w}, W _..__——— ___, v —~—————~ ".wwwwwmm .. , ,
A: - , , . _ . 7 7'
‘ . . , ‘ _. .

 

   

 

 

 

 

 

   

 

”i. ‘..
m ”A. '
,

..

B.‘

 

 

2&3

         
           
     

 

 

 

            
   

 

       

.A -. . "laud

llllllllllllllllllllllllllllllllIlllllHllllllllllllllll urge/3.2

3 1293 00788 1083

4 ‘ , LIBRARY
Mimi-"8 EMIChigan stage
.5 University

 

 

 

1 This is to certify that the

thesis entitled

CORRELATIONS 0F INDIVIDUAL POLYCHLORINATED
BIPHENYL CONGENERS IN EGGS OF LAKE MICHIGAN CHINOOK
SALMON' (Oncorlynchus tshafltscha) WITH REARING MORTALITY

 

presented by

Lisa Lynn Williams

has been accepted towards fulfillment
of the requirements for

 

 

Master of Science degree in Fisheries and Wildlife

    

Major professor

Date July 20, 1989

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

< " M‘h‘

_‘~_‘-~‘“"

PLACE IN RETURN BOX to remove this} eheckout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU I: An Affirmative Action/Equal Opportunity Institution
«Mullins

CORRELATIONS OF INDIVIDUAL POLYCHLORINATED BIPHENYL
CONGENERS IN EGGS OF LAKE MICHIGAN CHINOOK SALMON

(QEQQRHXNQHHS IEHAHXIEQHA) WITH REARING MORTALITY

BY

Lisa Lynn Williams

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Fisheries and Wildlife

1989

ABSTRACT

CORRELATIONS OF INDIVIDUAL POLYCHLORINATED BIPHENYL
CONGENERS IN EGGS OF LAKE MICHIGAN CHINOOK SALMON

(W W) WITH REARING MORTALITY

BY
Lisa Lynn Williams

Chlorinated hydrocarbons, including polychlorinated
biphenyls (PCBs), have been associated with poor
reproduction in Great Lakes salmonids. The purpose of this
study was to examine relationships between mortality of
chinook salmon fry and concentrations of PCB congeners in
the eggs from their respective clutches. Eggs from twenty
Lake Michigan chinook salmon were divided into groups for
rearing in the laboratory and for PCB analysis by GC-ECD.
PCB extracts were fractionated on a carbon/glass fiber
column to allow detection of fifteen coplanar congeners,
including IUPAC #77 and #126, among the 94 congeners
detected. The mode of action of the coplanar congeners is
the same as 2,3,7,8-TCDD, the model compound for this type
of toxicity. Mortality of partially hatched fry was
correlated significantly with total TCDD toxic equivalents
(r2-0.51, ps0.02) but not with total PCB concentration.
Cumulative mortality from fertilization to three weeks after
swim-up was not correlated significantly with either toxic

equivalents or PCB concentrations in the eggs.

Copyright by
LISA LYNN WILLIAMS

1989

ACKNOWLEDGMENTS

I would like to first thank my committee for all of
their support. Dr. Kevern encouraged me with his good humor
and enthusiasm. Dr. Zabik provided guidance, advice and
support and taught me how to approach broken equipment. Dr.
Giesy taught me about being a researcher and a scientist.
We have had many interesting and challenging discussions for
which I am grateful.

This research would not have been possible without the
help and support of the graduate students and technicians at
the Pesticide Research Center. I can only mention a few
here. In particular, I owe a great deal to John and Pam
Newsted for their help in raising the fry. Sue Erhardt-
Zabik and Mary Ann Heindorf helped me countless times in the
laboratory and were good friends during those times. The
many conversations with Don Tillitt helped. me keep my
perspective. Dave Verbrugge struggled with the GC ' s with
me, and Nicole DeGalan helped tie up some of the loose ends

at the end of the project.

iv

I owe a great deal to Pat Walz and my friends at
Movement Arts for their energy, enthusiasm, support and love
and to my parents for their love and patience. I owe a
special thanks to Pat for her assistance with making the
floating baskets and proofreading data sets.

.A portion of this work was completed at the 0.8. Fish
and Wildlife Service's Columbia National Fisheries
Contaminants Research Center. Ted Schwartz provided me with
the opportunity to work with his group and was very generous
with his time, equipment, supplies and perSonnel. I am very
grateful for the help that he and Kevin Feltz provided.

The Fisheries Division of the Michigan Department of
Natural Resources assisted in the collection of the chinook
salmon eggs at their weir. Approval was granted by Harry
Westers and Ludwig Frankenberger and assistance was provided
by Jan Sapak and Jim Allen.

This material is based upon work supported under a
National Science Foundation Graduate Fellowship. Any
opinions, findings, conclusions or recommendations expressed
in this publication are those of the author and do not
necessarily reflect the views of the National Science
Foundation. Additional support for this research was
provided by the Michigan Sea Grant College Program as
project RTS-21 under grant #NAOBOAA-D-O90072 from the
National Sea Grant College Program, National Oceanic and

Atmospheric Administration, U.S. Department, of Commerce.

Methods development was completed in conjunction with work
performed under grant #59-3204-6-61, 0RD No. 40400, from the

0.8. Department of Agriculture.

vi

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . . .
LIST OF FIGURES. . . . . . . . .
Imam“: ON C O O O O O O O O O

Rationale . . . . . . .
Test Organism. . . .
Test Chemicals . . .
Exposure Method. . .
PCB Analysis Methods
Research Objectives . . . .
Experimental Design . . . .

MODS O O O O O O O O O O O O 0

Sampling and Fertilization.
Rearing . . . . . . .

Egg Diameter Determination.
PCB Analysis. . . . . . .

Extraction and Clean-up.

Carbon Column Chromatogra

Quantitation . . .
Lipid Analysis. . . .
Calculation of TCDD-EQs
Statistical Analysis. .

RESULTS 0 I O O I O O O O O O O 0

Female Fish Dimensions and Egg Characteristics

Rearing Mortality . . . . .
Concentrations of PCBs. . .
2, 3, 7, 8-TCDD Equivalents. .
Correlations. . . .
Prediction of Rearing Morta

DISCUSSION . . . . . . . . . . .

Correlations. . .

0000.00.00.

lities

Relationship to Other Studies .

coucws IONS O O O O O O O O O 0 0

vii

hY

ix

xi

13

13
14
16
16
16
20
25
28
3O
30

31

31
31
34
37
37
43

46

46
47

SO

APPENDIX A . . . . .
APPENDIX B . . . . .

LIST OF REFERENCES .

viii

Table

Table

Table

Table

Table

Table

Table

LIST OF TABLES

TCDD-EQ Conversion Factors (KEQs)
Calculated from EC-SOs Reported by Safe
(1987) O O O O O O O O O O O O I O O O O O O O O 8

Recoveries of PCB Congeners in Fractions
Collected from Carbon Column
Chromatography . . . . . . . . . ...... . . . .24

Duplicate Analysis of Chinook Salmon
Eggs Following Carbon Column
Chromatography. Concentrations are in
nmol/g, wet weight of eggs. . . . . . . . . . . 29

variability in Chinook Salmon Egg Dimen-
sions and Lipid Content of Eggs Which
Was Explained by Among Clutch
Variability Rather Than Analytical
Error... . . . . . . . . . . . . . . . . . . . .32

Pearson Pairwise Correlation
Coefficients (r) for Correlations Among
Fish Dimensions and Egg Characteristics
for Twenty Female Chinook Salmon.
Probabilities of Greater Correlation
Coefficients Given that the Null
Hypotheses Is True Are in Parentheses. . . . . .33

Proportion of Variability in Rearing
Mortality Among Egg Rearing Groups Which
Was Explained by Among Egg Clutch
Variability. . . . . . . . . . . . . . . . . . .36

Mean Concentrations -of Coplanar PCB
Congeners Quantitated After Carbon
Column Chromatography Enrichment of
Primary PCB Extract of Chinook Salmon
Eggs. Concentrations are in ng/g, wet
weight. . . . . . . . . . . . . . . . . . . . . 38

ix

Table 8.

Table 9.

Table 10.

Table 11.

Concentrations and TCDD-EQs of Coplanar
PCB Congeners Measured in Lake Michigan
Chinook Salmon Eggs. KEQs Are Based on
Molar Ratios of Aryl Hydrocarbon
Mydroxylase Induction Potency in Rat
Liver Cells (Sawyer and Safe, 1982). . . . . . .39

Chlorine Substitution Pattern and
Percent Contribution to Total TCDD-EQ
for Coplanar PCB Congeners Measured in
Lake Michigan Chinook Salmon Eggs . . . . . . . 39

Pearson Pairwise Correlation
Coefficients (r) for Correlations Among
Rearing Mortality and PCB Content
Expressed as Both Concentrations and
TCDD-EQs for Eggs from Nine Chinook
Salmon. Probabilities of Greater
Correlation Coefficients Given that the
Null Hypotheses Is True Are in
Parentheses. . . . . . . . . . . . . . . . . . .41

Mean values for Characteristics of Fish
and Eggs, Rearing Mortality and
Concentrations of Polychlorinated
Biphenyls (PCBs) and 2,3,7,8-TCDD
Equivalents (TCDD—EQs) in the Eggs for
Twenty Individual Chinook; Salmon from
Lake Michigan, 1986. . . . . . . . . . . . . . .53

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

LIST OF FIGURES

Structures of Polychlorinated
Biphenyls, Dibenzo-prdioxins and
Dibenzofurans. . . . . . . . . . . . .

Extraction and Clean-up Method for
Polychlorinated Biphenyls and
Pe‘tiCides O O O O O O O O O O O O O O 0

Carbon Column Chromatography for
Separation of Coplanar Polychlorinated
Biphenyl Congeners from Other
Congeners. . . . . . . . . . . . . . .

Survival of Chinook Salmon from
Fertilization to Swim-up. . . . . . . .

Concentration of Total PCBs Versus
Tatal TCDD-BQ e e e e e e e e e e e e 0

Total Rearing Mortality of Chinook
Salmon Eggs and Fry Versus Total TCDD-
BQB in the Eggs 0 O O O O O O O O O O 0

Mortality of Partially Hatched Chinook
Salmon Fry Versus Total TCDD-EQs in
the Eggs 0 O O O O O O O O O O O O O O 0

Calibration Curve for 3,3',4,4',5-
Pentachlorobiphenyl (IUPAC #126). . . .

xi

.17

22

.35

4O

42

INTRODUCTION

Polychlorinated biphenyls (PCBs) are a class of
compounds that are resistent to chemical, physical, thermal
and biological degradation and have chemical and physical
properties which make them well-suited to a variety of
applications. PCBs were produced commercially in the United
States under the tradename AroclorR. Commercial AroclorsR,
which were chlorinated at levels of 20% to 60% chlorine by
weight, were used as insulators in transformers and
capacitors, as hydraulic fluids, in carbonless copy paper
and in other applications for which highly stable compounds
were desired (DeVoogt and Brinkman 1989, Hutzinger et a1.
1974). Even though production of AroclorsR was stopped in
1972, PCBs are still in use and they continue to enter the
aquatic environment, where the very properties which made
PCBs so useful also make them. persistent environmental
pollutants.

