-
Title
-
Eng
Antimicrobial Resistance in Orcinus Orca Scat: Using Marine Sentinels as Indicators of Pharmaceutical Pollution in the Salish Sea
-
Date
-
2013
-
Creator
-
Eng
Potter, Sara Louise
-
Subject
-
Eng
Environmental Studies
-
extracted text
-
ANTIMICROBIAL RESISTANCE IN ORCINUS ORCA SCAT:
USING MARINE SENTINELS AS INDICATORS OF PHARMACEUTICAL
POLLUTION IN THE SALISH SEA
by
Sara Louise Potter
A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
December 2013
©2013 by Sara Louise Potter. All rights reserved.
This Thesis for the Master of Environmental Studies Degree
by
Sara Louise Potter
has been approved for
The Evergreen State College
by
________________________
Dr. Erin Martin
Member of the Faculty
________________________
Date
ABSTRACT
Antimicrobial resistance in Orcinus orca scat:
Marine sentinels as indicators of pharmaceutical pollution in the Salish Sea
by Sara Louise Potter
Antimicrobial drugs revolutionized health care, but drugs and antibiotic resistant
bacteria (ARB) entering surface waters may increase resistance, alter bacterial
populations, or create reservoirs of resistance genes transmittable to pathogens.
This study assessed risk factors for colonization with ARB in the scat of Orcinus
orca to determine if geographic, temporal, or animal traits related to resistance
prevalence and/or patterns. Samples were collected June-October 2012 and
August-October 2013, using a scat detection dog to locate feces. Eleven samples
were plated on agar infused with ampicillin, chloramphenicol, erythromycin, and
tetracycline for colony count and multidrug resistance assessment. PCR
amplification of the resistance-conferring genes ermB, mecA, tetB and tetM was
performed on 32 samples. The study area was divided into segments based on
watershed traits, and distance from shore, number of septic tanks, wastewater
treatment plants (WWTPs), land area and human population density were
analyzed for each sample based on segment. Animal age, sex, and pod were
studied as organism risk factors. Number of colonies, presence/absence of ARB
gene, and multidrug resistance (MDR) rates were independent variables. A total
of 1730 resistant colonies were cultured from 11 samples, with erythromycin
ranking first in prevalence and total resistant colonies, followed by ampicillin,
tetracycline, and chloramphenicol. The effect of sampling location on number of
colonies was significant (F2, 8= 6.78, p = 0.019), and a sample obtained from the
Southern Gulf Island site had more ARB colonies and was significantly higher in
bacteria resistant to ampicillin (F2, 8 =19.75, p=8.05×10-4), erythromycin (F2, 8
=5.36, p=0.03), and enteric bacteria resistant to ampicillin (F2, 8 =37.07,
p=8.99×10-5). This site has more WWTPs, including a plant that recently used
only primary treatment, but no environmental factors were statistically related to
colonies or MDR by regression analysis. Results showed 4 samples positive for
the tetM gene, and although all 4 were from J-pod females in the same geographic
segment, results were not significant. This study is the first to report positive
identification of ARB in the feces of Orcinus orca, and though no specific
environmental relationships were identified, the prevalence of ARB warrants
further research, particularly into factors such as time and travel before sampling,
to better understand the factors impacting ARB colonization in the SRKW and the
use of marine mammals as sentinel species for pharmaceutical pollution in the
Salish Sea.
TABLE OF CONTENTS
Abstract
Table of Contents………………………………………………………………....iv
List of Figures ……………..………………………………………...….…………v
List of Tables……………………….……………………………………...……..vii
List of Abbreviations……………………………………………………..……..viii
Acknowledgments……………………………………...………………...……....ix
Introduction……………………………………………………………………......1
Chapter One: Literature Review…………………………………...........................6
Chapter Two: Manuscript…………...……………………………………..……..52
Chapter Three: Discussion and Interdisciplinary Evaluation……….….…….…104
References……………………………………………………………..…..…....115
iv
List of Figures
Figure 1. Map of study site divided into six primary watershed units. Boat
docking location noted with red star………………………………....pg. 93
Figure 2. Map of culturing sample locations, 2013………………………….pg. 93
Figure 3. Map of genetic analysis sampling locations, 2012. Samples positive for
tetM gene are denoted with red pentagons…………………………..pg. 94
Figure 4. Sex distribution from DNA and hormone data for 2012 samples
(n=32)…………………………………………………………….…..pg. 94
Figure 5. Pod representation from DNA genotyping identification of 2012 PCR
assay samples (n=32)……………………………………………..….pg. 95
Figure 6. Total number of colonies grown on antimicrobial-infused plates by drug
type represented on logarithmic scale. Growth on TSA and MacConkey
plates differentiated by color, actual colony counts on bar………….pg. 95
Figure 7. Incidence of MDR in cultured samples 1-11. Four antimicrobials were
used in agar plates, thus four is the maximum MDR number……….pg. 96
Figure 8. Principle component analysis for samples separated by antimicrobial
and bacterial growth type (MacConkey or TSA) representing patterns in
abundance of ARB colonies cultured in samples 1-11 and control samples
1-2 (12 and 13). The relationship between samples is determined by the
similarity of ARB colony growth number for each antimicrobial plate, and
the direction determining spatial distance is demonstrated in the loading
plot below…………………………………………………………….pg. 97
Figure 9. Principle component analysis for samples separated by antimicrobial
representing patterns in abundance of ARB colonies cultured in samples 111 and control samples 1-2 (12 and 13). Loading factor map below shows
the dimensions between variables which determined the distance between
samples in the individual factors map………………………………..pg. 98
Figure 10. Averaged numbers of ARB colonies by study site segment……..pg. 99
Figure 11. Total number resistant isolates in each samples as function of distance
from shore (m)…………………………………………………….....pg. 99
Figure 12. Total ARB cultured as a factor of time using Julian date. No significant
relationship was found (y = -0.2561x + 217.48, R² = 0.0004)…..…pg. 100
Figure 13. ARB colony prevalence by drug resistance type related by Julian date,
2013. No significant relationships found and thus not reported.…..pg. 100
Figure 14. Scatterplot representing number of days the whales had been in basin
before sampling by total ARB count. Although the r2 value is low and
v
suggests no relationship, the small sample size (n=11) and uneven
distribution between the number of days before sampling may have had an
effect on the decreasing trend of ARB shedding in fecal as time in the
Salish Sea progressed. Further examination of this aspect of resistance
should be studied……………………………………………...…….pg. 101
Figure 15. DNA concentration in ng/L of extract used as PCR template…..pg. 101
vi
List of Tables
Literature Review
Table 1. Antimicrobials ranking in the top 200 drugs by number of prescriptions
in the United States for 2011……..………………………………...pg. 49
Table 2. Concentration range of six antimicrobials in WWTP influent and effluent
in the Puget Sound Basin, WA, 2008……………………………...pg. 49
Table 3. A brief summary of antimicrobial drug families, compounds,
mechanisms, and target bacteria………………………………...…pg. 50
Manuscript
Table 1. Names, doses, and abbreviations of antimicrobial drugs added to agar
plates in this
study………………………………………………………………..pg. 90
Table 2. Total growth from fecal sample cultures, summed in last row and column
by number of resistant colonies by antimicrobial and by sample. C1 and
C2 correspond to the two control samples. Refer to Table 1 for
antimicrobial abbreviations…………………………………….…..pg. 91
Table 3. Geographic segment division and environmental risk factors for
analysis……………………………………………………....……..pg. 92
Table 4. Functional metagenomic results for resistance genes in 3 SRKW fecal
samples from 2012 field season…………………………………... pg. 92
vii
List of Abbreviations
Amp: Ampicillin
ARB: Antimicrobial resistant bacteria
Cm: Chloramphenicol
ermB: Erythromycin-ribosomal methylase B
Erm: Erythromycin
km: Kilometers
m: Meters
mecA: Methicillin-resistance gene A
MDR: Multidrug resistance
NaCl: Sodium chloride
ng/L: nanogram per liter
PCPP: Personal care product and pharmaceutical pollution
PCR: Polymerase chain reaction
SDR: Single drug resistance
SRKW: Southern Resident Killer Whale
tetB: Tetracycline-resistance gene B
tetM: Tetracycline-resistance gene M
Tc: Tetracycline
TSA: Tryptic soy agar
µg/L: Microgram per liter
WWTP: Waste water treatment plant
viii
Acknowledgements
This research was truly an interdisciplinary and collaborative effort, and without
many supportive organizations and individuals, would not have been possible.
The support of University of Washington’s Department of Conservation Biology
and the Conservation Canines program was paramount to the completion of this
thesis. Program director Dr. Samuel Wasser permitted use of samples, and
primary researcher Jessica Lundin was a helpful mentor. I would like to thank
Washington SeaGrant and NOAA for sponsorship of the program. The guidance
of Dr. Marilyn Roberts of University of Washington’s Department of Public
Health was invaluable in protocol development and for laboratory analysis. Her
keen knowledge of environmental microbiology was critical to this work.
I would like to thank my reader, Dr. Erin Martin, for guiding me through the
thesis process, offering professional advice and marvelous revisions. Her tireless
support for my thesis research was crucial to my success in the MES program. I
am eternally thankful for her contribution to the program and to my professional
career.
The Moja field crew, Amanda Phillips, Elizabeth Seely, Deborah Giles, and Kari
Koski, worked together through adverse conditions to collect samples, and
without them, long hours in the field would have been more like work and less
like an adventure. I cannot express my gratitude for the experiences we’ve shared
and the chance to work, live, and learn from you amazing, intelligent women.
Additional thanks to Jim Rappold, Doug McCutchen, Paul Arons, Sharon Grace,
Sandy Buckley, Darvis Taylor, and the community of San Juan Island for support
and companionship. And of course, thanks to Tucker, Sadie, Pepsi and Waylon,
the Conservation Canines who tracked down samples for their endless love of the
ball.
I’m thankful for the friendship of Shelby Proie, whose devotion to the release of
captive whales brought her to Washington and taught me to love and respect these
amazing animals. Her activism for Lolita’s freedom is inspiring, and I hope to
contribute to research that will preserve the Salish Sea for Tokitae’s return.
I would like to thank my family for their continued support of my choices in
career and development as a scientist, even when this means living a continent
away and scooping feces for a living.
Gabriel, you made me fierce. Thank you for the love and support, and living with
whale poop in the freezer for much of my graduate school career. This book is for
you.
ix
"We have the ability to provide clean water for every man, woman and child on
the Earth. What has been lacking is the collective will to accomplish this. What
are we waiting for? This is the commitment we need to make to the world, now."
-Jean-Michel Cousteau
x
INTRODUCTION
Since the introduction of antimicrobials in the 1940’s, bacteria have
shown an increased response in resistance. Antimicrobials are the third-largest
group of medicines prescribed for humans, and the largest category of medicines
used in veterinary practices (Thiele-Bruhn, 2003). These antimicrobial drugs enter
the environment directly through flushing unused drugs down the drain, and
indirectly through unmetabolized compounds excreted by human and animal
waste through wastewater effluent and leaking septic tanks,, and runoff and
drainage from agricultural lands and aquaculture sites (Cabello, 2006; Kümmerer,
2004; Okeke and Edleman, 2001; Zhang et al., 2009). Increased input of
antimicrobial drugs creates opportunities for environmental and pathogenic
bacteria to develop selective resistance to pharmaceuticals due to the ease with
which bacteria exchange genetic material and the speed at which they reproduce.
More virulent and resistant bacteria are artificially selected by this increased
exposure, and also threaten water quality through altered environmental bacteria
populations. Given that increased levels of resistance in human gut bacteria can
persist for up to two years after treatment has stopped (Jakobsson et al., 2010), the
massive amount of antimicrobials that we are adding to our environment now may
have long term consequences.
Increased human population and decreased wild habitat has intertwined
the health of humans, wildlife, and the environment, breaking down divisions
between these disciplines in the assessment of ecosystem health (King et al.,
2008). This new paradigm, exemplified by the growing One Health movement,
1
has created a more collaborative and holistic approach to monitoring ecosystem
health through the use of animal sentinels (Rabinowitz and Conti, 2013). Marine
mammals are a good sentinel species in marine and aquatic research due to their
physiological relationship to humans, their position as apex predators, and their
visibility as a keystone species in the ecosystem (Bossart, 2006; Wong, 2002).
Fecal sampling can be employed as a non-invasive alternative to
capture/release in study of threatened species or in areas where tracking an animal
for other biological indicators is difficult. Feces contain a wealth of information
on the animal, including genetic, hormonal, and toxin loading, and the nutrientrich intestinal environment is particularly conducive to microbiological research
(Kohn and Wayne, 1997; Miranda and Zemelman, 2001; Taberlet et al., 1999).
Research using the intestinal bacterial flora of marine vertebrates as
indicator species for ARB pollution has shown varying degrees of resistance.
ARB presence has been studied in predatory fish, marine birds, pinnipeds, and
various whale and dolphin species by culturing bacteria from rectal or fecal swabs
(Blackburn, 2010; Grieg, 2007; Johnson et al., 2008; Lockwood et al., 2006;
Miller et al., 2008; Miranda and Zemelman, 2002; Rose et al., 2008; Schaefer et
al., 2011; Schroeder et al., 2009; Stoddard et al., 2002). The levels of resistant
samples and the complexity of resistance patterns has been positively correlated
by sampling site to suspected contamination points, ex. higher number of
wastewater treatment plant (WWTP) outflows (Grieg et al., 2007), number of
septic tanks (Schaefer et al., 2011), human population density (Blackburn et al,
2010; Grieg et al., 2007), and freshwater outflows (Johnson et al., 2008). Rainfall
2
and weather events have also been related to increase antibiotic resistance
findings in short-term temporal analyses (Grieg et al., 2007; Schaefer et al.,
2011).
The Salish Sea is an ecologically diverse and economically important
estuary in the Pacific Northwest region of the United States of America and
southwestern British Columbia, Canada (Fraser et al., 2006; Gaydos et al., 2009).
This glacial carved fjord inland marine ecosystem stretches along the inland
waters from Olympia, WA in the United States to the Campbell River in Canada,
and is the home to over 7 million people and many endemic and rare marine
species (Jackson and Kimerling, 2003). Like many of the world’s coastal zones,
population growth and increased toxin and fuel loading has amplified pressure on
the marine habitat and wildlife in the Salish Sea (Gaydos et al. 2009; Puget Sound
Partnership, 2011). Population growth has increased loading to septic and waste
treatment systems, increasing likelihood that harmful substances, including
pathogenic bacteria, toxins, and other chemical compounds will reach the water
column without being properly broken down (Take Back Your Meds, 2011).
Groundwater, storm water, and combined sewer overflows discharge untreated
and industrial wastewater and compounds leached from leaky septic drainage
systems when capacity is exceeded or water levels rise above the drainage fields
also contribute to surface water pollution (Puget Sound Partnership, 2011;
Dougherty et al., 2010). Research on marine bacterial resistance in Washington
shows that ARB can be found in marine sediment near aquaculture sites, the seasurface, and public beaches (Herwig et al., 1996; Roberts et al., 2008; Soge et al.,
3
2009). When analyzing sand samples from Washington beaches, five distinct
strains of multi-drug resistant methicillin-resistant Staphylococcus aureus
(MRSA) and 33 methicillin-resistant coagulase-negative Staphylococcus
(MRCONSA) spp were identified (Soge et al., 2009) Roberts et al. (2008) identified
18 resistant strains of vancomycin-resistant Enterococcus in Washington public
beaches. The presence of these resistant and pathogenic strains makes a case for
the possibility of ARB genes in offshore marine waters, and understanding the
transmission of these genes is critical to assessing the risk of the marine
environment serving as a reservoir of resistance genes transmittable to pathogens.
In the Salish Sea, abundant data is available the endemic ecotype of the
Orcinus orca, or Southern Resident Killer Whale (SRKW). Concerns over decline
in the SRKW population resulted in a large collection of data on the life history,
family lineage, geographic range, and individual identity of most whales. The
population consisted of 86 individuals at the 2012 spring census and declined to
80 by the fall of 2013, during the time of this study (Center for Whale Research,
2013). Declines in this population are primarily thought to be due to increased
toxin accumulation from human pollutants, stress from lack of nutrition with
declining salmon population, and underwater noise pollution, and other proposed
threats include pathogens (NOAA, 2011).The importance of this species as a
cultural icon, recreational draw, and keystone species make them ideal for the
study of antimicrobial resistance in marine mammals. Prior studies using marine
vertebrates as sentinel species have shown positive correlation between human
influence and ARB colonization to some degree (Blackburn, 2010; Grieg, 2007;
4
Johnson et al., 2008; Lockwood et al., 2006; Miller et al.,2008; Miranda and
Zemelman, 2002; Rose et al., 2008; Schaefer et al., 2011; Schroeder et al., 2009;
Stoddard et al., 2002), but no research has been able to draw robust conclusions
on anthropogenic effects on ARB acquisition due to the lack of a natural history
of the animals surveyed and a vague sense of their geographic range, confounding
variables not applicable to the study of ARB in this species.
In this study, those confounding variables were limited by studying
bacterial resistance caused by pollution of antimicrobials and wastewater by
analyzing prevalence and patterns of ARB colonization in the feces of the wellstudied Orcinus orca. By using a sentinel species for colonization with ARB and
pairing prevalence and patterns with demographic, geographic, and anthropogenic
risk factors, the relative influence of human impacts as well as traits intrinsic to
ARB susceptible wildlife can be assessed. Results indicate how other species
could be affected by exposure to waters affected by pharmaceutical pollution and
determine if ARB are a problem of relevance to public health in the Salish Sea
marine environment.
5
CHAPTER ONE
LITERATURE REVIEW
An evaluation of current literature is needed to frame the research question
in a manner that validates the problem and contributes to an explanation of the
chosen study design. For this, sufficient background on antimicrobial drugs,
complications of their entry into the environment, and how the phenomenon could
potentially impact the Salish Sea and wildlife therein is critical, as well as the
findings from similar research in order to adopt methods and establish anticipated
results for this study.
Pharmaceutical Pollution in the Salish Sea
Study Site Characteristics and Vulnerabilities
The Salish Sea is an ecologically diverse and economically important
estuary in the Pacific Northwest region of the United States of America and
southwestern British Columbia, Canada (Fraser et al., 2006; Gaydos et al., 2009).
