mann. 1999. Sediment quality values (SQVs) and ecological risk

18 assessment (ERA). Marine Pollution Bulletin 38(5):339-344.

19 de Bruijn, J., T. Crommentuijn, K. van Leeuwen, E. van der Plassche, D. Sijm, and M. van der

20 Weiden. 1999. Environmental Risk Limits in the Netherlands. Rijksinstituut voor

21 Volksgezondheid en Milieu (RIVM). Dutch Nationale Institute of Public Health and the

22 Environment.

23 Dillon, T.M., W.H. Benson, R.A. Stackhouse, and A.M. Crider. 1990. Effects of selected PCB

24 congeners on survival, growth and reproduction in Daphnia magna. Environmental Toxicology

25 and Chemistry 9:1317-1326.

26 Duke, T.W., J.I. Lowe, and A.J. Wilson Jr. 1970. A polychlorinated biphenyl (Aroclor 1254®) in

27 the water, sediment and biota of Escambia Bay, Florida. Bulletin of Environmental

28 Contamination and Toxicology 5:171-180.

29 Eaton, D.L. and C.D. Klaassen. 1995. Principles of Toxicology. In: Klaassen, C.D., Ed. Casarett

30 and Doulls Toxicology- The Basic Science of Poisons, 5th Edition. McGraw-Hill, New York,

31 NY, USA, pp. 13-33.

32 Efroymson, R.A., G.W. Suter II, B.E. Sample, and D.S. Jones. 1997. Preliminary Remediation

33 Goals for Ecological Endpoints. Prepared for the U.S. Department of Energy. 27 pp.

34 Environment Canada. 1995. Interim Sediment Quality Guidelines. Ecosystem Conservation

35 Directorate, Evaluation and Interpretation Branch, Soil and Sediment Quality Section,

36 Guidelines Branch, Ottawa, Ontario.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-110

-------
1 EPA (United States Environmental Protection Agency). 1996. ECO Update. Ecotox thresholds.

2 EPA 540/F-95/038. Office of Solid Waste and Emergency Response. Washington, DC, USA.

3 EPA (U.S. Environmental Protection Agency). 1999. National Recommended Water Quality

4 Criteria - Correction. April 1999. Office of Water. EPA 822-Z-99-001.

5 EPA (U.S. Environmental Protection Agency). 2000. Methods for Measuring the Toxicity and

6 Bioaccumulation of Sediment-Associated Contaminants with Freshwater Invertebrates. EPA

7 600/R-99/064. 192 pp.

8 EVS (Environment Consultants Ltd.). 2002. Global Overview of the Protocols Used in

9 Development of Water Quality Guidelines for the Protection of Aquatic Life for Metals. Prepared

10 for Canadian Council of Ministers of the Environment. Hull, Quebec, Canada.

11 EVS Environment Consultants. 2003. Assessment of in situ Stressors and Sediment Toxicity in

12 the Lower Housatonic River. Final Report, adapted from a study by G.A. Burton, 2001 Institute

13 for Environmental Quality, Wright State University, Dayton Ohio.

14 Fisher, S.W., S.W. Chordas III, and P.F. Landrum. 1999. Lethal and sublethal body residues for

15 PCB intoxication in the oligochaete, Lumbriculus variegatus. Aquatic Toxicology 45:115-126.

16 Hansen, D.J., P.R. Parrish and J. Forester. 1974. Aroclor 1016: toxicity to and uptake by

17 estuarine animals. Environmental Research 7:363-373.

18 Jarvinen, A.W. and G.T. Ankley. 1998. Linkage of Effects to Tissue Residues: Development of a

19 Comprehensive Database for Aquatic Organisms Exposed to Inorganic and Organic Chemicals.

20 SETAC Technical Publications Series.

21 Jarvinen, A.W. and G.T. Ankley. 1999. Linkage of Effects to Tissue Residues: Development of a

22 Comprehensive Database for Aquatic Organisms Exposed to Inorganic and Organic Chemicals.

23 SETAC Press, Pensacola, FL. 364 pp.

24 Krantzberg, G., J.H. Hartig and M.A. Zarull. 2000. Sediment management: Deciding when to

25 intervene. Environmental Science and Technology - News, 1 January 2000, pp. 22A-27A.

26 Landrum, P.F., H. Lee II, and M.J. Lydy. 1992. Toxicokinetics in aquatic systems: Model

27 comparisons and use in hazard assessment. Environ. Toxicol. Chem. 11: 1709-1725.

28 Lester, D.C. and A. Mcintosh. 1994. Accumulation of polychlorinated biphenyl congeners from

29 Lake Champlain sediments by My sis relicta. Environmental Toxicology and Chemistry. 13:1825-

30 1841.

31 Long, E.R., and P.M. Chapman. 1985. A sediment quality triad: measures of sediment

32 contamination, toxicity and infaunal community composition in Puget Sound. Marine Pollution

33 Bulletin 16:405-415.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-lll

-------
1 Long, E.R. and L.G. Morgan. 1990. The Potential for Biological Effects of Sediment Sorbed

2 Contaminants Tested in the National Status and Trends Program. National Oceanic and

3 Atmospheric Administration, Seattle, WA.

4 Long, E.R., D.D. MacDonald, S.L. Smith, F.D. Calder. 1995. Incidence of adverse biological

5 effects within ranges of chemical concentrations in marine and estuarine sediments.

6 Environmental Management 19(l):81-97.

7 Lowe, J.I., P.R. Parrish, J.M. Patrick Jr., and J. Forester. 1972. Effects of the polychlorinated

8 biphenyl Aroclor® 1254 on the American oyster Crassostrea virginica. Marine Biology 17:209-

9 214.

10 MacDonald, D.D., C.G. Ingersoll and T.A. Berger. 2000a. Development and evaluation of

11 consensus-based sediment quality guidelines for freshwater ecosystems. Arch. Environ. Contam.

12 Toxicol. 39:20-31.

13 MacDonald, D.D., L.M. DiPinto, J. Field, C.G. Ingersoll, E.R. Long and R.C. Swartz. 2000b.

14 Development and evaluation of consensus-based sediment effect concentrations for

15 polychlorinated biphenyls. Envir. Tox. Chem. 19(5): 1403-1413.

16 McFarland, V.A. 1994. Activity-Based Evaluation of Potential Bioaccumulation for Sediments.

17 In: Dredging and Dredged Disposal, Volume 1, R.L. Montgomery, J.W. Leach, Eds. American

18 Society of Civil Engineers, NY, pp. 461-467.

19 McLeese, D.W., C.D. Metcalfe, and D.S. Pezzack. 1980. Uptake of PCBs from sediment by

20 Nereis virens and Crangon septemspinosa. Arch. Environ. Contam. Toxicol. 9:507-518.

21 Menzie, C., M.H. Henning, J. Cura, K. Finkelstein, J. Gentile, J. Maughan, D. Mitchell, S.

22 Petron, B. Potocki, S. Svirsky, and P. Tyler. 1996. Special report of the Massachusetts Weight-

23 of-Evidence Workgroup: a weight of evidence approach for evaluating ecological risks. Human

24 and Ecological Risk Assessment 2:277-3 04.

25 Minshall, G.W. 1984. Aquatic insect-substratum relationships. In: Resh, V.H., and Rosenberg,

26 D.M. (eds.), The Ecology of Aquatic Insects. Praeger, New York. pp. 358-400.

27 Nebeker, A.V. and F.A. Puglisi. 1974. Effect of polychlorinated biphenyls (PCBs) on survival

28 and reproduction of Daphnia, Gammarus, and Tany tarsus. Transactions of the American

29 Fisheries Society 103:722-728.

30 Neff, J.M. and C.S. Giam. 1977. Effects of Aroclor® 1016 and Halowax® 1099 on juvenile

31 horseshoe crabs Limulus polyphemus. In: Physiological Responses of Marine Biota to Pollutants.

32 F.J. Vernberg, A. Calabrese, F.P. Thurberg and W.B. Vernberb, Editors. Academic Press, New

33 York, NY.

34 NJDEP (New Jersey Department of Environmental Protection). 1998. Guidance for Sediment

35 Quality Evaluations. New Jersey Department of Environmental Protection, Trenton, NJ. 29 pp.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-112

-------
1 NYSDEC (New York State Department Environmental Conservation). 1999. Technical

2 Guidance for Screening Contaminated Sediments. New York State Department Environmental

3 Conservation, Division of Fish, Wildlife and Marine Resources. January 1999. 39 pp.

4 Nimmo, D.R., J. Forester, P.T. Heitmuller, and G.H. Cook. 1974. Accumulation of Aroclor®

5 1254 in grass shrimp (Palaemonetes pugio) in laboratory and field experiments. Bulletin of

6 Environmental Contamination and Toxicology 11:303-308.

7 OMEE (Ontario Ministry of the Environment and Energy). 1996. Guidelines for Protection and

8 Management of Aquatic Sediment Quality in Ontario.

9 Parkerton, T.F. 1993. Estimating Toxicokinetic parameters for modeling the bioaccumulation of

10 non-ionic organic chemicals in aquatic organisms. PhD dissertation, Rutgers, the State

11 University of New Jersey, Graduate Program of Environmental Science, New Brunswick, New

12 Jersey. 335 pp.

13 Persaud, D., R. Jaagumagi, and A. Hayton. 1993. Guidelines for the Protection and Management

14 of Aquatic Sediment Quality in Ontario. Ontario Ministry of the Environment and Energy.

15 Toronto, Ontario, Canada.

16 Pinkney, A.E., G.V. Poje, R.M. Sansur, C.C. Lee, and J.M. O'Connor. 1985. Uptake and

17 retention of 14C-Aroclor® 1254 in the amphipod, Gammarus tigrinus, fed contaminated fungus,

18 Fusarium oxysporum. Archives of Environmental Contamination and Toxicology 14:59-64.

19 Plafkin, J.L., Barbour, M.T., Porter, K.D., Gross, S.K. and Hughes, R.M. 1989. Rapid

20 Bioassessment Protocols for Use in Streams and Rivers- Benthic Macroinvertebrates and Fish.

21 EPA/444/4-89/001. United States Environmental Protection Agency, Washington, DC.

22 Reynoldson, T. B., R.C. Bailey, K.E. Day, and R. H. Norris. 1995. Biological guidance for

23 freshwater sediment based on Benthic Assessment of Sediment (the BEAST) using a

24 multivariate approach for predicting biological state. Aust. J. Ecol. 20:198-219.

25 Sanders, H.O. and J.H. Chandler. 1972. Biological magnification of a polychlorinated biphenyl

26 (Aroclor 1254) from water by aquatic invertebrates. Bulletin of Environmental Contamination

27 and Toxicology 7(5):257-263.

28 Sodergren, A. and B. Svensson. 1973. Uptake and accumulation of DDT and PCB by Ephemera

29 danica (Ephemeroptera) in continuous-flow systems. Bulletin of Environmental Contamination

30 and Toxicology 9:345-350.

31 Swartz, R.C. 1999. Consensus sediment quality guidelines for polycyclic aromatic hydrocarbon

32 mixtures. Environ. Tox. Chem. 18(4):780-787.

33 Tidepool Scientific Software. 1994. TOXCALC: Comprehensive toxicity data analysis and

34 database software. Version 5.0.23. Tidepool Scientific Software, McKinleyville, CA.

35 Van den Berg, M, L. Birnbaum, A.T.C. Bosveld, B. Brunstrom, P. Cook, M. Feeley, J.P. Giesy,

36 A. Hanberg, R. Hasegawa, S.W. Kennedy, T. Kubiak, J.C. Larsen, F.X.R. van Leeuwen, A.K.D.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-113

-------
1 Liem, C. Nolt, R.E. Petersen, L. Poellinger, S. Safe, D. Schrenk, D. Tillitt, M. Tysklind, M.

2 Younes, F. Waern, and T. Zacharewski. 1998. Toxic equivalency factors (TEFs) for PCBs,

3 PCDDs, PCDFs for humans and wildlife. Environmental Health Perspectives 106:775-792.

4 Veldhuizen-Tsoerkan, M.B., D.A. Holwerda, and D.I. Zandee. 1991. Anoxic survival time and

5 metabolic parameters as stress indices in sea mussels exposed to cadmium or polychlorinated

6 biphenyls. Archives of Environmental Contamination and Toxicology 20:259-265.

7 Ward, J.V. 1992. Aquatic Insect Ecology. Chapter 1. Biology and Habitat. John Wiley and Sons,

8 Inc. New York, New York. 438 pp.

9 WDOE (Washington State Department of Ecology). 1997. Water Quality Standards for Surface

10 Waters of the State of Washington. Chapter 173-201A WAC. November 18, 1997.

11 WESTON (Roy F. Weston, Inc.). August 2000. Supplemental Investigation Work Plan for the

12 Lower Housatonic River. Prepared for U.S. Army Corps of Engineers and U.S. Environmental

13 Protection Agency. DCN: GEP2-020900-AAME.

14 Wildhaber, M. L. and Schmitt, C.J. 1996. Indices of benthic community tolerance in

15 contaminated Great Lakes sediments: relations with sediment contaminant concentrations,

16 sediment toxicity, and the sediment quality triad. Environ. Mon. Assess. 49:23-49.

17 Wogram, J. and Liess, M. 2001. Rank ordering of macroinvertebrate species sensitivity to toxic

18 compounds by comparison with that of Daphnia magna. Bull. Environ. Contam. Toxicol.

19 67:360-367.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD.DOC

D-114

-------
APPENDIX D

ASSESSMENT ENDPOINT—COMMUNITY STRUCTURE, SURVIVAL,
GROWTH, AND REPRODUCTION OF BENTHIC INVERTEBRATES

TABLES

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
LIST OF TABLES

Title Page

Table D.2-1 Correspondence of Sampling ID Codes and Samples from Different

Elements of the Benthic Invertebrate Risk Assessment 1

Table D.2-2 Summary of Central Tendency Measures for tPCB Concentrations by

Sampling Station, for Benthic Grab Samples 2

Table D.2-3 Concentrations of tPCB in Sediment Collected for Wright State
University Laboratory and In Situ Toxicity and Bioaccumulation Tests
(Burton 2001) 3

Table D.3-1 Results of Pairwise Statistical Tests Comparing Exposed Stations to
Negative Control (T-Ctrl) and Reference (Al, A3) Sediment (Water-
Only Exposures Excluded) 4

Table D.3-2 In Situ Evaluation of Toxicity in Housatonic River Sediment (Station-

by-Station Assessment) 6

Table D.3-3 Laboratory Evaluation of Toxicity in Housatonic River Sediment

(Station-by-Station Assessment) 7

Table D.3-4 Weight-of-Evidence Evaluation of Housatonic River Sediment

Toxicity, Relative to Background Responses 8

Table D.3-5 Summary of Endpoints Calculated for the 48-h and 10-d Acute In Situ

Toxicity Tests (Sediment Exposures) 9

Table D.3-6 Summary of Endpoints Calculated for the Hyalella azteca Chronic

(42-d) Laboratory Toxicity Tests 10

Table D.3-7 Summary of Endpoints Calculated for the Chironomus tentans Chronic

(43-d) Laboratory Toxicity Tests 12

Table D.3-8 Summary of Segmented Linear Regression Concentration Models
Applied to Toxicity Data, Relating Relative Performance Proportion
(RPP) to Sediment tPCB Concentrations 13

Table D.3-9 Summary of Available Invertebrate Tissue Residue Effects Data for

PCBs 14

Table D.3-10 Sediment Quality Values (SQVs) for Use in the Risk Evaluation of

Housatonic River Sediment COPCs 18

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC • • • 7/10/2003

-------
LIST OF TABLES
(Continued)

Title Page

Table D.3-11 Definitions for Environmental Quality Value (Sediment and Water)

Terms 25

Table D.3-12 Water Quality Benchmarks Used in the Risk Evaluation of Housatonic

River Benthic Invertebrates 27

Table D.3-13 Benthic Metric Candidates Considered for Multivariate Analyses 28

Table D.3-14 Summary of Vegetative Status in Vicinity of Benthic Community

Sampling Locations, at Exposed and Reference Locations 30

Table D.4-1 Weighting of Measurement Endpoints for Weight-of-Evidence

Evaluation 32

Table D.4-2 Evidence of Harm and Magnitude of Effects for Measurement

Endpoints Related to Maintenance of a Healthy Benthic Community 33

Table D.4-3 Weight-of-Evidence Risk Analysis Summary Indicating Concurrence

Among Endpoints for Coarse-Grained and Fine-Grained Sediment 34

Table D.4-4 Total PCB Data for Benthic Invertebrates Collected at West Cornwall,

Connecticut (1994-2001) 36

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC • 7/10/2003

-------
Table D.2-1

Correspondence of Sampling ID Codes and Samples from Different Elements of
the Benthic Invertebrate Risk Assessment

Benthic
IDa

Location
Description

River Mile

WESTON
Station Location
IDb

WSU Site
IDC

Benthic
Taxonomy
Samples'1

D-net Benthic Tissue
Samples6

Dalton Reference

144

HO-SEEC0011

011

12 replicates

Predators; Shredders

Upper West
Branch Reference

138

HW-SE000161

12 replicates

None

Lower West
Branch Reference

135.5

HW-SE000398

398

12 replicates

Predators; Shredders

Holmes Road

134.03

H3-SE000051

12 replicates

Predators; Shredders

0.25 Mile Below
Holmes Road

133.79

H3-SE000060

12 replicates

Shredders

1.25 Miles Below
Holmes Road

133.18

H3-SD043702

12 replicates

Predators

1.5 Miles Below
Holmes Road

132.34

H3-SEEC0019

019

12 replicates

Predators; Shredders

Immediately
Above WWTP

130.32

H3-SE000428

428 (in situ)

12 replicates

Predators; Shredders

0.25 Mile Below
New Lenox Road

128.7

H3-SE000116

12 replicates

Shredders

2 Miles Below
New Lenox Road

126.38

H3-SE000389

389

12 replicates

Predators; Shredders

0.5 Mile Above
Woods Pond

125.65

H3-SEEC0031

031

12 replicates

Predators; Shredders

0.5 Mile Above
Woods Pond

125.65

H3-SEEC0023

023 (lab)

None

Immediately
Above Woods
Pond

124.5

H4-SE000246

12 replicates

Predators; Shredders

Threemile Pond
Reference

H9-SE000615

12 replicates

Predators; Shredders

a Simplified ID format used throughout the benthic invertebrate portion of the ERA.

b Location of the sediment sample used to mark the center of the benthic station; exposure and effects data from
surrounding stations (within 5-meter radius) were also included in the ERA.

0 Label used by Wright State University to depict exposure and effects data collected in conjunction with toxicity
testing (EVS, 2003).

d At each sampled benthic station, 12 field replicates were collected within a diameter of a few meters. Each
replicate has a separate benthic taxonomy and chemistry ID. For example, at Station 5 (H3-SE000428), the 12
benthic taxonomy IDs range from H3-MI000015-0-9U30 to H3-MI000026-0-9U30 and the associated sediment
chemistry IDs range from H3-SE000517 to H3-SE000528.

e Table indicates benthic invertebrate tissue types evaluated for chemistry (PCBs and other COPCs).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC i 7/10/2003

-------
Table D.2-2

Summary of Central Tendency Measures for tPCB Concentrations by Sampling

Station, for Benthic Grab Samples

Benthic Station ID

tPCB (mg/kg)

TOC-Normalized tPCB (mg/kg OC)

Median"

Mean ± SD

Median

Mean + SD

Coarse-Grained Reference Locations (C/R)

NDb (0.25)

ND (0.26 + 0.01)

ND (127)

ND (115 + 56)

ND (0.25)

ND (0.25 + 0.02)

ND (95)

ND (116 + 39)

ND (0.30)

ND (0.33 +0.07)

ND (12)

ND (12 + 3.2)

Coarse-Grained Contaminated Locations (C/C)

9.82

42.2 + 70.6

3273

15,324 + 24,234

25.25

37.2 + 37.8

4599

8,709 + 9,124

12.2

22.8 + 29.9

4547

10,247 + 19,878

5.80

9.4 + 11.9

1359

3,572 + 6,926

5.06

7.7 + 8.4

821

1,960 + 2,576

Fine-Grained Contaminated Locations (F/C)

4.34

4.13 + 1.33

197

224 + 81

3.17

4.12 + 3.65

105 + 98

14.05

16.02+12.93

216

240 + 167

0.98

2.18 + 2.41

35 + 36

Fine-Grained Reference Location (F/R)

ND (0.45)

ND (0.43 +0.09)

ND (4.9)

ND (5.4+ 1.5)

aNote: tPCB concentrations were lognormally distributed, so median is superior central tendency measure to
arithmetic mean.

bND = majority of 12 replicate values not detected. Where tPCB was not detected, concentrations were
estimated using half the detection limit value.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.2-3

Concentrations of tPCB in Sediment Collected for Wright State University
Laboratory and In Situ Toxicity and Bioaccumulation Tests (EVS 2003)

Benthic

WSU

Laboratory

48-h In Situ

7-d In Situ

10-d In Situ

Site IDa

Samples

011

0.028

0.00011

0.0071

0.0014

398

0.28

0.38

5.40

0.082

019

8.7

0.95

0.67

14.0

428b

16.0°

7.30

17.0

1.36

389

213

139.3

7.09

52.3

031

72.0

521.7

16.88

112

023b

31.2

aLabel used by Wright State University to depict exposure and effects data combined in conjunction with toxicity
testing (EVS, 2003).

b Station 023 used only for laboratory tests; Station 428 used only for in situ tests. Whereas WSU considered tPCB
concentrations at these two sites to be similar, EVS (2003) separated the results due to potential for high small-
scale variability and the large physical separation of the locations.

0 Chemistry data reported in WESTON database, but no test conducted for this treatment.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.3-1

Results of Pairwise Statistical Tests Comparing Exposed Stations to Negative Control (T-Ctrl) and Reference

(A1, A3) Sediment (Water-Only Exposures Excluded)

Station

Pairwise Comparison

T-Ctri

T-Ctrl

28-d Hyalella survival

Yes

35-d Hyalella survival

Yes

42-d Hyalella survival

Yes

28-d Hyalella dry weight

42-d Hyalella dry weight

42-d Hyalella young per
female

Yes

42-d Hyalella mean young

Yes

20-d Chironomus survival

Yes

43-d Chironomus
emergence

Yes

20-d Chironomus dry
weight

Yes

20-d Chironomus ash-free
dry weight

Yes

48-h Hyalella survival
(sediment)

Yes

10-d Hyalella survival
(sediment)

Yes

48-h Chironomus survival
(sediment)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC a 7/10/2003

-------
Table D.3-1

Results of Pairwise Statistical Tests Comparing Exposed Stations to Negative Control (T-Ctrl) and Reference

(A1, A3) Sediment (Water-Only Exposures Excluded)

(Continued)

Station

Pairwise Comparison

T-Ctri

T-Ctrl

10-d Chironomus survival
(sediment)

Yes

48-h Daphnia survival
(sediment)

Yes

48-h Lumbriculus survival
(sediment)

Yes = Statistically different at alpha = 0.05
No = Not statistically different at alpha = 0.05

NC = Sublethal endpoint not calculable (due to zero survival in treatment)
NA = Not applicable; not tested for endpoint/station combination

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.3-2

In Situ Evaluation of Toxicity in Housatonic River Sediment (Station-by-Station Assessment)

Sampling Station
(ID, Location, WESTON ID)

Median

Bulk
Sediment
[PCB]

(mg/kg)

H. azteca

48-h
Survival
(Water -
Sediment)

C. tentans

48-h
Survival
(Water -
Sediment)

L. Varie-
gatus
48-h
Survival
(Water-
Sediment)

I), magna

48-h
Survival
(Water-
Sediment)

H. azteca

10-d
Survival
(Water-
Sediment)

C. tentans

10-d
Survival
(Water-
Sediment)

L. varie-

gatus
Residue
(mg/kg
lipid)

Overall
Assessment

Dalton Reference

011

0.018

O-O

5.3

Lower West
Branch Reference

398

0.28

o-o

O-O

20.7

1.5 miles below
Holmes Rd.

019

5.9

O-O

o-o

232.6

Near WWTP
Discharge

428

7.3

o-o

380.6

2 miles below
New Lenox Road

389

o-m

o-o

O-0t

O -

O-0t

128.3

•

Vi mile above
Woods Pond

031

o-o

O-0t

314.5

•

O = Negligible to low toxicity: less than 20% effect size relative to negative control. Overall assessment - negligible indication of ecological risk.
O = Moderate toxicity: 20 to 50% effect size relative to negative control. Overall assessment - ecological effects possible, but not conclusive.

• = High toxicity; greater than 50% effect size relative to negative control. Overall assessment - strong indication of potential ecological effects.

• = Very strong toxic response for individual endpoint; greater than 90% effect size relative to negative control.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.3-3

Laboratory Evaluation of Toxicity in Housatonic River Sediment (Station-by-Station Assessment)

Sampling Station
(ID, Location, WESTON ID)

Median Bulk
Sediment
[PCB]

(mg/kg)

H. azteca
Survival
(28d - 35d
- 42d)

H. azteca
28-42 day
Reproduction
(young/female)

H. azteca
Dry weight
(28-d - 42 d)

C. tentans
Survival

C. tentans
Growth
(Total Wt-
Ash Free)

C. tentans
Emergence

Overall
Assessment

A1 Dalton Reference

011

0.018

o-o-o

O - •

. , Lower West Branch
Reference

398

0.28

o-o-o

4 1.5 miles Below Holmes Rd.