PCBs have been detected in biota, water, sediment and
air samples collected over most of the globe. PCBs
constitute one of the major contaminants in sediments of
industrialized areas like the Great Lakes (Furlong et a1.

1988, Swackhammer and Armstrong 1988). The air over the

2
open ocean and snow and ice from polar regions are
contaminated by PCBs (Bidleman et al. 1981, Risebrough et
al. 1976, Tanabe et al. 1983). PCBs have been detected in
fish samples from pristine areas (Zell and Ballschmitter
1980) and have been measured at parts per million (ug/g)
concentrations in tissues and eggs of Great Lakes salmonids
(Baumann and Whittle 1988, Giesy et al. 1986, DeVault 1986).

The presence of PCBs in fish is of concern because PCBs
have been shown to be toxic to fish. In laboratory studies,
exposure to Aroclors results in detrimental effects to both
early and adult life stages of fish. These effects include
reduced survival , reproduction and growth (Mayer et a1 .
1985, Mauck et al. 1978, Halter and Johnson 1974, Nebeker et

al. 1974). The most sensitive responses to PCB toxicity by
I fish in the laboratory have been found to be reproductive.
PCBs have been found to reduce fecundity, hatchability and
fry survival.

Organic contaminants including PCBs may be impairing
reproduction of salmonids in the Great Lakes. Salmonids in
the Great Lakes have experienced unexplained increases in
rearing mortality in the past decade. Annual rearing
mortalities in lake trout fry as great as 97% were described
for hatchery-reared fish between 1978 and 1981 (Mac 1988).
Mortality in that study could not be attributed to disease
or nutrition and ‘was characterized by erratic swimming
behaviors and loss of equilibrium prior to death. Mortality
of lake trout fry was significantly greater when the eggs

3

and sperm were from southeastern Lake Michigan than when the
gametes were from other parts of the Lake Michigan, but
mortality was not correlated with the concentrations of
contaminants in the eggs or chemistry of the water in which
the eggs were reared (Mac et al. 1985). Rearing mortalities
of hatchery-raised, Lake Michigan chinook salmon in the
early 1980's were greater than those of Pacific Ocean salmon
(J .Hnath, Michigan DNR Fish Division 1981). The chinook
salmon mortalities were not the result of bacterial, fungal,
or viral infections (Flagg 1982). Xenobiotics were
suspected as the cause of the observed mortalities, but
correlations between. mortality and. regularly’ monitored
contaminants only' partially explained the observed poor
survival of salmonid early life stages (Giesy et al. 1986,
Mac 1988).

The poor correlations that have been observed between
reproductive success and concentrations of PCBs, expressed
as AroclorR or total PCB concentrations, in field-exposed
fish may be because these measures of PCB concentration do
not relate well to the toxicity of the mixture of PCB
congeners in these fish and their eggs. Concentrations of
the most toxic PCB congeners were not measured by Giesy et
al. (1986) or Mac (1988). The 209 theoretically possible
PCB congeners vary widely in their toxicity. In a bioassay
of enzyme induction, for example, 3,3',4,4',5-
pentachlorobiphenyl (#126 in the IUPAC numbering system

originated by Ballschmitter and Zell (1980)) is over 10,000

4

times more potent than 2,3',4,4,5,5'-hexachlorobiphenyl
(Safe 1987). The congener pattern changes among
environmental samples (Schwartz et al. 1987) as a result of
different sources and exposures and differential
partitioning, degradation, metabolism and excretion of PCB
congeners in environmental compartments and organisms. The
most toxic PCB congeners are also some of the congeners most
resistent to degradation and metabolism (Tanabe et al. 1987)
and may be selectively enriched relative to other PCB
congeners in organisms in the aquatic environment (Oliver
and Niimi 1988, Niimi and Oliver 1988). Therefore,
concentrations of PCBs expressed. as ‘total PCBs or some
mixture of AroclorRs may not relate well to the toxicity of
the PCB mixture to which an organism is exposed (Kubiak et
al. 1989).

Although the toxicities of PCB congeners vary widely,
the most potent, non-ortho- and mono-ortho-chlorinated
tri-, tetra- and pentachlorobiphenyl congeners all produce
the same pattern of symptoms in exposed animals and seem to
exert their toxic effect through the same mode of action.
The proposed mode of action involves transport by the Ah
cytosolic receptor to the nucleus of an exposed cell where
the PCB molecules interact with DNA to induce pleiotropic
effects (Poland et al. 1979, Poland and Khutson 1982). The
same mode of action and receptor binding is proposed for
compounds which are structurally similar to coplanar PCBs,

such as polychlorinated dibenzo-prdioxin (PCDDs) and

5

dibenzofurans (PCDFs) which elicit the same suite of
symptoms in exposed animals as do the coplanar PCBs (Long et
al. 1987, Poland and Knutson, 1982). Of the compounds from
these three classes of’ planar chlorinated. hydrocarbons,
2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) is the
most toxic (Poland et a1. 1979). PCB congeners that are the
most similar structurally to 2,3,7,8-TCDD are those that are
the most toxic. These congeners are those that have
chlorine atoms substituted in both para positions and at
least one meta position but have no chlorination in the
ortho (2, 2', 6, or 6', Figure 1) positions (Safe et al.
1985). The lack of ortho-chlorine substitution allows the
two rings to rotate around the sigma biphenyl bond and
assume a planar configuration very similar to that of
2,3,7,8-TCDD. The nonortho-PCB congeners include 3,3',4,4'-
tetrachlorobiphenyl (#77), 3,3',4,4',S-pentachlorobiphenyl
(#126) and 3,3',4,4',5,5'-hexachlorobiphenyl (#169).

A method has been developed to calculate toxic
potencies of mixtures of PCBs, PCDFs, and PCDDs based on the
assumption of additive toxicity through a similar mode of
action (Safe 1987, Colburn 1988). Conversion factors (KEQs)
have been calculated for each congener by dividing the
amount of 2,3,7,8-TCDD required to elicit a particular
response by the amount of that congener required to elicit
the same response. Therefore, all toxicities are normalized
to the toxicity of 2,3,7,8-TCDD in this method. The number

of 2,3,7,8-TCDD toxic equivalents (TCDD-EQs) of a given

3' 2' 2 3

 

 

P088 4' O 4 209 congeners
5' 6' a 5
9 1
. 8 . O 2
PCDDS 75 congeners
7 a
o w
e 4
9 1
a 2
PCDFS 135 congeners
7 a
, K0
6 4

Figure 1. Structures of Polychlorinated Biphenyls, Dibenzo-
prdioxins and Dibenzofurans.

7

congener are calculated by multiplying the concentration of
that congener in a sample by that congener's KEQ. In this
model, toxicities are assumed to be additive, and the
toxicity of a mixture of compounds is calculated by summing
the TCDD-EQs for each compound detected. The total TCDD-EQ
calculated for a sample is a measure of the concentration of
2,3,7,8-TCDD that would be required to produce the same
effect as the mixture of compounds.

Several of the effects elicited by this type of planar,
chlorinated hydrocarbons have been used to calculate KEQs
for this toxic equivalency model. Induction of cytochrome
P 450 enzyme activity both in yin and in mm in cultures
of liver cells has been used to compare relative potencies
of PCB, PCDD and PCDF congeners (Safe 1987). The cytochrome
P 450 enzyme activities induced by these compounds that are
most often measured are aryl hydrocarbon hydroxylase (AI-IN)
and ethoxyresorufin-O-deethylase (EROD) (Sawyer and Safe
1985, Melancon and Lech 1983) . Induction of AMI-I has been
highly correlated with thymic atrophy (r - 0.92) and body
weight (r - 0.93) in male Wistar rats (Safe 1987).
Conversion factors based on these latter two effects of
PCDDs, PCDFs and PCBs in whole organisms can also be
calculated for some congeners (Table 1).

This study was designed to assess the relationship of
rearing success of Lake Michigan chinook salmon to PCB

concentrations in the eggs of these fish. PCB analysis

included carbon column chromatography enrichment of the

 

Table 1. TCDD-EQ Conversion Factors (KEQs) Calculated from
EC-SOs Reported by Safe (1987)1
Sanand runner: EBQDKEQ SMMEMEM! FIBER
2,3,4,7,8-P5CDF 0.282 1.38 0.429 0.0481
1,2,3,4,7,8—H5CDF 0.203 0.488 0.180 0.0385
1,2,3,7,8-P5CDF 0.0285 0.0605 0.0511 0.0189
2,3,4,6,7,8-M6CDF 0.105 0.322 0.0968 0.0179
1,2,3,6,7,8-M5CDF 0.0492 0.149 0.0968 0.0156
2,3,7,8-T4CDF 0.0185 0.0916 0.0250 0.0156
1,3,4,7,8-P5CDF 0.0452 0.132 0.129 0.00192
2,3,4,7,9-P5CDF 0.00915 0.0319 0.0164 0.00227
2,3,4,7-T4CDF 0.00404 0.0125 0.0115 0.00147
1,2,3,7,9-P5CDF 0.000841 0.00215 0.00391 0.00101
1,2,4,7,8-P5CDF 0.000682 0.00125 0.00194 0.00101
1,2,3,7-T4CDF 0.00000268 0.00000294 0.000818 0.000575
2,3,4,8-T4CDF 0.00175 0.00492 NAZ 0.000365
2,3,7,8-T4CDD 1.00 1.00 1.00 1.00
1,2,3,7,8-P5CDD 0.00657 0.0109 0.529 0.0806
1,2,3,4,7,8-M6CDD 0.0344 0.0451 0.0841 0.0307
1,3,7,8-T4CDD 0.000123 0.000578 0.000900 0.000379
1,2,4,7,8-P5CDD 0.00344 0.0168 0.00818 0.00147
2,3,4,4',5-P5CB 0.0000743 0.000327 0.000450 0.000278
2,3,3',4,4',5'-H CB 0.00000544 0.00002056 0.000400 0.000227
2,3,3',4,4',5-H CB 0.0000349 0.000206 0.000500 0.000278
2,3,3',4,4'-P5CB 0.000826 0.00154 0.0000874 0.0000667
2,3',4,4',5-P50B 0.00000629 0.0000209 0.0000581 0.0000446
2',3,4,4',5-P5CB 0.0000185 0.000167 0.0000323 0.000135
3,3',4,4',5,5'-H6CB 0.00121 0.00768 0.0101 0.00333
3,3',4,4',5-P5CB 0.301 0.746 0.0947 0.0152

 

1Kros are calculated by dividing the sc-so for 2,3,7,8-TCDD
by the EC-50 for the same effect for each compound. The KEQ
values for aryl hydrocarbon hydroxylase (AME) induction,
ethoxyresorufin-O-deethylase (EROD) induction, thymic atrophy
(THEM) and inhibition of weight gain (WT) are based on molar
concentrations.

