-
extracted text (extracttext:extracted_text)
-
POLYCHLORINATED BIPHENYL CONCENTRATIONS IN ADULT CHINOOK
SALMON (Oncorhynchus tshawytscha) RETURNING TO COASTAL AND PUGET
SOUND HATCHERIES
by
Brian Missildine
A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
February 2005
This Thesis for the Master of Environmental Studies Degree
by
Brian Missildine
has been approved for
The Evergreen State College
by
r. Gerardo Chin-Leo
Member of the Faculty
Dr. John Perkins
Member of the Faculty
~
Fisheries Scientist
U. S. Fish and Wildlife Service
Date
ABSTRACT
Polychlorinated Biphenyl Concentrations in Adult Chinook Salmon (Oncorhynchus
tshawytscha) Returning to Coastal and Puget Sound Hatcheries.
Brian Missildine
Polychlorinated biphenyl (PCBs) concentrations were evaluated in the muscle tissue of 4year old Puget Sound and coastal hatchery Chinook salmon. Ten muscle tissue samples
per location were taken from two Puget Sound hatcheries and two Washington coast
hatcheries to determine PCB concentrations. Two trade name PCBs, Aroclor 1254 and
Aroclor 1260, were found. Aroclor 1254 was detected in all samples, while Aroclor 1260
was detected in 16 out of 40 samples. Generalized linear modeling (GLM) was used to
evaluate the influence of several variables on PCB concentrations. Twenty different
GLMs, representing multiple null hypotheses, were ranked using Akaike Information
Criterion (AIC). PCB concentrations were explained by region and lipids, followed by
region, and location (hatchery) and lipids. Region appears to be the variable that explains
most of the variation in PCB concentrations in Chinook salmon in the northwest. PCB
concentrations in Chinook salmon muscle tissue from Puget Sound hatcheries were
significantly greater (mean 49.26 µg/kg wet weight; standard deviation 40.55 µg/kg) than
those from coastal hatcheries (mean 17.41 µg/kg wet weight; standard deviation 6.8
µg/kg). This suggests that the primary source of PCBs observed in Puget Sound Chinook
salmon is contamination within Puget Sound. Four-227 gram (8oz) portions from Puget
Sound Chinook salmon and sixteen-227 gram portions from coastal Chinook salmon can
be consumed before the potential for adverse risk from PCB consumption becomes a
concern, based on EPA’s fish consumption guidelines and mean PCB concentrations
observed during this study. In addition, salmon carcass supplementation may need to be
reevaluated to determine if adding fish contaminated with PCBs to increase marine
derived nutrients is worth the environmental risk.
© Copyright 2005 by Brian Missildine
All rights reserved
TABLE OF CONTENTS
LIST OF FIGURES ........................................................................................................... iv
LIST OF TABLES...............................................................................................................v
ACKNOWLEDGMENTS ................................................................................................. vi
INTRODUCTION ...............................................................................................................1
Objective ..........................................................................................................................2
PCB LITERATURE REVIEW............................................................................................3
Physical and Chemical properties of PCBs .....................................................................5
Transportation mechanisms .............................................................................................9
Impacts to humans .........................................................................................................12
METHODS ........................................................................................................................15
Field Sampling ...............................................................................................................15
Laboratory Analysis.......................................................................................................19
Data Analysis .................................................................................................................20
RESULTS ..........................................................................................................................25
Factors influencing PCB concentrations........................................................................26
Evaluating Model Fit .....................................................................................................29
DISCUSSION ....................................................................................................................31
Management Implications..............................................................................................37
Human health implications ............................................................................................41
CONCLUSION..................................................................................................................43
Future Recommendations ..............................................................................................44
LITERATURE CITED ......................................................................................................45
APPENDIX 1. Combined lab results................................................................................53
LIST OF FIGURES
Figure 1. Structure and numbering pattern of PCB's in biphenyl ring system. ..................8
Figure 2. Sampling locations where Chinook salmon tissue was collected .....................17
Figure 3. Residual plots for the model REGION and LIPIDS (A), REGION (B),
LOCATION and LIPIDS (C), REGION and SEX (D), REGION, LENGTH and LIPIDS
(E), and REGION and LENGTH (F)…………………………………………………….30
Figure 4. Puget Sound and coastal Chinook salmon migration pattern............................35
Figure 5. Proper trimming preparation to remove fat from salmon before cooking.........43
iv
LIST OF TABLES
Table 1. The number of individual compounds in each of the different PCB categories...7
Table 2. Composition of Chlorinated biphenyls by Homolog............................................7
Table 3 Sample size (N), mean, standard deviation and range of total Aroclor
concentrations observed in Chinook salmon sampled at each of the hatcheries. Aroclor
concentrations were determined using ½ the detection limit for Aroclor1260…………..25
Table 4. Sample size (N), mean, standard deviation and range of percent lipids observed
in Chinook salmon sampled at each of the hatcheries .......................................................26
Table 5. Number of estimated parameters (k), log likelihood, AICc, scaled AICc, Akaike
weights, and the scaled Akaike weights for the generalized linear models evaluated for
determining the influence of different variables on PCB concentrations in Chinook
Salmon tissue (n = 40). Models with scaled AICc less than 3 have substantial support for
being the best model. Those with scaled AICc values greater than 3 and less than 7 have
moderate support. Those with scaled AICc values greater than 7 have no support for
being the best model of those evaluated (Burnham and Anderson 2002)……………….27
Table 6. Model averaged parameter estimates, standard errors, confidence limits for
parameters from the all-models generalized linear model evaluation of variables
influencing PCBs in adult chinook salmon........................................................................28
Table 7. Sample size (N), mean, standard deviation and range for the proportion of
Aroclor 1254 in the total PCB concentrations for each region ........................................31
Table 8. The recommended upper limit for the number of 8 oz. (227g) meals of Chinook
salmon which can be consumed per month from each hatchery before non-cancerous and
cancerous health concerns arise. Estimates are based on EPA’s recommendation using a
reference dose of 2.5 X 10-5 mg/kg per day………………………………………………32
v
ACKNOWLEDGMENTS
There are so many people to thank; I hope that I don’t leave anyone out. First and
foremost I would like to thank Doug Houck of King County for funding the laboratory
analysis (which can’t be cheap) and Fritz Grothkopp of the King County Environmental
Lab for making sure the samples were analyzed in a timely manner. I owe a big thank
you to Dave Zajac of the U.S. Fish and Wildlife Service for his invaluable help in the
field and aging fish. I would like to thank Roger Peters, Carrie Cook-Tabor and Joe
Polos in assisting me with the statistical design and Martin Liermann for reviewing the
statistical design. Thanks to Sandie O’Neil, Rich Eltrich, and Greg Lippart from the
Washington Department of Fish and Wildlife for showing me how to collect samples, and
John Sneva for verifying the age of the fish. I also owe thanks to Jay Davis of the U.S.
Fish and Wildlife Service who helped me understand PCBs and Howard Gearns for
assisting at the Makah National Fish Hatchery. I am gracious for the hospitality of the
staff from the Deschutes and Issaquah state fish hatcheries, Makah National Fish
Hatchery, and the Quinault Tribal hatchery for allowing me to intrude during their busiest
time. I want to thank the U.S. Fish and Wildlife Service for paying for sample supplies
and allowing me to collect samples on work time. I would also like to thank my thesis
committee Dr. Gerardo Chin-Leo, Dr. John Perkins, and Dr. Roger Peters. Lastly, thanks
to Peggy, Haleigh, and Rebecca for allowing me to continue my education at the sacrifice
of family time.
vi
INTRODUCTION
Chinook salmon (Oncorhynchus tshawytscha), also called king salmon, is distinguishable
from all other Pacific salmon by its large size. Chinook salmon have been a mainstay of
indigenous people of the Puget Sound and Washington coast for several thousand years
(Morishma and Henry 1999). Recently, studies have indicated that Puget Sound Chinook
salmon are contaminated with PCBs (PSAMP 2001; O’Neill et al. 1998). Human health
issues have been associated with exposure to PCBs from fish consumption (Jacobson and
Jacobson 1996; Korrick and Altshul 1998; EPA 2002). Humans have also been exposed
to PCBs through occupational exposure (Kilburn et al. 1989; Ojajarvi et al. 2000), and in
utero and breast milk (Jacobson and Jacobson 1996; Craan and Haines 1998). Although
cancerous and non-cancerous health hazards such as developmental neurotoxic effects,
reduced birth weights and immunotoxic effects have been associated with exposure to
PCBs (EPA 2002); the overall influence of PCBs on human health are mixed.
There are two potential sources of PCB contamination for Puget Sound Chinook salmon:
Puget Sound and the North Pacific Ocean. Present and past urban and industrial practices
have lead to environmental degradations in the Puget Sound region. Pollution and
contamination have become widespread in urban bays, estuaries, and rivers to the point it
is causing the decline of a variety of organisms within the Puget Sound (PSAMP 2001).
There are currently two Superfund sites in critical estuaries of the Puget Sound, the
Lower Duwamish Waterway in Seattle and Commencement Bay in Tacoma. Some
locations within Elliot Bay and Commencement Bay that were listed as Superfund sites
have since been remediated; however, there are still areas of concern. Furthermore, all of
1
these locations have been, and are currently subjected to heavy industrial uses that
commonly contributed to PCB contamination.
There is a potential that the North Pacific Ocean is a sink for contaminants (Bailey et al.
2000; McCain et al. 2000). There is speculation that contaminants are originating from
Asian countries and are being transferred by ocean and air currents into the North Pacific
Ocean where conditions are conducive for PCBs to settle. The north-migrating Chinook
salmon are potentially being exposed either through food sources or the water during
their migration through the North Pacific Ocean.
Objective
The primary objective of this research was to determine if PCB concentrations in
Chinook salmon differed between coastal and Puget Sound stocks. This information will
be useful for determining if the source of PCBs is occurring in the Puget Sound or the
North Pacific Ocean. The secondary objective was to determine if other variables such as
sex, weight or percent lipids influence PCB concentrations in Chinook salmon. These
objectives were met by testing multiple hypotheses for my thesis including: There is no
difference in PCB levels in Puget Sound versus coastal Chinook salmon; there is no
difference in PCB levels between sexes; and, PCB concentrations are not related to size
of fish. The results of this research will also be used to determine if PCB concentrations
in Chinook salmon returning to Washington are high enough to pose a risk to human
health based on EPA fish advisory guidelines.
2
PCB LITERATURE REVIEW
PCBs are one of the most touted chemical classes of concern in the Puget Sound. PCBs
are distributed by oceanic and atmospheric currents (Kanan et al. 1989; Bailey et al.
2000; Jaffe et al. 1999) and can be found in even the most remote locations throughout
the world (Hidaka et al. 1983). Some of the more common uses of PCBs were in
electrical transformers used in industrial and residential utilities, fire retardants, paint
additives, and immersion oils. Monsanto Corporation was the major manufacturer of
PCBs in the United States from the 1930's until 1976 when PCB use was banned (Davis
1993). Although PCBs are no longer produced in the United States, they are still used in
developing countries. PCB concentrations in muscle tissue of fishes have declined at
long-term monitoring stations in the Puget Sound since their use was banned in the U. S.