019

5.9

o-o-o

•

2 miles Below New Lenox
Road

389

- NA -

NA-NA

•

8 '/? mile above Woods Pond

031

•

8A mile above Woods Pond

023

4.6

•

O = Negligible to low toxicity: less than 20% effect size relative to negative controls. Overall assessment - negligible indication of ecological risk.
O = Moderate toxicity: 20 to 50% effect size relative to negative controls. Overall assessment - ecological effects possible, but not conclusive.

• = High toxicity; greater than 50% effect size relative to negative controls. Overall assessment - strong indication of potential ecological effects.

• = Very strong toxic response for individual endpoint; greater than 90% effect size relative to negative control.

NA = Sublethal endpoint not measured due to complete mortality in treatment.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.3-4

Weight-of-Evidence Evaluation of Housatonic River Sediment Toxicity, Relative to Background Responses

Sampling Station
(ID, Location, WESTON
ID)

Chronic Laboratory Endpoints
(20d, 42d)

Acute In Situ Endpoints (48h, lOd)

Overall
Assessment

H. azteca
Laboratory
(Survival,
Growth,
Reproduction)

C. tentans
Laboratory
(Survival,
Emergence,
Growth)

H. azteca
In situ
Survival
(Water,
Sediment)

C. tentans
In situ
Survival
(Water,
Sediment)

I). magna
In situ
Survival
(Water,
Sediment)

L. variegatus
In situ
Survival
(Water,
Sediment)

1.5 miles below
Holmes Rd.

019

o-o-o

O-O

5 Near WWTP
Discharge

428

NA-NA-NA

NA - NA - NA

o-o

O-O

2 miles below
7 New Lenox
Road

389

• - NA - NA

o - •

o-m

O-O

•

„ . Vi mile above
Woods Pond

023

• - O - •

NA-NA

•

g Vi mile above
Woods Pond

031

• - O - •

o-m

O - •

o-m

O-O

•

O = Negligible to low toxicity: less than 20% effect size relative to upstream background (Al, A3). Overall assessment - negligible indication of ecological risk.
O = Moderate toxicity: 20 to 50% effect size relative to upstream background (Al, A3). Overall assessment - ecological effects possible, but not conclusive.
• = High toxicity; greater than 50% effect size relative to upstream background (Al, A3). Overall assessment - strong indication of ecological effects.
NA = Endpoint not measured due to complete mortality in treatment (Location 7), or samples not collected at station (Locations 5, 8A).

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC q 7/10/2003

-------
Table D.3-5

Summary of Endpoints Calculated for the 48-h and 10-d Acute In Situ Toxicity Tests (Sediment Exposures)

Species

Endpoint

Compared To

Results (mg/kg tPCB)

NOAEL

LOAEL

LCS0 by Probit (95%
CL)a

LC20 by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

H. azteca

48-h survival

Control (Laboratory)

77.2

>77.2

22.1 (NC)* c

7.9 (NC)* c

11.5 (10.0- 13.3)c

Reference (Al)

8.3

54.1

20.4 (NC)* c

7.1 (NC)* c

8.2 (7.9 - 8.5)c

Reference (A3)

77.2

>77.2

19.4 (NC)* c

6.6 (NC)* c

8.2 (7.9 - 8.4)c

H. azteca

10-d survival

Control (Laboratory)

8.3

54.1

6.2 (NC)*

0.9 (NC)*

10.3 (9.0-11.9)

Reference (Al)

8.3

54.1

12.0(3.0-186.7)*

6.6 (0-15.0)*

13.0(12.3-13.8)

Reference (A3)

8.3

54.1

12.4

6.8 (0-16.7)*

13.4(12.7-14.2)

C. tentans

48-h survival

Control (Laboratory)

77.2

>77.2

Reference (Al)

77.2

>77.2

Reference (A3)

77.2

>77.2

C. tentans

10-d survival

Control (Laboratory)

8.3

54.1

26.2 (0.7 - 59.7)*

12.2(0-28.5)*

20.1 (18.5-21.7)

Reference (Al)

8.3

54.1

24.8 (0.8 - 66.2)*

11.3 (0-27.6)*

20.4(18.8-22.2)

Reference (A3)

8.3

54.1

30.6 (NC)*

15.8 (NC)*

22.0 (20.6 - 23.4)

D. magna

48-h survival

Control (Laboratory)

8.3

54.1

8.4(7.5-12.1)

6.6(3.2-7.4)

7.3 (2.3-23.4)

Reference (Al)

<5.9

5.9

8.1 (7.8-8.5)*

6.1 (5.8-6.3)*

10.0(9.2-10.7)

Reference (A3)

8.3

54.1

8.5 (8.2-8.9)*

7.0 (6.5-7.3)*

12.2(11.4-12.9)

L. variegatus

48-h survival

Control (Laboratory)

77.2

>77.2

>77.2 (NC)

>77.2

Reference (Al)

77.2

>77.2

>77.2 (NC)

>77.2

Reference (A3)

77.2

>77.2

>77.2 (NC)

>77.2

aLC50 or LC20 value calculated using Probit method; asterisks (*) indicate where Chi-squared value exceeded critical value.

bLC50 value calculated using Trimmed Spearman-Karber (TSK) method; method is preferred where Probit method Chi-squared value exceeds critical value.
°Highest PCB concentration (77.2 mg/kg) excluded because of anomalous concentration-response.

NC = not calculable

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.3-6

Summary of Endpoints Calculated for the Hyalella azteca Chronic (42-d) Laboratory Toxicity Tests

Endpoint

Compared To

Results (mg/kg PCB)

NOAEL

LOAEL

LCS0/ICS0 by Probit
(95% CL) a

LC20/IC20 by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

28-d survival

Control (Trout Farm)c

0.28

4.56

6.0 (NC)*

0.9 (NC)*

4.1 (3.8-4.5)

Reference (Al)

<4.56

4.56

6.6 (NC)*

4.2 (NC)*

8.4(7.4-9.7)

Reference (A3)

<4.56

4.56

4.4 (NC)*

0.6 (NC)*

35-d survival

Control (Trout Farm)

0.28

4.56

8.6 (NC)*

0.3 (NC)*

5.3 (4.1-6.9)

Reference (Al)c

<4.56

4.56

11.8 (NC)*

0.2 (NC)*

12.8 (7.2-22.8)

Reference (A3)

<4.56

4.56

7.9 (NC)*

0.4 (NC)*

42-d survival

Control (Trout Farm)

0.28

4.56

10.7 (NC)*

0.4 (NC)*

6.6(5.1-8.5)

Reference (Al)c

<4.56

4.56

13.9 (NC)*

0.2 (NC)*

14.1 (8.9-22.3)

Reference (A3)

<4.56

4.56

9.7 (NC)*

0.5 (NC)*

28-d dry weight

Control (Trout Farm)

77.2

>77.2

3.1

Reference (Al)

77.2

>77.2

Reference (A3)

77.2

>77.2

3.4

42-d dry weight

Control (Trout Farm)c

77.2

>77.2

44.0

Reference (Al)

77.2

>77.2

73.2

Reference (A3)

77.2

>77.2

61.7

35-d no. young

Control (Trout Farm)c

0.28

4.56

1.6 (0.0-3.5)

<0.018 (NC)

Reference (Al)c

<4.56

4.56

4.1 (2.8-41.6)

1.6(1.1-20.2)

Reference (A3)

<4.56

4.56

3.2(2.7-5.9)

1.4(1.3-2.0)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> rx 7/10/2003

-------
Table D.3-6

Summary of Endpoints Calculated for the Hyalella azteca Chronic (42-d) Laboratory Toxicity Tests

(Continued)

Endpoint

Compared To

Results (mg/kg PCB)

NOAEL

LOAEL

LCS0/ICS0 by Probit
(95% CL) a

LC20/IC20 by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

42-d total young

Control (Trout Farm)

0.28

4.56

0.8(0.0-2.2)

<0.018 (NC)

Reference (Al)c

<4.56

4.56

3.7 (2.8-40.6)

1.5(1.1-17.4)

Reference (A3)

<4.56

4.56

3.3 (2.8-4.2)

1.5(1.3-1.9)

42-d young/female

Control (Trout Farm)

0.28

4.56

2.1 (0.0-4.7)

<0.018 (NC)

Reference (Al)

5.9

77.2

16.4(0.0-52.3)

2.2(1.1-26.8)

Reference (A3)c

<4.56

4.56

4.5 (3.2-31.1)

2.0(1.4-3.3)

aLC50 or LC2o value calculated using Probit method; asterisks (*) indicate where Chi-squared value exceeded critical value.

' LCso value calculated using Trimmed Spearman-Karber (TSK) method; method is preferred where Probit method Chi-squared value exceeds critical value.
Interrupted dose-response observed.

NA = not applicable; NC = not calculated

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.3-7

Summary of Endpoints Calculated for the Chironomus tentans Chronic (43-d) Laboratory Toxicity Tests

Endpoint

Compared To

Results (mg/kg PCB)

NOAEL

LOAEL

LCSo/ICSo by Probit
(95% CL) a

LC20/IC20 by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

20-d survival

Control (Trout Farm)

0.28

4.56

0.93 (NC)*

0.20 (NC)*

0.8(0.7-0.9)

Reference (Al)

<4.56

4.56

<4.56 (NC)°

<4.56°

Reference (A3)

<4.56

4.56

<4.56 (NC)°

20-d dry weight

Control (Trout Farm)

0.28

5.9C

2.7(1.6-3.5)

0.7 (0.0-2.0)

Reference (Al)

<5.9°

5.9C

3.0 (3.0-3.2)

1.2(1.2-1.3)

Reference (A3)

<5.9°

5.9C

3.1 (3.1-3.3)

1.4(1.4-1.5)

20-d ash-free dry
weight (AFDW)

Control (Trout Farm)

o.ois46

0.018de

3.1 (1.9-3.3)

1.3 (0.0-1.6)

Reference (Al)

3.2 (3.0-3.8)

1.3 (1.2-1.5)

Reference (A3)

3.2 (3.1-3.2)

1.4(1.4-1.5)

43-d emergence

Control (Trout Farm)

0.28

4.56

<0.018 (NC)

<0.018

Reference (Al)

<4.56

4.56

2.5 (2.2-3.5)

1.0 (0.9-1.4)

Reference (A3)

<4.56

4.56

2.6 (2.3-3.7)

1.2(1.1-1.6)

aLC50 or LC2o value calculated using Probit method.

I:,LC\(J value calculated using Trimmed Spearman-Karber (TSK) method. Method is preferred where Probit method Chi-squared value exceeds critical value.
°Based on visual observations (model did not converge).

treatments of 5.9 and 77.2 mg/kg not included due to only a single replicate being available.

"Interrupted dose-response observed.

NA = not applicable; NC = not calculated

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.3-8

Summary of Segmented Linear Regression Concentration Models Applied to Toxicity Data, Relating Relative

Performance Proportion (RPP) to Sediment tPCB Concentrations

Endpoint Type

Model Statistics

Inflection Point (mg/kg)

Flat Intercept

Regression Intercept

Regression Slope

Statistically Significant
(a=0.05)?

Acute Endpoints

4.1

0.95

1.4

-0.75

Yes

42-d Hyalella
Survival

0.28

0.91

0.74

-0.32

Yes

42-d Hyalella
Growth

0.77

0.91

1.05

-0.18

Yes

Remaining Chronic
Endpoints

0.28

0.55

0.42

-0.24

Yes

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> ^ 7/10/2003

-------
Table D.3-9

Summary of Available Invertebrate Tissue Residue Effects Data for PCBs

PCB Type

Test Species

Life
Stage

Exposure
Duration

(days)

Tissue
Residue

(mg/kg wet
weight)

Description of Effect

Reference

Aroclor 1254

Crayfish, Orconectes nais
(freshwater)

Mature

0.04

Mortality - no effect

Sanders and Chandler
1972

Aroclor 1254

Grass shrimp, Palaemonetes pugio
(saltwater)

0.42

Mortality - no effect

Nimmo et al. 1974

Aroclor 1254

Midge, Corydalus cornutus
(freshwater)

Immature

1.02

Mortality - no effect

Sanders and Chandler
1972

Aroclor 1254

Pink shrimp, Penaeus duorarum
(saltwater)

Immature

1.3

Mortality - no effect

Duke et al. 1970

Aroclor 1254

Giant black stonefly, Pteronarcys
dorsata (freshwater)

Immature

1.4

Mortality - no effect

Sanders and Chandler
1972

Aroclor 1254

Glass shrimp, Palaemonetes
kadiakensis (freshwater)

Mature

3.2

Mortality - no effect

Sanders and Chandler
1972

Aroclor 1254

Pink shrimp, Penaeus duorarum
(saltwater)

Immature

3.9

Mortality -100% mortality

Duke et al. 1970

Aroclor 1254

Amphipod, Gammarus tigrinus
(saltwater)

Adult

4.64

Mortality - no effect

Pinkney et al. 1985

Aroclor 1254

Mosquito, Culex tarsalis
(freshwater)

Immature

5.4

Mortality - no effect

Sanders and Chandler
1972

Aroclor 1254

Grass shrimp, Palaemonetes pugio
(saltwater)

5.4

Mortality - no effect

Nimmo et al. 1974

Aroclor 1254

Midge, Chaoborus punctipennis
(freshwater)

Immature

6.0

Mortality - no effect

Sanders and Chandler
1972

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> a 7/10/2003

-------
Table D.3-9

Summary of Available Invertebrate Tissue Residue Effects Data for PCBs

(Continued)

PCB Type

Test Species

Life
Stage

Exposure
Duration

(days)

Tissue
Residue

(mg/kg wet
weight)

Description of Effect

Reference

Aroclor 1254

Mosquito, Culex tarsalis
(freshwater)

Immature

6.0

Development - reduced

Sanders and Chandler
1972

Aroclor 1254

Amphipod, Gammarus
pseudolimnaeus
(freshwater)

Mature

7.8

Mortality - no effect

Sanders and Chandler
1972

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

8.1

Growth -19% reduction in shell
growth rate

Duke et al. 1970

Aroclor 1254

Water flea, Daphnia magna
(freshwater)

Mature

10.4

Mortality - no effect

Sanders and Chandler
1972

Aroclor 1254

Pink shrimp, Penaeus duorarum
(saltwater)

Immature

Mortality - 72% mortality

Duke et al. 1970

Aroclor 1254

Grass shrimp, Palaemonetes pugio
(saltwater)

16.5

Mortality - no effect

Nimmo et al. 1974

Aroclor 1254

Grass shrimp, Palaemonetes pugio
(saltwater)

Mortality - no effect

Nimmo et al. 1974

Aroclor 1254

Blue crab, Callinectes sapidus
(saltwater)

Immature

Mortality - no effect

Duke et al. 1970

Aroclor 1254

Grass shrimp, Palaemonetes pugio
(saltwater)

Mortality - 45% mortality

Nimmo et al. 1974

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

Mortality - no effect

Duke et al. 1970

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> r 7/10/2003

-------
Table D.3-9

Summary of Available Invertebrate Tissue Residue Effects Data for PCBs

(Continued)

PCB Type

Test Species

Life
Stage

Exposure
Duration

(days)

Tissue
Residue

(mg/kg wet
weight)

Description of Effect

Reference

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

Growth - 41% reduction in shell
growth rate

Duke et al. 1970

Aroclor 1254

Pink shrimp, Penaeus duorarum
(saltwater)

Immature

Behavior - no effect on
equilibrium or behavior

Duke et al. 1970

Aroclor 1254

Grass shrimp, Palaemonetes pugio
(saltwater)

Mortality - 60% mortality

Nimmo et al. 1974

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

168

101

Growth - no effect on growth rate

Lowe et al. 1972

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

168

101

Cellular - no significant
histopathology in digestive
diverticular epithelium

Lowe et al. 1972

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

168

425

Growth - reduced growth rate

Lowe et al. 1972

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

168

425

Mortality - no effect

Lowe et al. 1972

Aroclor 1254

American oyster, Crassostrea
virginica
(saltwater)

Immature

168

425

Cellular - atrophy of digestive
diverticular epithelium;
degeneration of vesicular
connective tissues.

Lowe et al. 1972

PCBs

Bentnose clam, Macoma nasuta
(saltwater)

Adult

119

1.7

Behavior - no effect on
burrowing

Boese et al. 1995

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> s 7/10/2003

-------
Table D.3-9

Summary of Available Invertebrate Tissue Residue Effects Data for PCBs

(Continued)

PCB Type

Test Species

Life
Stage

Exposure
Duration

(days)

Tissue
Residue

(mg/kg wet
weight)

Description of Effect

Reference

PCBs

Bentnose clam, Macoma nasuta
(saltwater)

Adult

119

1.7

Growth - no effect

Boese et al. 1995

PCBs

Bentnose clam, Macoma nasuta
(saltwater)

Adult

119

1.7

Mortality - no effect

Boese et al. 1995

PCBs

Opossum shrimp, Mysis relicta
(freshwater)

1.9

Behavior - no effect on feeding

Lester and Mcintosh
1994

tPCB

Oligochaete, Lumbriculus
variegatus
(freshwater)

Mixed

4.41

Mortality - no effect (presumed)

Burton 2001

Clophen A50

Mayfly, Ephemera danica
(freshwater)

Immature

1.3

Growth - no effect

Sodergren and
Svensson 1973

Clophen A50

Mayfly, Ephemera danica
(freshwater)

Immature

1.3

Mortality - no effect

Sodergren and
Svensson 1973

Clophen A50

Mussel, Mytilus edulis
(saltwater)

Adult

3.0

Physiological - no effect on
ability to survive anoxic stress

Velduizen-Tsoerkan et
al. 1991

Clophen A50

Mussel, Mytilus edulis
(saltwater)

Adult

180

7.0

Physiological - significant effect
on ability to survive anoxic stress

Velduizen-Tsoerkan et
al. 1991

Clophen A50

Mussel, Mytilus edulis
(saltwater)

Adult

180

7.0

Physiological - no effect on
adenylate energy charge or
glycogen content

Velduizen-Tsoerkan et
al. 1991

* Results are reported in mmol/kg in this study.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> q 7/10/2003

-------
Table D.3-10

Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment COPCs

Benchmark Value

Benchmark

Contaminant

(mg/kg)

Code

Basis for Benchmark

Source

Metals

Antimony

MPC

Maximum Permissible Concentration (not TOC or clay dependent)

de Bruijnetal. 1999

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

Barium

67.7

MPC

Maximum Permissible Concentration, adjusted to 1% TOC and 1%
clay (typical of sediment above WWTP)

de Bruijn et al. 1999

154.8

MPC

Maximum Permissible Concentration, adjusted to 6% TOC and 10%
clay (typical of sediment below WWTP)

de Bruijn et al. 1999

Cadmium

0.6

LEL

From Persaud et al. (1993)

NYSDEC 1999

SEL

From Long and Morgan (1990)

NYSDEC 1999

Level I

AEL equivalent

BCMELP 1999

3.5

Level II

PEL equivalent

BCMELP 1999

TEC

Consensus-based TEC

MacDonald et al. 2000b

PEC

Consensus-based PEC

MacDonald et al. 2000b

6.5

MPC

Maximum Permissible Concentration, adjusted to 1% TOC and 1%
clay (typical of sediment above WWTP)

de Bruijn et al. 1999

9.1

MPC

Maximum Permissible Concentration, adjusted to 6% TOC and 10%
clay (typical of sediment below WWTP)

de Bruijn et al. 1999

1.5

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

Chromium

26.0

LEL

From Persaud et al. (1993)

NYSDEC 1999

110.0

SEL

From Persaud et al. (1993)

NYSDEC 1999

64.0

Level I

AEL equivalent

BCMELP 1999

90.0

Level II

PEL equivalent

BCMELP 1999

43.4

TEC

Consensus-based TEC

MacDonald et al. 2000b

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> q 7/10/2003

-------
Table D.3-10

Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment COPCs

(Continued)

Benchmark Value

Benchmark

Contaminant

(mg/kg)

Code

Basis for Benchmark

Source

Chromium

111.0

PEC

Consensus-based PEC

MacDonald et al. 2000b

159.0

PEC equivalent

Efroymson et al. 1997

197.6

MPC

Maximum Permissible Concentration, adjusted to 1% TOC and 1%
clay (typical of sediment above WWTP)

de Bruijn et al. 1999

266

MPC

Maximum Permissible Concentration, adjusted to 6% TOC and 10%
clay (typical of sediment below WWTP)

de Bruijn et al. 1999

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

370

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

Copper

16.0

LEL

From Persaud et al. (1993)

NYSDEC 1999

110.0

SEL

From Persaud et al. (1993)

NYSDEC 1999

120.0

Level I

AEL equivalent

BCMELP 1999

200.0

Level II

PEL equivalent

BC MF.I.P 1999

31.6

TEC

Consensus-based TEC

MacDonald et al. 2000b

149.0

PEC

Consensus-based PEC

MacDonald et al. 2000b

34.0

ER-L

From Long et al. (1995)

EPA 1996

77.7

PEC

PEC equivalent

Efroymson et al., 1997

32.9

MPC

Maximum Permissible Concentration, adjusted to 1% TOC and 1%
clay (typical of sediment above WWTP)

de Bruijn et al. 1999

49.9

MPC

Maximum Permissible Concentration, adjusted to 6% TOC and 10%
clay (typical of sediment below WWTP)

de Bruijn et al. 1999

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

270

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

Lead

35.0

ISQG

CCME 1999

91.3

PEL

CCME 1999

Level I

AEL equivalent

BCMELP 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC -> q 7/10/2003

-------
Table D.3-10

Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment COPCs

(Continued)

Benchmark Value

Benchmark

Contaminant

(mg/kg)

Code

Basis for Benchmark

Source

Lead

Level II

PEL equivalent

BCMELP 1999

LEL

Persaud etal. 1993

250

SEL

Persaud etal. 1993

31.0

LEL

From Persaud et al. (1993)

NYSDEC 1999

110.0

SEL

From Long and Morgan (1990)

NYSDEC 1999

LEL

From Persaud et al. (1993)

NJDEP 1998

250

SEL

From Persaud et al. (1993)

NJDEP 1998

ER-L

From Long et al. (1995)

EPA 1996

35.8

TEC

Consensus-based TEC

MacDonald et al. 2000b

128

PEC

Consensus-based PEC

MacDonald et al. 2000b

110

PEL

Efroymson et al. 1997

324.2

MPC

Maximum Permissible Concentration, adjusted to 1% TOC and 1%
clay (typical of sediment above WWTP)

de Bruijn et al. 1999

411.5

MPC

Maximum Permissible Concentration, adjusted to 6% TOC and 10%
clay (typical of sediment below WWTP)

de Bruijn et al. 1999

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

220

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

Mercury

0.17

ISQG

CCME 1999

0.486

PEL

CCME 1999

0.33

Level I

AEL equivalent

BCMELP 1999

0.49

Level II

PEL equivalent

BCMELP 1999

0.15

LEL

From Long and Morgan (1990)

NYSDEC 1999

1.3

SEL

From Long and Morgan (1990)

NYSDEC 1999

0.2

LEL

From Persaud et al. (1993)

NJDEP 1998

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC 7/10/2003

-------
Table D.3-10

Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment COPCs

(Continued)

Contaminant

Benchmark Value

(mg/kg)

Benchmark
Code

Basis for Benchmark

Source

Mercury

SEL

From Persaud et al. (1993)

NJDEP 1998

0.15

ER-L

From Long et al. (1995)

EPA 1996

0.018

TEC

Consensus-based TEC

MacDonald et al. 2000b

1.06

PEC

Consensus-based PEC

MacDonald et al. 2000b

0.7

PEL

Efroymson et al. 1997

6.8

MPC

Maximum Permissible Concentration, adjusted to 1% TOC and 1%
clay (typical of sediment above WWTP)

de Bruijn et al. 1999

8.1

MPC

Maximum Permissible Concentration, adjusted to 6% TOC and 10%
clay (typical of sediment below WWTP)

de Bruijn et al. 1999

0.15

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

1.0

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

Silver

1.6

Level I

AEL equivalent

BCMELP 1999

2.2

Level II

PEL equivalent

BCMELP 1999

LEL

From Long and Morgan (1990)

NYSDEC 1999

2.2

SEL

From Long and Morgan (1990)

NYSDEC 1999

ER-L;
ISQG-Low

From Long et al. (1995)

NJDEP 1998; ANZECC and
ARMCANZ 2000

3.7

ER-M;
ISQG-High

From Long et al. (1995)

NJDEP 1998; ANZECC and
ARMCANZ 2000

Polycyclic Aromatic Hydrocarbons (PAHs)

Total PAH

ER-L

Based on NSTPA (National Status and Trends Program Approach);
from Long and Morgan (1990)

BCMWLAP 2001

ER-M

Based on NSTPA (National Status and Trends Program Approach);
from Long and Morgan (1990)

BCMWLAP 2001

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC rs -i 7/10/2003

-------
Table D.3-10

Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment COPCs

(Continued)

Contaminant

Benchmark Value

(mg/kg)

Benchmark
Code

Basis for Benchmark

Source

Total PAH

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

100

SEL

When sediment contains 1% organic carbon; from Persaud et al.

(1993)

BCMWLAP 2001

9.2

Level I

AEL; calculated using values from MacDonald et al. (1996)

BC MF.I.P 1999

17.0

Level II

PEC; from MacDonald et al. (1996)

BCMELP 1999

LEL

Persaud etal. 1993

100(1%OC)
1,000 (10% OC)

SEL

Guideline values have been converted from "mg/kg OC" to bulk
sediment concentrations at 1% and 10% OC.

Persaud et al. 1993

LEL

From Persaud et al. (1993).