2MAI-not available.

9
toxic, coplanar PCBs. PCBs were reported as total PCBs, as
concentrations of individual congeners and as TCDD-EQs.
Pairwise correlations were used to examine the relationships
between rearing mortality and measures of PCB
concentrations. The predictive ability of the TCDD~EQ model

was evaluated relative to that of traditional PCB analysis.

Rationale
humanism
The chinook salmon (W: W) was

chosen as the test organism because it is a large fish known
to accumulate organic contaminants and because poor
reproductive success not associated with disease has been
observed for individuals of this species collected from the
Great Lakes in the past (e.g. Mac 1988, Giesy et al. 1986).
This species also comprised a large proportion of the Lake
Michigan sport catch and is supported by an extensive
hatchery program. Furthermore, information on
concentrations of PCBs in eggs of this species are available
from a 1982 study (Giesy et al. 1986).

W

PCBs are known to occur in eggs of Lake Michigan
chinook salmon at concentrations similar to those which
produce detrimental effects in laboratory exposures of other
species of salmonids. PCBs are present at much higher
concentrations than PCDDs and PCDFs in Great Lakes fish
(Kaczmar et al. 1985, Niimi and Oliver 1989) and, therefore,

may be responsible for most of the toxicity associated with

10

this class of compounds. TCDD-EQs of PCBs can be used to
evaluate the predictive power of the TCDD toxic equivalency
model.

mm

While controlled laboratory experiments are required to
demonstrate cause and effect relationships, exposure to
(contaminants in the field provides an ecologically relevant
measure of the possible impact of current concentrations of
PCBs within the milieu of additional contaminants and other
stresses to the female fish. Both laboratory and field
studies should be included in a comprehensive evaluation of
the effects of contaminants on fish in the Great Lakes.

IEELJHEHDHHULJEMBMELE

Traditional _congener-specific analysis provides
information on the pattern of PCB congeners present in the
eggs and. allows total concentrations of PCBs to \be
calculated. These measures of PCB concentrations can be
compared to previous studies. Carbon column chromatography
is required to separate the most toxic, coplanar PCB
congeners from other, interfering congeners present in the
chromatograms of traditional "congener-specific" PCB
analysis.
W

The purpose of this study was to examine the
relationships between concentrations of PCB congeners
in Lake Michigan chinook salmon eggs and rearing mortality

among clutches of those eggs. The specific objectives were

11
the following: .

1) Test for significant differences in mortality at
several stages of development among clutches of eggs and fry
from different females:

2) Test for significant differences in concentrations
of total PCBs and individual congeners among clutches of
eggs from different females:

3) Test for significant differences in egg lipid
content and size among female fish:

4) Compare tetal PCB concentrations determined in this
study with those of Giesy et al. (1986):

5) Determine AME-based TCDD-EQs for PCB congeners
quantitated in the eggs:

6) Test for significant differences in TCDD-EQs among.
egg clutches:

7) Determine the significance of correlations between
measures of PCB content and measures of rearing mortality
and evaluate the amount of among-clutch. variability in
rearing mortalities which can be explained by TCDD-EQs and
concentrations of PCBs:

'8) Develop a model to predict rearing mortality from
TCDD-EQs of PCBs and biological parameters of the eggs and
females.

W128

Twenty female chinook salmon with ripe eggs were

collected at one time and at one location. Relationships

among and between biological and chemical parameters of eggs

12
and females were examined by inference from correlation
analyses according to a Model II ANOVA design for random
treatments. Because individual females are dispersed during
their maturation in Lake Michigan they are exposed to
different quantities of contaminants. Thus, each female has
been exposed to a random ‘treatment. and females can. be
considered as treatments for egg clutches. variability in
measurements of residue concentrations within egg clutches
from individual females was assessed by triplicate
extractions. Power analysis (C. Cress, Department of Crop
and Soil Science, Michigan State University, 1986; Personal
Communication) and variances measured by Giesy et al. (1986)
were used to determine that three replicates would be
sufficient to distinguish differences in total PCB
concentrations in eggs from individual females as small as
15% of the mean and that five sets of eggs per clutch would
be sufficient to distinguish differences in rearing
mortality as small as 3% of the mean. The probablities of
Type I and. Type II errors ‘were set at 0.05 and 0.20,

respectively, for the power analysis.

METHODS
WW

Twenty female chinook salmon were collected on October
9, 1986, at the Little Manistee River weir operated by the
Michigan Department of Natural Resources (MDNR) . Ripe
females were anaesthetized with C02 and eggs were forced out
of the body by the injection of air into the body cavity.
The females were weighed and measured after the eggs were
removed. Belly flaps, livers, kidneys, and pyloric caeca
were removed for subsequent analysis. Subsamples of each
tissue were preserved for chlorinated hydrocarbon analysis
by freezing at -20 C. The remaining carcasses were tagged
and preserved in plastic bags at -20 C.

Two collections of milt pooled from five male salmon
were used to fertilize the eggs by the standard MDNR
procedure in which eggs from each female were mixed with 10
ml of the pooled milt for one minute and then rinsed
thoroughly with river water and allowed to water harden in
pro-rinsed plastic containers for at least one hour. During
water hardening, river water was continuously circulated
through the egg containers. Thereafter water was
periodically recirculated until all batches were ready for
transport. The plastic containers were sealed and placed on
newspaper over crushed ice for transport.

13

14
Eggs were distributed to incubators and preserved for

analysis at the Michigan State University (MSU) Fisheries
Research Facility. Eggs for chlorinated hydrocarbon
analyses were placed in foil packets which were frozen in
solvent-rinsed glass bottles. Twelve 30 g and six 15 9
packets were preserved for chlorinated hydrocarbon analysis,
including PCB analysis, and for determinations of egg
diameter and lipid content.
Bearing

Fertilized eggs were transported to MSU immediately
following water hardening and transferred to incubators for
rearing until just prior to yolk sac absorption. Since the
water temperature in the transport containers and the
incubators was within 1.5 C, eggs were transferred directly
from the containers into three Heath TecnaR vertical
incubators (F.A.L. Products, Inc., Tacoma, WA). Each of the
eight trays in each incubator was divided into five sections
and all trays were protected from direct light. Eggs from
one female were placed in a monolayer in all five sections
of a tray. The eggs in each section made up one rearing
group. Eggs and sac fry were observed daily and dead
embryos were removed, counted and frozen for possible later
analyses. Notes were made on abnormal 'fry, including
descriptions of ‘Siamese' twins, spinal abnormalities.

Water for the entire rearing period was well water
which had been filtered with activated carbon and aerated in

a head tank. This water had been shown to support normal

15

survival , growth and development of larval , juvenile and
adult salmonids (Giesy et a1. 1986, D. Garling, MSU
Department of Fisheries and Wildlife, 1986; Personal
Communication) . Water temperature was monitored daily and
was maintained at 12.0 i 0.4 C. The average dissolved
oxygen (DO) concentration in the incubators was 72.7 i 10.5%
of air saturation. Water flow through the incubators was
continuous and flow rates were maintained at 7-8 liters/min
until after the emergence of sac fry when the flow rates
were increased to 8-9 liters/min in order to maintain
adequate DO concentrations.

Free-swimming fry were reared in floating screen
baskets until the end of the rearing study. When sac fry
had nearly absorbed their yolk sacs, 100 fry from each
incubator compartment were transferred to floating baskets
in three rounded square molded tanks--100 fry per 5 liter
basket. Time of swim up was recorded as when 50% of the fry
in a basket were free-swimming above the bottom of the
basket at any given time. This behavioral endpoint
corresponded with button-up, the final absorption of the
yolk sac. Feeding was begun at first indication of swim up..
Fry were fed to satiation five times per day with Bio-Diet
(supplied by MDNR Wolfe Lake Hatchery) for the first seven
days. Thereafter fry were fed to satiation 2-3 times per
day. Dissolved oxygen (DO) and temperature were monitored
and bubblers and circulating pumps were added to improve D0

and reduce basket fouling” Fry were checked for aberrant

16

behaviors and dead fry were recorded and removed daily.
Egs_nianeter_neterninatien

Egg diameters were measured using an overhead projector
to project and magnify images of individual eggs. The 6.9x
magnification was calibrated by projecting the image of a
transparent ruler at the middle and edges of the viewing
area. Five eggs from each female were placed randomly on
the projector stage and the image of each was measured along

the horizontal plane.

W
W

Extraction and cleanup of egg samples for PCB analysis
was performed according to Giesy et al. (1986) after Ribick
et al. (1982) (Figure 2). All glassware used in the residue
analysis was washed in detergent, rinsed in hot tap water
and rinsed three times each with deionized, distilled water,
acetone, and hexane. All solvents used were pesticide grade
or better, unless noted otherwise, and all were purchased
from either Fisher Scientific (Fairlawn, NJ) or Mallinkrodt,
Inc. (Paris, KY).

Each 10 9 egg sample was extracted with dichloromethane
following internal standard addition. and. homogenization.
One egg in each aliquant was pierced and 0.375 ug of 2,4,6-
trichlorobiphenyl (#30) was injected in 50 ul of isooctane.
This PCB) congener"was selected. as an internal standard
because it does not occur in commercial AroclorR mixtures

nor has it been detected in environmental samples. Egg

17
Salmon eggs

WEIGH SAMPLE
ADD STANDARD

Spiked sample
DRY WITH Na2$O4
EXTRACT WITH MeCI2

 

Lipids and Unextractable
xenobiotics tissue

GEL PERMEATION
CHROMATOGRAPHY (GPO)

 

 

 

 

I 1
P088 and . Lipids
pesticides
FLORISIL PARTITIONING
PCBs and POW
some pesticides pesticides
Fraction I-A Fraction [-8

SILICA GEL \

PCBs Pesticides /
Fraction [-0 ’ Fraction [-0

Figure 2. Extraction and Clean-up Method for
' Polychlorinated Biphenyls and Pesticides.

 

Fraction I-E

 

18

aliquants were homogenized with four times their weight of
anhydrous Na2804 (stored at 130 C, J.T. Baker Chemical Co.,
Phillipsburg, NJ). The resulting dry powder was extracted
with 200 ml of dichloromethane at a flow of 3-5 ml/min in a
2 cm i.d. glass column. The volume of solvent in the
extracts was reduced under a vacumn with a Model 110
Rotovapor (Buchi, Flawil, Switzerland) at ambient
temperature and the volume was adjusted to exactly 9 . 5 ml
with 1:1 (v/v) mixture of cyclohexane and dichloromethane.