(PSAMP 2001). However, PCB concentrations appear to be increasing in pinnipeds and
cetaceans around the Northwest (PSAMP 2001).
Current research on PCB concentrations in Chinook salmon in the Pacific Northwest has
been limited to Puget Sound stocks (Arkoosk et al. 1998 a and b; O’Neill et al. 1998;
Stein et al. 1995). O’Neill et al. (1998) found that average concentrations in adult
Chinook salmon in the Puget Sound marine environment were 74.2 µg/kg wet weight.
O’Neill et al. (1998) estimated that only 1.1 percent of the adult Chinook salmon PCB
burden came from exposure as juveniles; the other 99 percent was accumulated during
saltwater migration. It is unclear whether Chinook salmon are accumulating PCBs in the
Puget Sound or throughout their migratory pathways in the Pacific Ocean.
3
Relatively little research has been conducted on exposure of juvenile Chinook salmon to
contaminants in estuaries or to determine if exposure could result in increased mortality
rates during their ocean life phase (Arkoosh et al. 1998b). Factors that negatively affect
the fitness of juvenile Chinook salmon may not only increase early life mortality, but
could create biological responses in the adult that may inhibit growth, impair locomotion,
or reduce fecundity.
Existing research suggests that juvenile Chinook salmon exposed to PCBs experience
adverse biological effects. Arkoosh et al. (1998a) found mean PCB concentration levels
in juvenile Chinook salmon ranged from 37 ng/g wet weight from the Kalama Creek
Hatchery (non-urban), to 270 ng/g wet weight in the Duwamish Waterway (urban), in
1993. Arkoosh et al. (1998a) also found juvenile Chinook salmon sampled from the
Duwamish Waterway were more susceptible to mortality induced by the marine pathogen
Vibrio anguillarum after four days than fish taken from a hatchery in a nearby river basin
(the control group). Vibrio anguillarum is a common bacterial pathogen of marine and
brackish water fishes and has caused severe losses in marine and estuarine aquaculture
(Post 1987). In addition, Arkoosh et al. (1998a) found an increase in mortality in juvenile
Chinook salmon that had greater levels of PCBs than the test group three months after the
fish were removed from their prospective estuary. This suggests that Chinook salmon
exposed to PCB contamination as juveniles would remain immunosuppressed, potentially
throughout their lives, even after they left the estuary. Stein et al. (1995) found mean
concentration levels of PCBs in juvenile Chinook salmon ranged from 22 ng/g wet
weight in the Nisqually River (non urban), to 300 ng/g wet weight in the Duwamish
4
Waterway (urban), in 1989. Stein et al. (1995) found that juvenile Chinook salmon
exposed to PCBs had an increased activity of the enzyme P450 that plays a central role in
the metabolism of toxins. Stein et al. (1995) also found higher levels of DNA damage in
juvenile Chinook salmon from urban estuaries than juvenile Chinook salmon from nonurban estuaries.
PCB mobility within the Chinook salmon is influenced by the amount of lipids within the
fish; hence, the higher the lipid content, the more the chemical is sequestered within the
fatty muscle reserves and is not readily available to the organs of the fish (Meador 2000).
Salmonid lipid content varies during their life cycle with the low points occurring during
the fry, smolt, and spawning stage, increasing the probability of the toxin mobilizing into
critical organs during crucial points in their life history. Jorgensen et al. (1999) found a
10-fold increase in PCB levels in the liver of an arctic char that was not fed for a
significant time period (starved). Jorgensen et al. (1999) observed that the decrease in
total body lipids due to starvation caused the PCB to mobilize to other lipid containing
organs such as the liver, kidney, and brain.
Physical and Chemical properties of PCBs
PCB is the generic name given to the entire group or a subset of the 209 different
chemical compounds having the formula of C12H10-nCln, where n =1-10 (Erickson 1997).
These 209 compounds occur in 11 groups or “homologs” based on the number of
chlorine atoms in the molecule. In turn, each homolog has from 1 to 46 isomers, based
5
on the position of the chlorine atoms (Tables 1 and 2) (Figure 1) (Erickson 1997).
PCBs are also defined as non-planar and planar. Non-planar PCBs are identified as
having at least one chlorine atom on the 2,2’ or 6,6’ position(s) on the biphenyl ring.
PCBs having the non-planar configuration are less stable than planar PCBs because the
chlorine atoms are too large to fit adjacent to each other. This configuration causes the
bond between the biphenyl rings to twist, causing the molecule to form at right angles
which weakens the bond. Planar PCBs form in the same plane, which creates a very
strong bond between the biphenyl rings. Planar PCBs are not easily broken down in the
environment due to the strength of their bond. Planar PCBs are less common than nonplanar PCBs with only 20 of the 209 congener’s able form the planar configuration.
PCBs used in manufacturing were a clear viscous liquid typically used for insulation,
especially in high voltage equipment such as capacitors and transformers. PCBs have
low water solubility and low vapor pressure (Erickson 1997). However, physical
properties such as boiling point, melting point, vapor pressure, bioconcentration factors in
fish and evaporation rate change throughout the homologous series. PCBs are lipophilic
(fat loving) and are soluble in organic solvents and in biological lipids.
Only half of the total 209 PCB congeners account for most of the environmental
contamination. Even fewer are prevalent and toxic. According to McFarland and Clarke
(1989), the number of congeners of concern is reduced to 36 if potential toxicity,
environmental prevalence, and relative abundance are used as criteria. Approximately 25
6
Table 1. The number of individual compounds in each of the different PCB categories
(Adopted from Erickson 1997).
Category
Congener
Homolog
Isomers/homolog
No. of individual compounds
209
11
1-46
Table 2. Composition of Chlorinated biphenyls by Homolog (Adopted from Erickson
1997).
Homologs
C12H9Cl
C12H8Cl2
C12H7Cl3
C12H6Cl4
C12H5Cl5
C12H4Cl6
C12H3Cl7
C12H2Cl8
C12HCl9
C12Cl10
Chlorine %, by
weight
19
32
41
49
54
59
63
66
69
71
No. of isomers
3
12
24
42
46
42
24
12
3
1
of the remaining 36 congeners account for 50-75 percent of the total PCBs in tissue
samples of fish, invertebrates, birds, and mammals.
PCBs in the United States are typically known by their trade name Aroclor. They are
further identified by a specific number. For example, Aroclor 1254 contains 12 carbon
atoms bound together within the biphenyl rings and contains 54 percent chlorine by
weight. Aroclor analysis is the method currently recognized by EPA. The Aroclor
analysis has been the most common method for measuring total PCBs, in part due to cost,
equipment availability, time, and technology. Congener specific data is much more
7
CLx
2
3
2’
1
4
5
6
3’
1’
CLy
4’
6’
Figure 1. Structure and numbering pattern of PCB's in biphenyl ring system.
5’
8
expensive and there is no single standard methodology for conducting congener specific
analysis. However, congener specific analysis is more detailed than Aroclor analysis and
may be more important in determining toxicity.
Transportation mechanisms
One of the main pathways PCBs entered into the environment was through disposal. No
literature searched indicated, or even attempted to predict the amount of PCBs that were
dumped into oceans, lakes and rivers. PCBs bind tightly to organic matter once dumped
into the environment. PCBs favor moist organic solids and will avoid the aqueous
portion of sediment. PCBs tend to mobilize toward the organic component where
sorption takes place by the carbon components of the sediment (Paya-Perez et al. 1991).
Movement (mobility) within the sediment can take years depending on the soil and PCB
characteristics (Girvin et al. 1993). PCBs, especially planar PCBs, are extremely stable
and may remain within the environment for many years and can be readily resuspended
during dredging, pile driving or other activities that disturb sediments (Eisler 1986).
Larsson and Okla (1987) theorized that oceanic suspended particles may provide
transportation mechanism for PCBs. Phytoplankton can readily absorb PCBs into their
lipid rich membranes and then into cellular material within the plankton (Broman et al.
1992). These lower order consumers store PCBs, are consumed by their predators where
they are again stored until eaten by yet another predator. EPA (1999) noted that PCB
concentrations could be 2,000 to more than a million times higher in an organism than the
9
concentrations found in surrounding waters, with the highest concentrations found at the
top of the food chain. The rate of PCB bioaccumulation varies among environmental
constituents. Factors such as age, species, gender, and lipid contents influence
bioaccumulation within organisms.
The circulatory system provides the primary transportation pathway for PCBs within an
organism (Duinkers et al. 1989). PCBs transported through the circulatory system are
dispersed into fatty tissue and critical organs which are lipid rich. Goldstein and Safe
(1989) indicated that one characteristic effect of planar PCBs is the induction of the
cytochrome P450 enzymes. This class of enzymes is primarily found in the liver, and to
a certain extent, the extrahepatic tissues in the lungs, kidney, intestines, spleen, and
testes. The P450 enzymes and its monooxygenase activities are responsible for the
metabolism of carcinogens, drugs, pesticides and other foreign compounds.
PCBs accumulate in the fatty tissue of organisms, including Chinook salmon, due to their
lipophilic nature. The inability of Chinook salmon to properly metabolize toxic
substances could lead to an increase in mortality in adult Chinook salmon, especially
during spawning migration. Adult salmonids stop feeding during their spawning
migration and depend on their fat reserves as a fuel source. The amount of fat reserve
used increases with the length of time without feeding. This results in PCBs moving to
other fatty tissues such as the internal organs and gametes (Hendry 1998). Jorgensen et
al. (1999) found an increase in PCB concentrations in the liver and kidneys of salmonids
that were starved. Miller (1993) found that concentrations of organochlorine compounds,
10
such as PCBs, found in the muscle tissue of adult Chinook salmon and lake trout
(Salvelinus namaycush) were significantly correlated to concentrations found in the eggs.
Miller’s research may indicate that PCBs are readily available in the gametes during all
the life stages of salmonids.
PCB transport in mammals is similar to fish with the main pathway through ingestion
then through the circulatory system to fat cells. However, not all PCBs ingested are
retained. This explains the difficulty of determining the impacts of PCB ingestion by
people since they are getting unknown and different concentrations from eating the same
contaminated food. Recent studies suggest that not all PCBs are absorbed once ingested
into the body. Juan et al. (2002) studied volunteers who were consuming a diet that had
some PCB contamination and found that congener type and body fat index influenced
PCB absorption. Fourteen of the congeners were absorbed by all volunteers. Other
congeners were absorbed in some volunteers but excreted by others. Volunteers tended
to excrete congeners with higher chlorination more often while those with lower
chlorination were more readily absorbed within the body. Volunteers with a higher body
fat index tended to absorb more of the congeners studied while volunteers with a lower
body fat index tended to excrete the congeners. This, along with exposures to other
potential toxic substances, makes it difficult to determine PCB impacts on humans.
PCBs can be transferred maternally in humans. Lanting (1999) found newborns that
were strictly breast fed for six weeks contained 4½ times the amounts of PCBs than
infants fed formula. Patadin et al. (1999) discovered that 12 to 14 percent of long-term
11
dietary exposure to PCBs is partly due to breast feeding. Korrik and Altshul (1998)
found that breast milk from four women contained PCB levels that were significantly
higher than the rest of the study group of 122 women. The greater PCB concentrations in
the four women were attributed mainly to fish consumption, and one was attributed to
occupational exposure. All four women also lived near a contaminated site. The infants
from these four women were delivered full term and healthy.