NJDEP 1998

100(1%OC)
1,000 (10% OC)

SEL

From Persaud et al. (1993). Guideline values have been converted
from "mg/kg OC" to bulk sediment concentrations at 1% and 10%

OC.

NJDEP 1998

ER-L

From Long et al. (1993)

EPA 1996

1.61

TEC

Consensus-based TEC

MacDonald et al. 2000

22.8

PEC

Consensus-based PEC

MacDonald et al. 2000

13.66

PEC

Efroymson et al. 1997

TEL

Consensus-based TEL

Swartz 1999

804

PEL

Consensus-based PEL

Swartz 1999

LP AH

0.1

No effects threshold based on background approach; from 1992
Environment Canada and Quebec Ministry of Environment interim
criteria for assessment of St. Lawrence River sediment.

BCMWLAP 2001

0.880

Level I

AEL; calculated using values from MacDonald et al. (1996)

BCMELP 1999

1.4

Level II

PEC; from MacDonald et al. (1996)

BCMELP 1999

3.369

PEC

Efroymson et al. 1997

TEL

Consensus-based TEL

Swartz 1999

153

PEL

Consensus-based PEL

Swartz 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC rsrs 7/10/2003

-------
Table D.3-10

Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment COPCs

(Continued)

Benchmark Value

Benchmark

Contaminant

(mg/kg)

Code

Basis for Benchmark

Source

LP AH

0.552

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

3.16

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

No effects threshold based on background approach; from 1992

HP AH

1.0

Environment Canada and Quebec Ministry of Environment interim
criteria for assessment of St. Lawrence River sediment.

BCMWLAP 2001

3.7

Level I

AEL; calculated using values from MacDonald et al. (1996)

BC MELP 1999

6.7

Level II

PEC; from MacDonald et al. (1996)

BCMELP 1999

4.354

PEC

Efroymson et al. 1997

TEL

Consensus-based TEL

Swartz 1999

651

PEL

Consensus-based PEL

Swartz 1999

1.7

ISQG-Low

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

9.6

ISQG-High

Adapted from Long et al. (1995)

ANZECC and ARMCANZ 2000

Polychlorinated Biphenyls

PCBs, total

0.034

ISQG

CCME 1999

0.277

PEL

CCME 1999

0.07

LEL

OMEE 1996

5.3 (1% TOC)
53 (10% TOC)

SEL

0.07

LEL

NJDEP 1998

5.3

SEL

NJDEP 1998

0.023

ER-L

Long et al. 1995; ANZECC and
ARMCANZ 2000

0.180

ER-M

Longetal. 1995

0.023

ER-L

EPA 1996

0.37

Level I

AEL; calculated using values from MacDonald et al. (1996)

BCMELP 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC ^ 7/10/2003

-------
Table D.3-10

Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment COPCs

(Continued)

Contaminant

Benchmark Value

(mg/kg)

Benchmark
Code

Basis for Benchmark

Source

PCBs, total

0.68

Level II

PEC; from MacDonald et al. (1996)

BCMELP 1999

0.04

TEC

Consensus-based value

MacDonald et al. 2000a

0.4

MEC

Consensus-based value

MacDonald et al. 2000a

1.7

EEC

Consensus-based value

MacDonald et al. 2000a

0.06

TEC

Consensus-based value

MacDonald et al. 2000b

0.676

PEC

Consensus-based value

MacDonald et al. 2000b

0.18

PRG

Preliminary Remediation Goal

Efroymson et al. 1997

Dioxins/Furans

PCDDs and
PCDFs

0.85 ng TEQ/kg

ISQG

Values are expressed as toxic equivalency (TEQ) units, based on
WHO 1998 TEF values for fish.

CCME 1999

21.5 ng TEQ/kg

PEL

Values are expressed as toxic equivalency (TEQ) units, based on
WHO 1998 TEF values for fish.

CCME 1999

2,3,7,8-TCDD
TEQ

0.10 (J-g/kg

Level I

AEL equivalent

BC MELP 1999

0.19 (J-g/kg

Level II

PEL equivalent

BCMELP 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC rs a 7/10/2003

-------
Table D.3-11

Definitions for Environmental Quality Value (Sediment and Water) Terms

Definition

Source

AEL

Average effect level; arithmetic mean of the TEL (ISQG) and PEL.

BCMELP 1999

ccc

Criteria continuous concentration. An estimate of the highest
concentration of a material in surface water to which an aquatic
community can be exposed indefinitely without resulting in an
unacceptable effect.

EPA 1999

CMC

Criteria maximum concentration. An estimate of the highest
concentration of a material in surface water to which an aquatic
community can be exposed briefly without resulting in an unacceptable
effect.

EPA 1999

ER-L

Effects range - low. The ER-L is the lower 10th percentile of the
effects studies for a chemical. Concentrations below the ER-L
represent a minimal effects range where effects would be rarely
observed.

Long and Morgan
1990

ER-M

Effects range - median. The ER-M is the median, or 50th percentile, or
the effects data for a chemical. Concentrations above the ER-M
represent a probable effects range, where effects would frequently
occur.

Long and Morgan
1990

ISQG

Interim sediment quality value. Refer to TEL.

CCME 1999

ISQG-
High

Comparable to ER-M; developed from Long et al. (1995)

ANZECC/ARMCA
NZ 2000

ISQG-
Low

Comparable to ER-L; developed from Long et al. (1995)

ANZECC/ARMCA
NZ 2000

Level I

Applies to contaminated sites. Level I criteria define the concentrations
of contaminants in bulk sediment and porewater that are unlikely to
cause unacceptable impacts on sediment dwelling organisms. Level I
criteria, also termed AELs, are calculated as the mean of CCME TEL
and PEL values. If TEL and PEL values were not available, the AEL
equivalent values were derived from other comparable guidelines that
were philosophically similar to the Canadian TELs and PELs.

BCMELP 1999

Level II

Applies to contaminated sites. Level II criteria define the
concentrations of contaminants in bulk sediment and porewater that are
likely to cause minor impacts on sediment dwelling organisms. Level
II criteria are PELs or equivalent values.

BCMELP 1999

LEL

Lowest effect level. Indicates a level of contamination which has no
effect on the majority of sediment-dwelling organisms; the sediment is
clean to marginally polluted.

Persaud et al. 1993

MPC

Maximum Permission Concentration, considering background
exposures. MPC values for metals can be generic, or can be modified,
based on site-specific TOC and clay content. The latter option was
chosen for this ERA, using two grain-size types reflecting typical
upstream and downstream portions of the PSA.

de Bruijn et al. 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC rs c 7/10/2003

-------
Table D.3-11

Definitions for Environmental Quality Value (Sediment and Water) Terms

(Continued)

Definition

Source

NEL

No effect level. The level at which the chemicals in sediment do not
affect fish or sediment-dwelling organisms. At this level no transfer of
chemicals through the food chain and no effect on water quality is
expected.

Persaud et al. 1993

PEC

Probable effect concentration. Concentration above which adverse
effects are expected to occur more often than not.

MacDonald et al.
1996, 2000

PEL

Probable effect level. Concentration above which adverse effects are
expected to occur frequently. Derived from BEDS, the PEL is the
geometric mean of the 50th percentile of the effects data set and the 85th
percentile of the no effects data set.

CCME 1999

SEL

Severe effect level. Sediment is considered heavily polluted and likely
to affect the health of sediment dwelling organisms.

Persaud et al. 1993

TEC

Threshold effect concentration. Concentration below which adverse
effects are not expected to occur.

MacDonald et al.
2000

TEL

Threshold effect level. Concentration below which adverse effects are
expected to occur only rarely. Derived from BEDS, the TEL is the
geometric mean of the 15th percentile of the effects data set and the 50th
percentile of the no effects data set.

CCME 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC 7/10/2003

-------
Table D.3-12

Water Quality Benchmarks Used in the Risk Evaluation of Housatonic River Benthic Invertebrates

Contaminant

Benchmark Value

Og/L)

Benchmark
Code

Basis for Benchmark

Source

Polychlorinated Biphenyls (PCBs)

tPCBs

0.014

ccc

EPA 1999

0.014

Chronic

Long-term average not to be exceeded

WDOE 1997

0.001

Provincial water quality objective

OMEE 1996

0.19

Tier II value

EPA 1996

0.0019

From river otter NOAEL. The lowest available
concentration for the protection of piscivores from
any Aroclor

Efroymson et al. 1997

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC rsq 7/10/2003

-------
Table D.3-13

Benthic Metric Candidates Considered for Multivariate Analyses

Category

Metric

Definition

Predicted Response
to Increasing
Perturbation

Abundance and
richness

Total number of taxa"

Measures overall variety of the benthic
assemblage

Decrease

measures

Total abundance3

Measures total abundance of animals
(surrogate for productivity)

Decrease

Number of EPT taxa

Total number of taxa in the "EPT"
orders (Ephemeroptera, Plecoptera,
Tricoptera)

Decrease

Ephemeroptera taxa
(Mayflies)

Total number of taxa in the "E" order

Decrease

Plecoptera taxa
(Stoneflies)

Total number of taxa in the "P" order

Decrease

Tricoptera taxa
(Caddisflies)

Total number of taxa in the "T" order

Decrease

Composition

Percent EPTa

Relative abundance of "EPT"

Decrease

measures

Percent "E"

Relative abundance of "E"

Decrease

Percent "T"

Relative abundance of "T"

Decrease

Dominant taxa

Relative abundance of dominant taxa

Increase

Individual taxa

Relative abundance

Decrease

Dipteranb abundance

Relative abundance: Tolerant families;
Intolerant families

Increase in tolerant
families; decrease for
intolerant

Gastropodb abundance

Relative abundance: Tolerant families;
Intolerant families

Increase in tolerant
families; decrease for
intolerant

Bivalve abundance

Relative abundance; Tolerant families;
Intolerant families

Increase in tolerant
families; decrease for
intolerant

01igochaeteb abundance

Relative abundance: Tolerant families;
Intolerant families

Increase in tolerant
families; decrease for
intolerant

Leechb (Hirudinea)
abundance

Relative abundance of leeches

Increase

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC 7/10/2003

-------
Table D.3-13

Benthic Metric Candidates Considered for Multivariate Analyses

(Continued)

Category

Metric

Definition

Predicted Response
to Increasing
Perturbation

Feeding
measures

Shredders and scrapers

Total and/or relative abundance

Decrease (considered
most sensitive to
pollution)

Filterers, collectors,
gatherers, and
omnivores

Total and/or relative abundance

Increase (considered
most tolerant to
pollution, as they
have a broader range
of acceptable feeding
materials)

Predators

Total and/or relative abundance

Decrease

Tolerance /
Intolerance

Oligochaete tolerance3
Bivalve tolerance
Dipteran tolerance3
Gastropod tolerance3
Leech tolerance

Calculated total number and relative
abundance measures for various taxa,
but subdivided into "sensitive" (i.e.,
taxa with tolerance values of 1 to 3)
and tolerant (i.e., taxa with tolerance
values of 8-10)

Increase in tolerant
taxa; Decrease in
sensitive taxa

aDenotes endpoint selected for rank and MDS analyses.

bDenotes orders that have considerable variability in taxa tolerance values.
Notes: Intolerant taxa: tolerance value = 1 to 3.

Facultative taxa: tolerance value = 4 to 7.

Tolerant taxa: tolerance value = 8 to 10.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC ~q 7/10/2003

-------
Table D.3-14

Summary of Vegetative Status in Vicinity of Benthic Community Sampling
Locations, at Exposed and Reference Locations

River Mile

Station
ID

Surrounding Bank Vegetation

Aquatic Macrophytes and Algae in
Vicinity

Coarse-Grained Stations

Thin strip of shrub and transitional
floodplain forest bordering the
river. Region is bounded to the
south by predominately residential
areas. Northeast of this location is
a cultural grassland. Northwest of
the location are several pools
surrounded by shrub wetlands and
floodplain forest.

Macrophytes not quantified, but not
believed to be present in significant
numbers, based on substrate, hydrology,
and recollections of field samplers.
Algae coverage not quantified.

Transitional floodplain forest to the
north and east. To the south of this
location is a shallow emergent
marsh surrounded by shrub
wetlands.

Macrophytes not quantified, but not
believed to be present in significant
numbers, based on substrate, hydrology,
and recollections of field samplers.
Algae not quantified.

Transitional floodplain forest and
shrub wetlands to the south of the
station. To the north of this
location is a residential area with
open lawn and scattered trees
along the river edge.

Macrophytes not quantified, but not
believed to be present in significant
numbers, based on substrate, hydrology,
and recollections of field samplers.
Algae coverage not quantified.

134.03

Transitional floodplain forest, with
neighboring cultural grasslands

Algae comprise 75-100% of the surface
area. Adjacent to this station,
downstream, are beds of Myriophyllum
spicatum and Vallisneria americana.

133.79

Transitional floodplain forest

Algae comprise 75-100% of the surface
area. Adjacent to this station,
downstream, is a bed of Myriophyllum
spicatum.

133.18

Transitional floodplain forest,
minor point bar beach nearby

Algae comprise 50-75% of the surface
area.

132.34

Transitional floodplain forest on
one bank; shallow emergent marsh
on opposite bank

Algae comprise 75-100% of the surface
area.

130.32

Agricultural field on one bank;
transitional floodplain forest on
opposite bank

Algae comprise 50-75% of the surface
area.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC 7/10/2003

-------
Table D.3-14

Summary of Vegetative Status in Vicinity of Benthic Community Sampling
Locations, at Exposed and Reference Locations
(Continued)

River Mile

Station
ID

Surrounding Bank Vegetation

Aquatic Macrophytes and Algae in
Vicinity

Fine-Grained Stations

128.70

Wet meadow on one bank;
transitional floodplain forest on
opposite bank

Macrophyte bed along east bank consists

of Elodea canadensis, Potamogeton cf.
epihydrus, and Sparganium fluctuans.

126.38

Shrub swamp on one bank;
transitional floodplain forest on
opposite bank

Macrophyte beds along both east and
west banks consist of Myriophyllum
spicatum, Elodea canadensis, and
Potamogeton sp.

125.65

Shrub swamp on one bank;
transitional floodplain forest on
opposite bank

Macrophyte beds along both east and
west banks consist of Myriophyllum
spicatum, Elodea canadensis, and
Potamogeton sp. A dense macrophyte
bed with similar species also dominates
the adjacent downstream cove along east
bank.

124.50

Shallow emergent marsh in
vicinity of sampling location;
nearby riverbank vegetation
consists of black ash-red maple-
tamarack calcareous seepage
swamp.

Macrophyte beds along both banks
consist of Myriophyllum spicatum,
Elodea canadensis, and Potamogeton sp.
Dense, pervasive macrophyte beds are
present throughout the upstream portion
of Woods Pond.

The area surrounding this location
is red maple swamp.

Not available.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC ^ -> 7/10/2003

-------
Table D.4-1

Weighting of Measurement Endpoints for Weight-of-Evidence Evaluation

Attributes

Endpoint Group C: Chemistry

Endpoint Group T: Toxicity

Endpoint Group B: Benthic
Community

C-l
(Water)

C-2
(Sediment)

C-3
(Tissue)

T-l
(Lab)

T-2
(in situ)

T-3 (TIE)

B-l

(Metrics)

B-2
(Multivar)

B-3
(MHBI)

I Relationship Between Measurement and Assessment Endpoints

1. Degree of Biological Association

Low

Mod

Mod/High

Mod

Low/Mod

2. Stressor/Response

Low

Low/Mod

Mod

Mod/High

Low/Mod

3. Utility of Measure for Judging Risk

Low

Mod

High

Mod/High

Low/Mod

H. Data Quality

4. Data Quality

High

EX Study Design

5. Site Specificity

Low/Mod

Mod/High

High

6. Sensitivity to Detecting Changes

Low/Mod

High

Mod/High

High

Low/Mod

7. Spatial Representativeness

Mod/High

High

Mod

Mod/High

Low

Mod/High

8. Temporal Representativeness

High

Mod/High

Mod

9. Quantitativeness

Mod/High

High

Mod/High

Mod

10. Standard Method

Mod

High

Mod/High

High

Overall Endpoint Value

Low/Mod

Mod

Mod/High

Mod

C. Chemical Measures

C-l. Concentration of PCB in overlying water in relation to concentrations reported to be harmful to benthic invertebrates.

C-2. Concentration of PCB in the sediment in relation to concentrations reported to be harmful to benthic invertebrates.

C-3. Concentration of PCB in invertebrate tissues in relation to concentrations reported to be harmful to benthic invertebrates.

T. Toxicological Measures

T-l. Sediment toxicity to multiple invertebrate species, as measured in laboratory toxicity tests.

T-2. Sediment toxicity to multiple invertebrate species, as measured in the in situ toxicity tests.

T-3. Indications of PCB as toxicity driver in TIE investigations.

B. Benthic Community Measures

B-l. Abundance, richness, and biomass of invertebrates, relative to reference stations of comparable substrate and habitat (ANOVA analysis).
B-2. Benthic community structure, as assessed using multivariate assessment of key benthic metrics (rank analysis and multidimensional scaling).
B-3. Water quality assessment using modified Hilsenhoff Biotic Index (MHBI) indicator of organic pollution.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC 7/10/2003

-------
Table D.4-2

Evidence of Harm and Magnitude of Effects for Measurement Endpoints Related to Maintenance of a Healthy

Benthic Community

Measurement Endpoints

Weighting
Value
(High,
Moderate,
Low)

Coarse-Grained Sediment

Fine-Grained Sediment

Evidence of
Harm
(Yes, No,
Undetermined)

Magnitude
(High,
Intermediate,
Low)

Evidence of
Harm (Yes, No,
Undetermined)

Magnitude (High,
Intermediate,
Low)

C. Chemical Measures

C-l. Concentration of PCB in overlying water in relation to
concentrations reported to be harmful to benthic invertebrates

Low /
Moderate

Yes

Intermediate

Yes

Intermediate

C-2. Concentration of PCB in the sediment in relation to
concentrations reported to be harmful to benthic invertebrates

Low /
Moderate

Yes

High

Yes

High

C-3. Concentration of PCB in invertebrate tissues in relation to
concentrations reported to be harmful to benthic invertebrates

Moderate

Yes

Intermediate

Yes

Intermediate

T. Toxicological Measures

T-l. Sediment toxicity to multiple invertebrate species, as
measured in laboratory toxicity tests

Moderate/
High

Yes

High

Yes

High

T-2. Sediment toxicity to multiple invertebrate species, as
measured in in situ toxicity tests

Moderate/
High

Yes

Intermediate

Yes

High

T-3. Indications of PCB as toxicity driver in toxicity identification
evaluations

Moderate

Undetermined

—

Yes

Intermediate

B. Benthic Community Measures

B-l. Abundance, richness, and biomass of invertebrates, relative to
reference stations of comparable substrate and habitat (ANOVA)

Moderate

Yes

Intermediate

—

B-2. Benthic community structure, as assessed using a multivariate
assessment of key benthic metrics (rank analysis and MDS)

Moderate

Yes

Intermediate

—

B-3. Water quality assessment using modified Hilsenhoff Biotic
Index (MHBI) indicator of organic pollution

Moderate

—

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC ^ 7/10/2003

-------
Table D.4-3

Weight-of-Evidence Risk Analysis Summary Indicating Concurrence Among Endpoints for Coarse-Grained

and Fine-Grained Sediment

(a) Coarse-grained contaminated (C/C) sediment

Harm/Magnitude

Weighting Factors (increasing confidence or weight)

Low

Low/Moderate

Moderate

Moderate/High

High

Yes/High

C-2

T-l

Yes/Intermediate

C-l

B-l, B-2, C-3

T-2

Yes/Low
Undetermined
No Harm

_______

T-3
B-3

_______

(b) Fine-grained contaminated sediment (F/C)

Harm/Magnitude

Weighting Factors (increasing confidence or weight)

Low

Low/Moderate

Moderate

Moderate/High

High

Yes/High

C-2

T-l, T-2

Yes/Intermediate

C-l

T-3, C-3

Yes/Low
Undetermined
No Harm

_______

B-l, B-2, B-3

_______

Note: See Tables D.4-1 and D.4-2 for definitions of endpoint codes.

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC

-------
Table D.4-4

Total PCB Data for Benthic Invertebrates Collected at West Cornwall, Connecticut (1994-2001)

Average Lipid
(%)

tPCB Concentration (mg/kg)

Year

Species

Number of
Composite Samples

Range

Median

Arithmetic Mean

Standard Error

1994

Caddisfly larvae

1.42-2.42

2.07

1.97

0.29

1994

Dobsonfly larvae

2.24-3.83

3.03

0.79

1994

Stonefly nymphs

1.09

1996

Caddisfly larvae

1.89-3.84

2.87

0.98

1996

Dobsonfly larvae

2.78-3.53

3.16

0.37

1996

Stonefly nymphs

2.43

1998

Caddisfly larvae

0.38

1.05

1998

Dobsonfly larvae

6.22

3.94

1998

Stonefly nymphs

1.18

0.54

2001

Caddisfly larvae

6.0

0.9

2001

Dobsonfly larvae

10.0

1.81

2001

Stonefly nymphs

1.0

0.53

Source: BBL and QEA (2003)
NA = Not available

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_TBL.DOC or 7/12/2003

-------
APPENDIX D

ASSESSMENT ENDPOINT—COMMUNITY STRUCTURE, SURVIVAL,
GROWTH, AND REPRODUCTION OF BENTHIC INVERTEBRATES

FIGURES

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
LIST OF FIGURES

Title Page

Figure D. 1-1 Conceptual Model Diagram: Exposure Pathways for Benthic Invertebrates

Exposed to Contaminants of Concern (COCs) in the Housatonic River PSA 1

Figure D. 1-2 Overview of Approach Used to Assess Exposure of Benthic Invertebrates to

COCs in the Housatonic River PSA 2

Figure D. 1-3 Overview of Approach Used to Assess the Effects of COCs to Benthic

Invertebrates in the Housatonic River PSA 3

Figure D. 1-4 Overview of Approach Used to Characterize the Risks of COCs to Benthic

Invertebrates in the Housatonic River PSA 4

Figure D. 1-5 Summary of Studies Conducted in Conjunction with Ecological Risk

Assessment for Benthic Invertebrates, and Linkage to ERA 5

Figure D.2-1 Benthic Invertebrate Sampling Locations and Simplified Station Identifiers 7

Figure D.2-2 Median Percent Fines and Percent TOC by Sampling Station, for Samples

Collected Synoptic with Benthic Community Grabs 9

Figure D.2-3 Cluster Analysis of Benthic Invertebrate Community Structure, Based on

Abundances of Eight Dominant Taxonomic Groups 10

Figure D.2-4 Non-Metric Multidimensional Scaling of Benthic Invertebrate Community

Structure, Based on Abundances of Eight Dominant Taxonomic Groups 11

Figure D.2-5 Particle Size Distribution (Cumulative Percent) for Each of the Four

Sediment Types Collected at Benthic Invertebrate Sampling Locations 12

Figure D.2-6 Histograms of Percent Fines Content for Each of the Four Sediment Types

Collected at Benthic Invertebrate Sampling Locations 13

Figure D.2-7 Total Organic Carbon (TOC) Within Each of the Four Sediment Types

Collected at Benthic Invertebrate Sampling Locations 14

Figure D.2-8 Total Organic Carbon (TOC) Content of Bulk Sediment Collected in the
Vicinity of Invertebrate Toxicity Stations, from Sampling Programs
Conducted in 1999 15

Figure D.2-9 Relationship Between Total Organic Carbon (TOC) Content and Percent
Fines Within Each of the Four Sediment Types Collected at Benthic
Invertebrate Sampling Locations (Logarithmic Axes) 16

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC • • • 7/10/2003

-------
LIST OF FIGURES
(Continued)

Title Page

Figure D.2-10 Concentrations of tPCB in Sediment by Sampling Location for Individual
Benthic Community Grab Samples, and Associated Measures of Central
Tendency (Logarithmic Scale) 17

Figure D.2-11 Histograms of tPCBs Concentrations for Samples Collected Synoptic with

Benthic Community Samples, Subdivided by Location Type 18

Figure D.2-12 Comparison of tPCB Concentrations in Sediment from Measurements

Taken in Conjunction with Wright State University Toxicity Tests (EVS
2003) 19

Figure D.2-13 Comparison of tPCB Concentrations in Sediment Collected at Benthic

Toxicity Stations, from Various Sampling Efforts Conducted in 1999 20

Figure D.2-14 Medians and Quartiles of PCB and TOC in the Housatonic River PSA,

Subdivided by River Reach and 0.25-Mile Subreaches 21

Figure D.2-15 Distribution of tPCB Concentrations Detected in Sediment Samples from

the GE Facility to Long Island Sound 22

Figure D.2-16 Concentrations of Total Dioxins in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 23

Figure D.2-17 Concentrations of Total Furans in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 24

Figure D.2-18 Concentrations of Dibenzofuran in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 25

Figure D.2-19 Concentrations of Total PAHs in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 26

Figure D.2-20 Concentrations of Antimony in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 27

Figure D.2-21 Concentrations of Barium in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 28

Figure D.2-22 Concentrations of Cadmium in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 29

Figure D.2-23 Concentrations of Chromium in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 30

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC • 7/10/2003

-------
LIST OF FIGURES
(Continued)

Title Page

Figure D.2-24 Concentrations of Copper in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 31

Figure D.2-25 Concentrations of Lead in PSA Sediment Samples Collected in Conjunction

with EPA Biological Effects Studies in 1999 32

Figure D.2-26 Concentrations of Mercury in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 33