This glacial carved fjord inland marine ecosystem stretches along the inland
waters from Olympia, WA in the United States to the Campbell River in Canada,
and is the home to over 7 million people and many endemic and rare marine
species (Jackson and Kimerling, 2003). Like many of the world’s coastal zones,
the Salish Sea is declining environmentally as human population pressure and
pollutants begin to accumulate (Gaydos et al. 2009; Puget Sound Partnership,
2011). Human population growth in the Pacific Northwest is putting pressure on
aquatic habitat and marine mammals and industrial effluent has resulted in
increased toxin and fuel concentration. About 100 municipal and industrial
6
wastewater treatment plants discharge over 430 million gallons of treated
wastewater to imperiled waters of Puget Sound and the surrounding area each day
(Puget Sound Partnership, 2011). Groundwater, storm water, and combined sewer
overflows can discharge untreated wastewater, and harmful compounds leach
from leaky septic drainage systems when capacity is exceeded or water levels rise
above the drainage fields (Dougherty et al., 2010; Puget Sound Partnership,
2011). Added input to septic and sewage systems from population growth
increases the chances that harmful substances, including pathogenic bacteria,
toxins, and other chemical compounds will reach the water column without being
properly broken down (Take Back Your Meds, 2011).
Pharmaceutical Pollution
Personal care product and pharmaceutical pollutants (PCPPs) are classified
as emerging contaminants of concern in water quality, and have become an
increasingly significant focus area in water quality research. Investigation of
pharmaceutical pollution is important because most drug wastes are biologically
active, readily mobile and not easily biodegraded (Kümmerer, 2008). A
biologically active compound can have direct physiological effects on living cells,
according to the theory that small molecules can change cellular phenotypes in
cells of non-target organisms, thereby altering cellular activity and function and
imitating the effects of genetic mutations (Klekota et al., 2005). The high mobility
of drugs and drug metabolites is due to a high water solubility, and results in high
dispersal rates of drugs once released to surface waters (Kümmerer, 2008). These
properties are exacerbated by the persistence certain drugs and their metabolites
7
in many environments, the unregulated use of many drugs in animal agriculture,
inefficient drug absorption by target species, the incomplete removal of excreted
or dumped compounds by water treatment facilities, and the myriad pathways of
indirect exposure of antimicrobials to the environment (Kümmerer, 2004;
Kümmerer, 2009). Pharmaceutical exposure to the environment can come from
WWTPs, untreated sewage, hospitals, runoff from croplands, animal feces, and
aquaculture activities (Bushman et al., 2002; Dougherty et al., 2010; Johnson et
al., 2004; Puget Sound Partnership, 2011; Take Back Your Meds, 2011).
The effectiveness of water treatment in removal of these contaminants
depends on the volume and type of chemicals received in waste as well as the
design on the treatment center (Lubliner, 2010). There are 96 publicly owned
waste water treatment plants (WWTPs) in the Puget Sound Basin, processing over
124 million gallons of sewage from over 3.5 million people each day (WA
Department of Ecology, 2010). A study conducted by the Washington State
Department of Ecology in 2008 analyzed 172 PCPPs in influent, effluent, and
holding tank grab samples from five WWTPs in the Puget Sound Basin (Lubliner,
2010). The study found that 56% of the analytes were detected in at least one
sample, and each sample had measurable concentrations of some combination of
PCPPs (Lubliner, 2010). About 21% of the compounds detected in the influent
were reduced below detection levels by secondary treatment alone, and 32% of
the remaining compounds were below detection levels after treatment with an
advanced tertiary technology (Lubliner, 2010). Three pharmaceuticals,
carbamazepine, fluoxetine, and thiabendazole, were identified as being untreated
8
by tertiary treatment and recommended as environmental tracers for PCPP
pollution (Lubliner, 2010). Most of the WWTPs in Puget Sound use secondary
treatment, which focus on “removing pathogens, biochemical oxygen demand,
toxic chemicals, and suspended solids, with the primary objective of protecting
human health” (Puget Sound Partnership, 2011). In other words, targeted removal
of pharmaceuticals is not typically conducted by the WWTPs emptying into the
Puget Sound Basin.
Examining the presence and concentration of PCPPs in urbanized areas and
comparing findings to a more remote “pristine” area helps to develop a sense of
the extent and severity of these pollutants in the Salish Sea. Researchers
compared concentration and incidence of detection of 37 dissolved anthropogenic
compounds such as in the surface waters of Barkley Sound and south Puget
Sound, where population density is approximately six times higher in the latter
site (Keil et al., 2011). In the control area of Barkley Sound, 28 of the 37
chemicals were detected at least once, but in the more populous Puget Sound
region all 37 chemicals were detected. Only two plasticizers were detected in
Barkley Sound more frequently than in Puget Sound, and average concentrations
of most the three most abundant compounds in both fjords was approximately 20
times higher in Puget Sound. The survey found the pharmaceuticals ibuprofen and
17a-ethynylestradiol, a hormone commonly found in birth-control pills in both the
natural and altered study sites. These pharmaceuticals were hypothesized to
originate from human sources, with higher concentrations of the chemicals
correlating to higher human population density (Keil et al., 2011).
9
A similar study conducted in Liberty Bay, WA found herbicides,
pharmaceutical and other personal care products in both ground water and surface
water in an area lacking a WWTP, supporting the hypothesis that non-point
source pollution of human waste contributes to this phenomenon (Dougherty et
al., 2010). Liberty Bay is a small embayment of the Puget Sound section of the
Salish Sea, with varying degrees of urbanization and use of 70% septic disposal.
The authors found that seven of the 12 compounds detected were detected in other
studies examining septic systems as potential sources for PPCPs, suggesting that
non-point pollution of surface waters in pristine areas could also contribute to
drug pollution in the Salish Sea. The concentrations of these drugs in both studies
were in sub-toxic levels, but the occurrence of such drugs in any quantity in
relatively pristine areas lacking a clear route for point-source contamination
speaks to the magnitude of the problem (Dougherty et al., 2010; Keil et al., 2011).
Antimicrobial Drug Pollution
Within the category of PCPPs, antimicrobial drugs are of special concern in
the environment. Antimicrobials are substances produced by microorganisms to
hinder growth of other microorganisms and thereby favor the propagation of their
own species (Hogg, 2005). Antimicrobials used in medical treatment are either
the metabolites of these organisms, known as allelochemicals, natural chemical
compounds analogous to these substances, or synthetic compounds mimicking the
inhibition of specific bacterial growth to target undesirable bacteria (Grunden,
2013).
10
Antimicrobials are the third-largest group of medicines prescribed for
humans and the largest group used in veterinary medicine (Thiele-Bruhn, 2003).
Worldwide, annual consumption of antimicrobials is estimated to be between
100,000 and 200,000 tons per year (Wise, 2002). In the United States alone, 1860
tons of antimicrobials are used in human medicine per year, resulting in 1.9 ug/L
appearing in sewage and 0.73 ug/L appearing in surface waters (Kümmerer,
2008). In the US, outpatient antimicrobial usage is relatively steady in the ranking
of top drugs (Bearden, 2013). The most commonly consumed antimicrobials are
the β-lactam family group of drugs which include penicillin, ampicillin,
amoxicillin and others, trailed by tetracycline, macrolide, fluoroquinolone and
sulfonamide family groups (see Table 1) (Bartholow, 2012; Kümmerer, 2008).
Although no data on prescription rate and use of antimicrobials is kept for a
state or regional basis, the Center for Disease Control is expanding its scope to
begin doing so (Bearden, personal communication). National data on most
common outpatient antibiotic prescriptions is available yearly, and hospitals do
track prescription rates, but outpatient antimicrobial use dwarfs that of the
hospital, making hospital estimates for a region an underrepresentation of true use
(Bearden, 2013).
Antimicrobial concentrations in the environment are found in some orders
of magnitude lower than in therapeutic use (Kümmerer, 2004). The environmental
persistence of an antimicrobial varies based on its chemical structure and the
environmental compartment where it is accumulating. Β-lactams are less stable in
the environment because of their easily hydrolyzed ring structure, but drugs like
11
the sulfonamides, fluoroquinolones and macrolides are especially severe
environmental contaminants because of their higher persistence and greater
stability (Cha, 2006; Segura et al., 2009). The primary method of pollution
elimination in most environmental environments is bacterial decomposition,
which is less effective against synthetic or semi-synthetic antimicrobial drugs or
in bacterial populations that have become less diverse due to unnatural selection
from the presence of resistant bacteria or pharmaceuticals (Barbosa and Levy,
2000; Kümmerer, 2004).
Continued exposure to most of these chemicals even at trace levels has
unknown consequences, and the combined action of this exposure in addition to
other routes is possibly cumulative (Smith, 2012). Continuous low doses to an
environmental system are thought to be more detrimental than high
concentrations, because the presence of antibiotics over a longer period of time
stimulates the transfer of resistance genes between bacterium and the propagation
of already resistant bacteria is favored by the decline in numbers and decreased
competition from susceptible species (Barbosa and Levy, 2000; Kümmerer,
2004). This concept is covered more thoroughly in the next section.
Water, sediment and fish tissue in Puget Sound have been previously shown
to have measureable concentrations of antimicrobials (Corcoran and Tyler, 2010;
Johnson, Carey and Golding, 2004; Nilsen et al., 2007). As was the case in the
Keil et al. (2011), there seems to be spatial relationships between urbanization and
higher chemical detection rates of antimicrobials in the environment (Dougherty
et al., 2010). In Liberty Bay, WA, data show that the community, which
12
predominately uses septic tanks rather than centralized WWTPs, is receiving
detectable levels of trimethoprim in surface and ground waters, with the
occurrence of detections increasing as population density increased (Dougherty et
al., 2010). The authors speculate that detections are likely to increase as
populations, prescriptions, and household product uses increases (Dougherty et
al., 2010).
Revisiting the Washington State Department of Ecology study, of the 172
compounds tested in WWTP influent and effluent, 24 made the department’s
short list for the highest priority PCPPs to test in environmental studies based on
use, toxicity, consumption, properties, and persistence (Lubliner, 2010). Six of
these 24 high-risk compounds have antimicrobial uses and included erythromycin,
sulfamethoxazole, tetracycline, triclosan, triclocarban, and trimethoprim
(Lubliner, 2010). Study results indicate that secondary and tertiary treatments
remove a majority of antimicrobial concentrations in waste water, but detectable
levels of these chemicals are still present in effluent released to surface waters
(see Table 2). Other antimicrobials not designated as high risk were also detected
in the effluent, including azithromycin, ciprofloxacin, clarithromycin,
clinafloxacin, cloxacillin, enerofloxacin, lyncomycin, lomefloxacin, norfloxacin,
ofloxacin, ormetoprim, oxacillin, oxolinic acid, penicillin G, penicillin V,
roxithromycin, sarafloxacin, virginiamycin, anhydrochlortetracycline,
chlortetracycline, demeclocycline, doxycycline, and other tetracycline
derivatives in effluent (Lubliner, 2010). Interestingly, veterinary drugs were
intentionally left off of this list with no reasoning given.
13
Introduction of antimicrobial resistance to the environment is thought to
only be partially caused by pharmaceutical pollution, with other resistance being
attributed to natural resistance or introduction of resistant bacteria through
wastewater or fecal pollution. A review of resistance is necessary to better
understand how bacteria transfer this trait in the natural world.
A Review of Antimicrobial Resistance
Antimicrobial Drugs
As noted above, antimicrobials are biochemical compounds derived from
or mimicking complexes designed to promote the growth of a particular
microorganism at the expense of another. The natural phenomenon of allelopathy
has been harnessed by scientists to produce a variety of drugs that suppress or
eliminate the growth of undesirable bacteria in human and veterinary medicine.
The five basic mechanisms of antimicrobial action are (Rollins and Joseph, 2000):
1. Inhibition of cell wall synthesis, caused by a compound
inhibiting peptidoglycan synthesis, precursors, or linkages,
which terminates cell wall growth and lyses the cell.
2. Alteration of cell membranes, Antimicrobials that employ this
method essentially injure the plasma cell wall membrane,
which disrupts the cell cross-membrane potential and causes
leaks and imbalances.
3. Inhibition of translation. This process disturbs protein synthesis
by binding proteins or peptides to ribosomal subunits, which
14
disrupts peptide elongation and results in the cell being unable
to correctly copy and pass on genes. This mechanism is
primarily bacteriostatic, except for the aminoglyceride class of
antimicrobials.
4. Inhibition of nucleic acid synthesis. Drugs that inhibit DNA
gyrases or DNA-dependent RNA gyrases, which prevent DNA
coiling and replication.
5. Antimetabolite activity. These drugs prohibit the synthesis of
folic acid, thereby preventing the synthesis and repair of DNA
within the cell.
The success or failure of an antimicrobial against a bacterium is dependent
on the natural defenses of the target cell, cellular metabolism, and cellular growth
processes. Therefore, not all antimicrobial mechanisms are effective against any
particular microbe, and different antimicrobial families using different
mechanisms can be used to target specific bacterial types (see Table 3 for a basic
summarization). This becomes important in the study of antimicrobial resistance
in the environment, as resistance rates can be overrepresented if the predominate
bacteria in the ecosystem are intrinsically resistant to antimicrobials used
(Kümmerer, 2004; Lorian, 1996).
Antimicrobial Resistance Acquisition and Mechanisms
Resistance is acquired by natural mutations or the transfer of genetic
material, known as vertical and horizontal transfer, respectively. Vertical transfer
of resistance genes is the result of spontaneous mutations for resistance within the
15
bacterial genetic code, which occur at an estimated frequency of 1 out of every
108-109 alleles (Todar, 2011). During DNA replication, resistance genes are
passed on to all descendants and the population gains resistance over time
(Barbosa and Levy, 2000; Khachatourians, 1998).
Horizontal transfer mechanisms are of primary importance to this study.
Horizontal transfer is the acquisition of heritable traits by the modification or
transfer of genetic material from one individual bacterium to another (Todar,
2011). The three methods of genetic material transfer are conjugation,
transduction, and transformation. Conjugation is the transfer of plasmids
containing DNA packets from one bacterium to another via direct cell-to-cell
contact (Todar, 2011). The transfer of resistance via plasmids is thought to be the
main mechanism of horizontal gene transfer, and bacterial species need not be
similar for conjugation to occur (Alekshun and Levy, 2007). Transduction is the
transmission of resistant DNA by bacteriophages, or bacterial viruses that infect
similar bacterial species (Todar, 2011). Transformation occurs when free DNA
from the environment is absorbed by a microbe. Resistance genes present in the
environment from the lysis or death of a bacterium can be assimilated into the
DNA of a completely different bacterial species, often through transposons or
gene cartridges (Alekshun and Levy, 2007).
The genetic codes for resistance manifest as a number of different
mechanisms that prepare a bacterium to survive antimicrobial treatment. Three
major mechanisms of bacterial adaptation have been identified to combat the five
mechanisms of antimicrobial action. These mechanisms are the enzymatic
16
breakdown or inactivation of antimicrobials, modification of receptor sites and
development of efflux pumps or revision of metabolic pathways (Hogg, 2005).
Spread of Resistance
Resistance is led into the environment by the introduction of resistant
bacteria or by unnatural selective pressure resulting from antimicrobial drug
pollution. Antimicrobial resistance to some drugs is natural in certain bacteria, but
clinical antimicrobial medication and human waste disposal methods favored the
selection and propagation of bacteria with genes encoding additional resistance
(Baquero et al., 2008; Kümmerer, 2004). Over time, genetic drift of microbial
populations subject to this selective pressure will increase the number of resistant
bacteria in the environment because only bacteria with these traits will survive to
multiply (Barbosa and Levy, 2000). The artificial selective pressure to maintain
resistance genes in a bacterium is documented to be increased with contact of
human wastes in environments, including water (Greig et al., 2007; Kümmerer,
2004; Kümmerer, 2008; Miller et al., 2008; Miranda and Zemelman, 2002). This
is because low and continued trace levels of antimicrobials, such as those emitted
from waste water treatment, are even more effective in conveying resistance,
which makes the emission of small doses even more concerning (Lorian, 1996).
This is known as subtherapeutic dosing, and this process kills out the weaker
strains of bacterium, naturally selecting for stronger and more virulent strains in a
very short amount of time (Lorian, 1996; Shnayerson and Plotkin, 2002).
Reversal of Resistance
17
There is evidence that by decreasing the input of antimicrobials to a
system, the bacteria will gradually shed the genetic material that makes resistance
advantageous, and order can be restored over time (Kümmerer, 2004). The
majority of studies on reduction of resistance have been conducted in clinical
environments, with some successful cases in decreasing the number of resistant
pathogenic bacterial strains after alterations in hospital policy (Klare et al., 1999;
Smith, 1999; van den Bogaard et al., 2000). Additionally, Corpet et al. (1998)
showed that after feeding volunteers a near sterile diet, the numbers of resistant
bacilli in their fecal flora decreased almost 1000-fold. These findings offer some
hope that means that by reducing antimicrobial use, it is possible to reverse some
of the effects of continued resistance.
The banning of agricultural antimicrobials in the European Union has
provided an interesting case study for tracking productivity in agricultural animals
and human pathogenic resistance after the removal of selective pressure. Sweden
banned of all food animal growth-promoting antibiotics by in 1986, and the
European Union banned avoparcin in 1997, followed by bacitracin, spiramycin,
tylosin and virginiamycin in 1999 (Casewell et al., 2003). Recent studies
performed in several European countries following the ban in the use of avoparcin
report an encouraging and sometimes dramatic decrease in the frequency of
certain pathogens resistant to the banned antimicrobials in animals and food
products as well as the microbial flora of humans (Klare et al., 1999, van den
Bogaard et al., 2000).
18
However, contrary evidence exists to discredit the theory that bacteria will
naturally shed genetic material coding for resistance as pressure is reduced.
Studies have shown that resistance can be found in bacteria in people and animals
without recent history of antimicrobial drug use (Calva et al., 1996; Gilliver et al.,
1999). And though there may be reductions in the frequency of resistance
detection, research suggests that the resistance does not return to a pre-exposure
level and that decreases in resistance are much slower to evolve than increases
(Austin et al., 1999; Barbosa and Levy, 1992). The delay in return to normal
susceptible bacterial flora provides opportunity for resistance to return if the
antimicrobial is reintroduced, as well as the development of cross-resistance to
other antimicrobial agents (Barbosa and Levy, 1992).