Impacts to humans
PCBs are ubiquitous and have been associated with human health issues ranging from
neurological impacts to cancer. Most humans in industrialized countries contain
measurable levels of PCBs. Contamination has occurred mainly from the consumption of
fatty sports fish (Jacobson and Jacobson 1996; Jensen 1984); however, occupational
exposure (Kilburn et al. 1989), and maternal transfer have also occurred.
The scientific community currently debates the human health impacts of PCB exposure.
Linking human health issues to PCB exposure is very difficult because humans are
subjected to many other environmental factors such as nutrition, pesticides, chemicals
and other pollutants, in addition to socio-economic stressors. Furthermore, results are
often difficult to interpret because of study design limitations or inconclusive results for
the most extreme health concerns (Shirai 1995). However, there are a numerous studies
that associate PCB exposure to human health issues. Jacobson and Jacobson (1996)
found an association between elevated PCB concentrations in children and poor verbal IQ
12
scores. Osius et al. (1999) found an association between PCB blood serum levels and
thyroid hormone levels. This association may be important because some toxins such as
PCBs share a similar molecular structure with thyroid hormones and may interfere with
endocrine function by imitating natural hormones (McKinney and Waller. 1998; Brouwer
et al. 1998). In addition, EPA (2002) has identified reduced birth weight, learning
disabilities, immunosuppression, and other non-cancerous heath effects to be associated
with exposure to PCBs. Although it is difficult to determine the influence of chronic
exposure to human health, case studies associate negative health effect from high
episodic exposure to PCB concentrations.
Ingestion of food contaminated with high concentrations of PCBs has been linked to
human health impacts. The most notorious exposure of PCBs occurred in Yucheng,
Taiwan in 1979 where approximately 2,000 people were exposed to rice oil contaminated
with PCBs during manufacturing. This exposure resulted in increased abnormal
menstrual bleeding and still births in exposed women compared to unexposed women
(Yu et al. 2000). Guo et al. (1999) also found that there was an increase in chloracne,
hyperkeratosis, abnormal nails, gum swelling and gum pigmentation, as well as an
increase in mortality from nonmalignant liver disease in those exposed at Yucheng
compared to an unexposed control unit. Similar effects were observed in a population
exposed to PCB contaminated rice oil during a similar incident in Yusho, Japan, 11 years
earlier (Guo et al. 1999).
13
The Michigan Maternal Infant Cohort Study (Fein et al. 1984; Jacobson et al. 1985,
1990a, b) found developmental deficiencies in the offspring of women who consumed
PCB contaminated fish from Lake Michigan in comparison to women who did not eat
fish from Lake Michigan. Developmental deficiencies such as head circumference,
gestational period, and birth-weight, as well as cognitive deficiencies were statistically
significant in the study. Most of these cognitive and developmental deficiencies were
still present up to age 4. These findings were subsequently subjected to scrutiny due to
the sampling and testing methodology. However, Swain (1991), using the epidemiologic
criteria of Susser (1986), found that there was strong epidemiological evidence to support
the initial findings.
Individuals who have worked in electrical component manufacturing such as capacitor,
light ballast, or transformer assembly in the past have likely been exposed to PCBs.
However, other occupational exposure has also occurred. Kilburn et al. (1989)
discovered 14 firefighters exposed to PCB fumes while fighting a fire in a transformer
room (the transformer contained PCBs) showed symptoms of extreme fatigue, headaches,
muscle weakness, and aching joints two days to three months after the fire. Several of
the firefighters had memory loss, impaired concentration, irritability and other
psychological impairments.
PCBs have also been implicated in breast cancer; however, the results are inconclusive.
Aronson et al. (2000) found a correlation between PCB concentrations in breast tissue
and cancer. Moysich et al. (1999) found an association between PCB concentrations and
14
cytochrome P4501A1 polymorphism and breast cancer risk. However, several authors
found no significant differences in blood serum PCB levels and breast cancer (Stellman
et al. 2000; Wolff et al. 2000; Laden et al. 2001). It is important to note that
methodologies differed between the studies that found correlations and those that didn’t.
Most of the studies that found no association between PCBs and breast cancer analyzed
PCBs in the blood serum while those studies that found an association analyzed adipose
tissue or breast tissue. Lipid concentrations in blood serum are much lower than lipid
levels in adipose or breast tissue. Aronson et al. (2000) suggests that the concentrations
measured in serum are not representative of the concentrations found in adipose tissue.
Aronson et al. (2000) also notes that PCB concentrations in serum and PCB
concentrations in adipose tissue vary. Furthermore, measurements in breast adipose
tissue may provide a better means for determining if PCBs cause breast cancer due to
tissues proximity and the cancer (Aronson et al. 2000).
Overall, PCBs probably impact human health; however, the exposure level impacting
human health and the human health impact are uncertain. Other variables such as
genetics and environment make it difficult to directly implicate PCBs to human health.
METHODS
Field Sampling
Salmon tissue samples for PCB analysis were collected from four western Washington
hatcheries, which represented two separate regions of western Washington; coastal and
Puget Sound. The coastal region was represented by salmon collected at the Quinault
15
Lake Tribal Hatchery (Quinault) and the Makah National Fish Hatchery (Makah). The
Puget Sound region was represented by salmon collected at Issaquah Creek (Issaquah)
and Deschutes River State Fish Hatcheries (Deschutes) (Figure 2). Tissue samples used
for PCB analysis were collected from fish that had been spawned (eggs removed and
fertilized) by hatchery personnel on the dates listed below. Samples were collected on
September 26 and October 8, 2003, at Deschutes; October 7 and 13, 2003 at Issaquah;
October 28, 2003, at Makah; and November 5, 2003, at Quinault.
Tissue samples were collected from ten 4-year-old ocean type (salmon that leave fresh
water for salt water as sub-yearlings) Chinook salmon at each hatchery for a total of forty
tissue samples during the spawning season. Samples were collected from five males and
five females from each hatchery. All fish were measured for fork length to the nearest
centimeter and weight to the nearest gram. All females were weighed without eggs, since
they had been spawned by hatchery staff. Scale samples were collected and read on-site
to verify that the fish were 4-year-old ocean type Chinook salmon. This was conducted
by placing the scales on a slide and viewing them with a microfiche reader (typically
used for reading film). Determining the age of the fish by reading the scale is similar to
counting the rings of a tree stump. Scale ring patterns can be close together or far apart,
depending on growth rate. Freshwater and saltwater growth can also be determined by
the ring spacing. In addition, the scales were taken to the Washington Department of
Fish and Wildlife for age verification.
16
Figure 2. Sampling locations where Chinook salmon tissue was collected.
17
Tissue samples were collected using sterilized stainless steel scalpels and forceps. Tools
were rinsed and stored in isopropyl alcohol (reagent-residue analysis grade) between
samples. Instruments were cleaned with soap and water and isopropyl alcohol after
sampling was complete. Two sets of instruments were used for sampling to minimize
any chance of contamination. One set was used for skin removal (outside set) and the
other set was for internal tissue collection (inside set).
The skin was removed by first making an incision just aft of the fish’s head. The incision
continued along the back of the fish just below the dorsal fin, to just in front of the tail.
The incision then continued down and back along the ventral side just above the pelvic
fin and back up to the origin of the incision using the outside scalpel. The skin was
peeled from the flesh and was discarded using the outside forceps.
Tissue samples were collected from each fish after the skin was removed. An incision
was made using the sterile inside scalpel at a minimum of ½ inch inside the cut that was
left by the skin removal procedure to ensure no outside to inside contamination occurred.
Tissue samples were collected by removing a fillet from the side of the fish. The tissue
was weighed to insure that a minimum of 250 grams of tissue was collected and was then
placed in a labeled, certified EPA clean, glass jar and stored at -0°C or below until
analyzed for PCBs. Isopropyl alcohol rinsed aluminum foil was used on the scale and
was changed between each tissue sample to eliminate contamination between samples.
18
Laboratory Analysis
King County Environmental Laboratory analyzed tissue samples for PCB concentrations
and percent lipids. King County Environmental Laboratory standard operating
procedures were used to analyze the samples. All of the individual samples were
homogenized using a commercial bar mixer. A sample was considered homogenized
when it was blended to a uniform consistency as determined by the analyst.
Homogenized samples were then extracted from the mixer and placed into labeled
specimen cups. Homogonized samples were mixed with methylene chloride and acetone
for extraction. Samples were then analyzed for PCB Aroclor 1016, 1221, 1232, 1242,
1248, 1254, and 1260 using a gas chromatograph equipped with electron capture
detectors (GC-ECD). A 1 to 2 µL aliquot was introduced into the GC via an
autosampler. A temperature program was used to separate the compounds as they moved
through two dissimilar phased capillary columns used to retard the elution of the
individual compounds. The compounds entered the separate ECDs as discrete
compounds. The detector response was transferred to a data system where the voltages
were charted and analyzed. Compounds were identified by their retention time matches
with standards and their presence on both the primary and confirmatory columns.
Quantitation was accomplished by integrating each compound to baseline and comparing
their responses with the responses of standards of known concentrations.
Percent lipid in each tissue sample was estimated using the following lipid extraction
method. A 15 to 30 gram portion of tissue was mixed to a sandy texture with anhydrous
sodium sulfate. A 1:1 mixture of methylene chloride (MeCl2)/acetone was added. The
19
samples were sonicated for four minutes with approximately 100 mLs of 1:1
(MeCl2)/acetone. The resulting extract was transferred to a specimen cup and dried. The
residue (the lipids) that was left over was weighed and percent lipids were calculated by
dividing the residue weight by the sample weight.
Data Analysis
PCB concentrations reported by the King County Environmental Laboratory were
reported in dry weight. Most PCB data in the literature is reported as wet weight. The
data was converted to wet weight using the formula:
PCB wet weight = (µg/kg PCB concentrations dry weight)*(percent total solids/100)
The method used to determine PCB concentrations in these samples has a minimum
detection limit ranging between 16µg/kg to 22 µg/kg. The detection limit variation was
due to the difference in the percent total solids between the samples. Concentrations
falling below these detection limits may range from zero to just below the detection limit.
Concentrations falling below the detection limit were given a concentration of ½ the
detection limit. This was conducted to minimize the under estimation of PCB
concentrations that would occur if all samples falling below the detection limit had been
given a value of zero. The concentrations of all the Aroclor detected in the sample were
summed to calculate total PCBs.
20
The influence of region, location (hatchery), sex, fish length, and percent lipids on PCB
concentrations were evaluated using Generalized Linear Modeling (GLM). GLM was
used since it can model categorical (region, location, sex) and continuous variables
(length, percent lipids) simultaneously. GLM is a predictor that is based on combinations
of measurements that are free to vary in response to other variables (Dobson 2002).
GLM is based on the formula y=B0+B1X1+B2X2…BnXn, where y is the predicted (or
probable) outcome (dependent variable), Bo is the y intercept, Bx is the slope of the
regression, which indicates the change in the mean of the probability distribution of Y per
unit increase in X. Xi represents the independent variables that were evaluated. All data
used in GLM was log transformed using the formula LOG(y+1).