Figure D.2-27 Concentrations of Silver in PSA Sediment Samples Collected in

Conjunction with EPA Biological Effects Studies in 1999 34

Figure D.2-28 Concentrations of Tin in PSA Sediment Samples Collected in Conjunction

with EPA Biological Effects Studies in 1999 35

Figure D.2-29 Distribution of tPCB in Benthic Invertebrate Tissues by Location and

Functional Feeding Group 36

Figure D.2-30 Aroclor Composition of Benthic Invertebrate Predator Tissues (Cumulative

Percent of Detected Values) 37

Figure D.2-31 Aroclor Composition of Benthic Invertebrate Shredder Tissues (Cumulative

Percent of Detected Values) 38

Figure D.2-32 Concentrations of PCBs Observed in Lumbriculus Body Tissues (mg/kg
ww) and BSAF Values Calculated from OC-Normalized Sediment and
Tissue Data (mg lipid/mg OC) 39

Figure D.2-33 Concentrations of PCBs (Sum of Measured and Detected Congeners, and

Sum of Detected Homologs) Observed in Water Column Samples Collected
Synoptic with the 7-d in Situ Toxicity Testing Program (EVS 2003) 40

Figure D.2-34 Concentrations of PCBs Observed in Water Column Samples Collected
Synoptic with the 48-h, 7-d, and 10-d in Situ Toxicity Testing Program
(EVS 2003) 41

Figure D.2-35 Concentrations of Furans (Tetra-, Penta-, And Hexa-Chlorodibenzofurans)
Observed in Water Column Samples Collected Synoptic with the 7-d In Situ
Toxicity Testing Program (EVS 2003) 42

Figure D.3-1 Survival of Hyalella azteca in Chronic Laboratory Toxicity Tests, at Three

Time Periods (28-d, 35-d, 42-d) 43

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC 7/10/2003

-------
LIST OF FIGURES
(Continued)

Title Page

Figure D.3-2 Growth of Hyalella azteca in Chronic Laboratory Toxicity Tests, at Two

Time Periods (28-d, 42-d) 44

Figure D.3-3 Reproduction of Hyalella azteca in Chronic Laboratory Toxicity Tests,

Based on Mean Number of Young (Day 28-42) 45

Figure D.3-4 Survival and Emergence of Chironomus tentans in Chronic Laboratory

Toxicity Tests (43-Day) 46

Figure D.3-5 Growth Endpoints for Chironomus tentans in Chronic Laboratory Toxicity

Test (20-Day) 47

Figure D.3-6 Survival of Hyalella azteca in 48-H Low Flow In Situ Toxicity Tests

Conducted 14-16 June 1999 48

Figure D.3-7 Survival of Hyalella azteca in 10-D Low Flow In Situ Toxicity Tests

Conducted 17-27 June 1999 49

Figure D.3-8 Survival of Chironomus tentans in 48-H Low Flow In Situ Toxicity Tests

Conducted 14-16 June 1999 50

Figure D.3-9 Survival of Chironomus tentans in 10-D Low Flow In Situ Toxicity Tests

Conducted 17-27 June 1999 51

Figure D.3-10 Survival of Daphnia magna in 48-H Low Flow In Situ Toxicity Tests

Conducted 14-16 June 1999 52

Figure D.3-11 Survival of Lumbriculus variegatus in 48-H Low Flow In Situ Toxicity

Tests Conducted 14-16 June 1999 53

Figure D.3-12 Statistical Endpoints for Toxicity Data, with Comparisons to Negative

Control (Sorted by LC5o/EC5o Value) 54

Figure D.3-13 Statistical Endpoints for Toxicity Data, with Comparisons to Negative

Control (Sorted by Test Duration) 55

Figure D.3-14 Statistical Endpoints for Toxicity Data, with Comparisons to Station A1

(Sorted by LC5o/EC5o Value) 56

Figure D.3-15 Statistical Endpoints for Toxicity Data, with Comparisons to Station A3

(Sorted by LC5o/EC5o Value) 57

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC • 7/10/2003

-------
LIST OF FIGURES
(Continued)

Title Page

Figure D.3-16 Segmented Linear Regression Models Applied to Toxicity Data, Relating
Relative Performance Proportion (RPP) to Bulk Sediment tPCB
Concentrations (mg/kg) 58

Figure D.3-17 Combined Effects and No-Effects Levels for PCB Concentrations (mg/kg

wet) in Benthic Invertebrate Tissue Samples - tPCBs and Aroclor 1254 59

Figure D.3-18 Physiological Tolerance of Macroinvertebrate Orders to Organic

Compounds Relative to Daphnia magna Sensitivity 60

Figure D.3-19 Average Ranks Analysis for Six Benthic Community Metrics, with Equal

Weighting Assigned to Each Metric 61

Figure D.3-20 Multidimensional Scaling (MDS) for Benthic Community Health Metrics,

Showing Metric Medians on MDS Plot 62

Figure D.3-21 Multidimensional Scaling (MDS) for Benthic Community Composition
Data at Fine-Grained Stations Only, Showing Centroid and Individual
Replicate Results on MDS Plot 63

Figure D.3-22 Multidimensional Scaling (MDS) for Benthic Community Composition
Data at Coarse-Grained Stations Only, Showing Centroid and Individual
Replicate Results on MDS Plot 64

Figure D.3-23 Mean Modified Hilsenhoff Biotic Index (MHBI) Values 65

Figure D.3-24 Biomass (g) of Benthic Invertebrates Measured in Grab Samples, with

Subdivision into Major Taxonomic Groups, for Fine- and Coarse-Grained
Sediment 66

Figure D.3-25 Relationship Between Taxonomic Richness and tPCB, by Substrate Type

and Sampling Location 67

Figure D.3-26 Relationship Between Abundance and tPCB, by Substrate Type and

Sampling Location 68

Figure D.3-27 Relationship Between PCB Concentrations in Coarse-Grained Locations and

Taxonomic Richness in Individual Replicates 69

Figure D.3-28 Relationship Between tPCB Concentrations in Coarse-Grained Locations

and Organism Abundance in Individual Replicates 70

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC •• 7/10/2003

-------
LIST OF FIGURES
(Continued)

Title Page

Figure D.4-1 Hazard Quotients (Median and Range) Based on Median Sediment

Chemistry for tPCBs, for Samples Collected in 1999 Close to Sediment
Quality Triad Stations 71

Figure D.4-2 Hazard Quotients (Median, Range) Based on Sediment Chemistry for Other
COPCs, for Samples Collected in 1999 Close to Sediment Quality Triad
Stations 72

Figure D.4-3 Hazard Quotients (Median, Range) Based on Sediment Chemistry for Other
COPCs, for Broad EPA Sampling Results Not Specifically Connected to the
Triad Studies 75

Figure D.4-4 Hazard Quotients (Median, Range) Based on Overlying Water PCB

Concentrations, Measured Synoptic with In Situ Toxicity Tests 78

Figure D.4-5 Hazard Quotients for tPCB Tissue Residues in Benthic Invertebrates,

Relative to Two Effects Thresholds Derived from Literature Studies 79

Figure D.4-6 Weight-of Evidence Evaluation of Housatonic River Benthic Sampling

Locations, with Indications of Alteration/Risk Relative to Background 80

Figure D.4-7 Assessment of Risk to Benthic Invertebrates Exposed to tPCBs Downstream

of Woods Pond 82

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC • • • 7/10/2003

-------
CO

CO
CO

a)
a)
o£

a)
o

3
O

-------
Exposure

Figure D.1-2 Overview of Approach Used to Assess Exposure of Benthic
Invertebrates to COCs in the Housatonic River PSA

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC rs 7/10/2003

-------
Effects

Figure D.1-3 Overview of Approach Used to Assess the Effects of COCs to
Benthic Invertebrates in the Housatonic River PSA

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC ^ 7/10/2003

-------
Risk Characterization

Figure D.1-4 Overview of Approach Used to Characterize the Risks of COCs to
Benthic Invertebrates in the Housatonic River PSA

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC a 7/10/2003

-------
Figure D.1-5 Summary of Studies Conducted in Conjunction with Ecological
Risk Assessment for Benthic Invertebrates, and Linkage to ERA

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC r 7/10/2003

-------
oS j

River Mile: N/A

-SE000161

River Mile: N/A

HW-SE000398

River Mile: 133.79

H3-SE000060

River Mile: 132.34 "

H3-SEEC0019

Reai

/20\

) / 0 \ —v n .

River Mile: 128.7

H3-SE000116

x/v/ >

(each 5E

(Reach 5C

River Mile: 125.65

H3-SEEC0023

Reach 51

(backwater areas)\

River Mile: 125.65

H3-SEEC0031

Reach 6

I Approximately 13 miles,

~tns~x

jch 1

LEGEND. ^ Benthic Community Stations Without Toxicity Testing
A Benthic Community Stations With Toxicity Testing
Reach Breaks
H Roads
[^j Hydrography

I | Housatonic River Basin Boundary

Notes: - Code in left box of each ID represents "simplified" nomenclature used for benthic invertebrate ERA.
- Station 8A was positioned 12 meters from Station 8, and was tested for laboratory toxicity only.

2500

5000 Feet

0.5

1 Miles

Ecological Risk Assessment
GE/Housatonic River Site
Rest of River

FIGURE D.2-1
BENTHIC INVERTEBRATE
SAMPLING LOCATIONS
AND SIMPLIFIED STATION
IDENTIFIERS

| O:\gepitt\aprs\macro_int.apr I layout - mac Iocs D.2-11 o:\gepitt\epsfiles\plots\in\mac_int_d2-1 .eps 111:58 AM, 7/1/20031

-------
100

10%

o
o

-------
<£>
e

£ ®
O)

"a5

CM
<

r n
< <

Coarse - Grained
References

i r

<0 is.

o> >r

Fine - Grained
Sites (All)

1- CM

m m

Coarse - Grained
"Exposed" Sites

Figure D.2-3 Cluster Analysis of Benthic Invertebrate Community Structure,
Based on Abundances of Eight Dominant Taxonomic Groups

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
More
Dipterans

-------
c
0

0
CL

C/C

C/R

F/C

F/R

n CLAY
D SILT
D SAND
D GRAVEL

Groupings: C/C (Coarse Contaminated) = Stations 1-5; C/R (Coarse Reference) = Stations A1-A3; F/C
(Fine Contaminated) = Stations 6-9; F/R (Fine Reference) = Station R4

Figure D.2-5

Particle Size Distribution (Cumulative Percent) for Each of the
Four Sediment Types Collected at Benthic Invertebrate Sampling
Locations

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
c/c

C/R

0.9

0.8

0.7 7

C 20

0.6 o
a.

0.5 o

° 15

i per
o

0-3 j?

0.2 "

0.1
0.0

B .

0 20 40 60 80 100
Percentage Fines (Silt, Clay)

F/C

F/R

20 40 60 80 100
Percentage Fines (Silt, Clay)

0 20 40 60 80 100
Percentage Fines (Silt, Clay)

Groupings: C/C (Coarse Contaminated) = Stations 1-5; C/R (Coarse Reference) = Stations A1-A3; F/C
(Fine Contaminated) = Stations 6-9; F/R (Fine Reference) = Station R4

Figure D.2-6

Histograms of Percent Fines Content for Each of the Four
Sediment Types Collected at Benthic Invertebrate Sampling
Locations

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
c/c

C/R

5 10

TOC (%)

0.0
15

5 10

TOC (%)

0.0
15

F/C

F/R

15"

10"

5 10

TOC (%)

0.4

0.3 7
o

0.2 §

CD
0)

0.1

0.0
15

-------
-

;

8 i °

;

¦7

~ ~

« 1 «

X
~

A1 A3 4 5 7 8 8A

Station Location

A48h in situ WSU toxicity (June 1999)

~ 7-day in situ WSU toxicity (June 1999)

• 10-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

~ Benthic Macroinvertebrate (Median of 12; June 1999)
O Porewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

~ Mussel Study (May 1999)

+ Mussel Study (July 1999)

— Discrete River Sampling (October 1999)
APre-Toxicity Screening (March-May 1999)

-Tree Swallow Study (March 1999)

Figure D.2-8 Total Organic Carbon (TOC) Content of Bulk Sediment Collected
in the Vicinity of Invertebrate Toxicity Stations, from Sampling
Programs Conducted in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
c/c

C/R

10.C"

o
o

0
o

0
CL

1.0 10.0 100.0

Percent Fines

10. C"

o
O

0
o

1.0 10.0

Percent Fines

100.0

F/C

F/R

10.C"

o
o

-------
1000

m
o

0)
o
c
o
o

100

0.1

& *

A
S

i 1

$ *

I
i

A2 A3

3 4 5
Location ID

A Median of 12 replicates
— Mean of 12 replicates

• Replicate concentration, assuming non-detected values equal to
half MDL

Figure D.2-10 Concentrations of tPCB in Sediment by Sampling Location for
Individual Benthic Community Grab Samples, and Associated
Measures of Central Tendency (Logarithmic Scale)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
c/c

C/R

16 —¦ ¦ ¦ i |—i—i 1111111—i—i 1111111—7—

5 8

1.0 10.0 100.0

Total PCB Concentration (mg/kg)

0.2 -a

o
~a
o

o
=3

0.1 ?

CD
Q)

0.0

¦ 11111

—1—1 1111111—1—1 1111111—1—

0.9

0.8

0.7

C 20

0.6

0.5

0 15

0.4

CO
0

0.2

0.1

* 1 * 1 *

0.0

o
3-

1~£

1.0 10.0 100.0

Total PCB Concentration (mg/kg)

F/C

F/R

5 8

I I I I III 1 1 I I I I III 7 1 I I I I III T~

1.0 10.0 100.0

Total PCB Concentration (mg/kg)

0.3

0.2 o
g-
o

0.1 g>

0.0

-1—1

§ 5

o 4
3
2
1
0

I I I I III 7 1 I I I I III

I I I III 1 I I I I I II

1.0 10.0 100.0

Total PCB Concentration (mg/kg)

0.7
0.6

0.5°

0.4 i

0.3®

0.2 9>

0.1
0.0

Groupings: C/C (Coarse Contaminated) = Stations 1-5; C/R (Coarse Reference) = Stations A1-A3; F/C
(Fine Contaminated) = Stations 6-9; F/R (Fine Reference) = Station R4

Figure D.2-11 Histograms of tPCBs Concentrations for Samples Collected
Synoptic with Benthic Community Samples, Subdivided by
Location Type

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

U)
O)

2
•*->
c

0)
o
c
o
o
m
o

100

0.1

0.01

Station ID

~ Laboratory

148-h in-situ

17-d in-situ

~ 10-d in-situ

Notes: Values represent analyses conducted on single composite samples.

Values for in situ samples at A3 were detected, but with concentrations below 0.01 mg/kg.

Figure D.2-12 Comparison of tPCB Concentrations in Sediment from

Measurements Taken in Conjunction with Wright State University
Toxicity Tests (EVS 2003)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

O)
O)

m
o
o.

3
o

100

0.1

0.01

+
6

~
X

4 5 7

Station Location

A48h in situ WSU toxicity (June 1999)

~ 7-day in situ WSU toxicity (June 1999)

• 10-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)
xBenthic Macroinvertebrate (Median of 12; June 1999)
O Porewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

~ Mussel Study (May 1999)

+ Mussel Study (July 1999)

— Discrete River Sampling (October 1999)
APre-Toxicity Screening (March-May 1999)

Figure D.2-13

Comparison of tPCB Concentrations in Sediment Collected at
Benthic Toxicity Stations, from Various Sampling Efforts
Conducted in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

S 100

m
o
o.

-1—i—i—|—i—i—i—[-

Reach 5A, N = 499

-i—i—|—i—i—i—|—i—i—r

-|—i—i—i—|—i—i—r-

Reach 5B, N = 218

Total PCB

l—i—i—i—i—i—i—i—r

Reach 5C, N = 286
Reach 5D, N = 63

"<—r

—I TTT

Reach E,
N = 22?

0.1
135
1000000

32 45 22 30 54 52 23 37 14 24 14 39 12 21 16 10 7 6 5 36 : 31 12 19 34 .38 39 20 25 H5 23 14 9 28 18 21 42 23 19 42 21 11.:

y 24 14 39 Y

134

133

132

131

130

129

128

127

126

125

124

^100000

£ 10000

o
o

1000

1—1 1 I 1—1 1 I

Reach 5A, N = 472

1—1—i—i I i

Reach 5B, N = 207

1—i—1 1 I—'—r

Reach 5C, N = 288
Reach 5D, N = 69

Reach
N = 174

Total Organic Carbon

100
135
100000

23 40 21 28 53 46 26 36 14 23 15 43 7 15 13 11 8 7 6 37

y 23

,3 21 13 9 25 18 20 46 23 24 51 15 10

134

133

132

131

130

129

128

127

126

125

124

1—i—I—i—i—

: Reach 5A, N = 440

10000

|> 1000
m

y 100

t-i—i—i i i

Reach 5B, N = 190

"il—i—'—'—I—1 r

I Reach 5C, N = 243
i Reach 5D, N = 63

Carbon Normalized PCB

72 40 21 25 49 45 23 34 13 22 14

135 134

LEGEND:

River Reach Median

River Reach Quartile

— Sub-reach Median and Quartile
J Reach 5D Median and Quartile

y 22

14 13 10 Z 6 4 34

L1 19 11 9 22 17 t

Reach 6,
N = 156

20 17 40 11

133 132 131 130 129

Mile Point (miles)

128

127

126

125

124

0.25 MILE SUB-REACH AVERAGING
Median and Quartiles of PCB and TOC in the Housatonic River (NOTE: Reach 5D is Backwaters)

(All Data, Half DL, Surface Layer: 0 - 6")

Figure D.2-14 Medians and Quartiles of PCB and TOC in the Housatonic River PSA, Subdivided by River Reach
and 0.25-Mile Subreaches

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
Massachusetts

Connecticut

400

350

300 <=;

O)
O)

250 E
c

200 S

1—

150 S

c
o

100 O

WPD

RPD

GFD

BBD

RDD

SD DSD

i I

150 140 130 120 110

100 90 80 70 60 50
River Miles from Long Island Sound

30 20

GE = General Electric facility; WPD = Woods Pond Dam; RPD = Rising Pond Dam; GF= Great Falls Dam; CB = Cornwall Bridge; BBD = Bulls Bridge Dam;
RDD = Bleachery Dam; SD (River Mile 25) = ShepaugDam; SD (River Mile 15) = Stevenson Dam; DSD = Derby-Shelton Dam

Figure D.2-15 Distribution of tPCB Concentrations Detected in Sediment Samples from the GE Facility to Long
Island Sound

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
100000

10000

1000

100

o
~

O
+

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)
O Laboratory Toxicity - WSU (May 1999)
oPorewater TIE (AD) - WSU (Sept. 1999)
X Porewater TIE (AW) - WSU (Sept. 1999)
+ Mussel Study (July 1999)

Figure D.2-16 Concentrations of Total Dioxins in PSA Sediment Samples

Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
100000

10000

5> 1000

100

~
o

0
o

~ £
£

A1 A3 4 5 7

Station Location

o
+

~ 7-day in situ WSU toxicity (June 1999)
O Laboratory Toxicity - WSU (May 1999)
OPorewater TIE (AD) - WSU (Sept. 1999)
X Porewater TIE (AW) - WSU (Sept. 1999)
+ Mussel Study (July 1999)

Figure D.2-17 Concentrations of Total Furans in PSA Sediment Samples

Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
10

s 0.1

0.01

¥
X

9 a

A1 A3 4 5 7

Station Location

o
+

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

OPorewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-18 Concentrations of Dibenzofuran in PSA Sediment Samples
Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

100

J*
O)

X
<
Q.

"5
o

O
~

o
¥

X
~

o
~

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

OPorewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-19 Concentrations of Total PAHs in PSA Sediment Samples

Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
10

> 1 -
c
o
E

0.1

o
o

~
X

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)
O Laboratory Toxicity - WSU (May 1999)
O Porewater TIE (AD) - WSU (Sept. 1999)
X Porewater TIE (AW) - WSU (Sept. 1999)
+ Mussel Study (July 1999)

Figure D.2-20 Concentrations of Antimony in PSA Sediment Samples

Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

100

J*
O)

ra
m

~
O

~
o

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

OPorewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-21 Concentrations of Barium in PSA Sediment Samples Collected
in Conjunction with EPA Biological Effects Studies in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
100

E
¦o
ra
O

0.1

0.01

¥
0

A3 4 5 7

Station Location

o
o

o
+

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

O Porewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-22 Concentrations of Cadmium in PSA Sediment Samples

Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

— 100

E
S

o 10

o
o

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

OPorewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-23 Concentrations of Chromium in PSA Sediment Samples

Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

100

U)
J*

0)
Q.
Q.
O
O

O
~

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

OPorewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-24 Concentrations of Copper in PSA Sediment Samples Collected
in Conjunction with EPA Biological Effects Studies in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

100

¦a

A1 A3 4 5 7 8 8A

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

OPorewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-25 Concentrations of Lead in PSA Sediment Samples Collected in
Conjunction with EPA Biological Effects Studies in 1999

OCX

o
o

: ~

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
10

£
0)

0.1

o
m

o
~

0.01

-n-

-B-

A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

O Porewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-26 Concentrations of Mercury in PSA Sediment Samples

Collected in Conjunction with EPA Biological Effects Studies
in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
10

0)
>

0.1

0.01

~
O

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

O Porewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-27 Concentrations of Silver in PSA Sediment Samples Collected
in Conjunction with EPA Biological Effects Studies in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
100

o
~

~
o

X
$

0.1

A1 A3 4 5 7

Station Location

~ 7-day in situ WSU toxicity (June 1999)

O Laboratory Toxicity - WSU (May 1999)

O Porewater TIE (AD) - WSU (Sept. 1999)

X Porewater TIE (AW) - WSU (Sept. 1999)

+ Mussel Study (July 1999)

Figure D.2-28 Concentrations of Tin in PSA Sediment Samples Collected in
Conjunction with EPA Biological Effects Studies in 1999

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
50

O)
CT>

if)

8 20

0.30/0.22 0 35

A1 A3 1

j ^ 0.05/0.50

8 9 R4

Predator

~ Shredder

Note: Text labels indicate detected values close to zero. Missing values with no text labels indicate stations
where no analysis was completed due to insufficient sample volume.

Figure D.2-29 Distribution of tPCB in Benthic Invertebrate Tissues by
Location and Functional Feeding Group

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
-------
100%

80%

o 60%
o>

o
o

| 40%

20%

Figure D.2-31 Aroclor Composition of Benthic Invertebrate Shredder Tissues
(Cumulative Percent of Detected Values)

A1 A3 1 2 3 4 5 6 7 8 9 R4
¦ AROCLOR-1260 n AROCLOR-1254 n AROCLOR-1248 n AROCLOR-1242

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
7.0

6.0

S 5.0
c
-------
300

250

200

~ 150

100

4 5

Station ID

¦ Sum of Homologs

~ Sum of Individual Measured Congeners

Figure D.2-33 Concentrations of PCBs (Sum of Measured and Detected
Congeners, and Sum of Detected Homologs) Observed in
Water Column Samples Collected Synoptic with the 7-d in Situ
Toxicity Testing Program (EVS 2003)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
350

4 5

Station ID

¦ 48-h Sum of Homologs

~ 7-d Sum of Homologs

~ 10-d Sum of Homologs

Figure D.2-34 Concentrations of PCBs Observed in Water Column Samples
Collected Synoptic with the 48-h, 7-d, and 10-d in Situ Toxicity
Testing Program (EVS 2003)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
140

120

100

D
LL

4 5

Station ID

¦ Detected TCDF

~ Detected PCDF

~ Detected HxCDF

Figure D.2-35

Concentrations of Furans (Tetra-, Penta-, And Hexa-
Chlorodibenzofurans) Observed in Water Column Samples
Collected Synoptic with the 7-d In Situ Toxicity Testing
Program (EVS 2003)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
0.28

¦ 28-day (24 June 99) ~ 35-day (1 July 99) ~ 42-day (8 July 99)

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean Trout Farm reference sediment).

100

^ 80

re
>

-4-i

a>
*2
a>
o.

c
re
a>

0.018

0.028

0.28

5.9

8.7

213

4.6

T-Control

Station Location

Figure D.3-1 Survival of Hyalella azteca in Chronic Laboratory Toxicity Tests,
at Three Time Periods (28-d, 35-d, 42-d)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
T-Control A1 A3 4 7

Station Location

128-day (24 June 99)

~ 42-day (8 July 99)

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean Trout Farm reference sediment).

Figure D.3-2 Growth of Hyalella azteca in Chronic Laboratory Toxicity Tests, at
Two Time Periods (28-d, 42-d)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------

T-Control

0.018

0.028

0.28

5.9

8.7

213

(no reproduction
data due to zero
survival rate)

A3 4 7

Station Location

4.6

28-42 day mean young per female ~ 28-42 day mean young (unstandardized)

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean Trout Farm reference sediment).

Figure D.3-3 Reproduction of Hyalella azteca in Chronic Laboratory Toxicity
Tests, Based on Mean Number of Young (Day 28-42)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
NA 0.28

T- C- F- A1 A3 4 7 8A 8

Control Control Control

Station Location

¦ 20-day Survival ~ 20-day Emergence

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

T-Control, C-Control, and F-Control are negative laboratory controls ("Trout Farm", "Cellulose", and
"Florissant", respectively).

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory reference sediment).

Figure D.3-4 Survival and Emergence of Chironomus tentans in Chronic
Laboratory Toxicity Tests (43-Day)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
5.0

+ 4.0 -

T- C- F- A1 A3 4 7 8A 8

Control Control Control

Station Location

¦ Dry weight ~ Ash-free Dry weight

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

T-Control, C-Control, and F-Control are negative laboratory controls ("Trout Farm", "Cellulose", and
"Florissant", respectively).