Antimicrobial Resistance in the Environment
Antimicrobial misuse is common and considered a serious problem (CDC,
2011; FDA, 2012; Harris, 2009; Shnayerson and Plotkin, 2002; Union of
Concerned Scientists, 2012). The misuse by humans in physicians (Harrison and
Svec, 1998) and patients (CDC, 2011) in human clinical medicine has created
virulent strains that are not able to be combated with typical, or even multiple,
antimicrobials. Commonplace products used in cleaning and health products also
contribute to increased antimicrobial exposure to the environment, in addition to
unregulated antimicrobial use in animal agriculture and aquaculture.
Resistance in Humans and Health Implications
19
Antimicrobial resistance bacteria are most often considered a problem
from the standpoint of clinical medicine. The formation of resistant bacterial
strains pose a serious threat to human health (da Costa et al., 2006; Kunin, 1993,
Shnayerson and Plotkin, 2002). Virulent pathogens are transmitted through
community infections (particularly in hospitals), infected resistant bacterialcontaminated foods, and through contact with water, soil or wildlife that is
contaminated. Any number of these situations can lead to infection of the
consumer, which is more difficult and costly to treat because of the
ineffectiveness of readily available antimicrobials.
A recent study focused on costs in medical treatment and mortality from
antimicrobial resistant bacterial infections in a Chicago, Illinois hospital. The
researchers concluded that 13.5% of 1391 patients surveyed were infected with an
antimicrobial resistant microbe (Roberts et al., 2009). Medical costs incurred by
each patient due to their resistance ranged from $18,588 to $29,069, and the
hospital stay for these patients was 6.4 – 12.7 days longer (Roberts et al., 2009).
Most alarmingly, mortality rates in this hospital alone due to antimicrobial
resistant bacteria were 6.5% which is twice the rate for patients without
antimicrobial resistant infections (Roberts et al., 2009). The anticipated societal
costs, or costs for the families of the ARB infected patients was estimated to fall
between $10.7 and $15 million (Roberts et al., 2009). Roberts extrapolated this
cost to hospitals nationwide, and conservative estimates of antimicrobial resistant
infection were at 900,000 in the year 2000, equating to $16.6 to $26 billion
20
dollars spent on the treatment of these preventable infections (Roberts et al.,
2009).
This data is from the year 2000, and the rise of antimicrobial resistant
infections implicates that this problem has only grown. A more recent report from
the World Health Organization found that death toll from the pathogen
methicillin-resistant Staphylococcus aureus (MRSA) is approximately 18,000
people per year in the United States alone (Braine, 2011). Other emerging
resistant infections include vancomycin-resistant Enterococcus, vancomycinintermediate/resistant Staphylococcus aureus, carbapenem-resistant
Enterobacteriacaea, carbapenum-resistant Klebsiella pneumonia,
fluoroquinolone-resistant Neisseria meningitides, isoniazid, rifampicin and
floroquinolone resistant Mycobacterium tuberculosis, multi-drug resistant
Acinetobacter baumannii, Bacillus anthracis, Neisseria gonorrhoeae, Group
B Streptococcus, Shigella spp., Streptococcus pneumonia, Salmonella spp. (CDC,
2013).
The discovery of the New Delhi Metallo-β-Lactamase group of enzymes is
especially disturbing, and has attracted significant scientific and media attention.
This finding deserves consideration, as the gene encoding for resistance to the
beta-lactam family of drugs is very mobile, complex, and adaptable to many
different pathogens (Dortet, 2012). The promiscuity of the gene known as blaNDM1
has eliminated the use of carabapenems, which in the 1980s and 1990s were
considered a last resort against extended-spectrum β-lactamase gram-negative
bacteria (Bonomo, 2011).
21
Concern over growing antimicrobial-resistant pathogen strains, the loss of
emergency antimicrobials for resistant infections, and the lack of development of
new antimicrobial drugs has resulted in a panic over global ‘superbugs’. Policies
opposing the USDAs lenient rules regarding agricultural antimicrobials, the
imprudent prescription of antimicrobials for unrelated illnesses in clinical
medicine, and encouraging research and development of new antimicrobials that
would be more strictly regulated have been proposed to help ameliorate the
growing health crises (Interagency Task Force on Antimicrobial Resistance,
2011).
Resistance in Animal Agriculture
The inappropriate use of antimicrobials in animal agriculture is a major
contributor to the emergence of ARB. The annual quantity of antimicrobials in
agriculture is 100 to 1000 times the use in human populations. In agriculture,
antimicrobials are added in trace amounts to food or water to prevent illness in
animals kept in close quarters or as a growth promoter (Khachatourians, 1998;
Levy, 1998; Peak et al., 2007). Ninety percent of the drugs administered are given
at continual subtherapuetic levels rather than to treat illness, which promotes the
selection of ARB in the gut of the animals and promotes the spread of resistance
(Chadwick et al., 1997; Khachatourians, 1998). In the United States, current
regulations have restricted the use of certain antimicrobials that are used in human
medicine to quell the rate that the medicine becomes ineffective for human use,
but the majority of drugs used to treat animals remain unregulated and do not
require a prescription from a veterinarian to dispense (FDA, 2012).
22
Many agricultural animals have been shown to harbor antibiotic resistant
bacteria, including cattle, poultry, and swine (Chee-Sanford et al., 2001; Peak,
2007; van den Bogaard, 2002). Moreover, the soil and waste water in pastures
where treated animals are raised has been shown to have increased resistance
levels, and runoff of animal waste from farms is a suspected contributor of ARB
to surface waters (Dougerhty et al., 2010; Santamaria et al., 2011; Yang et al.,
2010). Additionally, it has been shown that meat produced from animals treated
with antibiotics can transfer genetic resistance (Teubner, 1999).
Resistance in Aquaculture
Prophylactic antimicrobials are used in aquaculture to compensate for
overcrowding of fish farming sites, increased density of fish numbers in
aquaculture, unsanitary conditions, and the failure to isolate sick fish (Cabello et
al., 2006; Naylor et al., 2000). It has been shown that the heavy use of
antimicrobials in aquaculture causes residual traces of chemicals to remain in the
local sediment (Kruse and Sorum, 1994; Miranda and Zemelman, 2002; Sorum,
2006) and can increase the proportion of resistant bacteria in nearby sediment
(Herwig et al., 1997; Huys et al., 2000; Miranda and Zemelman, 2002; Schmidt et
al., 2000; Sorum, 2006). The most concerning impact of antimicrobial use in
aquaculture is that unconsumed food reaches the sediment and drugs can be
leached directly from the food pellets in addition to passing through fish feces,
and they can be washed by currents to distant sites (Cabello, 2006). Additionally,
research is showing decreased diversity of bacteria in sites with continuous use of
23
antimicrobials (Cabello, 2006; Davies et al., 2009; Herwig et al., 1997; Smith,
2008; Sorum, 2006).
Resistance in the Environment and Ecosystem Implications
Antimicrobial resistance is increasingly framed as an ecological problem.
Understanding and documenting the transmission of nonpathogenic ARB is
important because these bacteria can serve as reservoirs of resistance genes to
pass on to pathogens (Okeke and Edelman, 2001). Additionally, bacteria are an
ecologically critical group of organisms. They are essential in the biochemical
cycling of nitrogen, phosphorus, sulphur and oxygen, and are key decomposers,
with the resulting nutrients released often fueling primary production in the ocean
and stimulating the global carbon cycle (Kümmerer, 2008). Documentation of
resistant strains and presence of antimicrobials in the environment is an important
step in understanding the impact human wastes are having on bacterial mutation
rates, and consequently the health of the marine environment and public health in
the region.
ARB in the Marine Environment
This problem is significant in marine and aquatic environments in
particular because the persistence of chemicals in the sediment or in small
concentrations in the water column has the potential to alter the normal bacterial
flora and plankton in the affected areas, shifting the diversity of the
microorganisms in a way that potentially disrupts other processes necessary for
ecosystem health. For instance, decreased diversity has the potential to alter
24
trophic and metabolic functions and promote anoxic environments, contributing to
algal blooms and fish kills (Cabello, 2003; Valiela, 1995). These and other
environmental disturbances are not as well defined by media sources when
discussing antimicrobial pollutants, but are as, if not more, critical to the health of
humans and the ecosystem as a whole.
Research indicates that antimicrobial resistant bacteria can be found not
only in marine sediment near aquaculture sites (Herwig et al., 1996), but also in
public marine beaches on the west coast. Soge et al. (2009) collected water and
intertidal sand samples from 10 public beaches in Washington State and
optimized growth for Staphylococcus spp. The resultant 51 isolates from 9 of the
10 beaches were exposed to chloramphenicol, trimethoprim/sulfamethoxozole,
erythromycin, and tetracycline by disk diffusion analysis. PCR assays were then
used to identify ermA, ermB, ermC, msrA, tetM and tetK, specific genes
encoding for antimicrobial resistance in Staphylococcus exposed to the
aforementioned antimicrobials (Soge et al., 2009). Outcomes identified five
distinct strains of multi-drug resistant MRSA, which were more phylogenetically
similar to hospital-acquired MRSA infections than community acquired
infections, as compared to 4 methicillin susceptible strains of Staphylococcus
(Soge et al., 2009). At the time of press, this was the first report of MRSA and
MRCONSA isolated from marine water and intertidal beach sand, suggesting that
the marine environment may serve as a reservoir for antimicrobial resistant genes
(Soge et al, 2009). Nine of the MRSA strains had characteristics commonly
associated with hospital MRSA, though the beach sites were not near hospitals
25
and the cool water temperatures make infected human swimmers unlikely as the
source of the contamination, and the cause of contamination is unknown but not
ruled out as coming from a single source (Soge et al, 2009).
This study was expounded upon in 2012, with water and sand samples
collected from two marine water beaches [A and B] and one freshwater beach [C]
in the Seattle WA area (Levin-Edens et al., 2012). A total of 31 MRSA isolates
representing 21 different strains were identified, 71% from the fresh water
drainages and creeks surrounding marine Beaches A and B and/or fresh water
Beach C (Levin-Edens et al., 2012). MRSA isolates in this study were 67.7%
SCCmec IV, which often originates in untreated wastewater, while in the Soge et
al. study 83.3% of the isolates were SCCmec type I indicating hospital MRSA,
which implies a change in the relative contribution of methicillin resistance genes,
or possibly an artifact from changes in sample size and the addition of a
freshwater beach (Levin-Edens et al., 2012). The two studies agree that the
diversity of MRSA isolates support the hypothesis that MRSA is progressively
distributed in the environment via multiple sources including human WWTP
effluent, hospitals, human contact, and wildlife (Levin-Edens et al., 2012).
This conclusion is supported by research from Roberts et al. (2009), who
used a similar sampling protocol to identify vancomycin-resistant Enterococcus
(VRE) on beaches in Washington and California. This study detected 18 isolates
with vanA and/or vanB genes, in addition to tetM genes capable of plasmidmediated transfer (Roberts et al., 2009). The VRE strains also expressed
resistance to clindamycin, minocycline, tetracycline and teicoplanin (Roberts et
26
al., 2009). VRE were isolated from the samples collected in Washington in 2002,
2003 and 2008, but not from 2001, suggesting a temporal accumulation of
resistance genes in the environmental pool (Roberts et al., 2009). Samples from a
WWTP emptying into Puget Sound in 2001 were not positive for vanA or vanB,
but without retesting effluent in the years of positive identification of VRE there
is no relationship attributable to human influence (Roberts et al., 2009). A later
study conducted by Roberts et al., in 2013 evaluated 296 recreational beach
samples for MRSA, of which 31 (10.5%) were positive for MRSA with 22
isolates (71%) coming from fresh water streams running into the marine and
freshwater beaches. Again, the MRSA strains had characteristics commonly
associated with hospital MRSA but with no nearby hospitals, suggesting that
upstream freshwater influx into marine environments significantly contributes to
the pool of resistance in marine near-shore waters (Roberts et al., 2013).
A separate study examined differences in all antibiotic resistance gene
signals across the surface water samples of the Salish Sea as related to a WWTP
water sample. Seven samples were taken from six locations in the surface waters
of Puget Sound and a single effluent sample was collected from a WWTP
emptying into the basin of Puget Sound (Port et al., 2012). Eighteen resistance
genes were identified from the samples, with tetracycline resistance being most
common, and the abundance of genes increase in resistance from open waters to
the WWTP, suggesting a spatial relationship between resistance gene abundance
and human impacts (Port et al., 2012).The WWTP effluent resistance gene profile
showed similarities to resistance gene taxonomy from human infections,
27
suggesting human impact, and the spatial differences across the open water
samples did not alter their taxonomic similarity or profiles of mobile genetic
elements (Port et al., 2012). The authors suggest that resistance gene signals were
underestimated, which they attribute to the methodology of pyrosequencing (Port
et al., 2012).
The findings from these studies show that antimicrobial resistance in the
marine environment exists on the West Coast and waters of the Salish Sea,
including multi-drug resistant pathogens of clinical importance. This is
concerning because evolved bacterial resistance to an antimicrobial usually
manifests as a single resistance gene, and the abundance of multi-drug resistant
bacteria identified along public beaches implies that horizontal transfer of mobile
resistance genes is occurring in Washington’s marine waters (Miller et al., 2008).
ARB Colonization in Animals
Terrestrial Animals
Concern over antimicrobial resistant bacteria has led to studies on the
presence of resistant strains present in human and livestock, but resistance of
strains in wildlife are slower to be studied (Blackburn et al., 2010). Antimicrobial
resistance has been previously documented in various animals primarily from
fecal or intestinal sampling, due to the richness of animal enteric systems being a
prime site for bacterial colonization and the simultaneous presence of many
pathogenic or zoonotic bacteria being commonly found in the gut.
28
Ahmed et al. (2007) studied the occurrence of gram negative AR genes in
bacteria from zoo animal fecal and anal swabs and found that captive animals are
a potential reservoir for clinically important resistance genes. This study
highlighted the incidence and wide range of species affected, testing a variety of
birds, turtles, tortoises, monkeys, snakes, salamanders, foxes, giraffes, badgers,
tigers, and other reptilian and mammalian species (Ahmed et al., 2007). These
results show that feces from animal species can be used to isolate antimicrobial
resistant bacteria to monitor resistance in a contained environment. However, this
research was not intended to apply resistance as an indicator of ecosystem health,
expand upon possible reasons for varied findings in or between species, or relate
the findings to any environmental phenomenon.
To relate resistance to environmental factors, Edge and Hill (2005) studied
the relative contribution of fecal pollution sources to resistance in surface waters.
They studied resistant E. coli strains in surface waters and used a discriminant
function to distinguish the bacteria from the two sources. They concluded that
feces from seabirds had lower levels of resistance, but contributed more to fecal
pollution in the surface waters, highlighting the importance of considering wild
fecal sources in environmental analysis of resistance. Conversely, Parveen et al.
reported a lower incidence in fecal samples from terrestrial wild animals than
found in the surrounding waters. In agreement with this finding, Rose et al.
reported a higher resistance occurrence in seabirds than in marine mammals and
fish from nearby waters, speculating that the greater contact of the seabirds with
coastal communities and their diet of refuse may contribute to this phenomenon.
29
However, the sampling method depended on bycaught animals to represent a
healthy standard population, which by their own admission needs more
investigation. This study was limited in the number of samples, and swabbed
different areas for different species depending on their provenance (live, bycaught
or stranded), further confounding results and accuracy in measuring the
occurrence of ARB in coastal waters.
Marine Animals
To further explore the distribution of ARB in marine waters, study has
expanded to include different types of bacteria in off- and near-shore marine
species, in an attempt to elucidate reasons behind varied patterns of resistance
with other phenomenon.
Studies on penguins and polar bears have been used to spatially relate
resistance to human population, since they are primarily located in remote areas.
Sea turtles, fish, sharks, marine birds, pinnipeds and cetaceans have also been
used as environmental indicators for antimicrobial and fecal pollution, and
resistant bacteria have been detected in all of these animals (Al-Bahry et al.,
2010;Blackburn et al., 2010; Glad et al., 2010; Greig et al.,2007; Lockwood et al.,
2006; Miller et al., 2008; Miranda and Zemelman, 2001;Rose et al., 2008;
Schroeder et al., 2009; Stoddard et al., 2008).
Antimicrobial resistance in bacteria from seawater and penguin fecal
samples was studied in Antarctica, where international treaties have limited
antimicrobial contamination. Thus, this study serves as a measure of human
30
impacts on the environment and a potential snapshot of the natural resistance in
the pre-antimicrobial world (Miller et al., 2008). The authors found that drug
resistance was higher in introduced than endemic microbes, and that the ratio of
introduced to endemic bacterial spp. and resistance bacteria increased with
proximity to Palmer Station (Miller et al., 2008). The research concludes that
even in relatively pristine areas, the frequency of drug and multi-drug resistance is
low among native bacteria, and can be increased by even the most regulated of
human habitation (Miller et al., 2008).
Another study of resistance in an area of relatively little impact was
conducted on the microbiome of Polar bear (Ursus maritimus) fecal (Glad et al.,
2010). The researchers amplified the beta-lactam resistant blaTEM gene, and
discovered that only 4 out of 144 isolates were positive for blaTEM genes from the
wild polar bears (Glad et al., 2010). The blaTEM gene is increasingly found in
clinical and commensal bacteria, suggesting that proximity to development may
contribute to resistance in fecal samples, along with phylogeny and diet (Glad et
al., 2010). However, culturing fecal material from the polar bears on ampicillininfused agar plates resulted in no growth, which signifies that all bacteria were
susceptible to ampicillin. The use of molecular techniques in addition to bench
culturing was necessary to elucidate the exact proportion of bacteria capable of
resistance, which is important to consider in the development of further research.
Additionally, it was discovered that that more fecal samples than rectal swabs
showed evidence of blaTEM genes, suggesting that fecal samples are superior for
the study of resistance (Glad et al., 2010).