The multiple hypotheses testing method proposed by Anderson et al. (2000) was used to
determine which variables had the greatest influence on PCB concentrations in the
salmon tissue evaluated. This method is based on Kullback-Leibler (1951) Information
Theory, Akaike Information Criteria, along with likelihood-based inference for data
analysis (Anderson et al. 2000). This method relates the structure of relationships,
estimates of model parameters and components of variance in order to make appropriate
inferences about the data while separating out the noise (Anderson et al. 2000).
21
The following 20 models were evaluated to determine which variables influenced PCB
concentrations:
a) y = B0+B1(SEX)
b) y = B0+B1(SEX)+B2(PERCENT LIPIDS)
c) y = B0+B1 (PERCENT LIPIDS)
d) y = B0+B1 (PERCENT LIPIDS) + B2 (LOCATION)
e) y = B0+B1 (REGION)
f) y = B0+B1 (REGION) + B2 (PERCENT LIPIDS)
g) y = B0+B1 (REGION) + B2 (SEX)
h) y = B0+B1 (REGION) + B2 (LENGTH)
i) y = B0+B1 (LENGTH)
j) y = B0+B1 (LOCATION)
k) y = B0+B1 (LOCATION) + B2 (REGION)
l) y = B0+B1 (LOCATION) + B2 (LENGTH)
m) y = B0+B1 (LOCATION) + B2 (SEX)
n) y = B0+B1 (LOCATION) + B2 (PERCENT LIPIDS)
o) y = B0+B1 (LENGTH) + B2 (SEX)
p) y = B0+B1 (REGION) + B2 (LENGTH) + B3(LIPID)
q) y = B0+B1 (REGION) + B2 (LENGTH) + B3(LIPID) + B4 (LENGTH * LIPID)
r) y = B0+B1 (REGION) + B2 (LENGTH) + B3(LIPID) + B4 (LENGTH * REGION)
s) y = B0+B1 (REGION) + B2 (LENGTH) + B3(LIPID) + B4 (LIPID * REGION)
t) y = B0+B1 (LENGTH) + B2 (LIPID) + B3 (LENGTH * LIPID)
22
The number of models evaluated was limited due to sample size (n=40). There needs to
be a minimum of 10 samples per variable when using GLM (Harrell 2001). Data
analyses were conducted using PROC GENMOD (version 8.02) with gamma distribution
and log link functions using SAS statistical package (SAS Institute, Inc. 1999).
The resulting models were ranked using Akaike’s information criterion (AIC) (Burnham
and Anderson 2002). The AICC (AICc=corrected for small sample size) and scaled
Akaike weights (wi) were calculated for each model. Scaled Akaike weights range from
0 to 1, with values closer to one representing models with more support for being the best
model. For example, a model with a scaled Akaike weight of 0.85 has more support for
being the best model in the set evaluated than a model with a scaled Akaike weight of
0.60. A set of models in which no model has a scaled Akaike weight of 0.9 or greater
suggests a high level of model uncertainty (Burnham and Anderson 2002). Parameter
estimates were model averaged because no models possessed a scaled Akaike weight of
0.90 or greater. This process weights the predicted value by the scaled Akaike weight
(Burnham and Anderson 2002). The influence of variables on PCB concentrations was
determined by using model averaged 95 percent confidence intervals developed for each
parameter estimate. Important differences between the variables existed if the 95 percent
confidence intervals did not overlap for categorical variables and if the confidence
interval did not include zero for continuous variables.
Model fit was evaluated using deviance residual plots. Plots were visually analyzed for
patterns which may suggest that assumptions were violated (e.g. Increasing variance with
23
increases in the independent variable). Data that was precise around the y axis but was
scattered was considered a good fit. Data that was spurious or showed specific patterns
was either over fit or under fit by the model, or the assumptions (e.g. equality of
variances) were violated.
The proportion of Aroclor 1254 was compared to the total PCB load to determine if the
source of PCBs differed by region. This was conducted by dividing Aroclor 1254 by the
total PCB concentration. The data was non-normal (P<0.01) and would not normalize
using the arcsine transformation. Therefore, the data was analyzed using the KruskalWallis test.
In addition to evaluating factors that may influence PCB concentrations in Chinook
salmon, I also determined the amount of salmon individuals could consume without
increasing health risks. This was determined by using the Environmental Protection
Agency (EPA 1999) criterion, which is based on both non-cancer health endpoints and
cancer health endpoints. Mean PCB concentrations for each hatchery were converted
from µg/kg (parts per billion or ppb) to mg/kg (ppm) by dividing by 1000. The mean
PCB concentrations were then divided by two based on Anderson et al. (1993) who states
that a 50 percent reduction in PCB concentrations is reasonable based on proper
preparation and cooking. Half the mean PCB concentrations were compared to EPAs
recommended fish consumption limits (EPA 1999) which were based on a 227 gram (8
oz) portion over a month time frame for both non-cancerous and cancerous health
endpoints.
24
RESULTS
Aroclor 1254 and 1260 were the only Aroclors detected. Aroclor 1254 was detected in
all 40 samples, while Aroclor 1260 was detected in 16 out of 40 samples. Mean Aroclor
concentrations were similar in salmon tissue samples from the Deschutes and Issaquah
hatcheries, which were both greater than those from the coastal samples. Mean Aroclor
concentrations at Makah were slightly greater than the concentration levels at Quinault
(Table 3 and Table 6).
Tissue samples from adult Chinook salmon collected at Deschutes had less lipid
concentrations than samples collected at all other locations. Tissue samples collected at
Quinault had more lipid concentrations than all other locations. Overall, Chinook salmon
tissue samples collected from Puget Sound had less lipids than those collected from the
coast (Table 4).
Table 3. Sample size (N), mean, standard deviation and range of total Aroclor
concentrations observed in Chinook salmon sampled at each of the hatcheries. Aroclor
concentrations were determined using ½ the detection limit for Aroclor 12601,2,3.
Hatchery
N
Deschutes
Issaquah
10
10
Makah
Quinault
10
10
Mean µg/kg Std deviation
Puget Sound
48.67
38.57
49.85
44.53
Coastal
18.97
5.14
15.86
8.10
Range µg/kg
14-135
20-170
10-27
9-38
1
Data was not lipid normalized.
Raw data is contained in Appendix 1.
3
Data reported as wet weight.
2
25
Table 4. Sample size (N), mean, standard deviation and range of percent lipids observed
in Chinook salmon sampled at each of the hatcheries.
Hatchery
N
Deschutes
Issaquah
10
10
Makah
Quinault
10
10
Mean %
Std deviation
Puget Sound
0.097
0.037
0.57
0.30
Coastal
1.5
0.54
1.8
1.07
Range %
0.4-1.49
0.27-1.28
0.86-2.45
0.76-3.81
Factors influencing PCB concentrations
Six of 20 models evaluated had substantial support for being the best model (Table 5).
No single model had enough support to be considered the single best model to describe
PCB concentrations in Chinook salmon. The models containing REGION and LIPIDS;
REGION; LOCATION and LIPIDS; REGION and SEX; REGION, LENGTH AND
LIPIDS; and REGION and LENGTH, were the only models to have substantial support
(i.e., scaled AICc <3) as the best model of those evaluated. The models with the
variables REGION LENGTH/LENGTH*LIPIDS; LOCATION and LENGTH;
LOCATION; REGION LENGTH/LIPID*REGION; LOCATION and SEX; and
LOCATION and REGION; had moderate support as the best model in the set evaluated.
The remaining models had very little support as the best model in the set evaluated.
The parameters estimates were model averaged using all models since no one model had
enough support for being the best model. Region and location were the only two
variables that influenced PCB concentrations in Chinook salmon (Table 6). PCB
concentrations were greater in Chinook salmon tissue from the Puget Sound than from
26
Table 5. Number of estimated parameters (k), log likelihood, AICc, scaled AICc, Akaike weights, and the scaled Akaike weights for
the generalized linear models evaluated for determining the influence of different variables on PCB concentrations in Chinook Salmon
tissue (n = 40). Models with scaled AICc less than 3 have substantial support for being the best model. Those with scaled AICc
values greater than 3 and less than 7 have moderate support. Those with scaled AICc values greater than 7 have no support for being
the best model of those evaluated (Burnham and Anderson 2002).
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
Intercept
REGION
REGION
LOCATION
REGION
REGION
REGION
REGION
LOCATION
LOCATION
REGION
REGION
LOCATION
LOCATION
LENGTH
LENGTH
LIPIDS
LENGTH
LIPIDS
LENGTH
SEX
Model Variables
LIPIDS
LIPIDS
SEX
LENGTH
LENGTH
LENGTH
LENGTH
LENGTH
LENGTH
SEX
REGION
LIPIDS
SEX
SEX
LIPIDS
LIPIDS
LENTH*LIPIDS
LIPID*REGION
LENGTH*REGION
LENGTH*LIPIDS
k
3
2
5
3
4
3
5
5
4
5
5
5
5
4
2
2
3
3
3
2
loglikelihood
10.53
8.88
12.21
9.66
10.59
9.13
11.48
10.96
9.63
10.93
10.82
10.46
9.63
0.37
-2.54
-2.67
-1.84
-1.93
-2.13
-3.40
AICc
-14.38
-13.44
-12.66
-12.65
-12.05
-11.60
-11.19
-10.15
-10.12
-10.09
-9.87
-9.16
-7.50
8.41
9.40
9.67
10.35
10.53
10.92
11.12
Scale
AICc
0.00
0.94
1.73
1.73
2.34
2.78
3.20
4.23
4.26
4.29
4.52
5.23
6.88
22.79
23.78
24.05
24.73
24.92
25.31
25.51
Akaike
weights
1.00
0.62
0.42
0.42
0.31
0.25
0.20
0.12
0.12
0.12
0.10
0.07
0.03
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Scaled
Akaike
weights
0.26
0.16
0.11
0.11
0.08
0.07
0.05
0.03
0.03
0.03
0.03
0.02
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
27
Table 6. Model averaged parameter estimates, standard errors, confidence limits for parameters from the all-models generalized
linear model evaluation of variables influencing PCBs in adult Chinook salmon.
Estimated
Model Variable
REGION
REGION
LOCATION
LOCATION
LOCATION
LOCATION
REGION LENGTH LIPID
REGION LENGTH LIPID
LENGTH LIPIDS
LIPIDS
LENGTH
SEX
SEX
1
Reference variables
Class
PUGET SOUND
COAST1
DESCHUTES
ISSAQUAH
MAKAH
QUINAULT1
LENGTH*REGION
LIPID*REGION
LENGTH*LIPIDS
FEMALE
MALE1
Parameter
Estimate
0.27
0.34
0.36
0.10
<0.01
-0.07
<0.01
0.03
<0.01
-0.05
-
Standard
Error
0.07
0.06
0.07
0.06
<0.01
0.08
<0.01
0.06
<0.01
0.04
-
95 %Confidence
Interval
Lower
Upper
0.13
0.42
0.22
0.47
0.23
0.49
-0.02
0.22
<-0.01
0.01
-0.24
0.09
<-0.01
<0.01
-0.09
0.14
<-0.01
<0.01
-0.14
0.13
-
Effect
Yes
No
Yes
Yes
No
No
No
No
No
No
No
No
No
28
the coast. PCB concentrations in Chinook salmon tissue from Deschutes and Issaquah
were greater than Quinault and Makah, but were not different from each other.