Lipids were removed from the extract by gel permeation
chromatography (GPC) . Five ml of the extract was loaded
quantitatively onto an automated GPC apparatus (ABC, Inc. ,
Columbia, MO) equipped with a column packed with 60 g of SX-
3 Bio-Beads (Bio-Rad Laboratories, Richmond, CA). The 1:1
cyclohexane and dichloromethane mobile phase was pumped
through the column at 5 ml/min. A 110 ml fraction was
collected after a 150 ml dump cycle. The volume of solvent
of the collected fraction was reduced as above and the
volume was adjusted to 1 ml with hexane.

Interfering compounds were removed and PCBs were
separated from pesticides using FlorisilR followed by silica
gel chromatography.) The 1 ml extract from the GPC was
applied, with small volume hexane rinses, to a FlorisilR
column composed of 1 g Na2804, 5 g of FlorisilR (activated
at 130 C, Fisher Scientific, Fair Lawn, NJ), 1 g Na2804 and
solvent-extracted glass wool in a 1 cm i.d. reservoir

column. The column was rinsed with 20 ml of hexane before

19
the extract was applied. The first 5 ml of each eluent
described below was used to rinse the column walls before
the remaining volume was added to the reservoir column. A
50 ml elution of 6% (v/v) diethyl ether in petroleum ether
eluted most nonpolar pesticides and PCB congeners (Fraction
I-A). Elution of the same column with 50 ml of 20% (v/v)
diethyl ether in petroleum ether yielded Fraction I-B which
contained pesticides more polar than those eluting in
Fraction I-A. A second column was prepared with 1.0 g
Na2804, 5.0 g of silica gel (70-230 mesh, activated at 130
C, Sigma Chemical Co., St. Louis, MO), an additional 1.0 g
Na2804 and solvent-extracted glass wool in a 1 cm i.d.
reservoir column. The column was rinsed with hexane and the
1 ml concentrate of the Fraction I-A was applied to the
column. PCBs and p,p-DDE were eluted in Fraction I-C with
50 ml of 0.5% benzene in hexane. Nonpolar pesticides
including toxaphene, chlordanes and DDT-complex pesticides
were eluted in Fraction. I-D ‘with 25% diethyl ether in
hexane. The two pesticide fractions, Fractions I-B and I-D,
were combined and the total volume of solvent was reduced as
above to approximately 1 ml. An internal standard of 0.195
ug of 2,2',3,3',4,4',5,5',6-nonachlorobiphenyl was added in
15 ul of isooctane and the final volume was adjusted to 1.00
ml in a calibrated conical vial under a gentle stream of
purified nitrogen. This combined pesticide extract was
archived as Fraction I-E. The volume of Fraction I-C, which

contained the PCBs, was adjusted to 1.00 ml the same as was

20
Fraction I-E after 0.998 ug of mirex was added in 9.6 ul of
isooctane.

Wang

Coplanar PCB congeners were selectively enriched in the
PCB extract of eggs from nine fish using carbon column
chromatography (CCC). This method separates PCB congeners
on the basis of planarity using particulate carbon dispersed
on glass fiber filters (Smith 1981, Schwartz and Smith
1987) . Coplanar PCBs are retained more strongly by the
carbon than are the non-coplanar PCB congeners. The carbon
is dispersed on glass fibers to reduce the number of
theoretical plates so that coplanar congeners can be
recovered as a group.

The carbon column apparatus was designed to allow both
forward and reverse elution of a series of solvents under
positive pressure by N2. The column itself consisted of a
1.3 cm i.d. glass column with Teflon plungers secured with
threaded caps. The column was packed with an homogenized
mixture of 150 mg of Px-21 carbon (2-10 um particle size,
Amoco, Inc., Chicago, IL) and 750 mg of Whatman GF/D filters
(Whatman International Ltd., Maidstone, England). Two 1.3
mm diameter Whatman GF/F filters and 3 Whatman GF/D filters
were placed at each end of the column. The packed column
length was 23.5 cm. Both ends of the column were connected
with Teflon tubing to fittings on a 4-port, 90° plug valve
(Hamilton Co., Reno, NV).. Tubing from the other two ports

of the valve attached to a reservoir column and an effluent

21

line. Solvent flowed forward through the column when the
valve connected the reservoir column line to the line from
the bottom of the column and the line from the top of the
column to the effluent line. When the valve position was
switched, solvent flowed from the reservoir column to the
top of the column and from the bottom of the column to the
effluent line. The effluent line was placed manually in
either a collection flask or a waste container. Flow was
maintained at 2-3 ml/min. by a low range (less than 10 psig)
gas meter in line between a N2 cylinder and the top of the
sealed reservoir column.

PCB congeners were separated into three fractions using
a combination of forward and reverse elutions with Burdick 8
Jackson solvents (Burdick; & Jackson Laboratories, Inc.,
Muskegon, MI) (Figure 3). A 1 ml aliquant of Fraction I-C
was loaded onto the column in the forward direction with
three hexane rinses (total volume of 2 ml) of the reservoir
column and tubing through which the extract was loaded. A
relatively weak solvent system consisting of 90 ml of 10%
(v/v) dichloromethane in hexane was added to the reservoir
column and eluted through the carbon column at 2-3 ml/min in
the forward direction. This solvent mixture eluted non-
coplanar PCB congeners quantitatively as Fraction II-A.
Fraction II-B containing the mono- and diortho-chlorinated
congeners was then collected following elution in the
forward direction with 100 ml of 30% (v/v) methylene

chloride in hexane. The non-ortho-chlorinated congeners

22

FOB Fraction I-C

LOAD ON

 

CARBON COLUMN

 

 

i

l

10% MeCl2/
Hexane
(forward)

Bulk PCBs

Fraction II-A

Figure 3 .

 

ii

30% MeCI2/
Hexane
(forward)

Mono— and di-
ortho PCBs

Fraction ”-3

Toluene
(reverse)

 

i

Nonortho PCBs

Frac lion ”-0

Carbon Column Chromatography for Separation of

Coplanar Polychlorinated Biphenyl Congeners from
Other Congeners.

23
were eluted in the reverse direction with 60 ml of toluene
(Fraction II-C). Recoveries of congeners in these fractions
were at least 92% (Table 2). Concentrations of PCB
congeners quantitated following CCC were not corrected for
recovery through that system.

The CCC system was washed between samples with the
following sequence of solvents: 50 ml of toluene in reverse
direction, 50 ml of 10% dichloromethane in methanol in the
forward direction. and 100 ‘ml of‘ hexane in the forward
direction. Flow was increased to 4-5 ml/min during the wash
cycle.

The Fractions II-B and II-C were prepared for GC
analysis by reducing the solvent volume and then adding
internal standard as 1.23 ug of congener #30 in 5 ul of
isooctane. The volume of solvent was reduced on a custom-
made, vacumn rotary evaporator at ambient temperature. When
the solvent volume reached approximately 2 m1, an additional
2 ml of isooctane was added to the flask. The volume was
again reduced to 2 ml before the extract was transferred to
a calibrated test tube with three hexane rinses. The final
volume was adjusted to 2.00 ml under a gentle stream of N2.
A 1000 ul GastightR syringe (Hamilton Company, Reno, NV) was
used to transfer 1.00 ml of the extracts to an autosampler

vial.

24

Table 2. Recoveries of PCB Congeners in Fractions
Collected from Carbon Column Chromatography

 

 

 

sPersent_Beeexeriea
IUPAC Fraction Fraction Fraction Total
Consensr_i __II:A__ __II:B__ __II:C__ Becoxerx
s1 - - 9s 9s
77 - - 92 92
123 - - 96 96
11a - - 99 99
114 - - 99 99
153 <1 95 3 9a
105 - - 93 93
137 2 93 2 95
13s - 93 3 96
153 <1 65 27 92
126 - - 93 93
166 - 32 12 94
128 - 96 3 99
167 - - 93 93
156 - - 95 95
157 - - 92 92
130 <1 90 6 97
169 - - 93 93
170 - 94 6 100

189 - - 96 96

 

25

91131111151121:

zxggtign__1;g. Concentrations of non-coplanar PCB
congeners and total PCBs were determined in Fraction I-C by
gas chromatography' with electron capture detection (GC-
ECD) . The GC used was either a Perkin-Elmer 8500 or a
Tracor 560. Both GC's were equipped with a 63Ni detector at
320 C and a split injector at 225 C. The same column was
used for all analyses: 30m x 0.25 mm i.d. DB-5 column (J&W
Scientific, Folsom, CA) with 0.25 um film thickness. Helium
at 20.0 psig was used as the carrier gas. The Perkin-Elmer
oven was held at 120 C for 1 min and then temperature
programmed from 120 C to 260 C at 2 C/min. The Tracor oven
was held at 125 C and programmed from 125 C to 270 C at 2
C/min.

All quantitation was based on peak areas relative to
the area of congener #30, the procedural internal standard.
Integration was performed by Perkin-Elmer software for
extracts analyzed using the Perkin-Elmer GC and by an SP4270
integrator (Spectra-Physics, San. Jose, CA) for extracts
analyzed using the Tracor GC. Area integration accuracy and
precision for both integration methods were confirmed by
comparison to the cut-and-weigh integration method.

Response factors for individual PCB congeners which
were detected in the Fraction I-C, the complete PCB extract,
were based on those developed at the USFWS Columbia National
Fisheries Contaminants Research Center (T.R. Schwartz 1987:

Unpublished data). ECD response factors relative to 2,4,6-

26

trichlorobiphenyl were determined based on analytical
standards or on FID analyses of peaks in AroclorR standards
of known proportions. ECD response factors were linear over
a .lnmited range, which. was determined for’ each of the
detectors used by injecting AroclorR standards and plotting
relative areas of individual peaks against individual
AroclorR concentrations. The concentration of total PCBs
was determined by summing the concentrations of all of the
individual congeners detected in Fraction I-C.

Quality control for the analysis of Fraction I-C
included procedural blanks, an internal standard for
extraction and injection and extraction of standards. Egg
samples were extracted in groups of four to eight. A
procedural blank was extracted with each group. No PCB
congeners were detected in the blanks. All peak areas were
corrected for extraction efficiency and variations in
detector sensitivity by comparison with the internal
standard peak area. Recoveries of individual congeners
through the extraction procedure were confirmed by comparing
relative areas of individual peaks before and after
extraction of an AroclorR standard containing all of the
peaks detected in the egg samples. The mean recovery of
individual peaks was 112% based on internal' standard
quantitation.

Fractigng_11;B_gng_11;§. The fractions containing the
di-, mono-, and non-ortho-chlorinated congeners were

analyzed by GC-ECD with a Varian 3700 GC, which was equipped

27

with a low volume 63N1 ECD detector at 330 C and modified
splitless injector at 250 C. The carrier gas was hydrogen
at 17 psig with N2 make-up at 30 ml/min. Samples of 2 ul
were injected by a Varian 8000 autosampler onto a 30 m x
0.252 mm i.d. J 8 W Scientific (Folsom, CA) DB-l column with
a 1 m heatractivated retention gap. The oven was
temperature-programmed from 60 C to 240 C at 2 C/min. The
GC was interfaced to chromatography software (Model 2600,
Rev. 5.0, Perkin-Elmer Nelson Systems, Cupertino, CA) with a
Nelson Analytical 900 series interface. This software
allowed confirmation and correction, if necessary, of all
peak area definitions.