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory reference sediment).

Figure D.3-5 Growth Endpoints for Chironomus tentans in Chronic Laboratory
Toxicity Test (20-Day)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
0.018

Lab Control A1 A3 4 5 7 8

Station Location

~ Water Column ¦ Against Sediment

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory control).

Figure D.3-6 Survival of Hyalella azteca in 48-H Low Flow In Situ Toxicity Tests
Conducted 14-16 June 1999

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
120

100

— 80

£° 60

3
CO
C

re
a>

0.018

0.0014

8.3

Lab Control A1

ff2

A3 4 5 7

Station Location

~ Water Column ¦ Against Sediment

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory control).

Figure D.3-7 Survival of Hyalella azteca in 10-D Low Flow In Situ Toxicity Tests
Conducted 17-27 June 1999

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
120

q 100

CO
+

¦—1 80 -

™ 60

W 40
c
re
a>

20 -

0.018 Q.28

Lab Control A1 A3 4 5

Station Location

~ Water Column

I Against Sediment

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory control).

Figure D.3-8 Survival of Chironomus tentans in 48-H Low Flow In Situ Toxicity
Tests Conducted 14-16 June 1999

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
120

Lab Control A1

A3 4 5 7

Station Location

~ Water Column ¦ Against Sediment

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory control).

Figure D.3-9 Survival of Chironomus tentans in 10-D Low Flow In Situ Toxicity
Tests Conducted 17-27 June 1999

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
120

q 100

CO
+

" 80

re
>

3
CO
C

re
a>

Lab Control

0.018

0.0001

A3 4 5

Station Location

739

522

~ Water Column

I Against Sediment

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory control).

Figure D.3-10 Survival of Daphnia magna in 48-H Low Flow In Situ Toxicity
Tests Conducted 14-16 June 1999

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
120

q 100

CO
+

¦—1 80 -

0.028

re
>

3
CO
C

re
a>

60 -

40 -

20 -

Lab Control A1

A3 4 5

Station Location

~ Water Column

I Against Sediment

Notes: Labels represent tPCB concentration (mg/kg) in sediment.

Value in bold represents median tPCB concentration (from all measurements made within 5 meters of
station in 1999).

Value in italics represents "most synoptic" tPCB concentration; single concentration measured closest to
toxicity test in space/time.

NA = No PCB concentration data available for negative control, but PCB contamination assumed to
be low (i.e., clean laboratory control).

Figure D.3-11 Survival of Lumbriculus variegatus in 48-H Low Flow In Situ
Toxicity Tests Conducted 14-16 June 1999

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
Figure D.3-12 Statistical Endpoints for Toxicity Data, with Comparisons to Negative Control (Sorted by
LC50/EC50 Value)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
=/ S y ^ *>* j/ j/ / ^ t? ^ # n>v° ^ fi? n/

jf ^ jf ^ ^ ^ jf ^ *f ^ *f V ¦/J>

V- o- ^' . ®k Q.

Test Type - Species

Figure D.3-13 Statistical Endpoints for Toxicity Data, with Comparisons to Negative Control (Sorted by Test
Duration)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
>77 >77 73.2 >77

Test Type - Species

Figure D.3-14 Statistical Endpoints for Toxicity Data, with Comparisons to Station A1 (Sorted by
LC50/EC50 Value)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
Figure D.3-15 Statistical Endpoints for Toxicity Data, with Comparisons to Station A3 (Sorted by
LC50/EC50 Value)

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
All Acute End points

42-d Hyalella Survival

Q.
£

-1 0 1

log10(tPCB median concentration)

o
o

-1

log10(tPCB median concentration)

Remaining Chronic Endpoints

O
V

—1.

I -

s^.Ji

-1 o 1

log10(tPCB median concentration)

42-d Hyalella Growth

stc

. . J . . ..

*
K*
X

-1 o 1

log10(tPCB median concentration)

Note: Like symbols on plots represent individual replicates for a given test species.

Figure D.3-16 Segmented Linear Regression Models Applied to Toxicity Data,

Relating Relative Performance Proportion (RPP) to Bulk Sediment
tPCB Concentrations (mg/kg)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1000

100

Open Bar = no significant effect (NOEL)
Solid Bar = significant effect (LOEL)

0.01

& ir

2- 2

m ffl fll ^ (C (Q

c
o

2
0

co 10

t: tr

o o

2 2

P i£
E

s 2
0 o

>- i*=
nj >>

2 |

§ 1

ct» O

^ E 1

¦§ °

-S 2

^ ^ ^ ^ & >

o
Q.

-tr
o
2

w
(0

Q.
£

fe O

2 "o
0 2

a M M
o o =
o> .2
° ° ?

"92

0) (O y>

>» >*
x: _c

o ro

2 tr

0 J
a>

— WW

2. w w

ST 3 3

g 2 2
2

E
<

CO CO

tr -e

o o

2 2

a) q_

I I

aj
tr
o
2

CO CO

tr tr

o o

2 2

c/> w w
w >>

>: o

€ E

JZ c
in jc

-------
Decapoda
Copepoda
Diptera
Tricoptera
Odonata
Heteroptera
Coleoptera
Tricladida
Isopoda
Oligochaeta
Gastropoda
Hirudinea
Lamellbranchia

Recoptera

Amphipoda

Ostracoda

Cladocera

Ephemeroptera

-1

-0.5

0.5 1 1.5

Tolerance Relative to Daphnia

2.5

Adapted from: Wogram and Liess (2001).

Note: Tolerance is expressed on a logarithmic scale, and higher tolerance scores indicate lower sensitivity. Hence,
Daphnia sensitivity is standardized to zero, rather than one. A (log) tolerance value of 2 represents a factor of 100 in
differential sensitivity (i.e., 100 times less sensitive).

Figure D.3-18

Physiological Tolerance of Macroinvertebrate Orders to Organic
Compounds Relative to Daphnia magna Sensitivity

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
1 2 3 4 56 7 8 9 A1A2A3R4

Station

Notes: The box for each station is the interquartile range (1st through 3rd quartiles), and the whiskers are

the entire range.

The median of the average ranks is shown as a black dot within the box.

Descriptions of simplified IDs used here and elsewhere in the benthic invertebrate portions of the
risk assessment are provided in Section D. 1.1.2.

Figure D.3-19 Average Ranks Analysis for Six Benthic Community Metrics, with
Equal Weighting Assigned to Each Metric

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
MDS Axis

Note: Shaded ovals are for presentation purposes only and have no statistical meaning.

Figure D.3-20 Multidimensional Scaling (MDS) for Benthic Community Health
Metrics, Showing Metric Medians on MDS Plot

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
-2 0 2 4

MDS Axis 1

Figure D.3-21 Multidimensional Scaling (MDS) for Benthic Community

Composition Data at Fine-Grained Stations Only, Showing
Centroid and Individual Replicate Results on MDS Plot

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
-2

-1

0 1

MDS Axis 1

Note: Rays for extreme values have been truncated.

Figure D.3-22 Multidimensional Scaling (MDS) for Benthic Community

Composition Data at Coarse-Grained Stations Only, Showing
Centroid and Individual Replicate Results on MDS Plot

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
9.5

£ 8.5

re
s

3 7.5

«« 6.5

o
a>
Q

5.5 -

4.5

A1 A2 A3 1 2 3 4 5 6 7 8 9 R4

Notes: Reference location sediment: white bars.
Coarse-grained sediment: grey bars.
Fine-grained sediment: black bars.

Figure D.3-23 Mean Modified Hilsenhoff Biotic Index (MHBI) Values

MK01|O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
Fine-Grained Locations

3500

3000 -

2500 -

2000 -

m 1500 -
1000 -

500 -

6 7 8 9 R4

250

Coarse-Grained Locations

~ BIVALVIA

~ DIPTERA

~ ODONATA

~ GASTROPODA ~OLIGOCHAETA

~ PHARYNGOBDELLIDA ~TRICHOPTERA

~ OTHER

Figure D.3-24 Biomass (g) of Benthic Invertebrates Measured in Grab Samples,
with Subdivision into Major Taxonomic Groups, for Fine- and
Coarse-Grained Sediment

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC ^ 7/10/2003

-------
Coarse-G rained Sites

Fine-Grained Sites

A1 A2 A3 1 2 3 4 5
Station ID

~ Taxonomic Richness (+ SD)
¦ Median tPCB (mg/kg)

35
30
25
20
15
10
5
0

~ Taxonomic Richness (+ SD)
¦ Median tPCB] (mg/kg)

Note: Reference Stations are Al, A2, A3, and R4.

Figure D.3-25 Relationship Between Taxonomic Richness and tPCB, by
Substrate Type and Sampling Location

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
Coarse-Grained Sites

Fine-Grained Sites

200
180
160
140
120
100
80
60
40
20
0

A1 A2 A3

1 2
Station ID

~ Abundance (+ SD)

¦ Median tPCB Concentration

~ Abundance (+ SD)
¦ Median tPCB Concentration

Note: Reference Stations are Al, A2, A3, and R4.

Figure D.3-26 Relationship Between Abundance and tPCB, by Substrate Type
and Sampling Location

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------

0)
Q.
>

0)
¦Q

® -
x -2

£ 2 30
Z <3

o o

.2 = 20

Q £

4—

o
o

o qgp

<2> °
Oo0cPo^>0J BrO nU

° O°o oQxp §

O QqocD ® ®

II u 1 ¦ III

o
o

o °

o
cP

0.1

tPCB (mg/kg)

100

1000

Figure D.3-27

Relationship Between PCB Concentrations in Coarse-Grained
Locations and Taxonomic Richness in Individual Replicates

MK0110:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
tPCB (mg/kg)

Figure D.3-28 Relationship Between tPCB Concentrations in Coarse-Grained
Locations and Organism Abundance in Individual Replicates

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
0.001

Notes: Circle symbol and error bars represent the median and range of HQs calculated using literature-derived
sediment quality values.

Diamond symbols represent HQs calculated using site-specific sediment effects benchmark (MATC) of 3
mg/kg.

Figure D.4-1 Hazard Quotients (Median and Range) Based on Median Sediment
Chemistry for tPCBs, for Samples Collected in 1999 Close to
Sediment Quality Triad Stations

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
(a) Antimony

1 ^

in -Q
(5 ^

_Q >>
— (/)

ZZ,

O) 5
c

GJ "O
Q£ 0)

o s

0.1

o
£ £
5 ><

ZZ,

g\i

(5 "O

0£ (1)

o s

100
10
1

0.1
0.01
0.001

I I I I

Station

Figure D.4-2 Hazard Quotients (Median, Range) Based on Sediment Chemistry
for Other COPCs, for Samples Collected in 1999 Close to
Sediment Quality Triad Stations

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC qrs 7/10/2003

-------
(d) Chromium

(/) -Q

i5 ^
.q >*
it

a s

¦S 0.1

0.01

Station

(e) Copper

1 o

i/) -Q

™ I

£. 5T

C .2
ro T3
0£ o

Os
x

100

0.1

0.01

Mil

3=1=[

5 7

Station

(f) Lead

™ %
n

e I
a sr

-------
(g) Mercury

1 o

(/) -Q

i5 ^

.q >*

ii-

? s

re
0£

O
x

100
10
1

0.1
0.01

0.001 t

Station

(h) Silver

<° o

¦Q

re
0£

O
x

100 r

^ 1

0.1

0.01

Station

Figure D.4-2 Hazard Quotients (Median, Range) Based on Sediment Chemistry
for Other COPCs, for Samples Collected in 1999 Close to
Sediment Quality Triad Stations (Continued)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC nA 7/10/2003

-------
(a) Antimony

1 o

U) -Q

(5 =

_Q >>

ii in

a.
a

0.01

5a Main
Channel

5b Main
Channel

5c Main
Channel

5a SCOX 5b SCOX 5c SCOX

Reach and Substrate

(b) Barium

in -Q

S I
£. sr

— £

re
a.

0.1

0.01

0.001

5a Main
Channel

5b Main
Channel

5c Main
Channel

5a SCOX 5b SCOX 5c SCOX

Reach and Substrate

Figure D.4-3 Hazard Quotients (Median, Range) Based on Sediment Chemistry
for Other COPCs, for Broad EPA Sampling Results Not
Specifically Connected to the Triad Studies

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
(d) Chromium

1 o
(/) -Q

s £

.Q >*

it ^
o "T
s>

,2 0.1

0.01

5a Main
Channel

5b Main
Channel

5c Main
Channel

5a SCOX 5b SCOX 5c SCOX

Reach and Substrate

(e) Copper

O)
c
ro
on

1 O

I/) -Q

« I

¦Q >

.£S 0.1

0.01

5a Main
Channel

5b Main
Channel

5c Main
Channel

5a SCOX 5b SCOX 5c SCOX

Reach and Substrate

(f) Lead

1 o

in -Q

« I

£¦ 5T

-------
(g) Mercury

5a Main
Channel

5b Main
Channel

5c Main
Channel

5a SCOX 5b SCOX 5c SCOX

Reach and Substrate

(h) Silver

a.
o

T5
<

0.1

5a Main
Channel

5b Main
Channel

5c Main
Channel

5a SCOX 5b SCOX 5c SCOX

Reach and Substrate

(i) Total PAH

1 o
in -Q

(5 ^
>>
ii in

-------
(a) 48-in situ

~a
c

o ^

-Q {/>

£ S

> .q

t/> ^

T ®

.2 C

¦§ 2

1000
100
10

0.1
0.01

Station

¦a
c

O —

¦Q (/)

E (5
5T£
T <"

m 551
.2 =

¦§ 2

1000
100

0.1
0.01

Station

Figure D.4-4

Hazard Quotients (Median, Range) Based on Overlying Water PCB
Concentrations, Measured Synoptic with In Situ Toxicity Tests

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
(a) Relative to lower-bound effects threshold of 3 mg/kg (HQ > 1 in black)
Predators Shredders

A1 A3 1 2 3 4 5 6 7 8 9 R4

A1 A3 1

(b) Relative to upper-bound effects threshold of 10 mg/kg (HQ > 1 in black)

Predators

Shredders

5
4

+•»
c

'o 3

¦O

S 2

N
(3

1
0

- 1

1 1

• nsnl

O
Z

n I

A1 A3 1 2 3 4 5 6 7 8 9 R4

A1 A3 1

Figure D.4-5 Hazard Quotients for tPCB Tissue Residues in Benthic

Invertebrates, Relative to Two Effects Thresholds Derived from
Literature Studies

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC

-------
Endpoint

Stations

1. Sediment Toxicity

42-d Hyalella (lab)

Survival

•

42-d Hyalella (lab)

Growth

42-d Hyalella (lab)

Reproduction

•

43-d Chironomus (lab)

Survival

•

43-d Chironomus (lab)

Emergence

•

43-d Chironomus (lab)

Growth

•

10-d Hyalella {in situ)

Survival

•

10-d Chironomus {in situ)

Survival

•

48-h Daphnia {in situ)

Survival

•

48-h Lumbriculus {in situ)

Survival

TIE Treatments with Ceriodaphnia

Survival effect linked to PCBs

•

2. Benthic Community

Multivariate - Average Rank Plots

Equal Endpoint Weighting

•

Multivariate - MDS

Separation in 2-dimensional plot

•

Modified Hilsenhoff Bio tic Index

MUBI scores (ANOVA)

Taxa Richness

ANOVA versus references

•

Total Abundance

ANOVA versus references

•

Figure D.4-6 Weight-of Evidence Evaluation of Housatonic River Benthic Sampling Locations, with Indications
of Alteration/Risk Relative to Background

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC qa 7/10/2003

-------
Endpoint

Stations

3. Chemistry

Toxicity - Sediment tPCB

Synoptic with sediment toxicity tests

•

Benthos - Sediment tPCB

Synoptic with benthos collection

•

PSA Data - Sediment tPCB

Reach-wide sampling (median)

•

Water column tPCB

Synoptic with toxicity tests

•

Tissue tPCB in predators

Relative to literature benchmark

•

Tissue tPCB in shredders

Relative to literature benchmark

•

Tissue tPCB in oligochaetes (lab)

Relative to literature benchmark

4. Integrated Assessment

Toxicity Endpoints

Combined Assessment

•

Benthic Endpoints

Combined Assessment

•

Chemistry Endpoints

Combined Assessment

•

OVERALL

•

Notes:

• = major impact.

O = moderate impact.

O = negligible impact.

Figure D.4-6 Weight-of Evidence Evaluation of Housatonic River Benthic Sampling Locations, with Indications
of Alteration/Risk Relative to Background (Continued)

MK01 |O:\20123001.096\ERA_PB\ERA_PB_APD_FIGS.DOC q -i 7/10/2003

-------
NEW
YORK

MASSACHUSETTS

NOTES:

Risk to benthic invertebrates is based on a maximum acceptable threshold
concentration (MATC) of 3.0 mg/kg total PCB (tPCB) (dry weight) in surficial
sediments (0-6 inches).

LEGEND:

~ Town

Reach Break

BENTHIC INVERTEBRATE RISK

Roads

O <3 mg/kg

• >=3 mg/kg

1 Housatonic River Basin Hydrology

State Boundary

w^prE

0 2 4 6 8 Miles

Ecological Risk Assessment
GE/Housatonic River Site
Rest of River

FIGURE D.4-7
ASSESSMENT OF RISK TO
BENTHIC INVERTEBRATES
EXPOSED TO tPCBs

DOWNSTREAM
OF WOODS POND

| O:\gepitt\aprs\era_species.apr | layout - benthos d4-71 o:\gepitt\epsfiles\plots\in\benthos_risk_d4-7.eps 19:50 AM, 7/8/20031

-------
ATTACHMENT D.1

DESCRIPTIONS OF STUDIES CONSIDERED IN DETERMINATION OF

PCB TISSUE BENCHMARKS

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC

-------
1

ATTACHMENT D.1
DESCRIPTIONS OF STUDIES CONSIDERED IN DETERMINATION OF

PCB TISSUE BENCHMARKS

1. AROCLOR 1254

Sanders and Chandler (1972) exposed eight freshwater aquatic invertebrates to radiolabeled
solutions of Aroclor 1254 in a flow-through system for 4 days. Mortality was less than 10%
among all treatments, and the resulting mean dry weight (dw) tissue concentrations were as
follows: daphnid (Daphnia magna, 52 mg/kg dry), phantom midge (Chaoboruspunctipennis, 30
mg/kg dry), scud (Gammarus pseudolimnaeus, 39 mg/kg dry), mosquito larvae (Culex tarsalis,
27 mg/kg dry), glass shrimp (Palaemonetes kadiakensis, 16 mg/kg dry), stonefly (Pteronarcys
dorsata, 7.0 mg/kg dry), dobsonfly {Corydalus cornutus, 5.1 mg/kg dry), and crayfish
(Orconectes nais, 0.2 mg/kg dry). These results were reported in USACE ERED, after
conversion to wet weight (ww). There was one apparent error in USACE ERED: the converted
wet weight for C. punctipennis was reported as 1.2 mg/kg when it should have been 6.0 mg/kg.
This corrected value, as well as the other converted wet weights, was included for benchmark
derivation. Sanders and Chandler (1972) also reported limited data from prolonged exposures to
Aroclor 1254 for some of the above species. After a 7-d exposure, mosquito larvae accumulated
30 mg/kg dw and many of the pupae failed to metamorphose into adults. This "effect"
concentration was converted to wet weight (6.0 mg/kg wet) and used in the benchmark
derivation. The scud had a mean tissue concentration of 44 mg/kg (8.8 mg/kg ww) after 14 days
exposure, and the stonefly had a mean tissue concentration of 9 mg/kg (1.8 mg/kg ww) after 21
days exposure. No information was provided on effects associated with these exposures, so they
were not used in the benchmark derivation.

Lowe et al. (1972) exposed oyster spat (Crassostrea virginica) to continuous-flow solutions of
Aroclor 1254 for at least 24 weeks, and then transferred them to clean water for depuration.
After 24 weeks, oysters with a mean tissue residue concentration of 425 mg/kg (ww) showed no
significant mortality, but did show significant adverse effects on growth (weight and height) and
histopathology. Oysters with a mean tissue residue concentration of 101 mg/kg (ww) showed no
significant effects for these endpoints. These results were reported in USACE ERED, and

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC -> 7/10/2003

-------
1

considered in the benchmark derivation. Following transfer to clean water for depuration, the
oysters were purged of Aroclor 1254 after 12 weeks (for the 101 mg/kg treatment) and 28 weeks
(for the 425 mg/kg treatment), and recovery from tissue alterations occurred within 12 weeks.

Duke et al. (1970) investigated the toxicity of Aroclor 1254 to several marine aquatic
invertebrates. Following 48-h exposures, pink shrimp (Penaeus duorarum) showed no mortality
with a tissue residue of 1.30 mg/kg and 100% mortality with a tissue residue of 3.90 mg/kg. A
20-d exposure of pink shrimp to 5.0 |ig/L Aroclor 1254 resulted in 72% mortality; dead shrimp
had a mean tissue residue of 16 mg/kg, while surviving shrimp had a mean tissue residue of 33
mg/kg and exhibited no unusual behavior or loss of equilibrium. In the 96-h exposure of young
oysters, shell growth was completely inhibited at the highest test concentration (100 |_ig/L);
however, tissue residue was not measured. These oysters did survive and, within 3 weeks of
transfer to clean water, showed no significant difference in shell growth relative to controls.
After the 96-h exposure, tissue residues of 8.1 and 33 mg/kg were associated with decreases in
shell growth rate of 19 and 41%, respectively. Juvenile blue crabs (Callinectes sapidus)
accumulated 23 mg/kg of Aroclor 1254 after a 20-d exposure, with no adverse effects on
survival. These results were reported in US ACE ERED (although two were listed as "PCBs"
instead of Aroclor 1254) and Jarvinen and Ankley (1999). All these results were considered for
benchmark derivation.

Pinkney et al. (1985) exposed the estuarine amphipod, Gammarus tigrinus, to Aroclor 1254 in
feeding experiments with a contaminated fungus. After 24 h, the mean amphipod tissue
concentration was 23.2 mg/kg dw and mortality was only 3.4%. This result was reported in
USACE ERED (converted to 4.64 mg/kg ww), but the effect endpoint was reported as behavior.
This result was considered for benchmark derivation, but as a "no effect" concentration for
mortality. The authors also reported that virtually all the 24-h tissue concentration was
accumulated within the first 9 h, and also that the mean tissue concentration decreased to 8.42
mg/kg (dw) (1.68 mg/kg ww) after transfer to uncontaminated food for 144 h.

Nimmo et al. (1974) conducted laboratory exposures of the estuarine grass shrimp
(Palaemonetes pugio) to Aroclor 1254 for varying durations, and measured mortality and tissue
concentrations. In a 7-d exposure, mean mortality of 4% had a corresponding tissue

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC

-------
1

concentration of 5.4 mg/kg (ww), and 60% mortality occurred when the tissue concentration was
65.0 mg/kg. After 16 days, mean mortality associated with an 18-mg/kg ww tissue concentration
was 40%, but this was not statistically significant (p > 0.05). At 27 mg/kg (ww), mean mortality
was statistically significant at 45%. When grass shrimp were exposed to Aroclor 1254 for 35 d,
no significant mortality occurred and the mean tissue concentration was 16.5 mg/kg (ww). In
addition to these laboratory exposures, caged grass shrimp were also exposed to PCB-
contaminated sediment in the field. After 3 months, the mean tissue concentration was 0.42
mg/kg. Results from these studies were reported in USACE ERED and Jarvinen and Ankley
(1999), and were used in the benchmark derivation.

2. CLOPHEN A50

Clophen A is the trade name for German-manufactured PCB mixtures. Clophen A50 contains
44% pentachlorobiphenyls, and like Aroclor 1254, it has chlorine content of 54%. Four studies
listed in USACE ERED as using "PCBs" were conducted using Clophen A50; three of these
studies were listed as such in Jarvinen and Ankley (1999).

Sodergren and Svensson (1973) exposed mayfly nymphs (Ephemera danica) to a Clophen A50
solution for 9 days. There was no mortality and no significant change in organism weight
associated with a tissue residue concentration of 1.3 mg/kg ww. Because of the similarity of
PCB composition, this study was considered in the benchmark derivation.

Veldhuizen-Tsoerkan et al. (1991) exposed mussels (Mytilus edulis) to PCBs by feeding an algal
diet cultured in seawater spiked with Clophen A50, and then measured several physiological
parameters. Total PCB (tPCB) concentrations were calculated based on measurement of eight
PCB congeners (52, 87, 101, 105, 118, 138, 153, and 180). After 3 months there was no
significant reduction of anoxia tolerance (tissue concentration 3.0 mg/kg ww); however, there
was a significant reduction after 6 months (tissue concentration 7.0 mg/kg ww). There were no
significant changes in adenylate energy charge (AEC) after a 6-month exposure. The original
paper reported PCB tissue concentrations as wet weight, but reported cadmium tissue
concentrations as dry weight. In USACE ERED, the dry-to-wet weight conversion factor was

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC

-------
1

applied as if the PCB data were dry weight. We used the original wet weight concentrations for
benchmark derivation.

3. TOTAL PCBs OR PCB MIXTURES

Wright State University (EVS, 2003) conducted 7-d in situ exposures to Housatonic River
sediment using the oligochaete, Lumbriculus variegatus. Although mortality data were not
reported, it was presumed that no significant effects were observed, because sufficient tissue was
collected for chemical analyses. The highest tissue concentration was 4.41 mg/kg ww for tPCBs.
This result was considered for benchmark derivation.