Evaluating Model Fit
Residual plots for the top six models are shown in Figure 3. These plots show that the
variation in PCB concentrations was greater for the Puget Sound than the coast. Thus,
estimates of mean PCB concentrations for the coast are likely precise, while estimated for
the Puget Sound are likely imprecise. For the model REGION and LIPIDS (Figure 3A),
LOCATION and LIPIDS (Figure 3C), and REGION LENGTH and LIPIDS (Figure 3E),
the data fit the model fairly well up to a predicted value of 1.5. However, above a
predicted value of 1.5, the variance increases as the predicted data increased indicating a
poor fit and greater variability of data from the Puget Sound. However, these models fit
the data substantially better than the other models with substantial support. The residual
plots for the model REGION and SEX (Figure 3D) shows that much of the increase in
variability in PCB concentrations in Chinook salmon from Puget Sound samples was due
to males. There was a slight increase in variance from female Chinook salmon from the
Puget Sound samples relative to coastal samples, but a large increase in variance for male
Chinook salmon from the Puget Sound samples. Male and female Chinook salmon from
coastal samples had similar variation in PCB concentrations. The models with the
variable LOCATION and LIPIDS (Figure 3C) in the model were a good fit for the
coastal data but showed greater variability for Puget Sound data. However, the fit was
consistent for the hatcheries in each region. Similarly, the model REGION and LIPIDS
Figure 3A), and REGION LENGTH and LIPIDS (Figure 3E) fit the coast but showed
greater variability for Puget Sound.
29
Figure 3. Residual plots for the model REGION and LIPIDS (A), REGION (B),
LOCATION and LIPIDS (C), REGION and SEX (D), REGION, LENGTH and LIPIDS
(E), and REGION and LENGTH (F).
30
Aroclor 1254 is more prominent in Chinook salmon tissue; representing more than 60
percent of the total PCB burden in all samples. Aroclor 1254 contribution significantly
more to the overall PCB burden (Kruskall-Wallis P<0.002)) in tissue samples from the
coast compared to Puget Sound tissue samples (Table 7).
Table 7. Sample size (N), mean, standard deviation and range for the proportion of
Aroclor 1254 in the total PCB concentrations for each region.
Hatchery
Puget Sound
Coast
N
20
20
Mean
0.75
0.86
Std deviation
0.1
0.047
Range
0.6-0.93
0.75-0.93
The recommended amount of Chinook salmon to be consumed by humans varied by
region but was consistent within region (Table 8). In general, more coastal Chinook
salmon could be consumed compared to Puget Sound for both cancerous and noncancerous end points.
DISCUSSION
Mean PCB concentrations in Chinook salmon from the Puget Sound hatcheries were
almost 2.5 times greater than those in Chinook salmon from coastal hatcheries. PCB
concentrations observed in Puget Sound hatchery Chinook salmon in the current study
were similar to concentrations observed in samples collected by the Puget Sound
Ambient Monitoring Program (PSAMP) (O’ Neill WDFW, pers. comm. 2004).
31
Table 8. The recommended upper limit for the number of 8 oz (227g) meals of Chinook
salmon which can be consumed per month from each hatchery before non-cancerous and
cancerous health concerns arise. Estimates are based on EPA’s recommendation using a
reference dose of 2.5 X 10-5 mg/kg per day and half of the mean concentration of PCBs
found in this study.
Hatchery
Half of the mean
concentration of
PCBs in ppm
determined from
this study
(mg/kg)
Allowable consumption per month
Non-cancer health
Cancer health
end point
endpoint
Puget Sound
Deschutes
Issaquah
0.0251
0.0251
4
4
1
1
16
16
4
4
Coast
Makah
Quinault
1
1
0.01
0.011
Numbers were rounded up
The mean PCB concentrations in Puget Sound Chinook salmon collected in-river during
PSAMPs 10 year (1989-1999) monitoring program were slightly greater (53.88 µg/kg;
average of 4 sample rivers) than I observed (49.26 µg/kg; average of 2 sample rivers).
These data are relatively similar even though I sampled only four-year old fish, while
PSAMP sampled fish from two-to-five year olds. The PSAMP data showed greater PCB
concentrations in Chinook salmon from southern Puget Sound (south of the Tacoma
Narrows Bridge) than from northern Puget Sound (north of Admiralty Inlet) (PSAMP
2001). In addition, PSAMP (2001) in-river Puget Sound Chinook salmon had higher
lipid content (3.34 percent; average 4 sample rivers) than my samples (0.33 percent;
average of 2 sample rivers). The difference in the lipid content could be the result of how
long the fish had not been feeding (time in fresh water) compared to the time the samples
were collected.
32
These results suggest that regional differences exist in PCB concentrations in Chinook
salmon tissues. However, the residual plots suggest the fit of the models used to evaluate
these differences were poor. The models were fairly robust when describing the coastal
data, but tended to overestimate predicted values for the Puget Sound. The spuriousness
of the Puget Sound predicted values reflects a large difference in the variances which
could be due to the rearing, feeding or migration patterns of Puget Sound Chinook
salmon. Even though the predicted values for Puget Sound were spurious, four out of the
five best models have region as one of the variables, which indicates that Puget Sound is
the primary driver of PCB concentrations in this study. In addition, even though the
residual plots showed large variances, other data suggest that PCB concentrations
decrease in northern Puget Sound (PSAMP 2001). Data from this study, along with
PSAMP (2001) data indicates that Puget Sound Chinook salmon are more contaminated
with PCBs than coastal hatchery Chinook salmon.
There are several estuaries within Puget Sound that have been targeted as Superfund sites
where juvenile Chinook salmon tend to reside before migrating to sea. Juvenile Chinook
salmon migrating out of the estuarine environment and through Puget Sound must also
travel up to 100 miles (from the Deschutes sample site) through several urbanized and
industrial sites that could lead to additional exposure. Several authors have shown that
juvenile Chinook salmon are accumulating PCBs through their prey base while they are
in the estuarine environment in Puget Sound (Arkoosh et al. 1998b; McCain et al. 1990;
Meador 2000; Stein et al. 1995). PSAMP (2001) found a decrease in PCB concentration
in Chinook salmon in northern Puget Sound compared to southern Puget Sound.
33
However, this does not fully explain why returning adult Chinook salmon have a greater
body burden in comparison to the juvenile life stage upon returning to Puget Sound to
spawn (O’Neill, pers. comm. 2004). Variation in residency time in Puget Sound and/or
migration routes may explain the large variance observed in the Puget Sound data.
Puget Sound Chinook salmon have diverse ocean migratory patterns. In general, most
Puget Sound Chinook salmon travel through the inland waterway between the Canadian
mainland and Vancouver Island. Earlier returning adult Chinook salmon travel farther up
the west coast of Canada and into southeast Alaska than later returning adult Chinook
salmon (Figure 4) (NOAA 2002; Pacific Salmon Council 2003). There is also a segment
of the Chinook salmon that may reside in Puget Sound throughout their life. These
groups of fish obviously have different food sources which may have lead to the
increased variance in PCB concentrations observed in the current study The Puget Sound
fish I sampled returned several weeks earlier than the coast fish which suggests a longer
migratory route and potentially additional exposure to PCBs in the ocean. Currently,
there is no way of determining whether or not the fish in my samples resided in Puget
Sound or migrated to the Pacific Ocean. Coded wire tag data recoveries of sub-adults are
limited and biased by fishing and sampling effort.
PCB concentrations in coastal Chinook salmon were less variable and were explained
well by the models. Less variation may be the result of a less contaminated migratory
route or a more consistent migratory route. The estuarine environment on the coast is
34
Figure 4. Puget Sound and coastal Chinook salmon migration pattern (Courtesy of Dave
Galvin).
35
more pristine than that of the Puget Sound. The coastal environment in Washington State
is very rural and has a very limited industrial/urbanized component. Migration patterns
of coastal Chinook salmon also differ slightly from the Puget Sound Chinook salmon.
Coded wire tag data indicate that coastal Chinook salmon migrate on the west side of
Vancouver Island sometimes migrating to southeast Alaska (Pacific Salmon Council
2003; USFWS, unpublished data).
The observations of greater PCB concentrations in adult Puget Sound Chinook salmon
raises concerns that PCBs may be contributing to the declining population in the Puget
Sound, but it is unclear if PCBs impact Chinook salmon survival. Arkoosh et al. (1998a)
suggests that pollutants (which include PCBs) adversely affect the environment and may
affect certain life stages of salmonids. Arkoosh et al. (1998b) also suggests that exposure
to PCBs suppresses the immune system in juvenile Chinook salmon. The juvenile and
sub-adult life stage would be the most affected because it typically occurs within an
urbanized setting of the Puget Sound region.
PCBs are maternally transferred to the eggs (Missildine, unpublished data; Miller and
Amrhein 1995; Miller 1993) and most likely impact egg to fry survival (Walker et al.
1991; Ankley et al. 1991). Ankley et al. (1991) found that egg-to-fry survival was
inversely related to PCB concentrations in Lake Michigan Chinook salmon. Miller
(1993) found a direct correlation between PCB concentrations in muscle tissue (4.3 +/-0.4
mg/kg) and PCB concentrations in eggs (8.3 +/-0.9 mg/kg) of salmonids. Miller’s (1993)
PCB concentration levels were higher than the PCB concentrations found in the current
36
study. It is important to note that Miller’s (1993) samples were processed with the skin
intact. It would be extremely difficult to determine how much of an impact PCBs are
having on egg-to-fry survival in the wild. Healy (1991), found egg-to-fry survival for
Chinook salmon in the wild is approximately 30 percent, but can vary due to riverine
conditions. Variables such as floods, scour, entombment, and superimposition of redds
by other salmon contribute to egg mortality.
Research on salmon and other animals indicates the abnormal presence of genetic and
hormonal markers that would typically be seen in the opposite gender (Carlson et al.
2000; Nagler et al. 2001). Researchers suggest that endocrine disrupting chemicals such
as PCBs may be having this affect on salmon and other species. It is unknown whether
or not the presence of these markers is having an impact on reproductive success.
Carlson et al. (2000) suggest that embryos were more susceptible to toxicants than older
adult fish but that maternal transfer, especially in fish that live in heavily contaminated
areas, could affect embryo survival and local fish populations. Carlson et al. (2000) also
found that embryos injected with contaminants caused the embryo to hatch early, which
could be detrimental to survival, especially in the Northwest where freshets are common
during winter when most salmonid eggs are still in the gravel incubating.
Management Implications
The observation of PCBs in Chinook salmon returning to Washington hatcheries may
impact the way hatchery fish are managed after they have been spawned. Currently, fish
from the Makah and Quinault hatcheries are distributed to tribal members for
37
consumption. The State hatcheries sell their fish to a buyer, who typically processes the
fish for a variety of animal feeds and fertilizers, and/or they give the fish away to
conservation groups that place the carcasses into streams to increase the input of marine
derived nutrients. All of these methods reintroduce PCBs into the environment with little
knowledge of the overall impact.