31
ARB were studied as bio-indicators of contaminated effluent in green
turtles in Oman using oviductal fluid as the biological test agent. To test for
exposure of turtles to pollution, fluid was acquired from forty nesting turtles,
resulting in 132 species of bacteria from 7 genera (Al-Bahry et al., 2010). Of
these bacteria, the Kirby-Baur disk diffusion method was used to test resistance to
15 antimicrobials, and 60.6% showed multi-drug resistance, mostly expressed for
ampicillin, trailed by streptomycin and sulphamethoxazole (Al-Bahry et al.,
2010). The authors concluded that testing exposure to polluted effluents using
wildlife bacteria as bio-indicator is a valuable way to assess endangered species
and ecosystem health (Al-Bahry et al., 2010).
It was also observed that green turtles are exposed to many different types
of effluent and pollutants in their wide geographical and migratory regions, and
since it is not practical to investigate all along the migratory route, monitoring of
ARB is useful to find the extent of pollution (Al-Bahry et al., 2010). This
assumption is important, because the authors accept that resistance travels with
the turtles during migration rather than being acquired at the nesting sites, though
the time period of travel and spatial range of each individual turtle is unknown.
This assumption is unproven and unbacked in the text.
Another study using marine species as an indicator for effluent pollution
used fish caught in Concepcion Bay, Chile (Miranda and Zemelman, 2001). This
study was designed to determine the frequency of resistant bacteria in fish, to
evaluate potential differences in the frequencies based on pelagic or demersal
habitat, and to determine resistance patterns of some selected specific bacteria of
32
human health interest. Miranda and Zemelman (2001) showed that bacteria
isolated from demersal fish showed higher multi-resistance levels than fish from
pelagic habitats, demonstrating resistance to up to 10 antimicrobials while pelagic
fish samples yielded bacterial strains resistant to seven antimicrobials or less, with
high frequencies of antimicrobial resistance for ampicillin, streptomycin, and
tetracycline. (Miranda and Zemelman, 2001). This difference was not
significantly significant, but suggests that feeding and ecological habits, as well as
exposure to the high content of resistant bacteria in polluted sediment (Herwig et
al., 1997) may lead to higher contact of antimicrobial agents.
Another study concentrating on antimicrobial resistance in marine
vertebrates by Blackburn et al. (2010) focused on predatory fish and sharks off the
coast of Florida. The study goals were to determine if prevalence of ARB
decreases as distance from shore increases, if animals from the same species in
different locations exhibit different resistance patterns, and if resistance patterns
differ within different species in the same area. Anal swabs of fish were cultured
and resulting isolated subjected to the Kirby-Baur test for drug resistance to
antimicrobials (Blackburn, 2010). Results showed that resistance was ubiquitous
in the marine environment and multidrug resistance was common in areas with
larger human populations (Blackburn et al., 2010). Older fish had higher
incidences of resistance, and bacteria were most often resistant to penicillin,
ticarcillin, cefitofur, doxycycline, and chloramphenicol (Blackburn et al., 2010).
The most common resistances were not the same in fish from Chile to fish
in Florida; in fact, chloramphenicol resistance was uncommon in Chile and the
33
most common in Florida, suggesting that local intrinsic resistance or antimicrobial
use may play a role in the development of resistance.
Both the Miranda and Zemelman paper and the Blackburn study found
resistance in marine fish through culturing in agar and testing isolates for
resistance using the standard clinical Kirby-Baur disk diffusion test. The
exclusion of anaerobic bacteria and lack of investigation into the genes encoding
for resistance is likely misrepresenting the proportion of resistant bacteria in
marine fish. Anaerobic bacteria account for up to 99% of bacteria in the intestinal
tract of most animals (Moore, 1969), thus improvements in both studies would be
to attempt to culture both aerobic and anaerobic bacteria or utilize biochemical
methods like Glad et al. (2010) to identify resistance genes so as to not
underrepresent resistant bacterial numbers. However, investigating the expression
of resistance through the Kirby-Baur method is also important, as genes for
resistance may not always be expressed, although they can still be passed on to
other bacterium.
Another shortcoming in both of these studies in respect to drawing
conclusions about human impacts is the lack of a natural history of the animals
and a vague sense of their geographic range. Small sample sizes and lack of
repeated sampling of an individual fish or the coupling of resistance findings with
a life history that would support age approximations, sex determination and actual
geographic range of the fish would support the analysis of fish risk-factors for
colonization with ARB. These confounding variables makes geographic analysis
34
less robust and weakens the ability of the author to draw conclusions on the
human contribution to the resistance rates.
Marine Mammals
Additional research has been conducted on the antimicrobial resistance of
bacteria gathered from marine mammals. Revisiting Rose et al. (2008), swabs
from the anus, feces, and tissue of marine fish, birds, pinnipeds and cetaceans
were gathered along the coast of the Northeastern United States. The goal of the
study was to determine differences in ARB bacteria type and occurrence among
different marine vertebrates and to determine if the provenance of the animal
affects the ABR pattern. From 64 marine mammals, 174 isolates were collected,
50% of which showed resistance to at least one antimicrobial. While this is lower
than the 68% of resistant isolates collected from seabirds, and marine mammal
sampling may be biased due to sampling bycaught and stranded animals, the
study indicates that there is a high incidence of ARB in the marine environment.
Carbenicillin, augmentin, ampicillin, and cepholothin were ineffective against
more that 25% of the isolates, denoting a large environmental pool of ARB in the
marine environment. This study did make use of anaerobic culturing, providing a
more accurate estimation of the culturable ARB in the samples.
Several studies have concentrated on ARB in pinnipeds. Stoddard et al.
(2008) used elephant seals on the coast of California to determine potential
environmental and demographic factors associated with ABR in three specific
pathogenic bacteria. Sex, weight, county, month, coastal human population,
35
exposure to sewage or freshwater outflow, and precipitation in the previous 24hr,
7 d, 30 d, 90 d, and 180 d were variables considered in risk analysis using rectal
swabs from live and stranded seals as the source of bacteria. Broth microdilution
methods to test for resistance against many antimicrobials were employed,
resulting in high resistance to ampicillin and tetracycline (Stoddard et al., 2008).
Elephant seals that were stranded, closer to a freshwater outflow, or in areas of
high human population density demonstrated higher rates of resistance and were
resistant to more drugs.
In a study from Lockwood et al. (2006), bacterial cultures collected over a
period of 12 years from stranded harbor seals in Puget Sound were evaluated to
define common pathogens and their ARB patterns. Cultures from wounds,
umbilici, ears, conjunctiva, nares, oral lesions, and feces bore 134 pathogenic
isolates, which were most resistant to amikacin (99%) and gentamicin (97%), and
least affected by ampicillin (26%). Again, the possible bias related to the
provenance of the seals and the use of culturing rather than biochemical analysis
may muddle the results, but the findings of resistant pathogens in marine
mammals in Puget Sound is encouraging for the prospect of using marine
mammals as an indicator species for resistance in the study site basin.
Study of cetacean antimicrobial resistance patterns is more relevant to this
study. Greig et al. (2007) studied E. coli isolates from bottlenose dolphins in
Florida and South Carolina to determine if resistance between sampling sites is
homogenous and applied a population genetic analysis to estimate within-animal
isolate diversity, and identify reason for different resistance patterns in different
36
study sites. Findings indicated that in ARB, prevalence and complexity of ARB
patterns increased in rectal samples taken near more developed areas compared to
rural areas, implying WWTP outflows influence ARB presence in cetaceans
(Greig et al., 2007). Resistance was detected in 19 of 25 antibiotics, with
resistance to penicillin being most common followed by cephalothin, ampicillin,
and amoxicillin (Greig et al., 2007). These antimicrobials represent the betalactam and cephalosporin drug classes, some of the oldest and most commonly
used drugs in human medicine.
An important assumption made by these authors is that the dolphins that
were captured had spent considerable time in the locale in which they were
sampled. The range of the bottlenose dolphin is considerable, and there is no
reason to believe that the sampled dolphins were native to the Florida coast or
South Carolina bay. As seen in all of the papers regarding resistance and marine
vertebrates, the range and natural history of the animals were assumed, which
weakens links between geographic location and human influence to ARB carriage
in these animals.
The lack of information on geographic range of the individuals prior to the
sampling event calls into question the relationship between time and the retention
of resistant bacteria. It is currently unknown how long ARB may colonize a
cetacean and if the effects of human influence are long-term or immediately
apparent once an animal enters ARB or antimicrobial-poisoned waters.
Arguments for the quick loss of resistance genes in a bacterium once the artificial
pressure is removed are viable, because a bacterium wants to expend all possible
37
energy and genetic material on the ability to reproduce, and can shed these genes
as they become obsolete (Kümmerer, 2004). Contrarily, ARB colonization and
resistance patterns may be long lasting due to pathogenic nature and research
indicating the specificity of resistance patterns. Firstly, if an animal is a carrier of
ARB genes and those genes have taken hold in a pathogen, the virulence of the
pathogen is increased and potentially make the animal sick. Once the animal is
infected, bacterial infections can spread throughout the pod, and whether or not
this results in sickness or mortality the genes for resistance may continue to be
passed back and forth between family groups (Gaydos et al., 2004). Second,
researchers have been able to trace the sources of fecal contamination in
subtropical and other water bodies to specific sources by analyzing the resistance
patterns of enteric bacteria collected from the water (Harwood et al., 2000). This
was possible due to the astounding specificity of the genetic patterns for
resistance genes in contamination from a specific site. If these resistance genes
were able to be analyzed and traced from water in an oligotrophic system, a more
nutrient-rich environment, such as the enteric system of a cetacean, would retain
more bacterial populations and thus may contain more bacterial density and/or
diversity.
The critique of studies above suggests that improvements to the research
of antimicrobial resistant drugs and bacteria using marine sentinels would begin
by choosing an animal with a known geographic range, a well-studied natural
history, known dietary preferences, and that can be resampled over time. An ideal
marine sentinel for in the Salish Sea would be the southern resident killer whale
38
(SRKW), an ecotype of Orcinus orca with annual census reports dating back to
the 1970s. This ecotype has a long lifespan similar to humans, eat almost strictly
salmon, with high preference for Chinook, and spend most of the summer months
in the Salish Sea basin from the Strait of Georgia south to Bainbridge Island and
from the San Juan Islands west to the mouth of the Strait of Juan de Fuca (Ford et
al., 2000). Thus, to study the phenomenon of spreading resistance due to human
pollution in the south Salish Sea, the similarities between humans and the SRKW
and the degree of information available on their life history makes them an ideal
indicator species.
One experiment on ARB in SRKW has been conducted, concentrating on
the blow (exhalation upon rising to the surface to breathe) respiratory bacteria.
Schroeder et al. (2009) collected orca blow samples in Puget Sound by closely
following individual whales and holding exposed petri dishes with various agars
above surfacing orcas, catching the exhalation as they rose (Schroeder et al.,
2009). Instances of ARB were discovered in the respiratory tract, comprised of
different bacteria of clinical and veterinary importance, including Salmonella,
Vibrio, Pseudomonas, and Bacillus spp (Schroeder et al., 2009). The most
common drugs to which resistance was expressed were lincomycin, sulfatrimethoprim, and ampicillin-sulbactum (Schroeder, 2009).
Animal Sentinels in Environmental Surveillance
Marine Mammals as Sentinel Species
39
A sentinel species is evaluated to identify negative trends in an
ecosystem’s health and to better manage the possible impacts of these effects on
human and animal health (Bossart, 2006). Marine mammals are probably one of
the best sentinel organisms in aquatic and coastal environments because of their
closer physiological relationship to humans, long life spans, their position as apex
predators in their food chain, have extensive fat stores that can serve as depots for
anthropogenic toxins and their visibility as a sentinel species for aquatic health
(Bossart, 2006).
Use of Scat as Biological Indicator
The use of scat samples as a method of biological sampling of wild animal
populations has become popular and even standard in the field of conservation
biology. While researching existing research on ARB in marine vertebrates, fecal
or anal swabs were used by the majority of researchers (Blackburn et al., 2010;
Glad et al., 2010; Grieg et al., 2007; Miller et al., 2009; Schaefer et al., 2011;
Stoddard et al., 2008), with gill (Miranda and Zemelman, 2001; Rose et al., 2009),
breath (Schroeder et al., 2009), or stomach contents (Miranda and Zemelman,
2001) used less often. Thus, fecal sampling is the most logical choice for this
study to make it comparable to current literature.
There are several benefits to using scat for study of a wild species. First,
fecal sampling can be employed as a non-invasive alternative to capture/release in
study of species that are threatened or endangered. Under Endangered Species Act
law, permitting for study of wild animals is required. These permits are specific
40
contracts with particular numbers (called “takes”) that limit the disturbance a field
crew may make to an animal population. Compliance with this permit is critical
for the completion of the study, and the right to research endangered species may
be revoked at any time. The non-invasive nature of scat sampling removes many
of these problems, and although Endangered Species Permits may still be
required, permit violation and more importantly animal harm may be reduced by
using non-invasive techniques.
Second, feces contains a wealth of information on the animal, including
genetic, hormone, and toxin levels (Kohn and Wayne, 1997; Taberlet et al., 1999).
Using advanced molecular biology techniques, population size, genetic variation,
kinship, paternity, sex determination, pathogen sequences from bacteria, viruses,
protists and other macroparisites, and food sources can be determined (Kohn and
Wayne, 1997). From hormone levels, physiological stress, reproductive status,
pregnancies, and aborted pregnancies can be determined for an animal (Wasser et
al., 2004). Toxin levels are also determinable through organic chemistry methods
and solid-liquid extraction of chemicals. The suite of biological information that
can be determined from a scat sample makes its collection useful for an array of
research goals, and thus this method is mutually beneficial to scientist of differing
disciplines.
Third, and particular to this study, the nutrient-rich intestinal environment
is particularly conducive to microbiological research (Kohn and Wayne, 1997;
Miranda and Zemelman, 2001; Taberlet et al., 1999). The nutrient-rich stomach of
marine fish has been shown to harbor more bacteria than surface waters, making
41
use of fecal samples more practical for identification of bacteria that may be in
low density in surface waters (Miranda and Zemelman, 2001).
The findings of Jakobsson et al., (2010) that increased levels of resistance
in human gut bacteria can persist for up to two years after treatment support the
use of scat as an indicator for contact with antimicrobials or ARB. Information
about the rate of shedding of resistant bacteria in the feces is not readily available,
but findings that long-term effects persist after antimicrobial treatment cessation
makes the analysis of ARB in scat worthwhile, and perhaps findings from this
study can even help answer the question of ARB residence time in mammalian
gut flora.
Characteristics of Orcinus orca SRKW Ecotype
The SRKW population that makes its home in the Salish Sea and the PNW
coast are some of the best-studied population of whales in the world, and
consequently their health and habitat is fairly well documented, which makes
them an ideal subject.
The SRKW is a subspecies of Orcinus orca that are genetically,
morphologically, and culturally distinct from other killer whale ecotypes of the
world, though a review of the endangered species listing did not consider them a
discrete population segment (Gaydos et al., 2004; Krahn et al., 2002). This subset
of a larger population of killer whales that inhabit all of the world’s oceans has
yet to receive a taxonomic distinction, which has led to the opposition of its
listing as an endangered species by some groups. However, this oversight is
42
apparent in the community of marine mammal scientists and currently some
researchers focus on cataloguing subspecies of the killer whale (Gaydos et al.,
2004). The SRKW population was devastated in the 1960’s, when capture of
animals for aquarium exhibits resulted in the death of 13 whales and the livecapture of 45 juveniles (Center for Whales research, 2013). The population
declined to approximately 70 animals in the mid-1970’s, spurning Endangered
Species Act listing of the SRKW. In line with recovery efforts, the population
rose as high as 100 individuals in 1995 after a trend of increasing population
throughout the 1980’s and 1990’s, but numbers are again falling (Center for
Whale Research, 2013). The SRKW population consisted of 86 individuals at the
2012 spring census, decreasing to 80 individuals by the fall 2013 census (Center
for Whale Research, 2013). The population is divided into three family groups,
called pods, dubbed J, K and L (Ford et al., 2000). This particular ecotype does
not breed or associate with the transient or offshore killer whales off the
Washington coast. The SRKW populations is also distinct in that the diet is
composed almost entirely of Chinook salmon, while off-shore and transient killer
whales are known to attack and consume harbor porpoise, seals, sea lions, and
occasionally other whales (Ford et al., 2000). This population are known as
‘residents’ because they return each summer and fall to the Salish Sea from
wintering habitats that are varied and less well defined (NOAA, 2011). The pods
spend 18-65% of their days in the San Juan Island area of the Salish Sea from
April to October, ranging to the Strait of Juan de Fuca, the Fraser River, further
south into Puget Sound, or in unknown waters for the remaining time (Hanson,
43
2010). Winter sightings of the SRKW since 2005 suggest a range from
Monterrey, California, USA, to Langara Island, British Colombia, CA (Center for
Whale Research, 2013). Toxin analysis of blubber biopsy samples have shown
differences in the proportion of toxins in the three pods, indicating that they likely
occupy different winter ranges (Krahn et al., 2007). California sightings of L and
K pods correspond to higher DDT/PCB ratios, suggesting that these groups travel
further south (Krahn et al., 2007). J pod has a higher PBDE/PCB ratio, which
suggests they winter where they consume prey closer to an urban source (Krahn et
al., 2007).
Family groups of whales are known as pods. Pods are all descended from
a central female ancestor, as the SRKWs are a matriarchal society and calves stay
with their mothers for life. A SRKW pod travels, feeds, and hunts together, and
their communication “language” is distinctive from other pods (Center for Whale
Research, 2012; Gaydos et al., 2004). Within pods, family groups are even more
tightly bound, and female descendants of the matriarch and their young will
virtually always be in close contact. Commonly, SRKWs are presumed dead if
they do not return to Puget Sound in the summer with their family group, since
deceased whales are rarely recovered.
One of the characteristics of the SRKW is the highly developed cultural
bonds (Gaydos et al., 2004). Scientists monitoring the behavioral ecology of the
whales have noted the complex community structure and familial attachments.
When all three pods converge, there is an elaborate greeting ceremony of sorts, in
which they form two lines facing one another and swim until they are nearly face
44
to face. This is followed by contact and touching, which seems to be similar to
play. The great matriarch and eldest of the whales, J-2, or “Granny”, is over 100
years old (Center for Whale Research, 2013).