Quantitation of the congeners detected in CCC Fractions
II-B and II-C was based on areas relative to the internal
standard. Nine-point calibration curves were developed for
each of the 17 coplanar PCB congeners using independent
dilutions of a standard solution of these congeners and
others at concentrations near 0.5 ug/ml. Concentrations of
congeners in. all of the fractions ‘were calculated from
second order polynomial equations which fit the calibration
curves with coefficients of determination greater than 0.997
(Appendix A).

The aaccuracy and. precision of 'the analysis of CCC
fractions was evaluated with authentic standards, procedural
blanks, and duplicate analysis. Recoveries of the 17
congeners analyzed averaged 95.3% with minimum and maximum

values of 92% and 100%, respectively (Table 2). No

28

concentrations of the coplanar congeners greater than 10% of
the lowest concentrations quantitated in samples were
detected in fractions of a procedural blank. The blank was
processed through both the preliminary PCB extraction and
cleanup and CCC and then injected in the same autosampler
batch as all of the sample fractions. The mean coefficient
of variation (CV) was 7.78 i 3.58 (mean 1 standard
deviation) for the concentrations of the 15 coplanar
congeners which were quantitated in the duplicate analyses
of an egg sample (Table 3).
W

Lipid contents of the eggs were determined
gravimetrically following a solvent extraction similar to
that used to extract PCBs from eggs. Six eggs from each
fish were ground with five times their weight of NaZSO4 and
eluted with 50 ml of dichloromethane in a 1 cm i.d. column.
Solvent was reduced to 2.00 ml and one-half (1.00 ml) of the
extract was transferred to a tared aluminum weigh boat. The
solvent was evaporated in a hood for at least 12 hours and
held in a desiccator for one hour prior to weighing of the
samples. Percent lipid in the egg samples was calculated as
follows: ,

% Lipid = LL121Q_!§IQDL1_1_Z_
Egg sample weight

Lipid analyses were performed in duplicate for all clutches.

29

 

Table 3. Duplicate Analysis of Chinook Salmon Eggs
Following Carbon Column Chromatography.
Concentrations are in nmol/g, wet weight of eggs.

some: Minimum Maxim Haas 539.091 911

31 N02 ND ND - -
77 0.0275 0.0281 0.0278 0.000423 1.52
123 0.0185 0.0193 0.0189 0.000507 2.68
118 0.903 0.997 0.950 0.0663 6.97
114 0.0332 0.0369 0.0350 0.00265 7.55
105 0.513 0.590 0.551 0.0543 9.84
138 1.09 1.15 1.12 0.0400 3.58
158 0.0846 0.0945 0.0895 0.00697 7.78
126 0.00614 0.00778 0.00696 0.00116 16.60
166 0.00433 0.00478 0.00456 0.000321 7.04
128 0.243 0.269 0.256 0.0186 7.27
167 0.0460 0.0519 0.0490 0.00418 8.54
156 0.0622 0.0706 0.0664 0.00594 8.94
157 0.0227 0.0257 0.0242 0.00214 8.84
169 ND ND ND - -

170 0.120 0.140 0.130 0.0141 10.83
189 0.00339 0.00384 0.00362 0.000318 8.78

 

lCV - the coefficient of variation:

2ND - not detected above 0.3 ng/g in the egg samples.

(Standard deviation/Mean) * 100%

30

QAIQHIALIQB_QI_TQDD:EQ§

TCDD-EQs were calculated from KEQs for AHH induction
because information on the AHH induction potency for all of
the coplanar congeners was available (Sawyer and Safe 1982)
and because AHH induction is highly correlated with organism
level effects of TCDD structural analogs in mammals. The
AHH-KEQs used in the calculation of TCDD-EQs ranged from
0.000007 to 0.40 for the coplanar PCB congeners quantitated
in the egg samples. All reported TCDD-EQs are based on AHH
induction. .
We

All statistical analyses were performed with the
Statistical Analysis System Version 6.02 for Personal
Computers (SAS Institute 1987) or SAS Version 5.18 (SAS
Institute 1982) for mainframe computers. The level of
significance that was used in all statistical tests was
p 5 0.05. Comparisons of means were performed by Tukey's
Studentized Range with the Generalized Linear Model (GLM)
procedures of SAS. The GLM procedure was also used to
perform analyses of variance (ANOVA) for fish as random
treatments for chemical and biological parameters of eggs.
Information on correlations among variables, including
Pearson product moments and p values, was obtained using the
CORR procedure in SAS. The REG procedure with the MAXR
option for model selection was used to determine the best

combination of variables for predicting embryo mortalities.

RESULTS

 

The female chinook salmon were all large, mature fish.
Fish lengths ranged from 80.6 cm to 94.6 cm and weights
ranged from 3.6 kg to 7.3 kg (Appendix B). The means and
standard deviations for length and weight were 88.1 1 4.6 cm
and 5.1 1 0.9 kg, respectively. Condition factors,
calculated from the English units of measure for length and
weight (CF-100,000 * weight(lb) / length3(in): Leitritz and
Lewis 1980) , were fairly uniform within the group of twenty
female fish with a mean and standard deviation of 26.5 3;
2.2.

Eggs from these females had a mean diameter of 8.6 mm,
a mean weight of 0.22 g, and mean lipid content of 8.5%.
These parameters all varied significantly among clutches
(Table 4). Egg diameter, egg weight, and the absolute mass
of lipid per egg were significantly and positively
correlated with female fish length and weight, but percent
lipid was not correlated with fish size, CF, egg size or
absolute lipid content per egg (Table 5).
W111}:

Rearing mortality from fertilization to four weeks
after swim up (FEMORT) ranged from 3.1% to 97.5% (Appendix
B) with the largest portion of the rearing mortality being

31

32

Table 4. Variability in Chinook Salmon Egg Dimensions and
Lipid Content of Eggs Which Was Explained by Among
Clutch Variability Rather Than Analytical Error.

 

W 021 __r_ .2121 _32__
Egg Diameter _19 15.29 0.0001 0.7840
Egg Weight 19 134.75 0.0001 0.9922
Mass Lipid 19 32.16 0.0001 0.9683
Percent Lipid 19 14.52 0.0001 0.9324

 

1DF-degrees of freedom for the model (clutches from
individual female fish), F-ratio produced by dividing the
mean square for the model by the mean square for error,
Pr>F-probability of exceeding the F ratio by chance,
Rz-proportion of variance in egg parameter explained by the
egg clutch.

33

Table 5. Pearson Pairwise Correlation Coefficients (r) for
Correlations Among Fish Dimensions and Egg
Characteristics for Twenty Female Chinook Salmon.
Probabilities of Greater Correlation Coefficients

Given that the Null Hypotheses Is True Are in
Parentheses.

 

Emlumnm mm mm 9r.

LIPERCT -0.135 0.335 0.101 0.205 0.109 0.263
(0.57) (0.093) (0.67) (0.39) (0.65) (0.26)

EGGWT 0.360 0.363 0.600 0.650 0.025
(0.0001) (0.0001) (0.005) (0.002) (0.92)

LIPWT ,0.340 0.661 0.663 0.150
(0.0001) (0.002) (0.001) (0.53)

DIAM 0.575 0.603 0.066
(0.003) (0.004) (0.73)

FWT 0.359 0.457
(0.0001) (0.04)

FLENG -0.057
(0.31)

 

1LIPERCT-percent lipid (w/w) in eggs, EGGWTsegg weights,
LIPWT-mass of lipid per egg, DIAMsdiameter of eggs, FWT-fish

length CF-condition factor of fish.
ungfid’I pkfiféfli -

34

observed during the hatching period (Figure 4). Fry were
observed dying partially hatched. These fry died when only
the head was free of the egg shell. A single swim up fry
exhibited the erratic, spinning behavior pattern that has
been associated with exposure to chlorinated planar
compounds. Total percent mortalities and subsets thereof
differed significantly among clutches (Table 6).
W

The mean concentration of total PCBs was 7.02 ug/g wet
weight and the mean concentrations of individual congeners
ranged from below detection limit to 0.73 ug/g. Detection
limits were different for different congeners, but all
reported concentrations were greater than 0.0005 ug/g.
Total PCBs varied significantly among clutches, based on an
ANOVA, and ranged from 3.21 to 13.64 ug/g (Appendix B).
Analysis of the PCB extract Fraction I-C resulted in
detection and quantitation of 94 congeners and unresolved
congener pairs, of which 92% of the concentrations varied
significantly among egg clutches. The pattern of congeners
observed in Fraction I-C appeared similar to a mixture of
AroclorsR 1248, 1254 and 1260 with AroclorR 1254
predominating.

The fifteen congeners quantitated following CCC were
congeners #77, #105, #114, #118, #123, #126, #128, #138,
#156, #157, #158, #166, #167, #170 and #189. Of these, only
#105, #118, #156 and #189 were able to be measured in the

PCB analysis before the use of CCC. Concentrations of

35

 

 

 

 

 

 

Hutch Swim-up
I l I l
100 A I r I t l I 1’ T I t I r '
r ‘ ‘____‘\ . J
90.. \ All Clutches Except {15 I
b . . \\._——-—_-~—_~ a
h. 50i- l fl -l
E _ Moon ole Clutches .
O I
a 70- -
o I- | d
5‘ 60- ' -
Q— . d
3 50+- I -
.g l- . .
E 4o— -
a . ‘ Clutch {15 l
m 30:— -i
'0',
3 l ‘ l
e 20- 2 -
0‘: b \ '1
{ 10- - -
. - - l
r l . 14 1-1'1-1'1' j'l'LI r
0 200 400 600 800 1000 1200 1400

Degree-Days after Fertilization(90-doy)

Figure 4. Survival of Chinook Salmon from Fertilization to

Swim-up .

36

Table 6. Proportion of Variability in Rearing Mortality
Among Egg Rearing Groups Which Was Explained by
Among Egg Clutch Variability.

 

322131112 01:1 _r_ .2:_r_> _22_
Hatching/A2 19 15.36 0.0001 0.7349
Hatching/H3 19 35.23 0.0001 0.3932
Partial Match 19 2.53 0.0021 0.3754
Total 13 23.47 0.0001 0.3723

 

1DF-degrees of freedom for the model (clutches from
individual female fish), Faratio produced by dividing the
mean square for the model by the mean square for error,
Pr>F-probability of exceeding the F ratio by chance,
Rz-proportion of variance in egg parameter explained by the
egg clutch.

2Hatching/A indicates percent mortality during hatch
relative to all eggs initially present.