Washington State’s salmon carcass distribution program is supported by local, state, and
federal agencies. Carcass planting has become an important component in salmonid
habitat restoration. Several authors have shown the benefits of carcass planting into
streams (Helfield and Naiman 2001; Cederholm et al. unpublished data; Cederholm et al.
1999; Bilby et al. 1995). Cederholm et al. (1999) stated that salmonids supply a major
source of marine derived nutrients to the aquatic and terrestrial landscape. Bilby et al.
(1995) found that many aquatic invertebrates and streamside plants are enriched with
marine derived nutrients from coho salmon (Oncorhynchus kisutch) carcasses. However,
none of the authors discuss the potential of recontamination of the environment with
PCBs or other persistent organic pollutants from planting salmon carcasses.
I could not find any published reports which evaluate the influence of salmon carcass
deployment on pollutant reintroduction to stream ecosystems. However, PCB
concentrations in the water column (O’Toole et al. In preparation), and in the sediment
(Krummel et al. 2003), have increased following natural spawning. O’Toole et al. (In
preparation) found an increase in PCB concentrations in the water column of the Credit
River from post-spawned Chinook salmon. Krummel et al. (2003) found that PCBs
38
levels in the sediment of Alaskan lakes increased seven-fold upon the return of adult
sockeye salmon. PSAMP (2001) data shows that coho salmon, a prevalent salmon
species in Washington State, are also contaminated with PCBs, although at lower levels
than Chinook salmon. These data indicate that PCB concentrations in rivers and streams
in Washington State may increase following carcass deployment.
Carcass deployment may also distribute PCBs to the terrestrial ecosystem. Decaying
salmon carcasses are an important component of the terrestrial food web. Willson and
Halupka (1995) indicate that over 20 mammalian and avian species combined are direct
consumers of salmonid carcasses. The consumption of salmon carcasses contaminated
with PCBs may impact the survival rate of species that feed on salmon carcasses.
Bowerman et al. (1995) indicated that bald eagle (Haliaeetus leucocephalus) birth rates
and adult mortality in the Great Lakes region may still be impacted by consuming postspawned salmon carcasses contaminated with PCBs. Buck et al. (1999) indicated that
bald eagles from the lower Columbia River were also experiencing lower birth rates than
bald eagles in other northwest locations in association with PCB contaminated prey
sources.
Currently, fish in the carcass distribution program are tested for bacteria and viruses but
not contaminants. No research to date is occurring to determine if the PCBs in salmon
carcasses are having an impact on other fish and wildlife species in Washington State that
feed on these carcasses. Testing the hundreds of carcasses that are planted into streams
and rivers each year would be cost prohibitive and time consuming for the entities
39
involved in the carcass distribution program and the labs that would be needed to
analyzed the samples. Other alternatives such as pellets or seeding the streams and rivers
could be an effective means of delivering marine derived nutrients to the ecosystem, but
may not be as efficient as carcass deployment. Either way, we need to stop the cycle of
PCBs, whether that means stopping the carcass deployment program or spending the
money to test each fish. Others may argue that the naturally spawning fish are already
contaminated so what difference does it make. I would argue that many of the salmon
deployment projects are located in streams with very few returning salmon and therefore
would have extremely low concentrations of PCBs.
Managers and scientists need to determine whether the benefits of marine derived
nutrients outweigh the detrimental PCB impacts. Other toxic chemicals are also being
transported by salmon including polybrominated diphenyl ether (PBDE; fire retardants)
and dichlorodiphenyltrichloroethane (DDT) (O’Neill, pers. comm.2004). Planting
carcasses will most likely result in the introduction of these toxic chemicals (or an
increase, for streams that already have naturally spawning salmon), along with PCBs,
into the environment where toxins might not currently be found. I would surmise that
contamination would still occur by the die-off of natural spawners if the carcass
deployment programs were discontinued. Furthermore, natural spawners will also
provide the marine derived nutrients to the ecosystem, albeit not at historic levels.
Technological advances may be able to duplicate the marine derived nutrients without the
increased risk of contamination, or the use of less contaminated stocks such as those from
the coast of Washington could be used for carcass deployment.
40
Human health implications
PCBs are usually associated but not specifically implicated with human health issues
because humans are subject to many other variables that can impact health. Schecter et
al. (1994) noted that there is a variance in toxicity between the 209 specific congeners
and that failure to report specific congeners can lead to two problems. Firstly, not all
PCBs are toxic and determining if the PCB is causing toxic response is problematic.
Secondly, total PCB levels within an exposed person may be within “normal” limits, but
toxic congeners may be masked within the total PCB concentrations reported.
Even though PCBs are difficult to implicate with human health issues, several authors
have found associated health issues. Brown (1987) found an increase in cancer in the
biliary tract, liver, and gall bladder from workers exposed to PCBs compared to the
national rates. Loomis et al. (1997) found an increase in malignant melanoma and brain
cancer from workers exposed to PCBs. In the Michigan Maternal Infant Cohort Study,
children of women who consumed PCB contaminated fish experienced developmental
deficiencies compared to children whose mothers did not consume contaminated fish
(Fein et al. 1984; Jacobson et al. 1985, 1990a, b).
The consumption of fish is one of the main vectors of human exposure to PCBs. Greater
exposure to PCBs may occur to some cultures in the Puget Sound because salmon are a
main staple in their diets. A literature search was conducted but did not find any
information relative to PCB levels in northwest indigenous people. However, Schell et
al. (2003) found increased PCB concentrations in blood serum from Akwesasne Mohawk
41
youth that consumed fish from the St. Lawrence River in New York. Schell et al. (2003)
also discovered that breast fed Akwesasne children had PCB concentrations 1.3 times
higher than non-breastfed Akwesasne children. Dellinger et al. (1997) reported higher
levels of PCBs in Ojibwa members who ate fish in comparison with the overall
population. Dellinger et al. (1997) also notes that the Ojibwa members may also be at a
higher risk for health effects.
The consumption of fatty fish, especially salmon, may decrease cardiovascular disease.
The U. S. Food and Drug Administration and the American Heart Association currently
emphasize the benefits of Omega-3 fatty acids typically found in salmon and other fish.
Stone (1996) reported that Omega-3 fatty acids may reduce the risk of heart disease. The
American Heart Association recommends eating two servings of fish a week. This is
double the recommended consumption rate calculated using EPA guidelines for the
consumption of Puget Sound salmon; however, it is within the consumption rate
calculated for coastal Chinook salmon.
Most individuals will have to weigh the benefits/risks for themselves. Those that have
low risks of cardiovascular disease may choose not to eat as much fish as recommended
where those at risk of cardiovascular disease may choose to eat more fish. There are
several recommendations for cooking salmon that will reduce the amount of PCBs in the
fish; 1) remove the skin; 2) trim away excess fat from the lateral line, ventral side and
backbone (Figure 5); and 3) grill, bake or broil the fish (WDOH 2004). Anderson et al.
(1993) states that a 50 percent reduction in PCBs occurs during cooking and preparation.
42
Figure 5. Proper trimming preparation to remove fat from salmon before cooking
(Courtesy of Washington State Department of Health).
CONCLUSION
Region appeared to be the primary factor influencing PCB concentrations in Chinook
salmon from western Washington, with greater concentrations observed in Puget Sound
Chinook salmon than coastal Chinook salmon. My data shows that inhabitants of the
Puget Sound and coast who are eating Chinook salmon are still being exposed to PCBs
long after they were outlawed in the United States. Many other species, such as bears
and eagles, which rely on Chinook and other salmon, may be experiencing health risks
associated with PCB exposure. Several Superfund clean ups within Puget Sound either
have been completed or are going to be completed over the next several years which will
help reduce exposure to PCBs. However, it appears that Chinook salmon may still be
exposed to PCBs during their open ocean migration. Unfortunately, little can be done to
minimize exposure without an international ban and clean up plan for PCBs worldwide.
43
Future Recommendations
This research project is just one piece of the PCB puzzle. Research being conducted by
PSAMP and USFWS will help develop a more complete picture of PCBs in salmon.
Additional research needs to be conducted to determine if PCBs are affecting Chinook
salmon survival throughout its life history. Evidence suggests that there is an inverse
correlation between egg-to-fry survival and PCB concentrations (Ankley et al. 1991).
Evidence also suggests that juvenile Chinook salmon contaminated with PCBs suffer
immunosuppression, which could lead to increased disease susceptibility (Arkoosh et al.
1998b). We don’t know if PCB concentrations in adult Chinook salmon are causing a
decrease in survival in the open ocean due to immunosuppression. Research also needs
to be conducted to determine if planting salmon carcasses in streams is recirculating
PCBs into the environment, which could adversely impact other species that feed on
salmon carcasses. We also need to determine the “hot spots” in the Pacific Ocean and
Puget Sound where PCBs are accumulating.
In conclusion, PCBs are ubiquitous, but can be found in greater concentrations in
urbanized areas such as the Puget Sound and Great Lakes. The chemical composition of
PCBs that makes them stable in the environment may hinder its total elimination in the
environment. International efforts must be put forth to eliminate the use and improper
disposal of PCBs in the environment.
44
LITERATURE CITED
Anderson, D. R., K. P. Burnham, and W. L. Thompson. 2000. Null Hypothesis testing: a
problem, prevalence, and an alternative. Journal of Wildlife Management 64:912923.
Anderson, H. A., J. F. Amrhein, P. Shubat, and J. Hesse. 1993. Protocol for a uniform
Great Lakes sport fish consumption advisory. Great Lakes Fish Advisory Task Force
Protocol Drafting Committee. 92pp.
Ankley, G. T., D. E. Tillitt, J. P. Giesy, P. D. Jones, and D. A. Verbrugge. 1991.
Bioassay-derived 2,3,7,8-tetrachlorodibenzo-p-dioxin equivalents in PCB containing
extracts from the flesh and eggs of Lake Michigan Chinook salmon (Oncorhynchus
tshawytscha) and possible implications for reproduction. Canadian Journal of
Fisheries and Aquatic Sciences 48:1685-1690.
Arkoosh, M. R., E. Casillas, E. Clemons, A. N. Kagley, R. Olson, P. Reno, and J. Stein.
1998a. Effect of pollution on fish diseases: potential impact on salmonid populations.
Journal of Aquatic Animal Health 10:182-190.
Arkoosh, M. R., E. Casillas, E. Clemons, P. Huffman, J. Evered, J. E. Stein, and U.
Varanasi. 1998b. Increased susceptibility of juvenile Chinook salmon from a
contaminated estuary to Vibrio anguillarum. Transactions of the American Fisheries
Society 127:360-374.
Aronson, K. J., A. B. Miller, C. G. Woolcott, E. E. Sterns, D. R. Mcready, L. A. Lickley,
E. B. Fish, G. Y. Hiraki, C. Holloway, T. Ross, W. M. Hanna, S. K. SenGupta, and
J.P. Weber. 2000. Breast adipose tissue concentrations of Polychlorinated Biphenyls
and other organochlorines and breast cancer risk. Cancer Epidemiology, Biomarkers
and Prevention 9:55-63.