Boese et al. (1995) conducted 119-d laboratory exposures using the marine clam, Macoma
nasuta. Three sediment with varying TOC content (0.8 to 2.5%) were spiked with equal
amounts of 13 PCB congeners (18, 52, 101, 105, 118, 128, 138, 151, 153, 170, 180, 194 and
209) plus hexachlorobenzene (HCB). After 119 days, total mortality among test and control
clams was 4.1%; there were no obvious differences in mortality, behavior (i.e., reburial) or
weight between test and control treatments. The mean tPCB (sum of 13 congeners) tissue
concentrations for clams sampled from Days 42 to 119 were 8.7 mg/kg (bulk; TOC = 0.8%), 4.6
mg/kg (medium; TOC = 1.3%), and 2.6 mg/kg (fine; TOC = 2.5%) for the three sediment types.
All tissue residue concentrations were reported as dry tissue weight. Tissue residues were
highest in the sediment with the lowest TOC (all three sediment contained the same nominal
PCB concentration). Tissue lipid data were available only for Days 0 and 42, due to a freezer
failure; mean dry weight lipid was 7.5% on Day 0 and 4.6% on Day 42. Data from this study
were included in US ACE ERED, after conversion to a wet weight concentration of 1.7 mg/kg.
This tissue residue value of 1.7 mg/kg was used in the benchmark derivation.

Lester and Mcintosh (1994) exposed mysids (Mysis relicta) to PCB-contaminated sediment from
Lake Champlain and a control sediment for 24 d. In this study, the mysids were either allowed
direct contact with the sediment or were screened to prevent such direct sediment contact (i.e.,
exposure was only through overlying water). Tissue samples were analyzed for 89 individual
congeners, and "total PCBs" referred to the sum of these congeners. The authors noted that there
was no unusual mysid behavior during the 24-d exposure period, but gave no information about

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC a 7/10/2003

-------
1

whether significant mortality had occurred. Mysids in the unscreened treatment accumulated 1.9
mg/kg (ww) tPCB, whereas organisms in the screened treatment accumulated only 0.24 mg/kg
(ww). Lipid content ranged from 5.8 to 7.7% (ww), and there were no significant differences
between treatments or time. The results from this study were reported in USACE ERED, and
considered in the tissue benchmark derivation.

Drouillard et al. (1996) exposed mayfly nymphs (Hexagenia limbata) to contaminated sediment
from the Detroit River for 32 d to measure uptake of selected organochlorine compounds. No
information was provided regarding effects associated with this exposure, so these data have not
been included in the tissue benchmark derivation. Total PCB tissue concentrations (based on the
sum of 13 PCB congeners) in two exposure systems were 16.1 and 10.7 mg/kg lipid, and lipid
content was 1.61 to 2.65%.

Ankley et al. (1992) exposed the oligochaete Lumbriculus variegatus to PCB-contaminated
sediment for 30 d. The resulting mean tPCB tissue concentration was 33.0 mg/kg lipid, and the
lipid content was 1.05%. Concentrations of PCB homologs were also reported, based on lipid
content. No information was provided regarding effects associated with this exposure, so these
data have not been included in the tissue benchmark derivation.

Brunson et al. (1998) measured tPCB concentrations in field-collected oligochaetes as well as
oligochaetes (Lumbriculus variegatus) that were exposed to the same sediment in 28-d
laboratory exposures. Total PCB concentrations in field-collected and laboratory-exposed
oligochaetes ranged from 0.045 to 0.697 mg/kg (ww) (geometric mean of 0.18 mg/kg ww). Test
sediment contained other organic contaminants and no information about effects was reported, so
these data have not been included in the tissue benchmark derivation.

4. OTHER AROCLORS OR INDIVIDUAL PCB CONGENERS

This section presents literature review results of Aroclors other than 1254/1260 and individual
PCB congeners. These studies were not included in the development of a PCB tissue threshold
for the Housatonic River because they were determined not to be the preferred compounds for
development of the benchmark, but they are included to provide a comparison of effects
concentrations across various PCB types.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC r 7/10/2003

-------
1

Hansen et al. (1974) investigated the acute (96-h) toxicity of Aroclor 1016 to three estuarine
invertebrates: American oyster (Crassostrea virginica), brown shrimp (Penaeus aztecus), and
grass shrimp. For oysters, tissue concentrations of 4.0, 32, and 95 mg/kg were associated with
10, 38, and 93% reductions in shell growth, respectively. For brown shrimp, tissue
concentrations of 3.8 and 42 mg/kg were associated with 8 and 43% mortality, respectively. For
grass shrimp, tissue concentrations of 1.1, 22, and 44 mg/kg were associated with mortalities of
33, 38, and 93%, respectively. Neff and Giam (1977; cf. Jarvinen and Ankley [1999]) exposed
juvenile horseshoe crab (Limulus polyphemus) to Aroclor 1016 solutions for 2 days. There was
no effect on survival at a tissue concentration of 11.2 mg/kg (ww), but >50%) survival occurred at
a tissue concentration of 31.9 mg/kg (ww). There was no effect on growth at a tissue
concentration of 92.9 mg/kg (ww).

Nebeker and Puglisi (1974; cf. Jarvinen and Ankley [1999]) exposed the freshwater amphipod,
Gammarus pseudolimnaeus, to solutions of Aroclor 1242 or 1248 for 60 days. For Aroclor 1242
(in terms of ww tissue concentrations), survival was reduced at 409 mg/kg, but there was no
effect at 246-387 mg/kg. Reproduction was reduced at 246-387 mg/kg, but not at 71-80 mg/kg.
For Aroclor 1248 (in terms of ww tissue concentrations), there was no effect on survival at 552
mg/kg and no effect on reproduction at 127 mg/kg, but reproduction was reduced at 552 mg/kg.

Borgmann et al. (1990) studied the chronic effects of Aroclor 1242 and two PCB congeners
(2,5,2',5'-tetrachlorobiphenyl and 3,4,3',4'-tetrachlorobiphenyl) using the freshwater amphipod,
Hyalella azteca. Each PCB compound was tested individually. Aroclor 1242 concentrations
were measured radiometrically while the two congeners were analyzed by gas chromatography.
For Aroclor 1242, there were no effects on survival, weight, or reproduction associated with a
mean tissue concentration of 30 mg/kg (ww) after 10 weeks. For 2,5,2',5'-tetrachlorobiphenyl,
there were no effects on survival, weight, or reproduction associated with a mean tissue
concentration of 53.9 mg/kg (ww) after 6 weeks or 74.5 mg/kg after 10 weeks. For 3,4,3',4'-
tetrachlorobiphenyl, there were no effects on survival, weight, or reproduction associated with a
mean tissue concentration of 141 mg/kg (ww) after 10 weeks. Complete mortality did occur at
higher concentrations of Aroclor 1242 and 2,5,2',5'-tetrachlorobiphenyl, but tissue
concentrations were not measured in those treatments. None of the 3,4,3',4'-tetrachlorobiphenyl
treatments tested showed significant effects. Although no significant effects were reported for

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC

-------
1 the tissue concentrations reported above, data from this study should be treated with caution

2 because of high replicate variability (potentially masking significant effects) and variable control

3 performance.

4 Dillon et al. (1990) conducted 21-d Daphnia magna tests with seven PCB congeners (52, 77,

5 101, 118, 138, 153, and 180), each tested individually. After 21 days there were no significant

6 decreases in survival, reproduction, or biomass, except for biomass with PCB-101. Some

7 increases in biomass and reproduction occurred. Mean tissue concentrations accumulated after

8 21 days ranged from 4.0 mg/kg (ww) for PCB-52 to 26.6 mg/kg (ww) for PCB-118.

9 Fry and Fisher (1990) exposed fourth instar Chironomus riparius to sediment spiked with 5,5',6-

10 trichlorobiphenyl for 24 h. However, all of the concentration data for water, tissue, and sediment

11 samples were reported as "|ig/L" so these data were not considered for the benchmark derivation

12 due to data quality concerns.

13 5. REFERENCES

14 Ankley, G.T., P.M. Cook, A.R. Carlson, D.J. Call, J.A. Swenson, H.F. Corcoran, and R.A. Hoke.

15 1992. Bioaccumulation of PCBs from sediments by oligochaetes and fishes: comparison of

16 laboratory and field studies. Canadian Journal of Fisheries and Aquatic Sciences 49:2080-2085.

17 Boese, B.L., M. Winsor, H. Lee II, S. Echols, J. Pelletier, and R. Randall. 1995. PCB congeners

18 and hexachlorobenzene biota sediment accumulation factors for Macoma nasuta exposed to

19 sediments with different total organic carbon contents. Environmental Toxicology and Chemistry

20 14:303-310.

21 Borgmann, U., W.P. Norwood, and K.M. Ralph. 1990. Chronic toxicity and bioaccumulation of

22 2,5,2',5'- and 3,4,3',4'-tetrachlorobiphenyl and Aroclor® 1242 in the amphipod Hyalella azteca.

23 Archives of Environmental Contamination and Toxicology 19:558-564.

24 Brunson, E.L., T.J. Canfield, F.J. Dwyer, C.G. Ingersoll and N.E. Kemble. 1998. Assessing the

25 bioaccumulation of contaminants from sediments of the upper Mississippi River using field-

26 collected oligochaetes and laboratory-exposed Lumbriculus variegatus. Archives of

27 Environmental Contamination and Toxicology 35:191-201.

28 Dillon, T.M., W.H. Benson, R.A. Stackhouse, and A.M. Crider. 1990. Effects of selected PCB

29 congeners on survival, growth and reproduction in Daphnia magna. Environmental Toxicology

30 and Chemistry 9:1317-1326.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC

-------
1 Drouillard, K.G., J.J.H. Ciborowski, R. Lazar and G.D. Haffner. 1996. Estimation of the uptake

2 of organochlorines by the mayfly Hexagenia limbata {Ephemeroptera: Ephemeridae). Journal of

3 Great Lakes Research 22:26-3 5.

4 Duke, T.W., J.I. Lowe, and A.J. Wilson Jr. 1970. A polychlorinated biphenyl (Aroclor 1254®) in

5 the water, sediment and biota of Escambia Bay, Florida. Bulletin of Environmental

6 Contamination and Toxicology 5:171-180.

7 EVS Environment Consultants. 2003. Assessment of In Situ Stressors and Sediment Toxicity in

8 the Lower Housatonic River. Adapted from a study by G.A. Burton. GE/Housatonic River

9 Project, Pittsfield, MA.

10 Fry, D.M. and S.W. Fisher. 1990. Effect of sediment contact and uptake mechanisms on

11 accumulation of three chlorinated hydrocarbons in the midge, Chironomus riparius. Bulletin of

12 Environmental Contamination and Toxicology 44:790-797.

13 Hansen, D.J., P.R. Parrish and J. Forester. 1974. Aroclor 1016: Toxicity to and uptake by

14 estuarine animals. Environmental Research 7:363-373.

15 Jarvinen, A.W. and G.T. Ankley. 1999. Linkage of Effects to Tissue Residues: Development of a

16 Comprehensive Database for Aquatic Organisms Exposed to Inorganic and Organic Chemicals.

17 SETAC Press, Pensacola, FL. 364 pp.

18 Lester, D.C. and A. Mcintosh. 1994. Accumulation of polychlorinated biphenyl congeners from

19 Lake Champlain sediments by My sis relicta. Environmental Toxicology and Chemistry. 13:1825-

20 1841.

21 Lowe, J.I., P.R. Parrish, J.M. Patrick Jr., and J. Forester. 1972. Effects of the polychlorinated

22 biphenyl Aroclor® 1254 on the American oyster Crassostrea virginica. Marine Biology 17:209-

23 214.

24 Nebeker, A.V. and F.A. Puglisi. 1974. Effect of polychlorinated biphenyls (PCBs) on survival

25 and reproduction of Daphnia, Gammarus, and Tany tarsus. Transactions of the American

26 Fisheries Society 103:722-728.

27 Neff, J.M. and C.S. Giam. 1977. Effects of Aroclor® 1016 and Halowax® 1099 on juvenile

28 horseshoe crabs Limulus polyphemus. In Physiological Responses of Marine Biota to Pollutants.

29 F.J. Vernberg, A. Calabrese, F.P. Thurberg and W.B. Vernberb, Editors. Academic Press, New

30 York, NY.

31 Nimmo, D.R., J. Forester, P.T. Heitmuller and G.H. Cook. 1974. Accumulation of Aroclor®

32 1254 in grass shrimp (Palaemonetes pugio) in laboratory and field experiments. Bulletin of
3 3 Environmental Contamination and Toxicology 11:303-308.

34 Pinkney, A.E., G.V. Poje, R.M. Sansur, C.C. Lee and J.M. O'Connor. 1985. Uptake and

35 retention of 14C-Aroclor® 1254 in the amphipod, Gammarus tigrinus, fed contaminated fungus,

36 Fusarium oxysporum. Archives of Environmental Contamination and Toxicology 14:59-64.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC

-------
1 Sanders, H.O. and J.H. Chandler. 1972. Biological magnification of a polychlorinated biphenyl

2 (Aroclor 1254) from water by aquatic invertebrates. Bulletin of Environmental Contamination

3 and Toxicology 7(5):257-263.

4 Sodergren, A. and B. Svensson. 1973. Uptake and accumulation of DDT and PCB by Ephemera

5 danica (Ephemeroptera) in continuous-flow systems. Bulletin of Environmental Contamination

6 and Toxicology 9:345-350.

7 U.S. Army Corps of Engineers (USACE), U.S. Environmental Protection Agency (EPA).

8 Environmental Residue Effects Database (ERED). http://www.wes.army.mil/el/dots.

9 Veldhuizen-Tsoerkan, M.B., D.A. Holwerda, and D.I. Zandee. 1991. Anoxic survival time and

10 metabolic parameters as stress indices in sea mussels exposed to cadmium or polychlorinated

11 biphenyls. Archives of Environmental Contamination and Toxicology 20:259-265.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD1_PB.DOC

-------
ATTACHMENT D.2

RATIONALE USED IN DEVELOPMENT OF SEDIMENT QUALITY

VALUES FOR PCBs

MK01 |O:\20123001.096\ERA_PB\ERA_ATD2_PB.DOC

-------
1

ATTACHMENT D.2
RATIONALE USED IN DEVELOPMENT OF SEDIMENT QUALITY

VALUES FOR PCBs

1. APPROACHES TO DEVELOPING SEDIMENT QUALITY VALUES

This attachment discusses the development of sediment quality values (SQVs) for
polychlorinated biphenyls (PCBs).

The apparent effects threshold (AET) approach is based on the use of empirical data (both field
and laboratory) to identify concentrations above which biological effects are nearly always
expected. More specifically, this approach is intended to define the concentration of a
contaminant in sediment above which significant (p < 0.05) biological effects are always
observed (MacDonald et al. 2000a). Appropriate biological effects data include toxicity tests on
benthic and water-column organisms, changes in the abundance of various benthic species, and
changes in benthic community structure.

The water-sediment equilibrium partitioning approach (EqP) is another commonly used
technique for the development of SQVs. The approach is based on the premise that the
distribution of contaminants among different compartments in the sediment matrix (including
solids and pore water) is predictable based on their physicochemical properties, particularly their
degree of hydrophobicity. The EqP theory holds that non-ionic chemicals in sediment partition
between sediment organic carbon, pore water, and benthic organisms. The underlying
assumption of the EqP approach is that, at equilibrium, if the concentration in any one phase is
known, then the concentration in the other phases can be predicted. Sediment quality values are
calculated using water quality criteria (formulated to protect water column species) in
conjunction with the sediment/water partition coefficients for specific contaminants.

The screening-level concentration approach is a biological effects-based approach with the
objective of determining SQVs for the protection of benthic organisms. Matching biological and
sediment chemistry data are used to calculate screening level concentrations, which are estimates
of the highest concentration of a contaminant that can be tolerated by a predetermined proportion
of benthic infaunal species.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD2_PB.DOC 1 7/10/2003

-------
1

Long et al. (1995) developed the effects range approach, which was used to determine two
sediment quality guideline values: the effect range-low and the effect range-median. Matching
biological data from various modeling, laboratory, and in situ studies in marine and estuarine
studies were evaluated to determine the two guidelines.

Although the effects range approach was developed based on data gathered in saltwater
environments, the SQVs have also been applied to freshwater environments. MacDonald et al.
(2000b), in their examination of consensus-based sediment effect concentrations, determined that
the freshwater effect concentrations were similar to the marine. Their review of various
toxicological data sets indicated that the range of acutely lethal or effective concentrations of
PCBs for saltwater species (0.001 - 16.0 mg/kg) fully encompasses the range for freshwater
species (0.002 - 2.4 mg/kg). In addition, the range of species mean acute values for saltwater
crustaceans (0.011 - 0.013 mg/kg) falls within the reported range for freshwater crustaceans
(0.01 - 0.046 mg/kg).

A preliminary remediation goal (PRG) is included in the table of PCB SQVs, Table D.3-6,
Sediment Quality Values (SQVs) for Use in the Risk Evaluation of Housatonic River Sediment
contaminants of concern (COCs). Remediation goals are not the same as SQVs and should not
be viewed as such. Instead, they are regulatory values or thresholds for significant effects and
are determined based on the desirable level of environmental protection (Efroymson et al. 1997).
The PRG is included here to further typify the range of low-effect sediment PCB concentrations
available in the literature.

To better capture the potential contaminant risks to the benthic community, and in
acknowledgement of some of the confounding effects on contaminant toxicity exerted by
ancillary parameters such as sediment total organic carbon (TOC) and grain size, it is best to
review the SQVs as a range, rather than as discrete point-estimates. The sediment quality
benchmarks provide a low-effects range for total PCBs of 0.023 to 0.070 mg/kg. Based on the
multitude of modeling, laboratory, and in situ biological effects data evaluated in the
determination of the various SQVs, the degree of relative agreement among the range of low-
effect values is encouraging and enables a fairly high degree of confidence in the range. The
probable effect range (including mid-range effects concentrations), where effects are frequently

MK01 |O:\20123001.096\ERA_PB\ERA_ATD2_PB.DOC 2 7/10/2003

-------
1 expected to occur, is 0.18 to 0.40 mg/kg. The values for the severe or extreme effect range,

2 where effects almost always occur, are 1.7 to 5.3 mg/kg, and up to 53 mg/kg at 10% organic

3 carbon.

4 2. REFERENCES

5 Efroymson, R.A., G.W. Suter II, B.E. Sample, and D.S. Jones. 1997. Preliminary Remediation

6 Goals for Ecological Endpoints. Prepared for the U.S. Department of Energy. 27 pp.

7 Long, E.R., D.D. MacDonald, S.L. Smith, F.D. Calder. 1995. Incidence of adverse biological

8 effects within ranges of chemical concentrations in marine and estuarine sediments.

9 Environmental Management 19(l):81-97.

10 MacDonald, D.D., L.M. Dipinto, J Field, C.G. Ingersoll, E.R. Long, and R.C. Swartz. 2000a.

11 Development and evaluation of consensus-based sediment effect concentrations for

12 polychlorinated biphenyls. Environmental Toxicology and Chemistry 19(5): 1403-1413.

13 MacDonald, D.D., C.G. Ingersoll, and T.A. Berger. 2000b. Development and evaluation of

14 consensus-based sediment quality guidelines for freshwater ecosystems. Archives of

15 Environmental Contamination and Toxicology 39:20-31.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD2_PB.DOC

-------
ATTACHMENT D.3

DISCUSSION OF METRICS ADOPTED FOR THE ASSESSMENT OF
BENTHIC COMMUNITY ASSEMBLAGES

MK01 |O:\20123001.096\ERA_PB\ERA_ATD3_PB.DOC

-------
1

ATTACHMENT D.3
DISCUSSION OF METRICS ADOPTED FOR THE ASSESSMENT OF
BENTHIC COMMUNITY ASSEMBLAGES

1. TAXA RICHNESS AND COMPOSITION MEASURES

1.1 Taxa Richness

Taxa richness represents the number of distinct taxa types and is one of the most widely used
benthic metrics. This metric measures the overall variety of the macroinvertebrate community.
The metric can be applied to an individual benthic grab (i.e., replicate richness) or to the entire
biological assemblage sampled at a given location (i.e., station richness). In general, a reduction
in taxa richness is the expected response to major environmental perturbations. Although
communities with competitively dominant species and/or climax communities may exhibit
increased richness with small to moderate environmental perturbations, larger perturbations
(including chemical contamination) typically reduce taxa richness.

Taxa richness is usually based upon species-level identifications, but also can be designated as
groupings of taxa, often as higher taxonomic groups (such as genera, families, orders). Although
some organisms in the grab samples have been identified to the species level, for the purposes of
the multivariate analyses, taxa are differentiated on the basis of the lowest practical taxonomic
level.

1.2Total Abundance

The total number of organisms in a replicate or at a station provides a simple and intuitive
assessment of biological productivity. Caution must be used in the interpretation of raw
abundances because total abundance can be heavily influenced by substrate type, and because
raw abundance does not take into consideration the sensitivity of the identified organisms (i.e.,
pollution-resistant or invasive species are assigned the same value as sensitive species). It is for
this reason that other benthic metrics were combined with total abundance using a multivariate
approach to provide a more robust indicator of potential ecological disruption.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD3_PB.DOC

-------
1 1.3 EPT Relative Abundance

2 The relative abundance of organisms from the orders Ephemeroptera, Plecoptera, and

3 Trichoptera (EPT) is one of the most widely used and accepted metrics for the evaluation of

4 pollution disturbance in aquatic habitats. Lydy et al. (2000) investigated the effectiveness of 11

5 indices (including diversity, similarity, and biotic indices) and the EPT metric in summarizing

6 the degree of aquatic pollution. They found that the EPT metric was the most descriptive metric

7 in the evaluation of water quality degradation.

8 In general, species from these three orders are considered the most pollution-sensitive

9 invertebrate groups; EPT relative abundance is expected to decrease with increasing levels of

10 habitat perturbation. All invertebrate biotic indices in the literature assign low tolerance values

11 to these orders, regardless of the level of taxonomic resolution used in the particular index. The

12 Housatonic River grab samples contained no plecopterans (stoneflies). This may be an

13 expression of the sampling design, which preferentially sampled finer grained sediment that may

14 be inappropriate habitat conditions for juvenile stoneflies. Of the three EPT orders, plecopterans

15 are the most sensitive, based on the tolerance scores assigned to taxa in this order (Bode et al.

16 1991).

17 2. POLLUTION TOLERANCE MEASURES

18 In addition to EPT taxa, other freshwater organisms have been evaluated for their sensitivity to

19 contaminant-induced stresses. Various rating scales or pollution tolerance indices have been

20 devised to describe the relative sensitivity or tolerance to pollution. The ratings also form a key

21 component of the modified Hilsenhoff Biotic Index (MHBI). For this project, the tolerance

22 scores were also summarized as follows:

23 ¦ Intolerant (sensitive) taxa: tolerance value = 0 to 3.

24 ¦ Facultative taxa: tolerance value = 4 to 7.

25 ¦ Tolerant taxa: tolerance value = 8 to 10.

27 The invertebrate orders Diptera, Oligochaeta, and Gastropoda were broken into these categories

28 for inclusion in the multivariate assessment. These specific orders were selected for the

29 following reasons:

MK01 |O:\20123001.096\ERA_PB\ERA_ATD3_PB.DOC

-------
1

¦ Taxa in these three orders comprise 80.3% of the organisms collected in the grab
samples. This provides sufficient sample size for adequate numerical representation
of the various levels of pollution sensitivity in all three orders.

¦ Pollution sensitivities for taxa in these three orders are highly variable, ranging from
sensitive (tolerance value of 0 to 3) to tolerant (tolerance score of 8 to 10).

¦ The relative pollution tolerance of these three orders selected was further affirmed in
an evaluation of macroinvertebrate species sensitivities to toxic organic compounds,
relative to Daphnia magna sensitivity (Wogram and Liess 2001; Figure D.3-1). The
authors conducted database and literature searches to obtain toxicity data on the
macroinvertebrate orders. Macroinvertebrate species sensitivities were then
compared to D. magna sensitivity to toxic organics and a sensitivity ranking
calculated. The addition of dipterans, oligochaetes, and gastropods to the EPT
assessment provides for a range of taxa sensitivities in the overall analysis.

14 Many taxa within the order Diptera are considered pollution-tolerant (Barbour et al. 1996). An

15 increase in dipteran abundance is expected to increase in the presence of habitat perturbation.

16 However, closer examination of tolerance values in this order shows species encompassing the

17 full range of tolerance values (0 to 10). We therefore evaluated the relative abundance of only

18 those taxa that could be termed "pollution-tolerant," defined as those having tolerance values of

19 8 to 10.

20 The rationale for the evaluation of pollution-tolerant taxa within Oligochaeta and Gastropoda

21 orders was similar to that of selection of tolerant Diptera as a metric. Both of these groups

22 contain species that span the entire range of pollution tolerance values, from 0 to 10.

23 3. REFERENCES

24 Barbour, M.T., J. Gerritsen, G.E. Griffith, R. Frydenborg, J.S. White, and Bastian, M.L. 1996. A

25 framework for biological criteria for Florida streams using benthic macroinvertebrates. J.N. Am.