3Hatching/H indicates percent mortality during hatch
relative only to eggs present at the beginning of the
defined hatch period (degree-day 440).

37

congeners #81 and 169 were not detected above 0.3 ng/g,_the
limit of quantitation for these congeners based on the
lowest concentration used in the calibration curves. Mean
concentrations of the congeners which were detected in the
nine clutches were in the ng/g range (Table 7). Among-
clutch variability could not be separated from analytical
variability for concentrations of these 15 congeners because
a complete set of replicates was not available.
ZillllfizTQDD_Eflnilil§nL§

The total AHH induction-based TCDD-EQs in the nine egg
samples ranged from 8808 pmol/kg to 3860 pmol/kg with a mean
of 2370 pmol/kg or 762 ng/kg (Table 8). Three congeners
together contributed greater than 99% of the total TCDD-EQs:
congeners #77, #105 and #126 (Table 9). Total TCDD-EQs were
significantly and positively correlated with the
concentration of total PCBs (Figure 5) . The concentration
of total PCBs explained 63% (r-0.792) of the variability in
total TCDD-EQs and 53% (r80.728) of the variability in TCDD-
EQs contributed by congener #126 (Table 10).
W

Rearing’ mortality from fertilization to four weeks
after swim up (FEMORT)-was not significantly correlated with
concentrations of total PCBs or any of the coplanar
congeners or with TCDD-EQs (e.g. Figure 6) . FEMORT was
inversely related to percent lipid in the eggs (r a -0.425)
‘but the relationship was only marginally significant
(p-0.062).

Table 7.

Mean Concentrations of Coplanar PCB Congeners
Quantitated After Carbon Column Chromatography
Enrichment of Primary PCB Extract of Chinook

Salmon Eggs.

weight.

38

Concentrations are in ng/g, wet

 

calamari-Minimalism

77
123
118
114
105
138
158
126
166
128
167
156
157
170
189

3.34
1.88
111.06
3.34
65.27
35.15
8.89
0.51
0.21
4.18
5.06
7.71
2.21
4.30
0.40

8.43
8.40
399.98
15.10
233.04
440.46
36.29
2.54
1.97
102.27
21.34
30.69
11.41
57.01
1.91

HEAD §L§_D§! §L§_EIIQI 92

6.79
5.02
250.84
9.06
141.52
276.56
20.99
1.48
1.07
59.70
12.80
17.22
5.98
30.16
1.01

1.64
1.89
84.65
3.60
49.30
127.60
10.01
0.67
0.58
31.03
5.37
7.13
2.99
18.16
0.48

0.52
0.60

26.77

1.14

15.59
40.35

3.17
0.21
0.18
9.81
1.70
2.25
0.94
5.74
0.15

24.2
37.7
33.7
39.7
34.8
46.1
47.7
45.5
53.9
52.0
41.9
41.4
49.9
60.2
47.7

 

1980) is used to denote PCB congeners.

1The IUPAC numbering system (Ballschmitter and Zell

39

 

 

 

Table 8. Concentrations and TCDD-EQs of Coplanar PCB
. Congeners Measured in Lake Michigan Chinook Salmon
Eggs. KEQs Are Based on Molar Ratios of Aryl
Hydrocarbon Hydroxylase Induction Potency in Rat
Liver Cells (Sawyer and Safe 1982).
IUPAC Concentration AHHKEQ TCDD-EQ TCDD-EQ
WEI—1 41111212159.)—

77 23.3 0.00274 63.7 20.5
105 433.6 0.00110 477.0 153.6
114 27.7 0.0000246 0.7 0.2
118 768.3 0.00000834 6.4 2.1
123 15.4 0.0000986 1.5 0.5
126 4.5 0.400 1811.0 583.1
156 47.7 0.0000464 2.2 0.7
157 17.9 0.000135 2.2 0.7
167 35.5 0.00000722 0.3 0.1
170 76.3 0.0000160 1.2 0.4
189 2.6 0.0000084 0.02 0.0

Totals: 2366.2 761.9
Table 9. Chlorine Substitution Pattern and Percent
Contribution to Total TCDD-EQ for Coplanar PCB
Congeners Measured in Lake Michigan Chinook Salmon
Eggs.
IUPAC
90093332.: chlorination W1
77 3,3',4,4' 2.7
105 2,3,3',4,4' 20.2
114 2,3,4,4',5 NS
118 2,3',4,4',5 0.3
123 2',3,4,4',5 0.1
126 3,3',4,4',5 76.5
156 2,3,3',4,4',5 0.1
157 2,3,3',4,4',5' 0.1
167 2,3',4,4',5,5' NS
170 2,2',3,3',4,4',5 NS
189 2,3,3',4,4',5,5' NS
Total: 100.0

 

1N8 indicates that the contribution of that congener
to the total TCDD-EQ was not significant (<0.1%).

40

 

 

 

50-
.‘2” 4o- '
4.;
E
c 01
g5? 30-
CI
53>
in O 20..
o E
a C
E 10
’2 e
C ' l ' . I ' I ' I
0.000 0.001 0.002 0.003 . 0.004
Total Toxic Equivalents
nmol TODD/g egg
Figure 5. Concentration of Total PCB Concentration Versus

Total TCDD-EQ.

41

Table 10. Pearson Pairwise Correlation Coefficients (r) for

Correlations Among Rearing Mortality and PCB Content
Expressed as Both Concentrations and TCDD-EQs for
Eggs from Nine Chinook Salmon.
Greater Correlation Coefficients Given that the Null
Hypotheses Is True Are in Parentheses.

Probabilities of

 

Total
PCB

Total
TCDD-EQ

IUPAC #126

'TCDD-EQ

HHOMCRT

Total

0.7917
(0.006)

IUPAC #126
1922239

0.7231
(0.017)

0.9922
(0.0001)

W1

0.2936
(0.410)

0.7169
(0.020)

0.7336
(0.007)

25110312

-0.4560
(0.217)

(0.290)

-0.3639
(0.336)

0.1087
(0.700)

 

lHHOMORT- percent of fry which died partially hatched
relative to the number of live eggs at the beginning of the

hatch period.

ZFEMORTi-I percent of eggs and fry which died between
fertilization and the end of the experiment

42

 

 

100- .

S?
v 80-
9:"

E

o 50"

:E

a 1

c .
0: 40d

0 e

O
m d

.3 20-
|—- . 0 s 9 e

T
8
0 . , . r . I
0.002 0.003 0.004

0.000 ' 0.001

Total Toxic Equivalents
nmol TCDD/g egg

Figure 6. Total Rearing Mortality of Chinook Salmon Eggs
and Fry Versus Total TCDD-EQs in the Eggs.

43

The only subset of rearing mortality that was
significantly correlated to PCB parameters was the mortality
of partially hatched fry during the hatch period (HHOMORT).
HHOMORT was significantly and positively correlated with
total TCDD-EQs (Figure 7) and TCDD-EQs from several
individual coplanar congeners. Total TCDD-EQs and the toxic
equivalents of congener #126 alone explained 51% and 61%,.
respectively, of the variability observed in HHOMORT (Table
10). f

TCDD-EQs, which are expressed in units of
concentration, were inversely related to both egg weight and
the mass of lipid per egg. The Pearson pairwise correlation
coefficients (r) for total TCDD-EQs versus egg weight and
lipid per egg were -0.705 (p-0.023) and -0.668 (p-0.035),
respectively. Thus, larger eggs with greater absolute
amounts of lipid had lower concentrations of TCDD-EQs.
MW

Total rearing mortality was not predicted better by

regressing mortality on fish and egg characteristics and PCB
concentrations than it was by simple correlations with PCB
concentrations alone. Percent lipid, egg weight, lipid per
egg, egg diameter, fish length, fish weight and either total
TCDD-EQ lor’ concentration. of total PCBs were entered as
possible parameters in a multiple regression. The best
models containing a given number of parameters were selected
by maximizing R? (meR in PROC REG of SAS Institute, Inc.

1987). When either total TCDD-EQ or concentration of total

ry

hatched F

. y of Partially: .
Z of live eggs at beginning of hatch)

Mortalit

(as

44

 

 

1.0-
0.8-
. /‘
0.6- //
z’
1’
1’
0.4”" . //
. ///
0.2- . // °
/ .
z’
I
0.0 I f l r r m
0.000 0.001 0.002 0.003 0.004
Total Toxic Equivalents
nmol TCDD/g egg

Figure 7. Mortality of Partially Hatched Chinook Salmon Fry
Versus Total TCDD-EQs in the Eggs.

45
PCBs was entered into the list of possible parameters, none
of the possible regressions was significant until the number
of parameters in the model approached the number of
observations. In both cases, however, the PCB content
parameter was the first parameter added to the model to
predict mortality.

Prediction of mortality of partially hatched fry
expressed as a percentage of those which died during the
hatch period (mlOMORT) was improved by adding parameters
other than those for PCB content into multiple regression
models. The best two- to four-parameter models for
prediction of HHOMORT did not include total TCDD—EQs or,
alternatively, concentrations of total PCB, however (Table
11). Multiple regression models containing more than four
or five parameters [were not considered to be truly
significant, even though they were statistically
significant, since only nine observations were available for

regressions which included TCDD-EQs.

DISCUSSION

Minions

The observed correlations between measures of rearing
mortality and of PCB content of eggs do (not demonstrate
cause and effect relationships and are complicated by the
fact that concentrations of PCBs are usually intercorrelated
with other contaminants (Giesy et al. 1986). The tests of
correlation that were performed indicated that
concentrations, of 'total PCBs did. not relate ‘to overall
rearing mortality. The concentrations of PCBs in the eggs
of fish exposed to PCBs in the field were not great enough
to elicit a response which could be distinguished from
responses to the many other variables which may have
contributed to rearing mortality. In addition, the range of
concentrations of PCBs in the eggs was restricted because
eggs were taken from one location at one time. The lack of
correlation between parameters such as concentration of
total PCBs and cumulative rearing mortality may be an
artifact of this restriction in range of concentrations of
PCBs in samples rather than because such a relationship does
not exist. The strong correlation of total TCDD-EQs with
HHOMORT indicated a toxicant-specific response ‘in the
chinook salmon early life stages, although laboratory
studies would be required to confirm such a relationship.

46

47

The significant correlation between the concentration
of total PCBs and total TCDD-EQs is not unexpected because
the coplanar congeners occur in the AroclorR mixtures which
are the source of PCBs in the environment (Kannan et al.
1987). However, 37% of the variability of one variable was
not explained by the other, which indicates that the
coplanar PCB congeners may accumulate differently in fish
eggs from the congeners which constitute to majority of the
concentration of total PCBs. TCDD-EQs were correlated with
the mortality of partially hatched fry while the
concentration of total PCBs was not: therefore, the portion
of the variation in TCDD-EQ which is not explained by its
relationship to the concentration of total PCBs may be
toxicologically significant. Fish monitoring programs which
measure concentrations of PCBs without including the
coplanar congeners may be missing important predictive .
information.