Bailey, R., L. A. Barrie, C. J. Halsall, P. Fellini, and D. C. G. Muir. 2000. Atmospheric
organochlorine pesticides in the western Canadian Arctic. Evidence of Transpacific
transport. Journal of Geophysical Research 105:805-811.
Bilby, R. E., B. R. Fransen, and P. A. Bisson. 1995. Incorporation of nitrogen and
carbon from spawning coho salmon into the trophic system of small streams:
evidence from stable isotopes. Canadian Journal of Fisheries and Aquatic Sciences
53:164-173.
Bowerman, W. W., J. P. Giesy, D. A. Best, and V. J. Kramer. 1995. A review of factors
affecting productivity of bald eagles in the Great Lakes region: implications for
recovery. Environmental Health Perspectives 103:51-59.
45
Broman, D., C. Naf, C. Rolff, Y. Zebuhr, B. Fry, and J. Hobbie. 1992. Using ratios of
stable nitrogen isotopes to estimate bioaccumulation and flux of polychlorinated
dibenzo-p-dioxin (PCDDs) and dibenzofurans (PCDFs) in two food chains from the
Northern Baltic. Environmental Toxicology and Chemistry 11:331-345.
Brown, D. P. 1987. Mortality of workers exposed to polychlorinated biphenyls - An
update. Archives of Environmental Health 42:333-339.
Brouwer, A., D. C. Morse, M. C. Lans, A. G. Schuur, A. J. Murk, E. Klasson-Wehler, A.
Bergman, and T. J. Visser. 1998. Interactions of persistent environmental
organohalogens with the thyroid hormone system: mechanisms and possible
consequences for animal and human health. Toxicology of Industrial Health 14:5984.
Buck, J., R. G. Anthony, and F. B. Issacs. 1999. Changes in productivity and
environmental contaminants in bald eagles nesting along the lower Columbia River.
Miscellaneous report. U.S. Fish and Wildlife Service, Portland, Oregon.
Burnham, K. P. and D. R. Anderson. 2002. Model selection and multimodel inference.
A practical information-theoretic approach. Second edition. New York. SpringerVerlag.
Carlson, D. B., L. R. Curtis, and D. E. Williams. 2000. Salmonid sexual development is
not consistently altered by embryonic exposure to endocrine active chemicals.
Environmental Health Perspectives 108:249-255.
Cederholm, C. J., H. Michael Jr., and N. Pittman. Unpublished. Considerations in
establishing ecologically based salmon spawning escapement goals for selected
Washington Rivers. Wild Salmon and Steelhead Conference 2001.
Cederholm, C. J., M. D. Kunze, T. Murota, and A. Sibatani. 1999. Pacific salmon
carcasses: essential contributions of nutrients and energy for aquatic and terrestrial
ecosystems. Fisheries 24(10):6-15.
Craan, A. G. and D. A. Haines. 1998. Twenty five years of surveillance for
contaminants in human breast milk. Archives of Environmental Contamination and
Toxicology 35:702-710.
Davis, J. W. 1993. An analysis of tissues for total PCB and planar PCB concentrations
in marine mammals stranded along the Gulf of Mexico. Master’s Thesis, Texas
A&M University.
Dellinger, J. A., S. L. Gerstenberger, L. K. Hansen, and L. L. Malek. 1997. Ojibwa
health study: assessing the health risks from consuming contaminated Great Lakes
fish. Health Conference ’97 Great Lakes and St. Lawrence.
46
Dobson, A. J. 2002. An introduction to Generalized Linear Models, second edition.
CRC Press.
Duinker, J., M. Hillebrand, T. Zeinstra, and J. Boon. 1989. Individual chlorinated
biphenyls and pesticides in tissues of some cetacean species from the North Sea and
the Atlantic Ocean: tissue distribution and biotransformation. Aquatic Mammals
15:95-124.
Eisler, R. 1986. Polychlorinated biphenyl hazards to fish, wildlife, and invertebrates: a
synoptic review. Miscellaneous Report. U.S. Fish and Wildlife Service, Patuxent
Wildlife Research Center, Laurel, MD.
Environmental Protection Agency (EPA). 2002. White paper no. 13. Hudson River
record of decision PCB non-cancer health effects.
Environmental Protection Agency (EPA). 1999. Polychlorinated Biphenyls (PCBs)
update: Impact of fish advisories. EPA-823-F-99-019
Erickson, M. D. 1997. Analytical Chemistry of PCBs, 2nd edition. Argonne National
Laboratory, Argonne, Illinois. Lewis Publishers, New York.
Fein G. G., J. L. Jacobson, and S. W. Jacobson. 1984. Prenatal exposure to
polychlorinated biphenyls: effects on birth size and gestation age. Journal of
Pediatrics 105:315-320.
Girvin, D. C., D. S. Sklarew, and C. C. Ainsworth. 1993. Polychlorinated biphenyls
sorption/desorption in soil-water systems, in Proceedings: 1993 PCB Seminar. New
Orleans, LA: Electric Power Research Institute.
Goldstein, J. A. and S. Safe. 1989. Mechanism of action and structure-activity
relationships for the chlorinated dibenzo-p-dioxins and related compounds. Pages
239-293 in Halogenated biphenyls, terphenyls, naphthalenes, dibezodioxins and
related products. R. Kimball and S. Jensen, editors. Elsevier Science Publishers,
England.
Guo, L. Y., M. L. Yu, C. C. Hsu, and W. J. Rogan. 1999. Chloracne, goiter, arthritis,
and anemia after polychlorinated biphenyl poisoning: 14-year follow-up of the
Taiwan Yucheng cohort. Environmental Health Perspectives 107:715-719.
Harrell, F. E. 2001. Regression modeling strategies with applications to linear models,
logistic regression, and survival. Springer, New York.
Healey, M. C. 1991. Life history of chinook salmon. Pages 311-393 in C. Groot and L.
Margolis, editors. Pacific Salmon Life Histories. University of British Columbia
Press, Vancouver.
47
Helfield, J. M. and R. J. Naiman. 2001. Effects of salmon-derived nitrogen on riparian
forest growth and implications for stream productivity. Ecology 82:2403-2409.
Hendry, A. P. 1998. Reproductive energetics of Pacific salmon: strategies, tactics and
tradeoffs. Doctoral disseration. University of Washington, Seattle, WA.
Hidaka, H., S. Tanabe, and R. Tatsukawa. 1983. DDT compounds and PCB isomers and
congeners in Weddel seals and their fate in the Antarctic marine ecosystem.
Agricultural Biology and Chemistry 47:2009-2017.
Jacobson, J. L. and S. W. Jacobson. 1996. Intellectual impairment in children exposed to
polychlorinated biphenyls in utero. New England Journal of Medicine 335:783-789.
Jacobson J. L., S. W. Jacobson, and H. E. B Humphrey. 1990a. Effects of in utero
exposure to polychlorinated-biphenyls and related contaminants on cognitivefunctioning in young children. Journal of Pediatrics 116:38-45.
Jacobson J. L., S. W. Jacobson, and H. E. B Humphrey. 1990b. Effects of exposure to
PCBs and related compounds on growth and activity in children. Neurotoxicology
Teratology 12:319-326.
Jacobson S. W., G. G, Fein, and J. L. Jacobson. 1985. The effect of intrauterine PCB
exposure on visual recognition memory. Child Development 56:856-860.
Jaffe, D., and 12 other authors. 1999. Transport of Asian pollution to North America,
Geophysical Research Letter 26:711-714.
Jensen, A. L. 1984. PCB uptake and transfer to humans by lake trout. Environmental
Pollution 34:73-82.
Jorgensen, E. H., B. E. Bye, and M. Jobling. 1999. Influences of nutritional status on
biomarker response to PCB in the Arctic char (Salvelinus alpinus). Aquatic
Toxicology 44:233-244.
Juan, C. Y., G. O. Thomas, A. J. Sweetman, and K. C. Jones. 2002. An input-output
balance study for PCBs in humans. Environmental International 28:203-214.
Kanan, N., S. Tanabe, and O. M. and R. Tatsukawa. 1989. Critical evaluation of
polychlorinated biphenyl toxicity in terrestrial and marine mammals: increasing
impact of non-ortho and mono-ortho coplanar polychlorinated biphenyls from land to
ocean. Archives of Environmental Contaminants and Toxicology 18:850-857.
Kilburn, K. H., R. H. Warsaw, and M. G. Shields. 1989. Neurobehavioral dysfunction in
firemen exposed to polychlorinated biphenyls (PCBs): possible improvement after
detoxification. Archives of Environmental Health 44:345-350.
48
Korrick, S. A. and L. Altshul. 1998. High breast milk levels of polychlorinated
biphenyls (PCBs) among four women living adjacent to a PCB contaminated waste
site. Environmental Health Perspective 106:513-518.
Krummel, E. M., R. W. MacDonald, L. E. Kimpe, I. Gregory-Eves, M. J. Eaves, M. J.
Demers, J. P. Smol, B. Finney, and J. M. Blais. 2003. Aquatic ecology: Delivery of
pollutants by spawning salmon. Nature 425:255-256.
Kullback, S. and R. A. Leibler. 1951. On information and sufficiency. Annals of
Mathematical Statistics 22:79-86.
Laden, F., G. Collman, K. Iwamoto, A. J. Alber, G. S. Berkowitz, J. L. Freundenheim, S.
E. Hankinson, K. J. Helzlsouer, T. R. Holford, H. Y. Huang, K. B. Moysich, J. D.
Tessari, M. S. Wolff, T. Zheng, and D. J. Hunter. 2001. 1,1-Dichoro-2,2-bis(pchlorophenyl) ethylene and polychlorinated biphenyls and breast cancer: combined
analysis of five U. S. studies. Journal of the National Cancer Institute 93:768-776.
Lanting, C. I. 1999. Effects of perinatal PCB and dioxin exposure and early feeding
mode on child development. Doctoral dissertation, University of Groningen,
Netherlands.
Larsson, P. and L. Okla. 1987. An attempt to measure the flow of chlorinated
hydrocarbons, such as PCBs, from water to air in the field. Environmental Pollution
22:219-225.
Loomis D., S. R. Browning, A. P. Schenck, E. Gregory, D. A. Savitz. 1997. Cancer
mortality among electric utility workers exposed to polychlorinated biphenyls.
Occupational Environmental Medicine 54:720-728.
McCain, B. B., D. W. Brown, S.-L. Chan, J. T. Landahl, W. D. MacLeod, Jr., M. M.
Krahn, C. A. Sloan, K. L. Tilbury, S. M. Pierce, D. G. Burrows, and U. Varanasi.
2000. National benthic surveillance project: Pacific Coast. Organic chemical
contaminants, cycles I to VII (1984-90). U.S. Department of Commerce, NOAA
Tech. Memo. NMFS-NWFSC-40, 121 p.
McCain, B. B., D. C. Malins, M. M. Krahn, D. W. Brown, W. D. Gronlund, L. K. Moore,
and S-L Chan. 1990. Uptake of Aromatic and chlorinated hydrocarbons by juvenile
Chinook salmon (Oncorhynchus tshawytscha) in an urban estuary. Archives
Environmental Contaminants and Toxicology 19:10-19.
McFarland, V. A. and J. U. Clarke. 1989. Environmental occurrence, abundance, and
potential toxicity of polychlorinated biphenyl congeners: considerations for a
congener-specific analysis. Environmental Health Perspective 81:225-239.