26 Benthol. Soc. 15:185-211.

27 Bode, R.W., M.A. Novak, and L.E. Abele. 1991. Quality Assurance Work Plan for Biological

28 Stream Monitoring in New York State. New York State Department of Environmental

29 Conservation. Albany, NY.

30 Lydy, M.J., C.G. Crawford, and J.W. Frey. 2000. A comparison of selected diversity, similarity,

31 and biotic indices for detecting changes in benthic invertebrate community structure and stream

32 quality. Arch. Environ. Contam. Toxicol. 39:469-479.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD3_PB.DOC

7/10/2003

-------
1 Wogram, J. and Liess, M. 2001. Rank ordering of macroinvertebrate species sensitivity to toxic

2 compounds by comparison with that of Daphnia magna. Bull. Environ. Contam. Toxicol.

3 67:360-367.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD3_PB.DOC

-------
ATTACHMENT D.4

GRAIN SIZE SENSITIVITY OF HOUSATONIC RIVER TOXICITY TEST

ORGANISMS

MK01 |O:\20123001.096\ERA_PB\ERA_ATD4_PB.DOC

-------
1

ATTACHMENT D.4
GRAIN SIZE SENSITIVITY OF HOUSATONIC RIVER TOXICITY TEST

ORGANISMS

This document presents an overview of information regarding the sensitivity to sediment grain
size distribution of four aquatic species (Daphnia magna, Hyalella azteca, Chironomus tentans,
and Lumbriculus variegatus) used for testing Housatonic River sediment. Two of the four
species (C. tentans and H. azteca) were tested using both chronic laboratory methods and in situ
methods. The remaining two species were tested using in situ methods only.

1. HYALELLA AZTECA

Ingersoll and Nelson (1990) reported that Hyalella azteca had an extremely wide tolerance to
sediment grain size, and that long-term tests with sediment ranging from >90% silt and clay to
100% sand showed no adverse effects on either survival or growth.

Ankley et al. (1994) exposed//, azteca to 50 uncontaminated sediment in 10-d exposures. These
sediment represented a wide range of grain size, organic carbon content, and mineralogical
compositions. The approximate ranges of grain size composition were as follows: sand (5 -
100%>), silt (0 - 75%>), and clay (15 - 90%>). H. azteca survival appeared to be independent of
sediment grain size distribution.

Suedel and Rodgers (1994) exposed H. azteca to formulated sediment and non-toxic field-
collected sediment encompassing sediment grain size ranges of 0 to 100%> sand, 0 to 100%> silt,
and 0 to 60%> clay. Selection of sediment grain sizes for testing was based on sediment types
where H. azteca had been reported, and the range of sediment grain sizes found through the
United States. Mean 10-d survival of H. azteca was >83%> in 17 of 18 formulated sediment. The
one exception was for 100%> mortality in a sample of 100%> formulated sand. All the formulated
sediment were initially conditioned in flowing water for 7 days to promote microorganism
colonization prior to their use in tests. Further testing of fine, medium and coarse fractions of
100%o formulated sand that were conditioned for 14 days yielded 10-d mean survival values of
>90%o, indicating that the initial mortality was due to a lack of conditioning rather than the
sediment particles themselves. As additional verification, two field-collected sediment were

MK01 |O:\20123001.096\ERA_PB\ERA_ATD4_PB.DOC -> 7/10/2003

-------
1

wet-sieved to yield fractions of >90% medium or fine sand, and the resulting 10-d mean
survivals were 88 to 100%. Mean 10-d survival in 16 field-collected sediment (grain size
ranging from 4 to 100% sand, 0 to 94% silt, and 0 to 5% clay) was 84 to 100%; the organic
matter content in these samples ranged from 0.12 to 7.8%.

Ingersoll et al. (1998) conducted 42-h life-cycle tests with H. azteca, and found that there were
no statistically significant correlations between survival, growth or reproduction and the
sediment grain size (ranging from silt to sand) and TOC (ranging from 0.3 to 9.6%). There was
a weak relationship between grain size and reproduction, but this was attributed to the presence
of higher contaminant concentrations in sediment with higher sand content.

Overall, these studies indicate that H. azteca lethal and sublethal endpoints are unlikely to be
affected significantly over the range of grain sizes tested in the Housatonic River toxicity testing
program.

2. CHIRONOMUS TENTANS

In addition to the studies performed with H. azteca, Ankley et al. (1994) also exposed
Chironomus tentans to 50 uncontaminated sediment in 10-d exposures. The results suggested
that C. tentans growth was better in slightly coarser substrates. However, Sibley et al. (1997)
determined that this correlation was influenced by the presence of inorganic material in the gut
contents, and that when ash-free dry weight (AFDW) was measured instead of dry weight the
correlation disappeared.

As with H. azteca, Suedel and Rodgers (1994) also exposed C. tentans to a wide range of
formulated and non-toxic field-collected sediment. When C. tentans were initially exposed to the
18 formulated sediment, mean 10-d survival ranged from 6 to 83%, and mean survival was only
>80% when the formulated sediment consisted of at least 80% silt; this poor survival was
attributed to a lack of particulate organic material. When the formulated sediment were amended
with 2.5%) organic material (to provide a food source and for tube building) mean 10-d survival
in all the formulated sediment was >80%. When the C. tentans were exposed to the non-toxic
field-collected sediment, mean 10-d survival ranged from 7 to 46% when the organic matter
content was <0.3%, and from 84 to 100% when the organic matter content was 0.91 to 7.8%.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD4_PB.DOC rs 7/10/2003

-------
1

Overall, C. tentans survival was positively correlated with silt and organic matter content, and
negatively correlated with sand and solids content. The Suedel and Rodgers (1994) study
indicated that reduced survival may occur in sediment with extremely low organic carbon
content and coarse particle sizes. However, the most pronounced toxicity in the Housatonic
River testing program was for fine-grained and high TOC sediment; therefore, particle size
considerations cannot explain the trends in toxicity observed.

Kemble et al. (1999) evaluated two formulated sediment using C. tentans. The formulated
sediment represented two sand:total organic carbon (TOC) combinations (FS-A = 74% sand +
2.2% TOC; and FS-C = 14% sand + 1.6% TOC) that were compared to a field-collected control
sediment containing 74% sand + 3.3% TOC. In 10-d exposures, mean survival was 73 and 78%
in the formulated sediment and 90% in the control, mean head capsule width was 0.59 to 0.62
mm in all treatments, and mean dry weight was 1.90 and 2.34 mg in the formulated sediment and
1.87 mg in the control. There were no statistically significant differences between the formulated
sediment and the control for these endpoints, except that dry weight in the FS-C treatment was
significantly higher than the control. Since the FS-A and FS-C sediment approximately bracket
the range of sediment particle size distributions found in the PSA, the lack of statistically
significant differences in effects between these formulated sediment suggests that particle size
differences do not explain the wide range of responses observed in Housatonic River test
sediment.

Overall, these studies indicate that C. tentans lethal and sublethal endpoints are unlikely to be
affected significantly over the range of grain sizes tested in the Housatonic River toxicity testing
program.

3. LUMBRICULUS VARIEGATUS

In addition to H. azteca and C. tentans, Ankley et al. (1994) also exposed Lumbriculus
variegatus to 50 uncontaminated sediment in 10-d exposures. L. variegatus appeared to tolerate
a wide range of physicochemical conditions, including sediment grain size and TOC.

Kemble et al. (1999) also evaluated two formulated sediment applied in adult L. variegatus
testing. The formulated sediment represented two sand:total organic carbon (TOC) combinations

MK01 |O:\20123001.096\ERA_PB\ERA_ATD4_PB.DOC ^ 7/10/2003

-------
1 (FS-A = 74% sand + 2.2% TOC; and FS-C = 14% sand + 1.6% TOC) that were compared to a

2 field-collected control sediment containing 74% sand + 3.3% TOC. In 10-d exposures, mean

3 survival in both formulated sediment was significantly higher than in the control sediment, but

4 mean dry weight in both formulated sediment was significantly lower than in the control

5 sediment. Although these tests began with the addition of 10 adult oligochaetes to each replicate

6 chamber, the final survival values were all reported to be >100%; there was no mention of

7 offspring being produced.

8 4. DAPHNIA MAGNA

9 Daphnia magna have been used for sediment toxicity testing in other studies, either for testing

10 sediment porewater, elutriates, or whole sediment. Unlike the other three species discussed here,

11 D. magna is planktonic and does not burrow into sediment. Therefore, the potential effects of

12 sediment grain size are not relevant to this species. D. magna may be exposed to sediment-

13 associated contaminants that are released into the water column, or during feeding at the

14 sediment surface (ASTM, 2001; Nebeker et al., 1984).

15 5. REFERENCES

16 Ankley, G.T., D.A. Benoit, J.C. Balogh, T.B. Reynoldson, K.E. Day and R.A. Hoke. 1994.

17 Evaluation of potential confounding factors in sediment toxicity tests with three freshwater

18 benthic invertebrates. Environ. Toxicol. Chem. 13:627-635.

19 ASTM (American Society for Testing and Materials). 2001. Test method for measuring the

20 toxicity of sediment-associated contaminants with freshwater invertebrates. Method E1706-00.

21 Volume 11.05. Annual Book of ASTM Standards. American Society for Testing and Materials,

22 Philadelphia, PA.

23 Ingersoll, C.G. and M.K. Nelson. 1990. Testing sediment toxicity with Hyalella azteca

24 (Amphipoda) and Chironomus riparius (Diptera), pp. 93-109. In: W.G. Landis and W.H. van der

25 Schalie (eds.). Aquatic Toxicology and Hazard Assessment: Thirteenth Symposium. STP 1096.

26 American Society for Testing and Materials, Philadelphia, PA, USA.

27 Ingersoll, C.G., E.L. Brunson, F.J. Dwyer, D.K. Hardesty and N.E. Kemble. 1998. Use of

28 sublethal endpoints in sediment toxicity tests with the amphipod Hyalella azteca. Environ.

29 Toxicol. Chem. 17:1508-1523.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD4_PB.DOC

-------
1 Kemble, N.E., F.J. Dwyer, C.G. Ingersoll, T.D. Dawson and T.J. Norberg-King. 1999. Tolerance

2 of freshwater test organisms to formulated sediments for use as control materials in whole-

3 sediment toxicity tests. Environ. Toxicol. Chem. 18:222-230.

4 Nebeker, A.V., M.A. Cairns, J.H. Gakstatter, K.W. Malueg, G.S. Schuytema and D.F.

5 Krawczyk. 1984. Biological methods for determining toxicity of contaminated freshwater

6 sediments to invertebrates. Environ. Toxicol. Chem. 3:617-630.

7 Sibley, P.K., P.D. Monson and G.T. Ankley. 1997. The effect of gut contents on dry weight

8 estimates of Chironomus tentans larvae: implications for interpreting toxicity in freshwater

9 sediment toxicity tests. Environ. Toxicol. Chem. 16:1721-1726.

10 Suedel, B.C. and J.H. Rodgers Jr. 1994. Responses of Hyalella azteca and Chironomus tentans

11 to particle-size distribution and organic matter content of formulated and natural freshwater

12 sediments. Environ. Toxicol. Chem. 13:1639-1648.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD4_PB.DOC

-------
ATTACHMENT D.5

ALTERNATIVE CONCENTRATION-RESPONSE ASSESSMENT FOR
PCB AND TOXICITY TEST ENDPOINTS

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC

-------
1

ATTACHMENT D.5
ALTERNATIVE CONCENTRATION-RESPONSE ASSESSMENT FOR
PCB AND TOXICITY TEST ENDPOINTS

As described in Appendix D, two separate types of exposure data, the "median" and the "most
synoptic," were used to evaluate concentration-response for sediment toxicity test endpoints.
These two approaches made different assumptions about the extent to which small-scale
variations in sediment chemistry would affect the choice of appropriate sediment chemistry for
use in dose-response modeling.

Two Approaches to Treatment of Exposure Data in Conjunction with

Sediment Toxicity Tests

¦ "Median," in which all sediment chemistry data located within a 5-meter radius
and over the 1999 summer season were combined and the median value
determined. This approach was the preferred method in the ERA since it
incorporated a range of PCB concentrations to which the test organisms may
have been exposed, and accounted for the small-scale variations in sediment
chemistry.

¦ "Most Synoptic," in which only the single PCB chemistry value taken closest in
space and time to the toxicity test, was used. Individual PCB measurements
have high uncertainty due to small sample size and large small-scale variations
in sediment chemistry, even for samples collected at the same location on the
same day. However, use of these samples does not introduce additional
uncertainty due to temporal variation in PCB concentrations.

The discussion in Section 3 and Appendix D focuses on the "median" approach. This attachment
presents a reanalysis using the "most-synoptic" data processing approach. The purpose of the
separate statistical assessments is to determine whether the threshold effects concentrations are
robust with respect to changes in data processing methods.

Methods for Evaluating Concentration-Response for Toxicity Data

¦ Individual Endpoint Analysis - Each toxicity endpoint was investigated
individually using conventional descriptive statistics that related degree of effect
to PCB concentrations (e.g., LC50, IC2o, NOAEL, LOAEL).

¦ Combined Endpoint Analysis - The toxicological endpoints were integrated using
a general linear modeling approach to identify similarities and differences in
concentration-response relationships across species and endpoints.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC -> 7/10/2003

-------
1

1. CALCULATION OF INDIVIDUAL TOXICITY TEST ENDPOINTS

Point estimates were calculated for each toxicity test, including LC20 and LC50 values for
survival endpoints, and IC20 and IC50 values for sublethal response endpoints (e.g., growth,
reproductive success). For each data set, test endpoints were calculated based on comparison to
both negative laboratory controls and reference sediment (Stations Al, A3). The results are
presented in Tables 1 though 3.

LC50 statistics in Tables 1 through 3 are presented using two methods, the probit method and the
Trimmed Spearman-Karber (TSK) method. A comparison of results from these methods
indicates overall similarity (i.e., generally values are within a factor of two). In some cases, the
probit model was considered suspect due to a failure to meet underlying distribution assumptions
of the test. EPA (1994a,b; 2000) indicates that TSK calculations are an appropriate replacement
for the probit method in such cases. Therefore, interpretation of the LC50 values (and
presentation in figures) was based on the TSK method if the probit method was found to be
inappropriate.

Although there are small differences in the toxicity threshold values calculated using different
statistical methods (i.e., choice of extrapolation model or choice of reference sediment), the data
indicate an increase in the frequency and magnitude of adverse biological responses with
increasing sediment tPCB concentration. This finding matches that of the analysis using
"median" sediment PCB exposure data.

Figures 1 through 3 portray the estimated 20% and 50% effects levels, sorted by EC50/LC50
value. The graphical analysis of individual endpoint data yields findings that are generally
similar to those from the "median" analysis:

¦ <3 mg/kg - Some sensitive endpoints exhibited apparent responses, but the
magnitude of responses were not large. These subtle responses were difficult to
evaluate precisely due to statistical power limitations, caused in part by the high
variability in some treatments.

¦ 3 to 10 mg/kg - Numerous endpoints indicated ecologically significant responses,
with many LC20/EC20 andLCso/ECso values falling in this range.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC

-------
1

¦ 10 to 30 mg/kg - Nearly all toxicity endpoints (14/18) indicated large (>50%)
responses relative to reference stations. The only endpoints that did not exhibit large
responses in this concentration range were either growth endpoints or were short-term
(48-hour) tests and/or with tolerant species.

¦ >30 mg/kg - The concentration-response analyses indicated that most survival and
reproduction endpoints exhibited very large reductions at these concentrations, with
complete mortality of some species.

To calculate threshold effects concentrations, the average of values from the six most sensitive
endpoints was calculated for both 50% effects and 20% effects levels. This approach was based
on the rationale that thresholds should be based on consideration of multiple sensitive endpoints,
but should not be based on the single most sensitive endpoint. The 50% effects level corresponds
to major impacts, for which there is a high degree of confidence in a significant biological
impact. The 20% effects levels correspond to lower but potentially biologically significant effect
sizes.

Calculations were performed for comparisons to negative control sediment and also to field
reference sediment. In general, comparisons to field references were preferred for derivation of
sediment MATC values, since field references account for physicochemical factors that may
mediate sediment toxicity.

Summary of 50% and 20% Effects Levels

¦ Comparison to Negative Control - The mean of the lowest six 50% effects levels
was 2.7 mg/kg tPCB. The mean of the lowest six 50% effects levels was 0.1
mg/kg tPCB.

¦ Comparison to Reference A1 - The mean of the lowest six 50% effects levels
was 6.7 mg/kg tPCB. The mean of the lowest six 50% effects levels was 2.5
mg/kg tPCB.

¦ Comparison to Reference A3 - The mean of the lowest six 50% effects levels
was 6.9 mg/kg tPCB. The mean of the lowest six 50% effects levels was 2.5
mg/kg tPCB.

Overall, the endpoint calculations made using "most synoptic" PCB chemistry data were on
average slightly higher than those made using "median" PCB chemistry data. For example, using
"median" exposure data, the average of the lowest six 50% response levels were 3.5 and 3.3
mg/kg for comparisons to A1 and A3, respectively, which are half the values calculated above

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC ^ 7/10/2003

-------
1 for the "most synoptic" exposure data set. Both sets of calculations indicate that significant

2 adverse effects can be expected in the 3-7 mg/kg range. Below 3 mg/kg, subtle impacts may still

3 be possible, especially if the "median" exposure data set provides a better indication of actual

4 average exposures to organisms throughout the toxicity tests.

5 2. GENERAL LINEAR MODEL OF CONCENTRATION-RESPONSE

6 The assessment of individual endpoints is sensitive to test variability, which can mask broader

7 trends in toxicity of PCBs. Therefore, a supplemental approach was applied that combined the

8 toxicity results from various endpoints to identify the overall trend(s) in concentration-response

9 observed. The endpoints for all toxicity tests were standardized so that the response variables

10 were equivalent (i.e., responses represented the proportion of their control mean response). This

11 transformation of all endpoints to the relative performance proportion (RPP) values standardized

12 results from different toxicological endpoints to similar ranges and facilitated the search for a

13 single unified model among all endpoints.

14 The methods used for the linear modeling are identical to those described in Appendix D for the

15 analysis using "median" PCB sediment chemistry.

16 The results of the general linear modeling are depicted in Figure 4 and summarized in Table 4.

17 The "hockey-stick" regression model failed to converge for the 42-d Hyalella growth endpoint;

18 this endpoint is smoothed in Figure 4 using a quadratic curve. Overall, the linear modeling

19 indicated that most toxicity endpoints evaluated were significantly correlated with log-

20 transformed PCB concentration. Differences between acute endpoints and chronic endpoints

21 were observed; these are likely related to the greater sensitivity of chronic endpoints in toxicity

22 tests.

23 The modeling procedure enabled the identification of threshold tPCB concentrations in sediment.

24 These results are in agreement with the summary of individual test endpoints provided in Section

25 2 above. The regression models indicated the following:

26 "1 mg/kg and lower - Generally insignificant effects. Responses are variable, but are

27 within range of RPP values from reference stations.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC

-------
1

¦ 10 mg/kg - Significant adverse effects for multiple endpoints; therefore,
concentration is clearly not protective against significant risk. Close to threshold (i.e.,
inflection point) for acute toxicity for multiple species; large increases in acute
mortality occur between 10 and 100 mg/kg. Chronic test endpoints at 10 mg/kg yield
RPP values that are approximately half those exhibited at reference station PCB
concentrations.

9
10

¦ 1-10 mg/kg - Intermediate effects range. This concentration range brackets the
threshold level for significant toxicity in Housatonic River sediment. The number of
endpoints exhibiting 20% and 50% response levels increases considerably toward the
upper end of this range.

12 3. REFERENCES

13 EPA (U.S. Environmental Protection Agency). 1994a. Short-term methods for estimating the

14 chronic toxicity of effluents and receiving waters to freshwater organisms. Third edition. EPA-

15 600/4-91 -002, Cincinnati, OH.

16 EPA (U.S. Environmental Protection Agency). 1994b. Short-term methods for estimating the

17 chronic toxicity of effluents and receiving waters to marine and estuarine organisms. Second

18 edition. EPA-600/4-91/003, Cincinnati, OH.

19 EPA (U.S. Environmental Protection Agency). 2000. Methods for measuring the toxicity and

20 bioaccumulation of sediment-associated contaminants with freshwater invertebrates. March

21 2000.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC

7/10/2003

-------
Table 1

Summary of Endpoints Calculated for the 48-h and 10-d Acute In Situ Toxicity Tests (Sediment Exposures):

Calculations Made Using "Most Synoptic" Exposure Data Set Only

Species

Endpoint

Compared To

Results (mg/kg tPCB)

NOEL

LOEL

LCS0 by Probit (95%
CL)a

LC2o by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

H. azteca

48-h survival

Control (Laboratory)

139.3°

>139.3°

14.6 (NC)*c

1.4 (NC)* c

9.5 (3.6 - 25.3)c

Reference (Al)

7.3°

139.3°

18.6 (NC)*c

3.0 (NC)* c

8.3 (3.4-20.6)c

Reference (A3)

139.3°

>139.3°

18.1 (NC)*c

3.2 (NC)* c

8.2 (3.4- 19.8)c

H. azteca

10-d survival

Control (Laboratory)

14.0

52.3

3.9 (NC)*

0.2 (NC)*

8.0 (4.1-15.8)

Reference (Al)

14.0

52.3

30.8 (NC)*

20.5 (NC)*

22.4 (19.3 -26.1)

Reference (A3)

14.0

52.3

15.2(15.1-15.2)*

22.9(20.0-26.3)

C. tentans

48-h survival

Control (Laboratory)

521.7

>521.7

>521.7d

>521.7 d

Reference (Al)

521.7

>521.7

>521.7d

>521.7 d

Reference (A3)

521.7

>521.7

>521.7d

>521.7 d

C. tentans

10-d survival

Control (Laboratory)

14.0

52.3

39.5 (NC)*

20.5 (NC)*

28.0 (24.6-31.9)

Reference (Al)

14.0

52.3

39.2 (NC)*

20.3 (NC)*

29.0 (25.7-32.7)

Reference (A3)

14.0

52.3

43.2 (NC)*

23.7 (NC)*

30.7 (27.6-34.3)

D. magna

48-h survival

Control (Laboratory)

7.3

139.3

5.5 (3.2-9.1)

1.5 (0.7-2.6)

4.8(2.6-8.7)

Reference (Al)

<0.9

0.9

5.0 (3.1-8.0)

1.3 (0.6-2.2)

6.4 (3.5-11.6)

Reference (A3)

7.3

139.3

8.1 (3.4-21.4)

3.7(0.0-6.0)

9.4 (6.0-14.8)

L. variegatus

48-h survival

Control (Laboratory)

521.7

>521.7

>521.7d

> 521.7 d

>521.7 d

Reference (Al)

521.7

>521.7

>521.7d

> 521.7 d

>521.7 d

Reference (A3)

521.7

>521.7

>521.7d

> 521.7 d

>521.7 d

aLC50 or LC20 value calculated using Probit method; asterisks (*) indicate where goodness-of-fit statistic exceeded critical value.
bLC50 value calculated using Trimmed Spearman-Karber (TSK) method.

°Highest PCB concentration (521.7 mg/kg) excluded because of anomalous concentration-response.
dNo observed effects >20%. NC = not calculable

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC s 7/10/2003

-------
Table 2

Summary of Endpoints Calculated for the Hyalella azteca Chronic (42-d) Laboratory Toxicity Tests: Calculations Made

Using "Most Synoptic" Exposure Data Set Only

Endpoint

Compared To

Results (mg/kg PCB)

NOEL

LOEL

LCS0/ICS0 by Probit
(95% CL) a

LC20/IC20 by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

28-d survival

Control (Trout Farm)c

0.28

8.7

17.9(0.1-45.2)*

5.0 (0.0-16.5)*

10.5 (7.7-14.4)

Reference (Al)

<8.7

8.7

25.5 (0.0-65.4)*

8.4 (0.0-26.1)*

23.9 (18.5 -30.9)

Reference (A3)

<8.7

8.7

14.5 (0.1-36.7)*

3.6 (0.0- 13.1)*

12.7(9.2-17.5)

35-d survival

Control (Trout Farm)

0.28

8.7

19.1 (NC)

2.7 (NC)

13.8(4.8-40.1)

Reference (Al)

<8.7

8.7

25.2 (7.0-53.7)

3.0 (0.0-9.3)

20.1 (15.0-26.9)

Reference (A3)

<8.7

8.7

15.9(6.0-27.8)

2.0 (0.1-5.5)

14.9 (10.7-20.7)

42-d survival

Control (Trout Farm)

0.28

8.7

21.3 (8.3-37.4)

3.1 (0.1-8.1)

15.6(4.6-53.4)

Reference (Al)

8.7

31.2

27.1 (8.9-57.6)

3.8(0.0-10.5)

20.4 (15.2-27.6)

Reference (A3)

<8.7

8.7

18.4(6.2-33.1)

2.4 (0.1-6.9)

16.2 (11.9-22.2)

28-d dry weight

Control (Trout Farm)

72.0

>72.0

17.1 (NC)

Reference (Al)

72.0

>72.0

28.5 (NC)

Reference (A3)

72.0

>72.0

16.2 (NC)

42-d dry weight

Control (Trout Farm)c

72.0

>72.0

53.0 (NC)

Reference (Al)

72.0

>72.0

69.7 (NC)

Reference (A3)

72.0

>72.0

62.9 (NC)

35-d no. young

Control (Trout Farm)c

0.28

8.7

4.4 (0.0 - 16.0)

<0.028

Reference (Al)

8.7

31.2

17.9(5.9-24.7)

7.3 (2.4-14.8)

Reference (A3)

<8.7

8.7

7.9 (5.4-19.3)

3.3 (2.3 - 10.1)

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC q 7/10/2003

-------
Table 2

Summary of Endpoints Calculated for the Hyalella azteca Chronic (42-d) Laboratory Toxicity Tests

(Continued)

Endpoint

Compared To

Results (mg/kg PCB)

NOEL

LOEL

LCS0/ICS0 by Probit
(95% CL) a

LC20/IC20 by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

42-d total young

Control (Trout Farm)

0.28

8.7

1.8(0.0-11.1)

<0.028

Reference (Al)

8.7

31.2

14.4(5.6-27.7)

4.5 (2.3 - 15.7)

Reference (A3)

<8.7

8.7

7.8 (5.6-19.3)

3.3 (2.4-10.7)

42-d young/female

Control (Trout Farm)

0.28

8.7

7.0 (0.0-23.6)

<0.028

Reference (Al)

31.2

24.0(1.7-51.3)

9.5 (0.7-31.3)

Reference (A3)

8.7

31.2

19.9(5.0-38.7)

5.8(2.8-17.4)

aLC50 or LC2o value calculated using Probit method; asterisks (*) indicate where goodness-of-fit statistic exceeded critical value.
l:,LC\u value calculated using Trimmed Spearman-Karber (TSK) method.