We:

PCB concentrations in the eggs were similar to those
reported in other studies, but no other studies are
available with which to compare total TCDD-EQs in fish eggs.
The mean total PCB concentration of 7.02 ug/g was less than
the 9.09 ug/g measured in eggs of chinook salmon from the
same locality in 1982 (Giesy et al. 1986). The mean total
1 TODD-E0 for PCBs of 762 ng/kg in the eggs was at least ten
times the concentrations of PCDDs, primarily 2,3,7,8-TCDD,

measured in eggs from the same locality (M. Zabik, MSU

48

Pesticide Research Center, 1989: Personal Communication) .
Thus, based on a strictly additive model, the coplanar PCB
congeners were responsible for approximately 90% of the
toxic potency of the 2,3,7,8-TCDD structural analogs.
Similarly, coplanar PCBs contributed five to ten times the
toxic potency of PCDDs and PCDFs in salmonids collected in
Lake Ontario (Niimi and Oliver 1989).

The observed pattern of mortality was different from
that observed in other studies. The greater mortalities
observed in hatcheries in the early 1980's occurred
following the absorption of the yolk sac and the onset of
active feeding (J. Hnath, MDNR Fish Division, 1981: Personal
Communication). In the current study, however, very few fry
died once the hatching period was completed. The mortality
observed during the hatching period in this study was less
than that observed in a similar study of chinook salmon eggs
in 1982 (Giesy et al. 1982). Mortality of partially hatched
lake trout fry has been observed in a study in which the
eggs and fry were exposed to 2,3,7,8-TCDD, but this
phenomenon occurred at greater doses than those doses which
produced "blue sac" disease (J. Spitzbergen, Department of
Veterinary Pathology, Cornell University, 1989: Personal
Communication) . None of the hemorrhaging or edema which
characterize "blue sac" disease were observed in the chinook

salmon fry in the present study.

49

The lack of a relationship between concentration of
total PCBs and rearing mortality endpoints is consistent
with other studies. Lake trout fry mortalities of 40.3 to
65.5% were not correlated with concentration in a study in
which eggs were collected from several locations on the
Great Lakes (Mac 1988). Concentrations of AroclorsR in eggs
of Lake Michigan chinook salmon eggs did not correlate well

with rearing success in a 1982 study (Giesy et al. 1986) .'

CONCLUSIONS

1) Among clutch variation was significant relative to
analytical error for rearing mortality measurements, egg and
fish characteristics, and PCB concentrations and TCDD-EQs
for most congeners.

2) The mean concentration of total PCBs in the eggs of
Lake Michigan chinook salmon in 1986 was 7.02 ug/g, which
was lower than what was measured in chinook salmon eggs from
the same locality in 1982.

3) The mean total TCDD-EQ was 762 ng/kg wet weight in
the eggs when calculated from AHH induction potency KEQs and
concentrations of individual PCB congeners. Estimates of
TCDD-EQs from PCDDs and PCDFs are an order of magnitude less
than that contributed by coplanar PCB congeners.

4) Total TCDD-EQ was correlated with the concentration
of total PCBs (r2-.627).

5) .Total TCDD-EQs in eggs were correlated with the
mortality of partially hatched fry which died during the
hatching period, but TCDD-EQs were not correlated with total
rearing mortality.

6) Concentration of total PCBs was not correlated with
any of the measures of rearing mortality.

7) Total rearing mortality was not predicted better by
models containing parameters for fish and egg size and egg

50

51
lipid content than by models containing concentrations of
total PCBs alone.

Thus, current concentrations of PCBs in Lake Michigan
chinook salmon eggs are not affecting total rearing success
within the range of variability observed for mortalities
from all causes and over the restricted range of PCB
concentrations observed in this egg sample. However, the
toxicity of coplanar PCB congeners may be contributing to
the specific loss of fry during the hatching period. The
relationship of PCBs to this teratogenic effect was only
apparent when TCDD-EQs were calculated for the coplanar PCB

congeners.

APPENDICES

1.0

0.8

0.6

0.4

0.2

Area Relative to Internal Standard

Figure 8. Calibration Curve for 3,3',4,4',5-

52

APPENDIX A

 

Ra=1 .000

 

L J1

0.08
Concentration of PCB Congener #126 (ug/ml)

I . I I
0.02 0.04 0.06

Pentachlorobiphenyl (IUPAC #126).

I'

r T r I r I ' I I '
_ CONC=0.0207(Rel. Area)’+0.104(Rel. Area)-0.00132

 

I
0.10

 

0.12

53

APPENDIX B

.couoc «on M o» coauouaaeuuou none suaaouhoa ucoouoduemosmm

«9.73:6 nouns 3.003 moon on scaussaaauuou noun unwasuuoa uneducatemozumm

Lemma oases oco uueuuaou anicaveuocod\moe..hdsucoeoaneov uouood coaflccoonsoH

 

 

Nm.n Hm.¢ Hm.w wHN.o om.m w.m~ om.n m.Hm on
on; $23 mug. 035 2.5 «.3 ové Ema 3
HN.h nv.cH on.o moa.o om.h o.m~ nm.n w.om ma
HN.H mH.oa up.» mvu.o no.» m.nm m~.m h.mm 5H
om.h cm.» oa.m mem.c bo.m m.m~ o~.m ~.om ma
mm.om om.hm h~.m «ha.o nm.m «.mm ve.m b.5m ma
n~.m om.NH em.» omm.o bo.m m.~n o~.h h.~m «a
hm.n H~.m Nm.m NbN.o ov.m w.v~ -.m ¢.Hm nH
mb.o mm.HH nv.m bom.c hv.m N.h~ o~.m o.vm NH
N¢.H Ha.n mH.HH nha.c om.» o.o~ vm.e H.mw AH
mm.m mn.NH mm.h nHN.c on.» m.¢~ mh.¢_ m.mm OH
me.e mv.HH on.m nv~.o om.m h.mm no.w o.¢m m
3.3 oaén ow.» man... 2.6 «.3 moé oéo m
hm.n No.m do.» ovN.o Sh.m H.m~ mm.v m.em h
om.HH mH.vN hm.» wuu.o cm.o h.n~ v¢.m o.¢m m
an.m No.0N ed.m mc~.o on.» n.m~ na.e m.nm m
bd.n mm.da «m.m an.o hn.h m.hN em.e o.nm e
¢N.m mm.HH vv.m Han.o mm.» n.m~ Nb.e m.vo n
Hn.h cm.HH «a.» mmn.c no.0 H.b~ ~o.m m.om N
oo.v em.nn No.m «nm.o ow.m «.mn ho.m n.nm A
Hmmmmu manque“ named. amquemdum 4mmqumummmda Hum amuqmomdmm demqmwmamq. u mean
a use use cues cues
.eeae .cooecoem oxen none condom
socceao Hoece>eccn spaces you noon on» ca .uosrooosv aucoao>eeom nous
Io.s.n.~ use 2393 93:23: mouscwuoHnOhHom uo acoausuucoocoo use
15:69.3: modicum down one an: no nodunauouosusso no... decade, and: :3 canon.

haunts... us0ou0nlamozm

.Aommav Haou

use n0uuasnonaaom so o0msn 0nd sown: snooze: OdmbH an o0unau 0mm 0u0c00cou mum dado

.oowu00 mowcousn 0:» no madcca00n 0:»

us 0>aas u0nasc 0:» :0 00002 any U0zousn aaaeauuom no muwasuuoa uc0ou0muamozomm soowu00
use» no 0.55200.“ 0:» us 02:0 .095: 05 so o0non 00.300 meanness >00 :0» 0:» 05350
au0uoa 03.003 .500 Gorge 0a....M00u 0.3 00 0:0 0:» on 051530
sown audaouMOI uc0ou0alamozmm «05Il«30 on mouse «on scum hueaeuuos uc0ou00namozmmn

 

54

amwummqmwdfi amuummqwddm

H0.0
ev.HH

Nh.HH
05.0

00.0
0H.0H
bd.b
V0.0

N0.0

0.00H
0.00H

0.00H
H.00

0.0HH
0.000
0.0NH
0.00

0.0NH

0H.0
HH.0

H0.h
00.0
00.0
00.0

00.0

0000
0000
5000
dude
v0Hh
00HOH
0000
00H0
0005
0000
000nd
H000
mama
H000
0000
0000
0N00
0000
0000
00N0

000.0
0v0.0
000.0
00H.0
~00.H
000.0
000.0
00H.0
0vH.0
end.0
000.0
000.0
0¢N.0
050.0
050.0
0bd.0
000.0
000.0
vvd.0
v0¢.0

0h.d
00.v
0>.H
00.0
00.0
«0.00
00.0
0H.n
05.0
hv.d
n0.v
00.0
00.0w
00.0
0N.h
Ah.0
HB.N
NH.v
0H.0
00.0

0N.0
00.0
00.0
00.H
00.0
00.0
0N.H
00.0
00.0
00.H
00.0
00.H
00.H
00.0
00.0
0¢.H
00.0
00.N
0¢.H
00.H

camauuquNfi amaummquum ammudmu Hudnu Hummuw
HUUOB

00.0
00.0
N0.N
00.0
0H.H
Hm.0
0H.N
00.N
0N.n
00.0
00.0
H0.0
00.0
00.0
00.0
«0.0
0n.N
H0.¢
0H.n
00.5

«amused

HNMQ‘M‘DI‘QG

1

 

...u.coov Ha canoe

55

 

00.00
00.00

00.00
00.00

00.00
00.00
00.00
00.0

0v.¢0

«.000
0.000

0.00
0.000

0.000
0.0vv
0.000
0.000

0.000

00.00
00.00

00.0
00.00

00.00
00.000
00.00
00.00

00.00

00.0
00.0

00.0
00.0
00.0
00.0

00.0

00.0
00.0

00.0
00.0

00.0
00.0
00.0
00.0

00.0

0.000
0.000

0.000
0.000

0.000
0.000
0.000
0.000

0.000

«MMumauowdm 00000000000 Ammuumquwdfl 000V0000000 Amuumquudfl «anyway 0000

HNO‘IQ’M‘OI‘QO‘

filmed“

 

.A.u.coov Ha canoe

56

 

00.0
00.0
00.0
00.0

00.0

00.00
00.00

00.0
00.00

00.00
00.00
00.00
00.00

00.00

00.00
00.00

00.00
00.0

00.0
00.00
00.0
00.0

00.00

00.0
00.0

00.9

00.0

00.0

00.00
00.00

00.0
00.00

00.00
00.00
00.00
00.0

00.00

00.0
00.00
00.0
00.0

00.0

Ammummqmw0fl Amauumqumdfl “0000900000

HNMV‘D‘OI‘QQO
HHHHHHHHHN

HNMfiD‘OI‘QO‘S

«Immwm

 

.1.u.coov He canoe

LIST OF REFERENCES

LIST OF REFERENCES

Ballschmiter, K., and Zell, M. 1980. Analysis of
polychlorinated biphenyls (PCB) by glass capillary gas

chromatography- Iresniua 3 Anal Chem 302: 20- 31.