49
McKinney, J. D. and C. L. Waller. 1998. Molecular determinants of hormone mimicry:
halogenated aromatic hydrocarbon environmental agents. Journal of Toxicology and
Environmental Health 1(1):27-58.
Meador, J. P. 2000. An analysis in support of tissue and sediment based threshold
concentration of the polychlorinated biphenyls (PCBs) to protect juvenile salmonids
listed by the Endangered Species Act. Technical Report, Northwest Fisheries Science
Center NOAA. Seattle, WA.
Miller, M. A. 1993. Maternal transfer of organochlorine compounds in salmonines to
their eggs. Canadian Journal of Fisheries and Aquatic Science 50:1405-1413.
Miller, M. A. and J. F. Amrhein. 1995. Maternal transfer of organochlorine compounds
in Lake Superior Siscowet (Salvelinus namaycush) to their eggs. Bulletin of
Environmental Contaminants and Toxicology 55:96-103.
Morishima, G. S. and K. A. Henry. 1999. The history and status of Pacific Northwest
Chinook and coho salmon ocean fisheries and prospects for sustainability. Pages
219-235 in E. E. Knutson, editor. Sustainable Fisheries Management. CRC Press,
N.Y.
Moysich, K. B., P. G. Shields, J. L Freudenheim, E. F. Schisterman, J. E. Vena, P.
Kostyniak, H. Greizerstein, J. R. Marshall, S. Graham, and C. B. Ambrosone. 1999.
Polychlorinated Biphenyls, cytochrome P4501A1 polymorphism, and
postmenopausal breast cancer risk. Cancer Epidemiology, Biomarkers and
Prevention 8:41-44.
Nagler, J. J., J. Bouma, G. H. Thorgaard, and D. D. Daubl. 2001. High incidence of a
male-specific genetic marker in phenotypic female Chinook salmon from the
Columbia River. Environmental Health Perspectives 109:67-69.
NOAA (National Oceanic and Atmospheric Administration). 2002. Federal Register
67(153).
Ojajarvi, I. A. and 11 coauthors. 2000. Occupational exposures and pancreatic cancer:
A meta-analysis. Occupational and Environmental Medicine 57:316-324.
O’Neill, S. 2004. Washington State Department of Fish and Wildlife. Personal
Communications.
O’Neill, S. M., J. E. West, and J. C. Hoeman. 1998. Spatial trends in the concentration
of polychlorinated biphenyls (PCBs) in Chinook (Oncorhynchus tshawytscha) and
coho salmon (O. kisutch) in Puget Sound and factors affecting PCB accumulation:
results from the Puget Sound Ambient Monitoring Program. Puget Sound Research.
Washington Department of Fish and Wildlife.
50
Osius, N. W., Karmaus, H. Kruse, and J. Witten. 1999. Exposure to polychlorinated
biphenyls and levels of thyroid hormones in children. Environmental Health
Perspectives 107:843-849.
O’Toole, S., M. Gross, I. Craine, and C. Metcalfe. In prep. Impact of polychlorinated
biphenyls and organochlorine contaminants from Lake Ontario Chinook salmon
(Oncorhynchus tshawytscha) on the Credit River. Trent University, Peterborough,
Ontario.
Pacific Salmon Council. 2003. Joint chinook technical committee report. Annual
exploitation rate analysis and model calibration. November 2003.
Patandin, S., P. C. Dagnelie, P. G. Mulder, E. Op de Coul, J. E. van der Veen, N.
Weisglas-Kuperus, and P. J. Sauer. 1999. Dietary exposure to polychlorinated
biphenyls and dioxins from infancy until adulthood: a comparison between breastfeeding, toddler and long-term exposure. Environmental Health Perspectives 107:4551.
Paya-Perez, A. B., M. Riaz, and B. R. Larsen. 1991. Soil sorption of 20 PCB congeners
and six chlorobenzenes. Ecotoxicology and Environmental Safety 21:1-17.
Post, G. 1987. Textbook of fish health. Revised and expanded edition. T. H. F.
Publications, Neptune City, N. J.
Puget Sound Ambient Monitoring Program (PSAMP). 2001. 2000 Puget Sound update:
Seventh report of the Puget Sound Ambient Monitoring Program. Puget Sound
Water Quality Action Team. Olympia, WA.
Schecter, A., J. Stanley, K. Boggess, Y. Masuda, J. Mes, M. Wolff, P. Furst, C. Furst, K.
Wilson-Yang, and B. Chisholm. 1994. Polychlorinated biphenyl levels in the tissues
of exposed and nonexposed humans. Environmental Health Perspectives
Supplements 102:149-158.
Schell, L. M., L. A. Hubicki, A. P. Decaprio, M. V. Gallo, J. Ravenscroft, A. Tarbell, A.
Jacobs. D. David, P. Worswoick, and the Akwesasne Task Force on the Environment.
2003. Organochlorines, Lead, and Mercury in Akwesasne Mohawk youth.
Environmental Health Perspectives 111:954-961.
Shirai, J. 1995. Variability in estimated physiological half-lives of PCBs in humans:
impact on exposure assessment. Master’s Thesis. University of Washington, Seattle,
WA.
Stein, J. E., T. Hom, T. K. Collier, D. W. Brown, and U. Varanasi. 1995. Contaminant
exposure and biochemical effects in out-migrant juvenile Chinook salmon from urban
and non-urban estuaries of Puget Sound, Washington. Environmental Toxicology
and Chemistry 14:1019-1029.
51
Stellman, S. D., M. V. Djordjevic, J. A. Britto, J. E. Muscat, M. L. Citron, M. Kemeny,
E. Busch, and L. Gong. 2000. Breast cancer risk in relation to adipose
concentrations of organochlorine pesticides and polychlorinated biphenyls in Long
Island, New York. Cancer Epidemiology, Biomarkers and Prevention 9:1241-1249.
Stone, N. J. 1996. Fish Consumption, Fish Oil, Lipids, and Coronary Heart Disease.
Circulation 94:2337-2340.
Susser M. 1986. Rules of interference in epidemiology. Regulatory Toxicology and
Pharmacology 6:116-28.
Swain W. R. 1991. Effects of organochlorine chemicals on the reproductive outcome of
humans who consumed contaminated Great Lakes fish: an epidemiologic
consideration. Journal of Toxicology and Environmental Health 33:587-639.
Walker, M. K., J. M. Spitsbergen, J. R. Olsen, and R. E. Petersen. 1991. 2,3,7,8 –
Tetrachlorodibenzo-p-dioxin (TCDD) toxicity during early life stage development of
lake trout (Salvelinus namaycush). Canadian Journal of Fisheries and Aquatic
Sciences 28:875-883.
Washington Department of Health. 2004. www.doh.wa.gov/ehp/oehas/EHA
Willson, M. F. and K. C. Halupka. 1995. Anadromous fish as keystone species in
vertebrate communities. Conservation Biology 9:489-4979.
Wolff, M. S., A. Zeleniuch-Jacquotte, N. Dubin, and P. Toniolo. 2000. Risk of breast
cancer and organochlorine exposure. Cancer Epidemiology, Biomarkers and
Prevention 9:271-277.
Yu, M. L., Y. L. Guo, C. C. Hsu, and W. J. Rogan. 2000. Menstruation and reproduction
in women with polychlorinated biphenyl (PCB) poisoning: long-term follow–up
interviews of the women from the Taiwan Yucheng cohort. International Journal of
Epidemiology 29:672-677.
52
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
Sound
Puget
720
4.11
Mercury, Total,
CVAA
Total Solids *
µg/kg
µg/kg
%
mg/kg
%
Sex
Percent Lipids *
Aroclor 1254
µg/kg
total pcb
mm
Aroclor 1260
Location
Deschutes
1
Deschutes
2
Deschutes
7
Deschutes
8
Deschutes
10
Deschutes
13
Deschutes
14
Deschutes
15
Deschutes
16
Deschutes
17
Issaquah
30
Issaquah
33
Issaquah
34
Issaquah
35
Issaquah
36
Issaquah
37
Issaquah
Fish weight
Region
Fish length
APPENDIX 1. Combined lab results (All results in dry weight).
50
8
58 1.5
0.204
24 m
880
80.1
9
89.1 1.2
0.235
23 m
820
394
180
574
1
0.556
23 m
830
103
41.8
145
1
0.563
23 m
840
90.2
50.4
141 1.4
0.366
24 m
800
5.8
320
154
474 0.5
0.695
21 f
800
5.6
113
29
142 0.8
0.496
19 f
845
5.5
79.5
40.5
120 0.4
0.533
21 f
800
4.8
206
66.5
273 1.2
0.78
19 f
840
4.6
163
66.3
229 0.7
0.559
20 f
790
3.7
180
61.2
241 0.4
0.854
18 f
820
4.7
98.6
53.1
152 0.9
0.483
21 f
950
6.4
104
67
171 0.4
0.362
18 f
760
5.1
169
80.7
250 0.3
0.727
18 f
820
4.2
133
34
167 0.5
0.659
21 f
810
845
4.7
4.8
92.8
274
10.5
120
103 0.5
394 0.4
0.307
0.633
20 m
19 m
53
39
Issaquah
41
Issaquah
47
Makah 70
Makah 72
Makah 73
Makah 75
Makah 76
Makah 77
Makah 79
Makah 80
Makah 81
Makah 82
Quinault
101
Quinault
102
Quinault
103
Quinault
104
Quinault
106
Quinault
107
Quinault
108
Quinault
109
Quinault
111
Quinault
113
Sound
Puget
Sound
Puget
Sound
coast
coast
coast
coast
coast
coast
coast
coast
coast
coast
830
5.6
128
10.5
850
880
866
833
804
800
826
834
894
810
880
5
7.3
8.9
6
6.4
6.9
6.9
6.1
8.2
5.4
7.3
114
69.3
77
43.4
102
112
109
62.8
81.5
63.1
73.8
coast
898
7.9
coast
845
coast
139 0.6
0.315
19 m
10
10
9
9.5
9.5
9.5
8.5
9.5
9.5
10
9
124
79.3
86
52.9
112
122
118
72.3
91
73.1
82.8
0.5
1.2
2
1.5
2.5
1
2
0.9
1.6
1
1
0.357
0.327
0.148
0.167
0.491
0.397
0.308
0.171
0.237
0.263
0.259
21
21
22
21
21
21
23
21
21
20
23
64.1
8.5
72.6 2.1
0.182
23 f
5.6
45
9.5
54.5 1.1
0.195
21 f
925
7.9
45.3
9.5
54.8 0.8
0.094
22 f
coast
900
6.8
36
9.5
45.5
1
0.121
21 f
coast
851
5.5
36
9.5
45.5 1.1
0.135
21 f
coast
975
9.6
65.3
9
74.3 1.3
0.181
22 m
coast
845
6.6
50
8
58 3.5
0.184
26 m
coast
1010
13.2
106
34.8
141 3.8
0.206
27 m
coast
960
9.3
69.9
9.5
79.4 1.3
0.209
21 m
coast
960
10.7
56
9
65 2.1
0.107
22 m
54
m
f
f
f
f
m
m
m
m
f
m
55