Interrupted dose-response observed.

NA = not applicable; NC = not calculated

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC

-------
Table 3

Summary of Endpoints Calculated for the Chironomus tentans Chronic (43-d) Laboratory Toxicity Tests: Calculations

Made Using "Most Synoptic" Exposure Data Set Only

Endpoint

Compared To

Results (mg/kg PCB)

NOEL

LOEL

LCSo/ICSo by Probit
(95% CL) a

LC20/IC20 by Probit
(95% CL) a

LCS0 by TSK
(95% CL) b

20-d survival

Control (Trout Farm)e

0.28

8.7

2.0 (NC)*

0.5 (NC)*

1.0(0.7-1.6)

Reference (Al)

<8.7

8.7

<8.7°

Reference (A3)

<8.7

8.7

<8.7°

20-d dry weight

Control (Trout Farm)

0.28

72.0d

3.9(2.0-5.1)

0.9(0.0-2.9)

Reference (Al)

<72.0d

72.0d

4.5 (4.4-4.7)

1.8(1.8-1.9)

Reference (A3)

<72.0d

72.0d

4.6 (4.5-4.7)

2.0 (2.0-2.1)

20-d ash-free dry
weight (AFDW)

Control (Trout Farm)

0.28d'e

8.7d'e

4.5 (2.9-4.8)

1.8(0.0-2.3)

Reference (Al)

NCd

4.7 (4.5-5.6)

1.9(1.8-2.3)

Reference (A3)

NCd

4.6 (4.6-4.7)

2.0 (2.0-2.0)

43-d emergence

Control (Trout Farm)

0.28

8.7

0.08 (0.01-0.26)

<0.028

Reference (Al)

<8.7

8.7

<8.7°

Reference (A3)

<8.7

8.7

<8.7°

aLC50 or LC2o value calculated using Probit method; asterisks (*) indicate where goodness-of-fit statistic exceeded critical value.

I:,LC\(J value calculated using Trimmed Spearman-Karber (TSK) method.

°Based on visual observations (model did not converge).

treatments not included due to only a single replicate being available.

"Interrupted dose-response observed.

NA = not applicable; NC = not calculated

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC

-------
Table 4

Summary of Segmented Linear Regression Concentration Models Applied to Toxicity Data, Relating Relative
Performance Proportion (RPP) to Sediment tPCB Concentrations: Calculations Made Using "Most Synoptic"

Exposure Data Set Only

Endpoint Type

Model Statistics

Inflection Point (mg/kg)

Flat Intercept

Regression Intercept

Regression Slope

Statistically Significant
(oc=0.05)?

Acute Endpoints

9.3

0.84

1.5

-0.65

Yes

42-d Hyalella
Survival

1.8

0.93

1.0

-0.43

Yes

42-d Hyalella
Growth

The hockey stick model failed to converge. No model specified.

Remaining Chronic
Endpoints

0.28

0.58

0.46

-0.22

Yes

Acute endpoints = 10-day Hyalella survival, 10-day Chironomus survival, 48-hour Daphnia survival.

Other chronic endpoints = 20-day Chironomus survival, 43-day Chironomus emergence, 42-day Hyalella survival, and 42-day
Hyalella young per female.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC -> rx 7/10/2003

-------
1

1000

100

~ LC50/IC50 Calculated Value

I LC20/IC20 Calculated Value
(lighter shade indicates poor
fitting probit model)

>521,7>521.7

0.1

0.01

y>vyy>vvy

^ A® ^

_ >v> a, _ >> _v* N ^ ' b \

& if ^ ^ „df .n;6 J* Jf

i> ^ ^ -o

r\? ¦? 6 6 t?> (?>'

f of ^ J*

O- v

*•

*$¦' ^ ^ *£* *£* -if *£* y >y >y >y >y ^

^ ^ ^ ^ ^ O- ir
-------
1

1000

100

~ LC50/IC50 Calculated Value

ILC20/IC20 Calculated Value
(lighter shade indicates poor
fitting probit model)

>521.7 >521./

jt / / / / / j> /¦ / / y / / / /¦ /• / /•

/ ^ V y V /V S y J? y „/ y./

f ^ ^ / J ^ ^ y J? $

A® <& <& n-6 ^

r y v •&* *$>' vX- & $ /p- v v ^ xe> J? ^

^ ^ O '^* o* ^ ^ ^ O* ^ ^ o * ^

o J> jf ^ if ^ JP ^ ^ / 4? . & Jf j? Jf j? J? J> J? J?

O *'

-------
1

1000

100

0.01

/ y / / y y / /

^ f ^ ^ ^ if if ^ ^ ^

66 J> n>v° ^ ^ ^ ff ^ ^ ^ ^ ^ ^

^ ^ & A® ef O® <& # <& O® <& n- <& <& ^ & J? *?

j? J? j3> <& ^ -i> <^ o% -£* -i> -£* ^ -£* ^ „<3> „<3> ^ <#

^ u
-------
All Acute Endpoints

+ 0

+ +g
—±?-
+o

O A

0.0001

0.01 0.1

100 1000

Total PCB (mg/kg)

Remaining Chronic Endpoints

O
V

o i

0.0001 0.01 0.1 1 10

Total PCB (mg/kg)

100 1000

42 day hyalella survival

X
X

—X-
X

^ X
\ X

\ X X
X\ X
X \x
x xx

X \
X

0.0001

0.01 0.1 1
Total PCB

100 1000

42 day hyalella growth

¦J"~

i
i

¦' *

• l\

0.0001

0.01 0.1 1
Total PCB

100 1000

Figure 4 Segmented Linear Regression Models Applied to Toxicity Data,
Relating Relative Performance Proportion (RPP) to Bulk Sediment tPCB
Concentrations (mg/kg): Calculations Made Using "Most Synoptic"

Exposure Data Set Only

MK0110:\20123001.096\ERA_PB\ERA_ATD5_PB.DOC

-------
ATTACHMENT D.6

RAW DATA FOR MULTIVARIATE ASSESSMENT OF BENTHIC

COMMUNITY METRICS

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC

-------
1

ATTACHMENT D.6
RAW DATA FOR MULTIVARIATE ASSESSMENT OF BENTHIC

COMMUNITY METRICS

Based on the rationale provided in Attachment D.3, the following six benthic community metrics
were included in multivariate statistical analyses:

¦ Organism abundance (number of animals per replicate or station).

¦ Taxonomic richness (number of unique taxa per replicate or station).

¦ "EPT" relative abundance (mayflies, caddisflies, stoneflies).

¦ Relative abundance of tolerant dipterans.

¦ Relative abundance of tolerant oligochaetes.

¦ Relative abundance of tolerant gastropods.

These metrics were used to generate summary graphics including relative rank plots and
multidimensional scaling (MDS) plots. This attachment contains the raw data used the generate
these graphics, plus summaries of the ordination data used in the MDS plots:

¦ Table 1 - Raw Scores for Benthic Community Variables.

¦ Table 2 - Summary of MDS Scores, for Analysis Using Median Station Data.

¦ Table 3 - Summary of MDS Scores, for Analysis Using Individual Replicate Data
(Fine-Grained Stations).

¦ Table 4 - Summary of MDS Scores, for Analysis Using Individual Replicate Data
(Coarse-Grained Stations).

For plotting MDS scores, a scalar of -1 was sometimes applied. This reversal of sign was
performed to reverse the plot axes for consistent presentation, and did not affect the distances
depicted on the plots.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC

-------
Table 1

Raw Scores for Benthic Community Variables

Station

Replicate

Grain Size
Category

Total
Abundance

EPT

Taxonomic
Richness

Diptera
Tolerant Taxaa

Gastropoda
Tolerant Taxa"

Oligochaeta
Tolerant Taxa"

Coarse-Grained

0.024

0.190

0.000

Coarse-Grained

105

0.095

0.133

0.010

0.067

Coarse-Grained

183

0.055

0.066

0.000

0.060

Coarse-Grained

352

0.131

0.156

0.000

0.091

Coarse-Grained

0.083

0.000

0.083

Coarse-Grained

0.000

0.231

0.000

Coarse-Grained

0.000

0.100

0.000

0.050

Coarse-Grained

0.000

0.460

0.000

0.160

Coarse-Grained

111

0.081

0.243

0.000

0.045

Coarse-Grained

0.000

0.217

0.000

0.043

Coarse-Grained

0.000

0.417

0.000

0.167

Coarse-Grained

0.000

0.450

0.000

145

Coarse-Grained

0.027

0.541

0.027

0.000

146

Coarse-Grained

0.023

0.535

0.000

147

Coarse-Grained

0.000

0.447

0.000

148

Coarse-Grained

0.000

0.577

0.000

149

Coarse-Grained

0.056

0.333

0.000

150

Coarse-Grained

0.077

0.487

0.000

151

Coarse-Grained

0.000

0.857

0.000

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC rs 7/10/2003

-------
Table 1

Raw Scores for Benthic Community Variables
(Continued)

Station

Replicate

Grain Size
Category

Total
Abundance

EPT

Taxonomic
Richness

Diptera
Tolerant Taxaa

Gastropoda
Tolerant Taxa"

Oligochaeta
Tolerant Taxa"

152

Coarse-Grained

0.016

0.203

0.016

0.031

153

Coarse-Grained

0.000

0.222

0.000

154

Coarse-Grained

0.014

0.171

0.000

155

Coarse-Grained

0.000

0.146

0.000

156

Coarse-Grained

0.025

0.225

0.025

0.000

133

Coarse-Grained

0.000

0.111

0.022

134

Coarse-Grained

0.000

0.157

0.039

0.098

135

Coarse-Grained

0.000

0.242

0.000

0.061

136

Coarse-Grained

0.000

0.105

0.000

0.105

137

Coarse-Grained

0.000

0.167

0.000

0.125

138

Coarse-Grained

138

0.007

0.196

0.000

0.014

139

Coarse-Grained

118

0.034

0.136

0.000

0.068

140

Coarse-Grained

0.000

0.167

0.000

0.063

141

Coarse-Grained

101

0.000

0.089

0.000

0.079

142

Coarse-Grained

0.000

0.069

0.000

0.069

143

Coarse-Grained

0.000

0.113

0.000

0.094

144

Coarse-Grained

0.000

0.158

0.000

0.053

Coarse-Grained

0.000

0.023

0.000

0.116

Coarse-Grained

0.000

0.222

0.000

0.111

Coarse-Grained

0.000

0.098

0.000

0.049

Coarse-Grained

0.000

0.067

0.000

0.067

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC ^ 7/10/2003

-------
Table 1

Raw Scores for Benthic Community Variables
(Continued)

Grain Size

Total

Taxonomic

Diptera

Gastropoda

Oligochaeta

Station

Replicate

Category

Abundance

EPT

Richness

Tolerant Taxaa

Tolerant Taxa"

Coarse-Grained

0.000

0.056

0.222

Coarse-Grained

0.000

0.138

0.000

0.034

Coarse-Grained

0.000

0.032

0.000

0.161

Coarse-Grained

0.000

0.056

0.000

0.056

Coarse-Grained

0.000

0.235

0.000

Coarse-Grained

0.000

0.042

0.000

0.042

Coarse-Grained

0.000

0.063

0.000

Coarse-Grained

0.000

Coarse-Grained

0.000

0.143

0.000

0.143

Coarse-Grained

0.061

0.000

0.091

Coarse-Grained

0.077

0.000

0.154

Coarse-Grained

0.036

0.000

0.036

Coarse-Grained

0.000

0.053

0.000

0.053

Coarse-Grained

0.000

0.033

0.000

0.067

Coarse-Grained

0.000

0.063

0.000

0.313

Coarse-Grained

0.000

Coarse-Grained

0.000

Coarse-Grained

0.000

0.200

0.000

Coarse-Grained

0.000

0.048

0.000

0.095

Coarse-Grained

0.000

0.108

Coarse-Grained

0.000

0.077

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC a 7/10/2003

-------
Table 1

Raw Scores for Benthic Community Variables
(Continued)

Grain Size

Total

Taxonomic

Diptera

Gastropoda

Oligochaeta

Station

Replicate

Category

Abundance

EPT

Richness

Tolerant Taxaa

Tolerant Taxa"

Coarse-Grained

0.125

0.000

0.250

Coarse-Grained

0.000

0.118

0.000

0.353

100

Coarse-Grained

0.000

0.077

0.000

0.154

101

Coarse-Grained

0.000

102

Coarse-Grained

0.000

0.333

103

Coarse-Grained

0.000

0.158

0.000

0.105

104

Coarse-Grained

0.000

0.083

0.000

105

Coarse-Grained

0.000

0.364

106

Coarse-Grained

0.067

0.200

0.000

0.133

107

Coarse-Grained

0.000

0.200

108

Coarse-Grained

0.000

1.000

0.000

Coarse-Grained

0.000

0.250

0.000

0.250

Coarse-Grained

0.000

0.429

0.000

0.071

Coarse-Grained

0.000

0.045

0.000

0.273

Coarse-Grained

0.000

0.100

Coarse-Grained

0.000

0.067

0.000

0.333

Coarse-Grained

0.000

0.095

0.000

0.429

Coarse-Grained

0.000

Coarse-Grained

0.000

0.083

0.000

0.083

Coarse-Grained

0.000

0.048

0.000

0.095

Coarse-Grained

0.000

0.095

0.000

0.095

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC r 7/10/2003

-------
Table 1

Raw Scores for Benthic Community Variables
(Continued)

Grain Size

Total

Taxonomic

Diptera

Gastropoda

Oligochaeta

Station

Replicate

Category

Abundance

EPT

Richness

Tolerant Taxaa

Tolerant Taxa"

Coarse-Grained

0.000

0.100

0.000

0.100

Coarse-Grained

0.000

0.077

0.000

0.154

Coarse-Grained

0.000

0.500

Coarse-Grained

0.000

1.000

0.000

Coarse-Grained

0.000

0.125

0.000

0.875

Coarse-Grained

0.354

0.114

0.000

0.152

Coarse-Grained

0.000

1.000

Coarse-Grained

0.071

0.000

0.786

Coarse-Grained

0.000

1.000

Coarse-Grained

0.000

0.250

0.000

0.250

Coarse-Grained

0.000

0.333

Coarse-Grained

0.000

0.500

Coarse-Grained

0.000

1.000

0.000

Fine-Grained

283

0.014

0.300

0.000

0.155

Fine-Grained

305

0.007

0.236

0.000

0.223

Fine-Grained

148

0.007

0.358

0.000

0.311

Fine-Grained

180

0.000

0.300

0.000

0.222

Fine-Grained

271

0.004

0.303

0.000

0.125

Fine-Grained

243

0.000

0.263

0.000

0.169

Fine-Grained

179

0.039

0.274

0.000

0.140

Fine-Grained

225

0.004

0.204

0.004

0.222

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC s 7/10/2003

-------
Table 1

Raw Scores for Benthic Community Variables
(Continued)

Grain Size

Total

Taxonomic

Diptera

Gastropoda

Oligochaeta

Station

Replicate

Category

Abundance

EPT

Richness

Tolerant Taxaa

Tolerant Taxa"

Fine-Grained

245

0.025

0.118

0.000

0.486

Fine-Grained

142

0.000

0.359

0.000

0.373

Fine-Grained

145

0.014

0.241

0.000

0.483

Fine-Grained

342

0.003

0.243

0.000

0.199

Fine-Grained

106

0.019

0.085

0.000

0.245

Fine-Grained

0.029

0.059

0.000

0.176

Fine-Grained

0.031

0.016

0.344

Fine-Grained

0.059

0.024

0.012

0.200

Fine-Grained

101

0.030

0.079

0.000

0.188

Fine-Grained

107

0.084

0.056

0.009

0.224

Fine-Grained

0.024

0.012

0.024

0.120

Fine-Grained

0.044

0.022

0.065

Fine-Grained

0.072

0.024

0.036

0.048

Fine-Grained

127

0.063

0.016

0.047

0.024

Fine-Grained

0.060

0.048

0.024

0.036

Fine-Grained

136

0.044

0.029

0.015

0.096

Fine-Grained

0.000

0.148

0.056

0.185

Fine-Grained

194

0.000

0.046

0.227

0.129

Fine-Grained

187

0.000

0.037

0.150

0.011

Fine-Grained

470

0.009

0.043

0.213

0.066

Fine-Grained

185

0.005

0.049

0.141

0.065

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC q 7/10/2003

-------
Table 1

Raw Scores for Benthic Community Variables
(Continued)

Station

Replicate

Grain Size
Category

Total
Abundance

EPT

Taxonomic
Richness

Diptera
Tolerant Taxaa

Gastropoda
Tolerant Taxa"

Oligochaeta
Tolerant Taxa"

Fine-Grained

364

0.006

0.047

0.242

0.077

Fine-Grained

203

0.015

0.069

0.222

0.039

Fine-Grained

0.000

0.089

0.071

Fine-Grained

0.026

0.211

0.026

0.053

Fine-Grained

0.032

0.290

0.000

0.129

Fine-Grained

0.015

0.162

0.015

Fine-Grained

0.000

0.024

0.146

0.012

109

Fine-Grained

222

0.005

0.153

0.045

0.180

110

Fine-Grained

161

0.000

0.075

0.056

0.286

111

Fine-Grained

172

0.000

0.047

0.081

0.151

112

Fine-Grained

192

0.005

0.026

0.182

0.130

113

Fine-Grained

318

0.003

0.050

0.072

0.110

114

Fine-Grained

287

0.004

0.077

0.070

0.063

115

Fine-Grained

323

0.003

0.031

0.068

0.056

116

Fine-Grained

224

0.000

0.063

0.161

0.121

117

Fine-Grained

153

0.000

0.052

0.170

0.124

118

Fine-Grained

207

0.005

0.053

0.174

0.039

119

Fine-Grained

368

0.005

0.060

0.147

0.046

120

Fine-Grained

145

0.000

0.166

0.076

0.103

121

Fine-Grained

100

0.060

0.260

0.020

0.400

122

Fine-Grained

0.016

0.484

0.000

0.172

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC q 7/10/2003

-------
Table 1

Raw Scores for Benthic Community Variables
(Continued)

Station

Replicate

Grain Size
Category

Total
Abundance

EPT

Taxonomic
Richness

Diptera
Tolerant Taxaa

Gastropoda
Tolerant Taxa"

Oligochaeta
Tolerant Taxa"

123

Fine-Grained

0.000

0.130

0.241

0.204

124

Fine-Grained

120

0.050

0.233

0.125

125

Fine-Grained

245

0.016

0.049

0.139

0.249

126

Fine-Grained

244

0.086

0.057

0.135

0.049

127

Fine-Grained

0.018

0.491

0.036

0.127

128

Fine-Grained

0.010

0.232

0.000

0.556

129

Fine-Grained

0.015

0.215

0.015

0.385

130

Fine-Grained

167

0.066

0.186

0.054

0.311

131

Fine-Grained

0.106

0.277

0.128

0.255

132

Fine-Grained

244

0.074

0.168

0.078

0.410

a Relative abundance.

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC q 7/10/2003

-------
Table 2

Summary of MDS Scores, for Analysis Using Median Station Data

Station

Grain Size Category

MDS1

MDS2

Coarse-Grained

1.430

-0.900

Coarse-Grained

1.720

-0.532

Coarse-Grained

1.954

0.014

Coarse-Grained

1.899

-0.569

Coarse-Grained

3.301

3.416

Fine-Grained

-2.555

0.162

Fine-Grained

-0.578

0.778

Fine-Grained

-1.992

1.000

Fine-Grained

-3.298

0.914

Coarse-Grained

0.191

-1.500

Coarse-Grained

-0.259

-2.890

Coarse-Grained

0.141

-1.145

Fine-Grained

-1.958

1.253

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> rx 7/10/2003

-------
Table 3

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Fine-Grained Stations)

Station

Replicate

MDS1

MDS2

-0.39

2.04

-0.49

2.19

1.37

2.35

0.72

1.71

-0.73

2.13

-0.15

1.7

0.36

1.21

-0.2

1.75

0.59

2.69

1.63

2.65

1.97

2.41

-1.1

2.54

1.04

0.15

2.03

-1.45

2.27

-0.78

1.87

-1.54

1.15

-0.47

2.29

-1.46

0.98

-1.64

1.41

-2.34

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> -> 7/10/2003

-------
Table 3

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Fine-Grained Stations)

(Continued)

Station

Replicate

MDS1

MDS2

1.11

-2.37

0.34

-2.12

0.73

-1.8

0.7

-1.41

1.08

-0.51

-1.47

-1.52

-1.78

-1.42

-4.33

0.35

-1.55

-0.97

-4.13

0.13

-1.72

-1.82

0.68

-1.74

1.59

-1.28

2.64

-0.71

-0.21

-2.96

-0.39

-2.61

109

-1.22

1.53

110

-0.31

0.96

111

-1.29

0.22

112

-2.23

-0.4

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> rs 7/10/2003

-------
Table 3

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Fine-Grained Stations)

(Continued)

Station

Replicate

MDS1

MDS2

113

-2.53

0.93

114

-2.33

0.61

115

-2.17

116

-1.84

-0.41

117

-1.37

-0.88

118

-2.94

-0.23

119

-4.31

1.22

120

-0.92

0.45

121

2.54

1.21

122

2.27

1.39

123

0.03

-1.6

124

-0.87

-1.97

125

-1.31

0.33

126

-1.18

-1.31

127

2.21

0.71

128

2.86

2.19

129

2.6

0.73

130

1.25

0.62

131

2.49

-0.94

132

0.64

1.54

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> ^ 7/10/2003

-------
Table 4

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Coarse-Grained Stations)

Station

Replicate

MDS1

MDS2

-1.47

0.71

-6.03

-0.46

-6.78

2.1

-13.73

4.05

-0.69

1.02

0.84

0.18

0.2

0.31

-0.12

0.43

-4.36

1.72

0.55

0.26

1.24

0.16

0.8

0.16

145

-2.21

-6.27

146

-0.6

0.56

147

0.47

0.27

148

0.57

0.18

149

-0.2

0.7

150

-1.19

1.01

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC

-------
Table 4

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Coarse-Grained Stations)

(Continued)

Station

Replicate

MDS1

MDS2

151

1.43

-0.01

152

-2.84

-3.16

153

0.06

0.35

154

-1.15

0.73

155

-0.45

0.46

156

-2.26

-5.68

133

-2.55

-5.07

134

-3.71

-9.31

135

0.12

0.35

136

-0.06

0.43

137

0.43

0.32

138

-4.01

1.24

139

-3.84

1.37

140

-0.57

0.49

141

-2.27

0.9

142

-0.21

0.4

143

-0.75

0.56

144

-1.52

0.78

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> r 7/10/2003

-------
Table 4

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Coarse-Grained Stations)

(Continued)

Station

Replicate

MDS1

MDS2

-0.07

0.47

1.3

0.16

0.1

0.41

1.17

0.21

-2.67

-13.7

0.48

0.31

0.5

0.38

0.87

0.25

1.47

0.13

0.39

0.32

1.16

0.19

0.74

0.28

1.6

0.15

-0.79

0.96

-1.35

1.19

-0.07

0.65

0.95

0.25

0.07

0.39

1.51

0.27

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> s 7/10/2003

-------
Table 4

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Coarse-Grained Stations)

(Continued)

Station

Replicate

MDS1

MDS2

1.1

0.22

1.12

0.2

1.32

0.12

0.96

0.27

0.58

0.39

1.33

0.2

-0.72

1.4

1.34

0.3

100

1.25

0.22

101

1.65

0.1

102

2.16

0.17

103

0.83

0.25

104

1.47

0.14

105

1.59

0.28

106

-0.37

0.87

107

0.94

0.3

108

2.08

-0.16

1.95

0.12

1.47

0.1

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> q 7/10/2003

-------
Table 4

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Coarse-Grained Stations)

(Continued)

Station

Replicate

MDS1

MDS2

0.99

0.34

1.19

0.25

1.67

0.26

1.35

0.35

1.54

0.12

1.48

0.17

0.96

0.27

0.86

0.27

0.9

0.26

1.46

0.2

2.53

0.2

2.08

-0.16

2.95

0.29

-9.05

4.2

3.39

0.29

1.12

1.04

3.36

0.3

1.95

0.12

2.16

0.17

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> q 7/10/2003

-------
Table 4

Summary of MDS Scores, for Analysis Using Individual Replicate Data (Coarse-Grained Stations)

(Continued)

Station

Replicate

MDS1

MDS2

2.53

0.2

2.08

-0.16

MK01 |O:\20123001.096\ERA_PB\ERA_ATD6_PB.DOC -> q 7/10/2003

-------