Baumann, P. C., and Whittle, D. M. 1988. The status of
selected organics in the Laurentian Great Lakes: an
overview of DDT, PCBs, dioxins, furans, and aromatic

hydrocarbons. Agnat. ngiggl. 11:241-257.

Bidleman, T. F., Christensen, E. J., Billings, W. N., and
Leonard, R. 1981. Atmospheric transport of
organochlorines in the North Atlantic Gyre. 1. Mar. Res.
39:443-464.

Colborn. T- 1988. Tna_nae_9f_10xicitx_nsnixalencx_as_a
§1§g§_Lakg§. USEPA Report, Environmental Results Branch.

 

DeVault, D. S., Willford, W. A., Hesselberg, R. J., Nortrupt,
D. A., Rundberg, E. G. S., Alwan, A. K., and Bautista, C.
1986. Contaminant trends in lake trout (Salyelinus
nanaygggh) from the upper Great Lakes. Arch. Enyirgg.
anggm. ngiggl. 15. 349- 356.

DeVoogt, P., and Brinkman, U. A. T. 1989. Production,
properties and usage of polychlorinated biphenyls. In
nal99enataa_BianenxlaI_Ternnenxl§1_Nantnaleneei

Dibenaaai0xin§_ana_Belated_2raanat§. edS- R. D- Kimbrough
and A. A. Jensen, pp. 3-46. Elsevier.

F1699. R. M. 1982 Qaaa_hiater!_infarmatian_fren_an

! 9'01 '. '1'! N.,-8V ' - - ‘9! .511
a ' ' ° . U. S.
Fish and Wildlife Report, CHN 82-78 and 82-81.

Furlong, E. T., Carter, D. 8., and Hites, R. A. 1988.
Organic contaminants in sediments from the Trenton
Channel of the Detroit River, Michigan. J. Great Lakes
33:. 14(4):489-501.

Giesy, J. P., Newsted, J., and Garling, D. L. 1986.
Relationships between chlorinated hydrocarbon

57

58

concentrations and rearing mortality of chinook salmon

(9029010521193 5203111132113) 9998 from Lake Hichigan- .1.
Great Lakes Bea. 12(1)=82-98-

Halter, M. T., and Johnson, H. E. 1974. Acute toxicities of
a polychlorinated biphenyl (PCB) and DDT alone and in
combination to early life stages of coho salmon

( kicntchbliiah-Beehcardcan.

31:543-1547.

Hansen, D. J., Schimmel, S. C., and Forester, J. 1975.
Effects of Aroclor 1016 on embryos, fry, juveniles and
adults of sheepshead minnows (syringggn yariggatus).

Trans. Amer. Elfin. figg. 104:584-588.

Hutzinger, 0., Safe, 8., and Zitko, v. 1974. Ihe_§hemi§;:y
91_Rgfl;§, Cleveland, OH: CRC Press.

Kaczmar, S. W., Zabick, M. J., and D'Itri, F. M. 1985. Part
per trillion residues of 2, 3, 7 ,8-tetrachlorodibenzo-p-
dioxin in Michigan fish. In gnlgzing;§g_pigxin§_gn§
WI. eds. L- H-
Keith, C. Rappe and G. Chaudhary, pp. 103- 110. Stoneham,
MA: Butterworth Publisher.

Kannan, N., Tanabe, S., Wakimoto, T. and Tatsukawa, R. 1987.
Coplanar Polychlorinated Biphenyls in Aroclor and Kanechlor
Mixtures. I- Macc- fo- Anal- cham-

70:451-454.

Kubiak, T. J., Harris, H. J., Smith, L. M., Schwartz, T. R.,
Stalling, D. L., Trick, J. A., Sileo, L., Docherty, D. and
Erdman, T. C. 1989. Microcontaminants and reproductive
impairment of the Forster' s tern on Green Bay, Lake

Michigan-1983 Arch Enxircn contain lexical 18: 706-727

Leitritz, E., and Lewis, R. C. 1980. TIQQL_QBQ_§§1an
WM. Berkely. CA: DiV- of AgriC-
Sci., Univ. of Ca.

Long, G., McKinney, J., and Pedersen, L. 1987.
Polychlorinated dibenzofurn (PCDF) binding to the Ah
receptor(s) and associated enzyme induction. Theoretical
model based on molecular parameters. fluent. Struct.-Ac .
3313;. 6:1-7.

Mac, M. J. 1988. Toxic substances and survival of Lake
Michigan salmonids: Field and laboratory approaches. In
., «u I2I : i!‘ -= - ‘u r e I: 1
Rogue, ed. M. S. Evans, pp. 389-401. New York: John Wiley
8 Sons.

Mac, M. J., Edsall, C. C., and Seelye, J. G. 1985. Survival
of lake trout eggs and fry reared in water from the upper

59

Great Lakes. J, Great Lake: Egg. 11(4):520-529.

Mauck, W. L., Mehrle, P. M., and Mayer, F. L. 1978. Effects
of the polychlorinated biphenyl Aroclor 1254 on the
growth, survival and bone development in brook trout

(Salxalinaa.fantinalia)- l.£iah. Baa. heard Can-
35(3):1os4-1ose.

Mayer, K. S., Mayer, F. L., and Witt, A. 1985. Waste
transformer oil and PCB toxicity to rainbow trout. Trans.

Angr. Elfin. Egg. 114:869-886.

Melancon, M. J., and Lech, J. J. 1983. Dose-effect
relationship for induction of hepatic monooxygenase
activity in rainbow trout and carp by Aroclor 1254. Agnat.
ngiggl. 4:51-61.

Nebeker, A. V., Puglisi, F. A., and DeFoe, D. L. 1974.
Effect of polychlorinated biphenyl compounds on survival
and reproduction of the fathead minnow and flagfish.

Trnng. Amer. Elfin. S99. 103:562-568.

Niimi, A. J., and Oliver, 8. G. 1989. Assessment of relative
toxicity of chlorinated dibenzo-p-dioxins, dibenzofurans,
and biphenyls in Lake Ontario salmonids to mammalian
systems using toxic equivalent factors (TEF). thngspngrg
18:1413-1423.

Niimi, A. J., and Oliver, 8. G. 1989. Distribution of
polychlorinated biphenyl (PCB) congeners and other
halocarbons in fish and muscle among Lake Ontario

salmonids. Enyirgn. S91. Iggnngl. 23:83-88.

Oliver, B. G., and Niimi, A. J. 1988. Trophodynamic analysis
of polychlorinated biphenyl congeners and other
chlorinated hydrocarbons in the Lake Ontario ecosystem.

Enyirgn. £21. Iggnngl. 22:388-397.

Poland, A., Greenlee, W. F., and Kende, A. S. 1979. Studies
onthe mechanism of action of the chlorinated dibenzo-p-
dioxins and related compounds. Arch. E-X- 593g. Sci. 320:
214-230.

Poland, A., and Knutson, J. C. 1982. 2,3,7,8-
tetrachlorodibenzo-p-dioxin and related halogenated
aromatic hydrocarbons: examination of the mechanism of

toxicity. Ann. Bey. Ennrnaggl. ngiggl. 22:517-554.

Ribick, M. A., Dubay, G. R., Petty, J. D., Stalling, D. C.,
and Schmitt, C. J. 1982. Toxaphene residues in fish:
indentification, quantification and confirmation at part

per billion levels. Enyirgn. S91. Iggnngl. 16:310-318.

6O

Risebrough, R. W., Walker, W., Schmidt, T. T., Delappe, B.
W., and Connors, C. W. 1976. Transfer of chlorinated
biphenyls to Antarctica. unsure 264:738-739.

Safe, 8. 1987. Determination of 2,3,7,8-TCDD equivalent
factors (TEFs): support for the use of the in vitro AHH
induction assay. ghgnggnngrg 16:791-802.

Safe, 8., Bandiera, S., Sawyer, T., Zmudzka, 8., Mason, G.,
Romkes, M., Denomme, M. A., Sparling, J., Okey, A.B. and
Fujita, T. 1985. Effects of structure on binding to the
2,3,7,8-TCDD receptor protein and AHH induction--

halogenated biphenyls. Enyirgn. Health Rgrapggt. 61:21-33.

SAS Institute Inc. 1982. mm- Cary.
NC: SAS Institute, Inc.

SAS Institute. Inc. 1987. W
WW- Cary. NC:
SAS Institute Inc.

Sawyer, T., and Safe, S. 1982. PCB isomers and congeners:
induction of aryl hydrocarbon hydroxylase and
ethoxyresorufin-o-deethylase enzyme activities in rat
hepatoma cells. T931991. Lgrr. 13:87-94.

Sawyer, T., and Safe, 8. 1985. In yirrg AHH Induction by
polychlorinated biphenyl and dibenzofuran mixtures:
Additive effects. gngngspngre 14:79-84.

Schwartz, T. R., and Smith, L. M. 1987. Determination of AHH-
active PCB isomers in PCB mixtures and environmental
samples. ACS Division of Environmental Chemistry Extended
Abstract. New Orleans, LA, September, 1987.

Schwartz, T. R., Stalling, D. L., and Rice, C. L. 1987. Are
polychlorinated biphenyl residues adequately described by
Aroclor mixture equivalents? Isomer-specific principal
components analysis of such residues in fish and turtles.

.Enxiran. Sci- leahnal- 21:72-76.

Smith, L. M. 1981. Carbon dispersed in glass fibers as an
adsorbent for contaminant enrichment and fractionation.

Anni. ghgm. 53:2152-2154.

Swackhamer, D. L., and Armstrong, D. E. 1988. Horizontal and
vertical distribution of PCBs in southern Lake Michigan
sediments and the effect of Waukegon Hargbor as a point

source. 1. Great Lakes Beg. 14(3):277-290.

Tanabe, 8., Hidaka, M., and Tatsukawa, R. 1983. PCBs and
chlorinated hydrocarbon pesticides in Antarctic
atmosphere and hydrosphere. Chenggnngrg 12:277-288.

61

Tanabe, 8., Kannan, N., Subramanian, A., Watanabe, S., and
Tatsukawa, R. 1987. Highly toxic coplanar PCBs:
Occurrence, source, persistency and toxic implications to
wildlife and humans. Enxirgn. 29113;. 47:147-163.

Zell, M., and Ballschmiter, K. 1980. Baseline studies of the
global pollution II. Global occurence of hexachorobenzene
(HCB) and polychlorocamphenes (Toxaphene)(PCC) in
biological samples. Frag. 1. Anal. Chan. 300:387-402.

If

3‘!

                    

TRTE UN V. L

VIIIW lli H 4| lililllulltlfllllliiwl
30 8 083

CTTEB 1

 

 

 

 

 

 

HICH?G?N s
IlMWllWlll
3129