Another interesting cultural distinction of this particular species of killer
whale is the return of the pods to Puget Sound from less well-defined wintering
habitats each summer. Historically, large salmon runs made the trip through the
Strait of Juan de Fuca energetically beneficial for the whales, when they could
feast on fat Chinook returning to the Fraser River and Hood Canal. The steep
decline in salmon populations in these runs have made the continued presence of
the whales in Puget Sound over summers seemingly unlikely, especially when
salmon stocks elsewhere are in boom. The return of the SRKW is thought to be a
cultural tradition more than a benefit to the whales, because the pods go different
places in the winter and only meet back up in Puget Sound during summers. It is
thought that most mating and calving is done in the relatively calm waters of the
Sound.
Unfortunately, this amazing tradition may be hurting the population. The
species is at a critically low number, and nutritional stress is proven to decrease
pregnancy rates (Ayres et al., 2011). Studies comparing birth rates to fish counts
have found more pregnancies occur when hormonal stress levels that indicate
nutritional deficiencies are absent, leading to the conclusion that more fish would
equal more whales (Ayres et al., 2011; Gaydos et al., 2004). Declines in this
population are primarily thought to be due to increased toxin accumulation from
human pollutants, stress from lack of nutrition with declining salmon population,
45
and underwater noise pollution (Taylor et al., 2013).The importance of this
species as a cultural icon, recreational draw, and ecosystem component makes
them ideal for the study of antimicrobial resistance in marine mammals.
Implications for Findings of Resistance
The significance of discovering antimicrobial resistance in the microbiome
of an indicator species are four-fold. Firstly, an increased number of resistant
bacteria suggest that human wastes are not being sufficiently treated in a way that
does not change the biological composition of the waters. A level of resistance
higher in developed or polluted areas than undeveloped or “pristine “areas would
show that human impacts are causing changes to the environment.
Secondly, the expression of resistance to anthropogenic antimicrobials
indicates genetic material in the ecosystem is actively encoding for resistance.
This becomes a natural reservoir of resistance genes that are capable of
transferring resistance to pathogenic or environmentally significant bacteria.
Similarly, if resistance is present in an indicator species, this organism has
become a vector for pathogenic resistance. Since the easy and quick genetic
exchanges between bacteria are so common, contact with the organism could
mean transfer of exotic resistant genes or pathogenic bacterial strains that are
harmful to human beings. Of the 1,461 recognized human diseases, roughly 60%
are pathogens able to move across species from animal to human lines and in the
past 30 years, approximately 75% of new emerging human infectious diseases are
zoonotic (King et al., 2008; Taylor et al., 2001; Torrey and Yolken, 2005).
46
The fourth implication of resistance in an indicator organism is the
possibility of bioaccumulation of antimicrobial compounds in the food web. If
antimicrobials are lipid-soluble, they may accumulate in small amounts in aquatic
primary producers, which are eaten by herbivorous, omnivorous, and then
predatory fish, such as the Chinook salmon. A study was conducted on the
bioaccumulation of tetracycline and oxolonic acid in blue mussels to assess their
potential for biomonitoring of antimicrobials in the marine environment (Le Bris
and Pouliquen, 2004). The discovery that these antimicrobials accumulated in the
soft tissues of the mussels indicates that biomagnification of antimicrobial
compounds in the marine ecosystem is a realistic possibility (Le Bris and
Pouliquen, 2004). Another study assessed the occurrence, distribution and
bioaccumulation of 22 antimicrobials in surface water, sediment and fish samples
and found that both ciprofloxacin and erythromycin exhibited potential
bioaccumulation in carp, with bioaccumulation factors of 3262 L/ kg and 4492 L/
kg (Gao et al., 2012). The study also found that sediments retained the majority of
antimicrobials, indicating that they could be a large reservoir of antimicrobial
compounds in marine environments (Gao et al., 2012).
Essentially, findings of resistance in the environment demonstrates the
presence of natural reservoirs of resistance and the phenotypic expression of
resistance in the environment, indicating modification of bacterial properties by
antimicrobial inputs to surface waters.
We can tell if antimicrobials are in the water column by using chemical
analysis, but to show a relationship between chemical occurrence and active
47
phenotypic expression of antimicrobial resistance, both chemical and biological
analysis from an indicator organism are necessary. The use of an indicator
organism is critical for creating this connection.
48
Table 1. Antimicrobials ranking in the top 200 drugs by number of prescriptions
in the United States for 2011. Repeat compounds produced by different
manufacturers not included. Some prescription totals unavailable (NA) (Bartholow,
2012).
Antimicrobial
Rank
Drug Name
Overall
Rank
Drug
Total
Prescriptions
1
Azithromycin
9
26,427,000
2
Amoxicillin
20
19,764,000
28
NA
Sulfamethoxazole/
3
Trimethoprim
4
Amoxicillin
Trihydrate/Clavulanate
Potassium
114
NA
5
Penicillin VK
166
NA
6
Cephalexin
173
NA
7
Clindamycin HCl
175
NA
Table 2. Concentration range of six antimicrobials in WWTP influent and effluent
in the Puget Sound Basin, WA, 2008 (Lubliner, 2010). All US WWTPs are
required to have secondary treatment of wastewater, which includes degradation of
the biological content of sewage, often through aerobic biological processes, as
opposed to primary treatment, which is merely screening of waste water. Tertiary
treatment includes some mechanism of water quality improvement before release
to surface waters, including removal of nutrients, additional filtration, or other
disinfection or odor-control methods.
Compound
Influent
Concentration
(ng/L)
Secondary
Effluent
Concentration
(ng/L)
Tertiary Effluent
of Reclaimed
Water
Concentration
(ng/L)
Erythromycin
255-556
154-327
nd-343
Sulfamethoxazole
2,770-4,010
2-1,830
2-104
Tetracycline
13-186
10-40
Nd
Triclosan
1,480-2,770
nd-805
nd-77
Triclocaraban
289-541
31-78
3-103
49
Trimethoprim
611-1,400
308-791
nd-294
Table 3. Antimicrobial drug families, key drugs, mechanisms and targets for action.
This list, while not all-encompassing, is demonstrative of the majority of drugs
covered in this research (adapted from EB Medicine, 2005).
Drug Family
Penicillin
Cephalosporins
Sub-families
Key drugs
Mechanism
Target
Bacteria
Aminopenicilllins,
antipsuedomonal
penicillins, antipsuedomonoal
penicillins with
beta-lactamase
inhibitor,
penicillinaseresistant
penicillins
Penicillin G,
methicillin,
ampicillin,
amoxicillin,
ticarcirillin,
pipracillin,
ticarcillin/clauvanate
Bactericidal
Gram +
Inhibits cell
wall synthesis
Some Gram
–
1st generation, 2nd
generation, 3rd
generation, 4th
generation
Cephalexin,
cefoxitin,
ceftriaxone,
cefepime
Bactericidal
Gram +
Interferes with
cell wall
synthesis
Some Gram
–
Some
Anaerobes
Some
Anaerobes
Imipenum,
meropenum
Carbapenems
Bactericidal
Gram +
Inhibit cell
wall synthesis
Gram –
Anaerobes
Floroquinolones
Extended
spectrum
floroquinolones
Ciprofloxacin,
norfloxacin,
levofloxacin,
moxifloxacin
Bactericial
Inhibit DNA
gyrase
Some Gram
+
Gram –
Atypicals
Macrolides
Erythromyin,
azithromycin,
clarithromycin
Bacteriostatic
Gram +
Inhibit protein
synthesis
Some Gram
–
Atypicals
Aminoglycerides
Gentamycin,
tobramycin,
amikacin
Bactericidal
Gram -
Inhibit protein
synthesis
50
Tetracycline
Tetracycline,
oxytetracycline
Bacteriostatic
Inhibit protein
synthesis
Some Gram
+
Some Gram
–
Atypicals
Some
Anaerobes
Clindamycin
Bacteriostatic
Gram +
Inhibits
protein
synthesis
Anaerobes
Bactericidal
Gram +
Inhibits cell
wall synthesis
and inhibits
RNA synthesis
Some
Anaerobes
Trimethoprim/
Bacteriostatic
Some Gram
+
sulfmethoxazole
Inhibits folate
synthesis
Vancomycin
Some Gram
–
Some
protozoans
Metronidazole
Chloramphenicol
Bactericidal
Anaerobes
Interferes with
electron
transport
Some
protozoans
and
parasites
Bacteriostatic
Gram +
Inhibits
protein
synthsis
Gram –
Anaerobes
51
CHAPTER TWO
MANUSCRIPT
Formatted and prepared for: Environmental Health Perspectives
First Choice: Environmental Health Perspectives
Second Choice: Marine Pollution Bulletin
Third Choice: Advances in Applied Microbiology
Fourth Choice: Journal of Environmental and Public Health
*NOTE: This manuscript is a preliminary draft submitted to fulfill graduation
requirements for The Evergreen State College Master of Environmental Studies
program. The following document has not been edited, reviewed, or otherwise
endorsed by any of the listed co-authors and serves only to exemplify the
potential final journal submission.
52
Influences on antimicrobial resistant bacteria colonization in Orcinus orca
scat: Using marine sentinel indicators for anthropogenic pollution in the
Salish Sea
Sara L. Potter1,2, Samuel K. Wasser2, Deborah A. Giles2,3, Elizabeth A. Seely2,4,
Amanda M. Phillips2, Jessica I. Lundin2, Erin E. Martin1, and Marilyn C. Roberts5
1
Department of Environmental Studies, Graduate Program on the Environment, The Evergreen
State College, Olympia, WA 98505
2
Center for Conservation Biology, Department of Conservation Biology, University of
Washington, Seattle, WA 98195
3
Department of Wildlife, Fish, &Conservation Biology, College of Agricultural and
Environmental Sciences, University of California, Davis, CA 95616
4
Conservation Canines, Center for Conservation Biology, University of Washington, Eatonville,
WA 98328
5
Department of Environmental &Occupational Health Sciences, College of Public Health,
University of Washington, Seattle, WA 98195
Author Contributions:
Correspondence should be addressed to: Sara L. Potter
615 North L St #7 Tacoma, WA 98403
Slp359@gmail.com,
potsar09@evergreen.edu
(253) 223-5544
Key words: Antimicrobial resistance, environmental health, fecal sampling,
marine mammals, Orcinus orca, pharmaceutical pollution, water
microbiology.
Acknowledgements: Thanks to Washington Sea Grant, NOAA’s Northwest
Fisheries Science Center, Canadian Consulate General, Center for
Conservation Biology, Conservation Canines, and Northwest Science
Association. We would like to thank Kari Koski, Doug McCutchen,
Sharon Grace, Paul Arons, Jim Rappold, Sandy Buckley, Lynn Minor,
Dave Ellifrit, Erin Heydenrich, Ken Balcomb, Joseph Gaydos, Jennifer
Hemplemann, Hilary Hayford, Emily Carrington, Friday Harbor Labs,
Center for Whale Research, The Whale Museum, San Juan Island
Conservation District, JISAO Internship Program, and the community of
San Juan Island. Special thanks to Tucker, Waylon, Sadie May and Pepsi,
for their noses and contribution to conservation research.
The authors declare no competing financial interests.
53
Abstract
Antimicrobial drugs revolutionized health care, but drugs enter surface waters
directly through drain disposal or indirectly through unmetabolized compounds in
human and animal wastes. This input of antimicrobials and resistant bacteria to
the environment may increase resistance, alter bacterial populations, or create
reservoirs of resistance genes transmittable to pathogens. This study assessed risk
factors for colonization with antimicrobial resistant bacteria (ARB) in the scat of
an Orcinus orca ecotype endemic of the Salish Sea, to determine if geographic,
temporal, or animal traits related to prevalence and/or patterns of resistance.
Samples were collected June-October 2012 and July-October 2013, using a scat
detection dog to locate whale feces as they floated in the water. Eleven samples
were plated on agar infused with ampicillin, chloramphenicol, erythromycin, and
tetracycline for colony count and multidrug resistance assessment. Polymerase
chain reaction (PCR) amplification of the resistance-conferring genes
erythromycin ribosomal-methylase B (ermB), methicillin-resistance gene A
(mecA), and tetracycline-resistance genes B and M (tetB and tetM), was
performed on 32 samples from at least 27 animals. The study area was divided
into segments based on watershed traits, and distance from shore, number of
septic tanks, wastewater treatment plants, land area and human population density
were analyzed for each sample based on segment. Animal age, sex, and pod were
54
analyzed as organism risk factors. Number of colonies, presence/absence of ARB
gene, and multidrug resistance (MDR) rates were independent variables. A total
of 1730 resistant colonies were cultured from all animals, with erythromycin
ranking first in prevalence and total resistant colonies, followed by ampicillin,
tetracycline, and chloramphenicol. The effect of sampling location on number of
colonies was significant (F2, 8= 6.78, p = 0.019), and a sample obtained from the
Southern Gulf Island site had more ARB colonies and was significantly higher in
bacteria resistant to ampicillin (F2, 8 =19.75, p=8.05×10-4), erythromycin (F2, 8
=5.36, p=0.03), and enteric bacteria resistant to ampicillin (F2, 8 =37.07,
p=8.99×10-5). This site has more WWTPs, including a plant that recently used
only primary treatment, but no environmental factors were statistically related to
colonies or MDR by regression analysis. Results showed 4 samples positive for
the tetM gene, and although all 4 were from females of the same pod, results were
not significant. The positive samples were also from the San Juan study area, in
contrast to culturing results. This study is the first to report positive identification
of ARB in the feces of Orcinus orca, and though no specific environmental
relationships were identified, the prevalence of ARB warrants further research.
55
1.
Introduction
Since the introduction of antimicrobials in the 1940’s, bacteria have
shown an increased response in resistance. Antimicrobials are the third-largest
group of medicines prescribed for humans, and the largest category of medicines
used in veterinary practices (Thiele-Bruhn, 2003). A 2002 estimate of global
antimicrobial consumption reports between 1 and 2 ×108 kg annually (Wise,
2002), and the FDA reported annual consumption in the United States as
approximately 1.6 x 107 kg, 80% of which is consumed by livestock for nontherapeutic treatment (FDA, 2009). These antimicrobial drugs enter the
environment directly through flushing unused drugs down the drain, and
indirectly through unmetabolized compounds excreted by human and animal
waste through wastewater effluent and leaking septic tanks,, and runoff and
drainage from agricultural lands and aquaculture sites (Cabello, 2006; Kümmerer,
2004; Okeke and Edleman, 2001; Zhang et al., 2009).
Increased input of antimicrobial drugs creates opportunities for
environmental and pathogenic bacteria to develop selective resistance to
pharmaceuticals due to the ease with which bacteria exchange genetic material
and the speed at which they reproduce. More virulent and resistant bacteria are
artificially selected by this increased exposure, and also threaten water quality
through altered environmental bacteria populations. Given that increased levels of
resistance in human gut bacteria can persist for up to two years after treatment has
stopped (Jakobsson et al., 2010), the massive amount of antimicrobials that we are
adding to our environment now may have long term consequences.
56
Antimicrobial resistance is increasingly framed as an ecological problem
in addition to a human health problem. The use of synthetic antimicrobials is
potentially devastating to natural populations of bacteria that provide important
ecosystem services because xenobiotic compounds are recalcitrant, especially
when they are broad spectrum and effective against more bacterial species, and
thus they tend to be more persistent in soils and waters (Kümmerer, 2004).
Understanding and documenting the transmission of non-pathogenic antimicrobial
resistant bacteria (ARB) is important because environmental bacteria can serve as
reservoirs of resistance, passing genes to pathogens through conjugation,
transduction, or transformation (Barbosa and Levy, 2000; Khachatourians, 1998).
Increased human population and decreased wild habitat has intertwined
the health of humans, wildlife, and the environment, breaking down divisions
between these disciplines in the assessment of ecosystem health (King et al.,
2008). This new paradigm, exemplified by the growing One Health movement,
has created a more collaborative and holistic approach to monitoring ecosystem
health through the use of animal sentinels (Rabinowitz and Conti, 2013). Marine
mammals are a good sentinel species in marine and aquatic research due to their
physiological relationship to humans, their position as apex predators, and their
visibility as a keystone species in the ecosystem (Bossart, 2006; Wong, 2002).
Fecal sampling can be employed as a non-invasive alternative to capture/release
in study of species that are threatened or endangered. In addition, feces contains a
wealth of information on the animal, including genetic, hormonal, and toxin
loading, and the nutrient-rich intestinal environment is particularly conducive to
57
microbiological research (Kohn and Wayne, 1997; Miranda and Zemelman, 2001;
Taberlet et al., 1999).
Research using the intestinal bacterial flora of marine vertebrates as
indicator species for ARB pollution has shown varying degrees of resistance.
ARB presence has been studied in predatory fish, marine birds, pinnipeds, and
various whale and dolphin species by culturing bacteria from rectal or fecal swabs
(Blackburn, 2010; Grieg, 2007; Johnson et al., 2008; Lockwood et al., 2006;
Miller et al., 2008; Miranda and Zemelman, 2002; Rose et al., 2008; Schaefer et
al., 2011; Schroeder et al., 2009; Stoddard et al., 2002). The levels of resistant
samples and the complexity of resistance patterns has been positively correlated
by sampling site to suspected contamination points, ex. higher number of
wastewater treatment plant (WWTP) outflows (Grieg et al., 2007), number of
septic tanks (Schaefer et al., 2011), human population density (Blackburn et al,
2010; Grieg et al., 2007), and freshwater outflows (Johnson et al., 2008). Rainfall
and weather events have also been related to increase antibiotic resistance
findings in short-term temporal analyses (Grieg et al., 2007; Schaefer et al.,
2011).
The Salish Sea is an ecologically diverse and economically important
estuary in the Pacific Northwest region of the United States of America and
southwestern British Columbia, Canada (Fraser et al., 2006; Gaydos et al., 2009).
This glacial carved fjord inland marine ecosystem stretches along the inland
waters from Olympia, WA in the United States to the Campbell River in Canada,
and is the home to over 7 million people and many endemic and rare marine
58
species (Jackson and Kimerling, 2003). Like many of the world’s coastal zones,
population growth and increased toxin and fuel loading has amplified pressure on
the marine habitat and wildlife in the Salish Sea (Gaydos et al. 2009; Puget Sound
Partnership, 2011). Population growth has increased loading to septic and waste
treatment systems, increasing likelihood that harmful substances, including
pathogenic bacteria, toxins, and other chemical compounds will reach the water
column without being properly broken down (Take Back Your Meds, 2011).
There are 96 publicly owned WWTPs in Washington State emptying into the
Salish Sea and processing over 124 million gallons of sewage from over 3.5
million people each day (WA Department of Ecology, 2010). Groundwater, storm
water, and combined sewer overflows discharge untreated and industrial
wastewater and compounds leached from leaky septic drainage systems when
capacity is exceeded or water levels rise above the drainage fields also contribute
to surface water pollution (Puget Sound Partnership, 2011; Dougherty et al.,
2010). In Liberty Bay, WA, data show that the community, which predominately
uses septic tanks rather than centralized WWTPs, is receiving detectable levels of
trimethoprim in surface and ground waters, with the occurrence of detections
increasing as population density increased (Dougherty et al., 2010).
Research on marine bacterial resistance in Washington shows that ARB
can be found in marine sediment near aquaculture sites, the sea-surface, and
public beaches (Herwig et al., 1996; Roberts et al., 2008; Soge et al., 2009). When
analyzing sand samples from Washington beaches, five distinct strains of multidrug resistant methicillin-resistant Staphylococcus aureus (MRSA) and 33
59
methicillin-resistant coagulase-negative Staphylococcus (MRCONSA) spp were
identified (Soge et al., 2009) Roberts et al. (2008) identified 18 resistant strains of
vancomycin-resistant Enterococcus in Washington public beaches. The presence
of these resistant and pathogenic strains makes a case for the possibility of ARB
genes in offshore marine waters, and understanding the transmission of these
genes is critical to assessing the risk of the marine environment serving as a
reservoir of resistance genes transmittable to pathogens.
In the Salish Sea, abundant data is available the endemic ecotype of the Orcinus
orca, or Southern Resident Killer Whale (SRKW). Concerns over decline in the
SRKW population resulted in a large collection of data on the life history, family
lineage, geographic range, and individual identity of most whales. The SRKW
population was devastated in the 1960’s, when capture of animals for aquarium
exhibits resulted in the death of 13 whales and the live-capture of 45 juveniles
(Center for Whales research, 2013). The population declined to approximately 70
animals in the mid-1970’s, spurning Endangered Species Act listing of the
SRKW. In line with recovery efforts, the population rose as high as 100
individuals in 1995 after a trend of increasing population throughout the 1980’s
and 1990’s, but numbers are again falling (Center for Whale Research, 2013). The
population consisted of 86 individuals at the 2012 spring census and declined to
80 by the fall of 2013, during the time of this study (Center for Whale Research,
2013). Declines in this population are primarily thought to be due to increased
toxin accumulation from human pollutants, stress from lack of nutrition with
declining salmon population, and underwater noise pollution, and other proposed
60
threats include pathogens (NOAA, 2011). The species is at a critically low
number, and compound factors of nutritional stress and toxin loading can increase
likelihood for infections in these animals (Ayres et al., 2011). The importance of
this species as a cultural icon, recreational draw, and keystone species make them
ideal for the study of antimicrobial resistance in marine mammals. Prior studies
using marine vertebrates as sentinel species have shown positive correlation
between human influence and ARB colonization to some degree (Blackburn,
2010; Grieg, 2007; Johnson et al., 2008; Lockwood et al., 2006; Miller et
al.,2008; Miranda and Zemelman, 2002; Rose et al., 2008; Schaefer et al., 2011;
Schroeder et al., 2009; Stoddard et al., 2002), but no research has been able to
draw robust conclusions on anthropogenic effects on ARB acquisition due to the
lack of a natural history of the animals surveyed and a vague sense of their
geographic range, confounding variables not applicable to the study of ARB in
this species. Here the issue of bacterial resistance caused by pollution of
antimicrobials and wastewater was examined by studying prevalence and patterns
of ARB colonization in the feces of Orcinus orca. By using a sentinel species for
colonization with ARB and pairing prevalence and patterns with demographic,
geographic, and anthropogenic risk factors, the relative influence of human
impacts as well as traits intrinsic to ARB susceptible wildlife can be assessed.
Results indicate how other species could be affected by exposure to waters
affected by pharmaceutical pollution and determine if ARB are a problem of
relevance to public health in the Salish Sea marine environment.
2.
Methods
61
2.1
Ethics Statement
Fecal samples from orcas were collected in United States waters under
National Marine Fisheries Service permits 532-1822-00, 532-1822 and 10045 and
in Canadian waters under Marine Mammal License numbers 2008–16 and 2009–
08 as well as Species at Risk Act permits numbered 91 and 102. Sample
collection methods were approved by the University of Washington’s Institutional
Animal Care and Use Committee (IACUC) although no permit was required,
because research was non-invasive.
2.2
Study Site
Sampling was conducted in the Salish Sea from May-October 2012 and
June-October 2013. The research team was based off the western coast of San
Juan Island, and sampled when SRKWs were confirmed within approximately 1
hour travel time from harbor. This limited the study site to approximately the
Strait of Georgia to Vancouver, BC in the north, the Strait of Juan de Fuca to
Sooke, BC in the west, Smith Island, USA to the south, and Rosario Strait through
the western shore of Washington’s interior coast to the east. This study range
breaks down into six major segments based on watershed and environmental
characteristics as determined by the Puget Sound Watershed Characterization
Project in the US and the Marine Environment Monitoring and Assessment
program in CA (Figure 1) (Capitol Region District, 2013; Stanley, 2010). The
Canadian regions are Juan de Fuca, Saanich Peninsula, and Southern Gulf Islands
and in the USA the study site breaks down into the San Juan Islands (including all
islands in the San Juan archipelago), Nooksack, and the Coastal Skagit Basin.
62
This classification was appropriate because of the prior use of these divisions in
water quality research, the grouping of major watersheds which help make nonpoint pollution assessment more accurate, and the extent of the study site which is
covered by these segments.
The land area, population density, number of WWTPs, and best current
estimate of the number of septic tanks in each region are the environmental
variables of relevance for this study and values for the six segments in the study
site are noted in Table 3 (Capital Regional District, 2013; ESRI, 2010; ESRI,
2013; Whatcom County Health Department, 2013; Skagit County Health
Department, 2013; Wiseman et al., 2000). The number of WWTPs and number of
septic tanks are important variables for relating the direct input of antimicrobials
or ARB to surface waters through sewage treatment or leaking septic systems.
The variables of land area and population density are to approximate indirect
anthropogenic pollution that could result in elevated ARB, including run-off from
large land masses, urban storm-water, and if the number of WWTPs or septic
systems is significant, if this is related to human population density or other
factors.
2.3
SRKW Profile
The SRKW population is a distinct ecotype of Orcinus orca, a sub-order
of Cetacea Odontoceti in the family Delphenidae, closely related to other toothed
whales such as pilot whales, sperm whales, and oceanic dolphins (Taylor et al.,
2013). The SRKW is genetically, morphologically, and culturally distinct from
63
other killer whale ecotypes of the world, though endangered species listings do
not yet consider the differing forms discrete population segments (Gaydos et al.,
2004; Krahn et al., 2002). There are three distinct killer whale groups in the north
eastern Pacific Ocean, commonly known as ‘resident’, ‘transient’ and ‘offshore’
Killer Whales, who maintain social isolation from each other despite overlapping
ranges (Ford, 2002). The SRKW population diet is composed almost entirely of
Chinook salmon, while off-shore and transient killer whales are known to attack
and consume harbor porpoise, seals, sea lions, and occasionally other whales
(Ford, 2002).
The population is divided into three pods, dubbed J, K and L, which are
descended from a central female ancestor, as the SRKWs are a matriarchal society
and calves stay with their mothers for life (Ford et al., 2000). A SRKW pod
travels, feeds, and hunts together, and their communication “language” is
distinctive from other pods (Center for Whale Research, 2012; Gaydos et al.,
2004). This population are known as ‘residents’ because they return each summer
and fall to the Salish Sea from wintering habitats that are varied and less well
defined (NOAA, 2011). The pods spend 18-65% of their days in the San Juan
Island area of the Salish Sea from April to October, ranging to the Strait of Juan
de Fuca, the Fraser River, further south into Puget Sound, or in unknown waters
for the remaining time (Hanson, 2010). Winter sightings of the SRKW since 2005
suggest a range from Monterrey, California, USA, to Langara Island, British
Colombia, CA (Center for Whale Research, 2013). Toxin analysis of blubber
biopsy samples have shown differences in the proportion of toxins in the three
64
pods, indicating that they likely occupy different winter ranges (Krahn et al.,
2007). California sightings of L and K pods correspond to higher DDT/PCB
ratios, suggesting that these groups travel further south (Krahn et al., 2007). J pod
has a higher PBDE/PCB ratio, which suggests they winter where they consume
prey closer to an urban source (Krahn et al., 2007).
2.4
Sample Collection
Samples were collected using a scat detection dog in the same procedure
using modifications outlined by Ayres et al., 2012, which was revised from
Wasser et al., 2004, and Rolland et al., 2006. Briefly, a detection dog rode on the
bow of a 6m fiberglass hull vessel with a professional dog handler. The boat was
positioned downwind from the whales, and the driver and handler assessed the
wind direction, strength, and water conditions to set the boat perpendicular to the
“scent cone” to optimize the ability of the dog to smell the feces. The dog was
selected for his obsessive tendencies to play with a ball, and his identification of a
sample was rewarded by play with the ball and the handler. This encouraged the
dog to associate sample detection with the play reward, and resulted in a change
in behavior when the target scent was perceived. This anticipation of reward
caused a change in behavior upon scent observation, which was noted by the
professional handler, who communicated with the driver as the scent changed
from high to low concentration. The handler and the driver worked together to
direct the course of the boat as the dog stood erect, turned, slobbered, and
whimpered as the concentration of the strength of the scent changed.
65
Simultaneously, crew members scanned the surface for whale fecal samples,
which were identified by algae-like appearance and distinct fishy smell.
Once identified, samples were scooped using 1 L polypropylene beakers
and brought back onboard to immediately discard the water while retaining the
fecal pellet in a 50 ml polypropylene screw-top tube. This process was repeated
until all floating sample was collected up to 30 mL. The tube was capped and
centrifuged on the boat at 1,000 rpm for five minutes and excess water poured off.
For culturing, 4-5 sterile cotton swabs were inserted into the fecal pellet
and stirred slightly to gain a heterogeneous sample of approximately 0.5 mL. The
swabs were stored in 10mL of buffered sterile peptone water and placed on wet
ice until they could be transferred to the lab, with times ranging from 1.5 to 6.5
hours and averaging 4 hours. Two control samples were taken by recreating the
sampling process with water scooped from the sea surface, passed through the
processing equipment, and swabbed in the same manner as the fecal pellets.
For PCR analysis of resistance genes, the same procedure was used with a
single swab, which was then carefully inserted into a 2 ml micro tube and the
stick sterilely broken off to close the lid. Samples were stored in 20% glycerol in
MilliQ water or dry for method comparison. Once back on land, samples were
immediately stored at -20º C until DNA extraction.
2.5
Plate Culturing Methods
All culturing was done at Friday Harbor Labs in Friday Harbor, San Juan
Island, WA, USA. Samples in peptone broth were vortexed for 10 seconds. A
66
1mL disposable pipette was used to add 0.1 mL of fecal/peptone mixture to
plates. When dilution was necessary, 0.1mL of fecal/peptone was added to 9.9mL
of sterile saline. The whole or diluted mixture was spread on tryticase soy agar
(TSA) + 2% NaCl plates infused with ampicillin (Amp), chloramphenicol (Cm),
erythromycin (Erm) and tetracycline (Tc), and MacConkey without Crystal Violet
plates with Amp, Cm, and Tc at the values shown in Table 1, using the plate
spreading method and standard bench techniques (Hurst, 2002). Plates were
incubated at 36.5°C for 24-48 hours under aerobic conditions. After the samples
had been incubated, the number of colonies on each plate was recorded and color
was noted for colonies on MacConkey agar. Plates were parafilmed and stored for
future research.
2.6
PCR Assay Methods
Validation for optimum rRNA extraction from fecal swabs was conducted
by comparing 16S rRNA content from two extraction kits and using two different
Taq polymerase master mixes for resistance gene amplification (see supplemental
material for additional methods validation procedures). The DNAeasy Blood and
Tissue Kit and Qiagen HotStarTaq reagent were used with primers amplifying the
resistance genes ermB, metA, tetB, and tetM (Wasser et al., 2011). These genes
were selected because the common use of erythromycin, methicillin, and
tetracycline in animal and human medicine, the promiscuity of these particular
genes in genetic exchange, and previous discovery of these resistance genes in
mammalian gut flora made them likely candidates for colonization of the SRKW.
Gel electrophoresis on 1% agarose gel with 5% TBE buffer at 100 volts for 45
67
minutes, and visualized with 2µL of stop mix and UV light excitation was used
for identification of the resistance genes.
Genotyping for whale identification, including pod and sex, was
performed by NOAA’s Northwest Fisheries Science Center at 2725 Montlake
Boulevard, Seattle, WA. Presumptive sex was identified for samples lacking
positive genotyping identification by comparing progesterone and testosterone
levels gained through radioimmune-assay analysis in the Department of
Conservation Biology Laboratory, University of Washington, Seattle, WA.
2.7
Data Management and Statistical Analysis
Culturing bacteria
The number of antimicrobial resistant colonies were totaled for each
antimicrobial and sample as an indicator of ARB density in fecal samples.
Approximate quantitative assessment of resistant colony forming units (CFU) per
milliliter of sample was calculated by dividing the number of colonies by the
volume plated and dividing that number by the total dilution factor. This number
was multiplied by total colonies on each plate to approximate the number of ARB
in each mL of whale feces. Multidrug resistance (MDR) was calculated by
summing the total number of drugs to which each sample expressed resistance.
Ordination was performed to analyze patterns between resistances within samples,
using freely available R software (The R Foundation for Statistical Computing,
2013). Principle component analysis (PCA) was selected because the primary
purpose was to identify and compute composite resistance numbers for any trends
in antimicrobial resistance patterns between samples.
68
Each sample was assigned a land location based on the GPS coordinates
taken at time of sampling and shortest distance from shore as calculated using
ArcMap 10.1 (ESRI, Redlands, Calif., USA).
Environmental risk analysis of culturing results
The environmental risk factors included in statistical evaluation were
distance from shore, population density (people/km2), watershed area size (km2),
number of septic tanks, and number of WWTPs in the geographic segment nearest
to the GPS location of each sample that was cultured for bacterial growth.
For each geographic segment, human population density was estimated
from and land watershed area was obtained using ArcGIS Online population
density maps (ESRI, 2010; ESRI, 2013). Number of septic tanks in each area
were obtained from San Juan, Skagit, and Whatcom County USA Health
Department Records, San Juan County Conservation District, and from the
Capitol Regional District in British Columbia (Capitol Regional District, 2013;
personal communications, 2013; Wiseman et al., 2000). Distance from shore and
number of resistant colonies was compared by correlation.
Total ARB for each sample were compared by ANOVA separated by
geographic segment to see if sampling area affected the number of colonies.
ANOVA was also performed on ARB colonies totaled for each antimicrobial and
each antimicrobial plate, separating the MacConkey bacteria from the TSA
growth to see if patterns in resistance were reflected in the type and amount of
resistant bacteria.
69
Total colonies were averaged by each area and linear regression analysis
against the environmental variables was conducted to assess if there is a
relationship between number of colonies and environmental parameters. The
MDR and environmental variables were analyzed for similarity trends using a
contingency table.
Temporal change in ARB prevalence based on total and number of
consecutive days the whales had been in the basin was assessed. The Julian date
of each sample was correlated to the total ARB and ARB for each antimicrobial to
look for trends in ARB prevalence as a factor of time. The total days the whales
were present in the study area before sampling occurred was calculated based on
field effort log entries. For days when new whales entered the study area and
joined previously sampled whales, the count was reset because sampling protocol
focused on alternating pods and family groups as much as possible. The number
of consecutive days was analyzed by correlating against total ARB and ARB by
antimicrobial to seek trends in number of resistant bacteria as days in the basin
increased, and an influence function plot was constructed to reduce the effect of
influential data points. ANOVA was conducted to compare means of total ARB
and ARB by antimicrobial by number of days in study site.
PCR genetic assay
For each sample undergoing genetic analysis, the sample distribution of
sex and pod was compared to the entire SRKW population by Χ2 goodness-of-fit
test to test for sample distribution bias. A representative age distribution of
70
samples was tested with Student’s t-test. When resistance genes were identified in
an assay, all samples from that assay were assessed for risk factors of sex and pod
by Χ2 testing, and using ANOVA for distance comparison.
3.
Results
3.1 Sample Collection Results
Genetic analysis samples were collected during July-October 2012, and
culturing samples were collected during August-October 2013. A total of 11 fecal
samples plus two control water samples were collected for culturing, sampling
from a single family group for 2 days and from the entire SRKW populations for
seven days. Samples were assumed to be independent due to the sampling focus
on collecting feces from all whale groups. Two water samples were also collected
as controls.
For PCR assay analysis, 32 fecal samples were collected for genetic
analysis in 2012. DNA genotyping by NOAA allowed for identification of
individual whales for 23 (71.9%) samples (Hemplemann, J, unpublished data).
From the 23 samples, 19 individuals were identified, so samples are not
considered independent. Age range was 13-79 years old, with a mean age of 28.5
and standard deviation of 16.5. Sex information on individuals lacking identify
confirmation was supplemented by hormonal analysis, bringing total sex
identified animals to 29 (90.6%) of samples (Wasser, unpublished data). The PCR
assay samples were representative of the known individuals in the SRKW
population by age distribution (t=1.34, p=0.18, d.f. =105), and sex (Χ2= 1.93, p=
71
0.38) (Figure 4), but the variance of individuals were not evenly distributed by
pod, with more samples attributed to the J family pod (Χ2= 34.85, p= 1.31 × 10-7)
(Figure 5).
3.2 Plate Culturing Results
Each of the 11 fecal samples resulted in bacterial growth on at least one
antimicrobial plate, while the two control water samples showed no growth on
any plate (Table 2). For this reason, bacteria are considered to be from the fecal
source rather than surface water or laboratory contamination, and the
environmental data for the control samples was not analyzed. Samples were
assumed to be independent in statistical analysis because of sampling protocol
dictating to move between whale family groups for maximum distribution of
samples and from track logs recording the specific whale groups followed at time
of sampling.
The total number of ARB colonies by antimicrobial is represented in
Figure 6, with TSA and MacConkey plates differentiated by color. Amp25 plates
colony growth ranged from 0-87 colonies (x¯ =14.91, stdev=26.22), Cm25 ranged
0-9 (x¯ =1.36, stdev=2.66), Erm10 ranged 0-520 (x¯ =138, stdev=167.57) and Tc25
0-16 colonies (x¯ =3, stdev= 5.46).
CFU per milliliter of fecal quantification by sample ranged from 2.10×102
CFU/mL fecal for sample 3 to 1.27×106 CFU/mL for sample 4, with x¯ =1.37×105
and stdev=3.78×105. The mean CFU/mL grouped by antimicrobial was 9.08×104
(stdev=1.53×105). Total CFU grouped by antimicrobials were: Amp25 =3.44×104,
72
Cm25 =3.15 ×103, Erm10 =3.19 ×105, and Tc =6.93 ×103. The CFU/mL notation
was not used in subsequent analysis because the homogenous nature of bacteria
within the fecal was not verified by replication of culturing; however, the
CFU/mL measurement is the best current estimation based on peer-reviewed data
on cetacean fecal.
ANOVA testing revealed that there was a significant effect on the number
of colonies by antimicrobial drug type with p<0.05 (F6, 84= 2.21, p = 1.38×10-5).
Tukey’s Post-Hoc results show Erm10 plates (x¯ =116.77, stdev=161.51) was
significantly higher in colony numbers than all other plates. ANOVA analysis of
effect on total colony growth by sample was not statistically significant at p<0.05
(F10, 66= 0.82, p = 0.61).
Single drug resistance (SDR) was expressed by three (27.2%) samples and
MDR on the remaining eight (72.8%), with x¯ = 2.45 and stdev=1.17 (Figure 7).
This is interesting because the maximum MDR is to all four antimicrobials.
Ordination analysis of similarities in number of ARB resistant bacteria for
each sample are represented by the spatial distance between sample numbers on
the PCA plots, and the weight and direction that each of the antimicrobial
variables was given is shown in the loading plots. For both PCA analyses,
samples 12 and 13 (Control 1 and Control 2) are overlapping.
PCA was run with the variable of colony count by antimicrobial and plate
type to examine similarities in samples of relative types of resistant bacteria in
addition to antimicrobial resistance (Figure 8). The loading plot shows that there
73
is similar ARB growth patterns in MacConkey and TSA plates for Amp and Tc,
but that the two plate types are not similar in ARB growth pattern for Cm (Figure
8). Samples are not spatially related by numerical order (i.e. date). Samples 4 and
5 are the most unrelated to any other samples. Sample 4 is influenced by the high
number of Erm resistant bacteria along with Amp resistant growth on both plate
types, while the distance of Sample 5 from other samples is best explained by the
resistance to Erm, Tc for both plates, and MacConkey Cm. This combination of
resistant growth is unique to this sample alone.
PCA for joined antimicrobial was also conducted to examine differences
in patterns among the resistance by drug disregarding agar type (Figure 9). Again,
Sample 4 is widely separated from other samples, but Samples 5 and 8 are also
outliers and closely related since the separation between Cm resistances by plate
is not recognized.
Aside from demonstrating relationships in resistance patterns between
samples, the comparison of loading charts shows the relationships between ARB
prevalence for antimicrobial. In Figure 8 it is demonstrated that Cm resistance is
often opposing to Tc and MacTc resistance and Erm, Amp and MacAmp
resistance are closely related. Figure 9 shows a weakened relationship between
Erm and Amp, and a more closely aligned behavior in resistance patterns of Tc
and Cm in the samples due to combining the two plate types.
3.3 PCR Assay Results
74
Four of 43 total samples (including duplicates) tested positive for the tetM
gene encoding resistance to tetracycline. No other resistance genes were detected.
All four tetM positives were from female J pod animals and in the San Juan Island
geographic segment. Analysis of the relationship between sex and
presence/absence of the tetM gene returned no significance (Χ2= 3.35, d.f. = 2, p=
0.18). Analysis of positive genetic identification of tetM by pod showed no
significant relationship (Χ2=6.68, d.f. = 3, p= 0.08) for p <0.05, but findings
would be significant at a p<0.15 acceptance of error. An age and ARB
colonization two-tailed student’s t-test for unequal variances also determined no
significant differences (t=3.18, 0.47).
The ANOVA conducted to test variance between positive and negative
groups and distance from shore resulted in no significant difference between
groups (F1, 39=0.73, p=0.38). Chi-square testing of the geographic segment
relationship to positive resistance gene identification resulted in no significant
relationship for p<0.05 (Χ2=1.05, d.f. = 2, p= 5.99).
3.4 Data and Spatial Analysis
Geographic assessment based on sample locations resulted in the use of
three of the six geographic segments in the study area: San Juan in the USA, and
Southern Gulf Islands and Juan de Fuca in CA. The tested environmental risk
factors related to each of these segments and sample numbers that correspond to
each is shown in Table 3. Of the 11 samples, eight were categorized as San Juan,
two as Juan de Fuca, and one as Southern Gulf Islands. The low representation of
75
geographic distribution makes spatial and environmental analysis less robust due
to low number of data points and lack of variability, but analyses still seek trends
in data.
Single-factor ANOVA was used to compare the effect of sampling
location on total number of colonies in Southern Gulf, San Juan, and Juan de Fuca
locations because the 2013 culture samples fell only within these three sectors
(Figure 10). There was significant effect on the mean number of colonies at the
p<0.05 level for the three locations (F2, 8= 6.78, p = 0.019). Tukey’s Post Hoc test
revealed that the mean number of colonies for the San Juan (x¯ =115.25,
stdev=128.44) and Juan de Fuca (x¯ = 100.5, stdev=126.57) areas were not
significantly different, but both differed widely from the Southern Gulf sample (x¯
=607, stdev=0).
Colony growth for each site was compared by one-way ANOVA to assess
if geographic segment had influence on an antimicrobial’s resistance prevalence.
This was done in two ways: by growth on each plate, and by growth numbers on
all plates for each antimicrobial. For growth by plate, TSA Erm10 (F2, 8 =5.36,
p=0.03) and MacConkey Amp25 (F2, 8 =37.07, p=8.99×10-5) were found to be
significant, both indicating that the sample from the Southern Gulf Islands
location produced more bacteria resistant to these antimicrobials. Accounting for
combined antimicrobial prevalence, Amp25 (F2, 8 =19.75, p=8.05×10-4) and
Erm10 (F2, 8 =5.36, p=0.03) were found to be significant, again indicating more
colony growth in the Southern Gulf Islands.
76
Total number of colonies by sample was compared to the environmental
variables of WWTP, septic tank, land area, and population density using linear
regression analysis. There were no statistically significant environmental factors
for α=0.05. The values for the insignificant relationships are as follows: number
of WWTPs (y=176.68x-138, r2=0.87, F1,2=7.09, p=0.23), number septic tanks (y =
-0.0521x + 772.74, r² = 0.106, F1,2=0.12, p=0.78), land area in km2 (y = -0.282x +
480.98, r² = 0.451, F1,2=7.09, p=0.23), and population density in inhabitants/ km2
(y = 3.9112x + 195.35, r² = 0.05, F1,2= 0.05, p=0.86).
Correlation of total number of resistant colonies as a function of distance
from shore showed no significant relationship (y = -0.0083x + 178.1, r² = 0.0175,
F1, 9=0.16, p=0.69) (Figure 11).
An analysis of the relationship between MDR rank (ranging from 1-4
drugs) and environmental risk factors was performed by creating contingency
tables and using the Williams (1976) adjusted Χ2-statistic for small sample size.
Analysis revealed no statistical significance between MDR and any
environmental variable for p < 0.05 (adjusted Χ2= 4.04, d.f. = 6, p= 0.30).
Revisiting the PCA analyses charts (Figure 8 and Figure 9), there is no
close relationship between Samples 10 and 11, the Juan de Fuca samples, over
their relationship to samples from the San Juan geographic segment. The single
Southern Gulf Islands sample, Sample 4, is an outlier in both PCA analyses, as
explained above. From the data, no relationship between ARB patterns and
geographic segment can be made.
77
Correlation of the relationship between time and total ARB colony number
growth assessed by Julian date showed no significant relationship (y = -0.2561x +
217.48, r² = 0.0004) (Figure 12). Analysis of time and ARB number by
antimicrobial was also not significant (Figure 13).
The relationship between the number of days the whales had been in the
basin before the sample was collected yielded no statistically significant results by
ANOVA by total or single drug resistance, or by regression analysis (Figure 14)
(y = -19.628x + 216.16, r² = 0.0935). However, due to the small sample size and
low variance in number of days before sampling, the decreasing trend in ARB
shedding as the whales remain in the basin warrants further investigation, as it
contradicts the hypothesis that the Salish Sea is a source of drug pollution and
resistant bacteria.
4.
Discussion
Antimicrobial resistance is ubiquitous in the environment and resistant
bacteria have been isolated from beaches, surface water, and SRKW blow in the
Salish Sea in prior research (Roberts et al, 2009; Schroeder et al, 2009; Soge et al,
2009). This research presents the first isolation of ARB from the feces of the
SRKW, with 11 (100%) of samples showing resistance to ampicillin,
chloramphenicol, erythromycin, or tetracycline and 8 (72.8%) samples growing
colonies of bacteria resistant to multiple drugs. The findings of no growth in
control samples support the assumption that all bacteria colonies originate from
78
the collected fecal rather than water surface, processing equipment, or laboratory
contamination.
Compared to other studies, resistance in SRKW fecal matter is at first
glance significantly higher. Stoddard et al studied ARB prevalence in Escherichia
coli from fecal bacteria of stranded and healthy elephant seals in California, and
found that only 6.7% of the free-ranging seals showed resistance, and a mere
1.2% showing resistance to more than one antimicrobial in an assessment of 12
different drugs (Stoddard et al., 2008). However, their study concentrated on a
specific bacterial species while this study was indiscriminate to bacterial
identification, so comparisons are hard to draw. Blackburn et al. (2010) study of
predatory fish returned a total of 130 bacterial isolates from 63 total animals
sampled, approximately 2.1 isolates per sample. Specific isolate identification was
not performed in either experiment, and this study shows and average of 157.3
colonies per sample, indicating more resistance. It must be noted that our sample
size of 11 is considerably smaller, which could alter the eventual outcomes.
However, the SRKW has been referred to as the most polluted cetacean in the
world (Ross et al., 2000), so the increased resistance in this study should be taken
seriously regardless of small sample size.
This study found that resistance to erythromycin was most common for
samples on non-differential media, followed by ampicillin and tetracycline, then
chloramphenicol. The differential media showed highest resistance to ampicillin,
chloramphenicol and tetracycline, in that order. Erythromycin was not used on
MacConkey plates. The high instance of ampicillin resistance is common in
79
papers researching marine fecal bacteria (Miranda and Zemelman, 2001; Rose et
al., 2008; Stoddard et al., 2008). However, other researchers have found higher
instances of tetracycline resistance, and there is no analogous explanation for the
high erythromycin resistance observed in this experiment (Blackburn et al., 2010;
Miranda and Zemelman, 2001; Stoddard et al., 2008). This study tested
significantly less antimicrobials for resistance, and this incongruity may be an
artifact of the laboratory analysis.
Growth of 1635 (94.5%) of the total colonies occurred on the TSA 2%
NaCl plates rather than MacConkey 2% NaCl plates. The TSA agar is a general
purpose growth medium used for cultivation of a wide variety of bacteria, while
MacConkey agar is selective for gram-negative bacteria and typically used in
investigation of mammalian gut enteric flora (BD Diagnostics, 2009). While the
SRKW is a marine mammal, the genetic analysis of the SRKW microbiome
shows a rather low number of Bacteroidetes and other gut flora common in
microbiota of terrestrial mammals, (Bik, Elisabeth, personal communication) so
less growth on media designed for human flora is not entirely unexpected
particularly when plated with salt solution.
A vast majority (87.7%) of colonies proliferated on TSA 2% NaCl Erm
plates, suggesting that bacteria within the fecal microbiome of the SRKW may
have some intrinsic resistance to erythromycin. Current research on the SRKW
microbiome shows a bacterial flora dominated by Clostridium sordelli,
Clostridium perfringens, Cetobacterium ceti, Fosobacterium mortiferum,
Photobacterium damselae, Escherichia coli, Edwardsiella tarda, and
80
Actinobacillus delphinicola, but without deeper investigation into the specific
resistance genes that these bacterial species commonly acquire, it cannot be ruled
out that this resistance has developed as a result of outside factors such as
pharmaceutical pollution or contamination with ARB (Bik, Elisabeth, personal
communication). In a study by the Washington State Department of Ecology,
erythromycin concentration in WWTP effluent was the second highest
antimicrobial in tertiary treated effluent, at concentrations of up to 343 ng/L, and
secondary treatment effluent showed concentrations of 154-327 ng/L (Lubliner,
2010). Tetracycline was also discovered in concentrations considerably lower
than erythromycin in both influent (13-186 ng/L) and secondary treatment
effluent (10-40 ng/L), and was non-detectible in tertiary effluent (Lubliner, 2010).
These patterns fit the incidence of ARB resistance found in SRKW scat, and the
contribution of large concentrations of unmetabolized erythromycin compounds
in WWTP is a possible explanation for the high rate of resistance seen in this
study.
The results from this study related to the 2009 study of orca blow
resistance show interesting comparisons. In 11 samples from male killer whale
breath, erythromycin resistance is low, expressed in only two samples (Schroeder
et al., 2009). However, the expanded macrolide-lyncomycin-streptogramin (MLS)
family of antimicrobials includes lyncomycin, and high resistance to this drug is
high in orca blow, with eight samples expressing resistance (Schroeder et al.,
2009). This suggests that there could be intrinsic resistance to antimicrobials with
the bacteriostatic mechanisms of protein-inhibiting synthesis on the 50s ribosomal
81
sub-unit, or that there is a gene that proliferates MLS resistance in the waters of
the Salish Sea. Tetracycline resistance was comparatively low, only demonstrated
in two of the 11 samples (18.2%), and resistance to beta-lactams was shown in 7
samples (63.6%) (Schroeder et al., 2009). The ratio of tetracycline to beta-lactam
resistance is similar in the study of feces, with 36.4% of samples showing
tetracycline resistance and 72.7% of samples resistant to ampicillin. The breath
study did not include chloramphenicol, but did include florfenicol, which is in the
same amphenicol family of antimicrobials. Resistance to these two drugs was not
similar within these two studies, with 27.2% of whale breath samples showing
resistance compared to 45.4% of scat samples (Schroeder et al., 2009). The
similarities in relationships for beta-lactams, tetracycline, and MLS drugs could
point to common drug resistance colonization in the waters or wildlife of the
Salish Sea and is an interesting trend worth further analysis.
The comparison of loading charts from the PCA ordinations demonstrate
resistance and Erm, Amp and MacAmp resistance are closely related. Increased
Amp resistance for differential and nondifferential media growth could be a result
of large numbers of Amp resistance genes or resistant bacteria. Amp falls within
the beta-lactam family of antimicrobials, which includes penicillin. These are the
oldest antimicrobial families, and as a result many bacterial species have
developed resistance, in addition to some naturally resistant species. The
alignment of macrolide bacterial growth with beta-lactam growth is not explained.
It has been shown that bacteria resistant to a single antimicrobial are more likely
to show resistance to other drugs after exposure (Livermore, 2003), and it is also
82
possible that there are genes resistance to Erm and Amp impacting the fecal
bacteria of the whales. Alternatively, beta-lactam and erythromycin are popular
human drugs, and the stability of erythromycin in wastewater treatment has
already been demonstrated, so this phenomenon could be a result of more
pollution of these two drugs in tandem in the Salish Sea. More research is needed
to confirm these explanations.
Again, the PCA loading charts indicating that Tc and MacTc have similar
patterns in resistance is expected and explained by tetracycline resistance genes.
However, the opposition of tetracycline resistance to chloramphenicol resistance
is unexpected. The above explanation that a bacterium developing resistance to
one drug is more likely to be multi-drug resistant would seemingly directly
oppose this finding. However, the overall low expression of chloramphenicolresistant bacteria may better explain this opposing ordination than any intrinsic
drug or bacterial properties. Again, this requires more study.
Although the relationship between time in basin and total ARB growth
was not significant, the low number of data points makes the likelihood of a
statistical relationship low. The downward trend is of interest because it directly
opposes the hypothesis that pollution in the Salish Sea increases ARB
colonization, and there are several possible explanations for increased ARB
shedding when animals reenter the basin. First, the long travel from the open
ocean to the basin could be taxing to the animal, resulting in an immune
compromised condition once they reach the basin, which is reflected by their
fecal. Secondly, this could be a factor of the water quality the whales encounter
83
on their travel from the ocean through the Strait of Juan de Fuca. Canada’s
WWTP regulations do not require secondary treatment, and two large WWTPs
that only screen sewage release to surface waters are located on Vancouver
Island, where the Strait of Juan de Fuca meets the Salish Sea and the whales must
pass to enter the basin. This theory supports the hypothesis that loading of
pollution over time might be affecting ARB more than immediate geographic
sampling area, important to note for other studies which relate indicator organism
sampling site to geographic location. A third explanation is that the fecal
composition is richer in ARB when the whales return from the ocean because of
food sources. Field observations show that scat samples are larger and appear
fattier in the first day after the whales return to the basin, decreasing in size and
fat content as the animals remain in the basin. This has been unofficially
attributed to differing food sources, and the pollution of oceanic salmon or
increased consumption of farmed salmon treated with antimicrobials could
explain the increased ARB when the whales re-enter the basin. The larger size of
the samples could also contribute to more ARB colonies, because the intestinal
system of the whale is voiding more completely more bacteria could be shed from
intestinal walls. Again, the relationship between return to basin and total ARB
colonization was not statistically significant, but the trend lacks data points and
warrants further study. Additionally, this theory opposes prior use of immediate
sampling location to assess water quality and geographic relationship to ARB
prevalence and suggests that the factors of time and recent geographic travel
should be considered in other studies of animal indicator organisms.
84
Genetic analysis revealed a smaller rate of resistance, yielding only tetM
positive samples in 4 (9.3%) of samples. Four of the cultured samples expressed
resistance to tetracycline (36.36%), indicating that the true resistance to
tetracycline could be higher. The variation in genetic resistance and cultured
resistance is curious, but could be a result of the time delay in DNA extraction for
the samples collected for genetic analysis. Multiple instances of freezing and
thawing occurred over the minimum 4 month period between sample collection
and DNA extraction, which denatures DNA and decreases the ability for
conventional PCR assays to detect genes (Qiagen, 2013). This theory is supported
by spectrophotometry of the nucleic extracts, which revealed that most samples
concentrations of DNA with the volume of template used for PCR reactions
resulted in less than the recommended 50-500ng DNA per reaction recommended
in general PCR guides (see Figure 13)(Palumbi et al., 2002). In contrast, the
samples collected for culturing were processed within hours of collection, leading
to the conclusion that quick transportation of samples results in better assessment
of ARB in SRKW fecal samples. Tetracycline and oxytetracycline are often used
in aquaculture, and the findings of tetM could be indicative of infiltration of these
antimicrobials to the food web.
Although there were no statistically relevant whale traits relating tetM
colonization to age, sex, or pod, all positives came from three female J pod
whales off the west side of San Juan island. The three females, J8, J17, and J31,
belong to different sub-family groups within J pod, and other J females sampled
in this region during this time did not result in positive resistance gene
85
identification. Two sampling incidents of J31 over a month apart resulting in two
positive identifications of the tetM gene confirm the hypothesis that using a wellstudied and identified marine mammal population is beneficial for research
seeking to study the residence time of ARB, and may indicate that effects are
more specific to individual health than outside factors. This is corroborated by the
study of marine animals on the east coast, which indicated that animal provenance
(healthy, stranded, or bycaught) was a better indicator of ARB colonization than
animal type or traits (Rose et al., 2009). However, J31 has not been known to
have health problems, while J8, who was also sampled, was missing and
presumed dead as of October 2013 (Center for Whale Research, 2013). Despite
limited explanations due to low positive results, the usefulness of a repeatable and
known population of animals in a study utilizing sentinel species is supported.
Examination of the microbiome for 3 of the SRKW fecal samples was
performed to examine the functional metagenomic resistance genes with more
sensitivity and precision. Qualitative polymerase chain reaction (qPCR) was
performed on a J pod female, a K pod female, and an unknown sample. Genes for
resistance to tetracycline, methicillin, extended-spectrum beta-lactam (ESBL),
vancomycin, and sulfonimides antimicrobial resistance were observed (Table 4)
(Roberts MC, personal communication). The detection of the mecA gene through
metagenomic analysis and not the PCR assay of this study indicates that low PCR
assay precision and time elapsed between sampling and gene amplification may
have flawed the genetic analysis in this study. The number and relative rarity of
some of the resistance genes, particularly the genes encoding for resistance to
86
ESBL genes, indicate more investigation into the array and scope of resistance
genes in SRKW fecal is of scientific interest, particularly if paired with
measurements of functional ARB genes in water or wastewater of the Salish Sea.
The Salish Sea is an interesting study site due to the discrepancies in
wastewater treatment and pharmaceutical disposal practices on either side of the
international border. Eight WWTPs empty into the Salish Sea in British
Columbia, CA, two of which use only primary treatment of wastewater before
discharging into surface waters (Capitol Region District, 2013). The remaining
Canadian WWTPs use secondary treatment, and all dispose of effluent directly
into the Salish Sea (Capitol Region District, 2013). In contrast, nine WWTP
facilities serve the US side of the Salish Sea, eight of which release effluent to the
Salish Sea. All US WWTPs have at least secondary treatment as required by the
1979 US Clean Water Act (Washington State Department of Ecology, 2007)
However, the Capital Regional District of British Columbia’s pharmaceutical
take-back program has been widely successful, with 95% voluntary participation
of pharmacies, collecting 60.32 kg of unused drugs in 2010 (Post-Consumer
Residual Stewardship Program, 2010). Drug take-back programs in Washington
State exist in 17 of 39 counties, and have been less successful, averaging just
18.96 kg/year in six years of surveillance (Take Back Your Meds, 2010).
Therefore, more unmetabolized drugs might be entering surface waters in WWTP
effluent from the US because of lack of medicine-disposal alternatives, despite
stronger water treatment standards.
87
Due to the lack of spatial variability in the cultured samples and the small
number of positives in the genetic samples, it is difficult to draw conclusions
based on human impact to the total resistance in SRKW fecal. To improve upon
this research, more detailed information on the environmental risk factors within
the study area, particularly the western coast of San Juan Island, would be
beneficial. The current breakdown of sample sites to six main land locations is too
broad a scale for the high concentration of samples collected in that specific area.
Additionally, other environmental risk factors that are thought to contribute to
increased ARB such as rainfall, agricultural lands, hospitals, freshwater outflows,
combined sewer overflows, and large on-site sewage systems were not included in
this study, partially due to difficulty obtaining records on the scale needed for
analysis and in part due to the difficulty of working with differing policies,
government divisions, and record keeping standards across international borders.
A future paper will include a more detailed spatial evaluation of these
environmental variables.
Further research is needed with complimentary genetic identification of
the SRKW to better identify trends in whale risk factors for colonization with
ARB. Other improvements would be a more thorough evaluation of ARB genes
for samples from both years, a larger suite of antimicrobial drugs tested, and
increased specificity on environmental risk factor data.
5.
Conclusion
88
The findings of resistance, both cultured and genetic, in SRKW fecal is of
importance to environmental and public health fields, conservation biologists, and
wildlife veterinarians. The prevalence of resistance in cultured samples is higher
than in previous studies of sand, water, and orca blow in the Salish Sea. Whether
this is a result of small sample numbers and a small suite of antimicrobials tested
or is evident of increasing resistance in the waters and wildlife is not clear, but
further research with finer-scale geographic analysis could improve our
understanding of the development of ARB in the Salish Sea, along with the
public, environmental, and veterinary health implications that could result from
increases in resistance.
89
Tables
Table 1. Names, doses, and abbreviations of antimicrobial drugs added to agar
plates in this study.
Drug Name
Dose Used (µg/mL)
Abbreviation
Ampicillin
25
Amp
Chloramphenicol
25
Cm
Erythromycin
10
Erm
Tetracycline
25
Tc
90
Table 2. Total growth from fecal sample cultures, summed in last row and column
by number of resistant colonies by antimicrobial and by sample. C1 and C2
correspond to the two control samples. Refer to Table 1 for antimicrobial
abbreviations.
TSA 2% NaCl
Sample
1
2
3
4
5
6
7
8
9
10
11
C1
C2
Total
Resistant
Colonies
MacConkey 2 %NaCl
Amp25
Cm25
Erm10
Tc25
Amp25
Cm25
Tc25
22
0
0
28
0
0
0
0
14
12
0
0
0
0
0
0
0
0
0
0
9
0
1
0
0
0
19
32
0
520
279
2
28
267
193
177
1
0
0
1
0
0
0
15
0
11
0
0
0
4
0
0
0
0
1
59
0
0
1
4
19
0
4
0
0
0
0
0
0
1
0
2
0
0
0
2
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
Total #
Resistant
Colonies
43
32
1
607
296
2
42
280
226
190
11
0
0
76
10
1518
31
88
5
2
1730
91
Table 3. Geographic segment division and environmental risk factors for
analysis.
Segment
Name
Cou
ntry
Coastal
Skagit Basin
Juan de Fuca
US
A
CA
US
A
Nooksack
Population Num
Density
ber
(people/km WW
2
)
TP
Land
Area
(km2)
Number
Permitted
Septic
Tanks
Sample
Number
s
Assigne
d
8029
14.60
2
25000*
NA
1512
2.96
2
11606
10,11
3651.9
55.72
1
30000**
NA
Saanich
Peninsula
CA
103
365.72
1
6345
NA
San Juan
US
A
471.38
33.94
1
8168
1,2,3,5,6
,7,8,9
Southern
CA
216
23.61
4
8946
4
Gulf Islands
*Best current estimate based on information from Skagit County Health
Department. **Best current estimate from Whatcom County Health Department.
Table 4. Functional metagenomic results for resistance genes in 3 SRKW fecal
samples from 2012 field season.
Whale ID
Sex
Age 2012
Resistance Genes
J31
F
17
sul2, sul3, blaCMY, ereA, oxa2, AAC6
K16
F
27
tetA, tetU, mecA, AAC6, catB8, aac31A, blaCMY,
vatE, vatA, blaMOX-CMY9
Unknown
NA
NA
ctxM1, ctxM2, mox, oxa9, ermB per2, cmy, mecA,
AAC6
92
Supplemental Material
Figure 1. Map of study site divided into six primary watershed units. Boat
docking location noted with red star.
Figure 2. Map of culturing sample locations, 2013.
93
Figure 3. Map of genetic analysis sampling locations, 2012. Samples positive for
tetM gene are denoted with red pentagons.
Figure 4. Sex distribution from DNA and hormone data for 2012 samples (n=32)
2012 PCR ASSAY SEX DISTRIBUTION
Unknown
9%
Male
31%
Female
60%
94
Figure 5. Pod representation from DNA genotyping identification of 2012 PCR
assay samples (n=32).
2012 PCR ASSAY POD DISTRIBUTION
L
9%
Unknown
28%
K
22%
J
41%
Figure 6. Total number of colonies grown on antimicrobial-infused plates by drug
type represented on logarithmic scale. Growth on TSA and MacConkey plates
differentiated by color, actual colony counts on bar.
95
Figure 7. Incidence of MDR in cultured samples 1-11. Four antimicrobials were
used in agar plates, thus four is the maximum MDR number.
Number of antimicrobials to which sample
demonstrated resistance
Incidence of Multidrug Resistance
4
3
2
1
0
1
2
3
4
5
6
7
8
9
10
11
Sample Number
96
Figure 8. Principle component analysis for samples separated by antimicrobial
and bacterial growth type (MacConkey or TSA) representing patterns in
abundance of ARB colonies cultured in samples 1-11 and control samples 1-2 (12
and 13). The relationship between samples is determined by the similarity of ARB
colony growth number for each antimicrobial plate, and the direction determining
spatial distance is demonstrated in the loading plot below.
97
Figure 9. Principle component analysis for samples separated by antimicrobial
representing patterns in abundance of ARB colonies cultured in samples 1-11 and
control samples 1-2 (12 and 13). Loading factor map below shows the dimensions
between variables which determined the distance between samples in the
individual factors map.
3
Individuals factor map (PCA)
8
1
7
0
11
10
13 6 2
12 3
1
4
9
-1
Dim 2 (29.92%)
2
5
-2
0
2
4
6
Dim 1 (44.34%)
98
Figure 10. Averaged numbers of ARB colonies by study site segment.
Averaged number of colonies by study segment
700
Mean colony number growth
607
600
500
400
300
200
115.25
100.5
100
0
San Juan
Southern Gulf Islands
Juan de Fuca
Sample Location Site
Figure 11. Total number resistant isolates in each samples as function of distance
from shore.
Total resistant colonies by sample
Total sample resistance as function of distance from land
650
600
550
500
450
400
350
300
250
200
150
100
50
0
y = -0.0083x + 178.15
R² = 0.0175
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Closest landfall (m)
99
Figure 12. Total ARB cultured as a factor of time using Julian date. No significant
relationship was found (y = -0.2561x + 217.48, R² = 0.0004)
Total ARB Cultured by Julian Date, 2013
700
Total ARB Colony Count
600
y = -0.2561x + 217.48
R² = 0.0004
500
400
300
200
100
0
220
225
230
235
240
245
250
Julian Date
Figure 13. ARB colony prevalence by drug resistance type related by Julian date,
2013. No significant relationships found and thus not reported.
ARB Colony Growth per Antimicrobial by Julian
Date
Number ARB Colonies
500
400
Am
p
300
Cm
200
Er
m
100
0
220
225
230
235
240
Julian Date
245
250
255
100
Figure 14. Scatterplot representing number of days the whales had been in basin
before sampling by total ARB count. Although the r2 value is low and suggests no
relationship, the small sample size (n=11) and uneven distribution between the
number of days before sampling may have had an effect on the decreasing trend
of ARB shedding in fecal as time in the Salish Sea progressed. Further
examination of this aspect of resistance should be studied.
Relationship between number of ARB colonies
and time in Salish Sea basin
700
Total ARB Colonies
600
500
y = -19.628x + 216.16
R² = 0.0935
400
ARB
Colony
Count
300
200
100
0
0
2
4
6
8
Time in Basin (days)
Figure 15. DNA concentration in ng/L of extract used as PCR template.
Genetic samples DNA averaged concentrations
120
DNA NG/ML
100
80
60
40
20
0
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43
SAMPLE NUMBER
101
PCR Methods Validation Protocol
To identify and validate the most efficient method of biochemically
identifying ARBs in preserved scat and swabs, methods of preservation, storage,
DNA extraction and PCR amplification methods had to be optimized. Duplicates
were taken in the field to compare recoveries in swabs preserved in 20% glycerol
in MilliQ water against swabs stored dry. Microtubes containing swabs were
placed immediately on wet or dry ice, which varies the temperature at which they
were stored. Additional swabs of whole fecal samples taken in the laboratory after
2-3 months of storage were taken as replicates to assess DNA degradation over
time.
A subsample of fecal swabs were extracted using Qiagen’s QiaAmp DNA
Stool Mini Kit and DNeasy Blood and Tissue Kit (catalog # 69506). Each swab
was extracted according to the directions of the manufacturer with tissue kit
modifications as noted in Wasser et al., 2011, and stool kit modifications found in
Wasser et al., 2004, which are optimized to run in 96-well plate formats rather
than individually. Blank negative controls were included in every extraction to
rule out contamination.
The extractions were compared on gel electrophoresis for whole DNA
extraction content by running 1 and 5 µL of extraction product and 2 µL of stop
dye for 45 minutes on 0.7% agarose gel with 5% TBE buffer at 100 volts for 45
minutes, and visualized by UV light. Bacterial content within extraction was
analyzed by quantifying bacterial genes after amplifying 3uL of each extraction.
The bacterial DNA was amplified using PCR with universal bacterial primers (63f
102
and 1387r) to obtain the 16S rRNA content, indicating the presence of
prokaryotes; PCR program set to denature for 15 min at 95°C, followed by 35
annealing cycles of 30 sec at 94°C, 90 sec at 55°C, and 1 min at 72°C, followed
by a 30 min extension at 60°C, and a 4°C hold.
Comparisons between the quality of PCR amplification was conducted
using two different Taq polymerase master mixes. Red Taq and Qiagen
HotStarTaq DNA Polymerase master mixes were tested in the amplification of
waste water treatment plant effluent, crow and cow feces along with a tetM
positive control to evaluate which master mix was better for amplifying the tetM
gene which encodes for tetracycline resistance. For the Red Taq, 12.5 µL of taq
was added to 10.5 µL of autoclaved PCR water and 0.5 µL of M4 and M6 primers
along with 1 µL of template for each reaction. For the optimized Qiagen
HotStarTaq, 5 µL of polymerase was added to 1 µL of PCR water along with 0.5
µL of each primer, and 3 µL of the DNA template (Qiagen, 2013). Both were run
on a PCR program set to denature for 15 min at 95°C, followed by 35 annealing
cycles of 30 sec at 94°C, 90 sec at 55°C, and 1 min at 72°C, followed by a 30 min
extension at 60°C, and a 4°C hold according to the Qiagen manual. Again, PCR
product in 1 and 5 µL concentrations was run on 1% agarose gel with 5% TBE
buffer at 100 volts for 45 minutes, and visualized with 2 µL of stop mix and UV
light excitation.
103
CHAPTER THREE
DISCUSSION AND INTERDISCIPLANRY STATEMENT
These results are the result of a multidisciplinary study approach and will
be of interest to professionals from differing academic and professional
spectrums. Conservation biology, veterinary science, microbiology, water
resources, and environmental and public health field research contributed to the
literature research and theory that made this project possible.
Field research on this project was made possible by the University of
Washington’s Center for Conservation Biology (CCB) and the non-profit
organization Conservation Canines (CK9). Their monitoring of the SRKW’s
health has occurred seasonally from 2007-2009 and 2010-2013 and uses
hormones extracted from the fecal samples of the killer whales to derive stress
levels for malnourishment, ambient noise, and pregnancy and testosterone rates to
understand relative impact of salmon population, commercial and recreational
boaters, and endocrine-disrupting compounds on the decline of SRKW
population. This research on ARB in the fecal of the killer whale is not geared
primarily to monitoring health of the population, but the data gathered can help
support recovery efforts nonetheless.
Veterinary science is intertwined with conservation efforts because the
evaluation of the health of the individual whales is both indicative and dependent
on the health of the entire population. The samples preserved for microbial
analysis are of interest to researchers from the National Oceanic and Atmospheric
104
Association (NOAA), the SeaDoc society, and researchers from the University of
British Columbia and University of California Davis that are highly engaged in
recovery efforts for the SRKW. Currently, a project is in planning that researches
the virus and whale pathogens in the fecal sample to assess infections that may be
endangering the population using the 2013 samples collected for this study.
An additional factor for veterinary science that has been incorporated in
this study is the use of scat detection dogs to local samples. Just as dogs can be
trained to sniff out drugs, bombs, and missing persons, they can learn to cue in to
the scat of a particular animal. CK9 was the first program to train dogs on the
scent of animal targets and use scat, scale and hair shedding, and other materials
from target animals. The data is used to monitor population numbers, extent of
habitat, sex distribution, family relationships, pregnancy and abortion rates,
reactions to environmental stress, and health of the species in question. The target
animals, which have included elephants, tigers, arboreal iguanas, moose, deer,
caribou, wolf, black bear, wolverine, Pacific fisher, marten, Northern Spotted
Owl, sharp-tail snake, cougar, and other animals, ultimately benefit from the
scientific research being conducted, as do the Conservation Canines that work for
the program. CK9 only accept dogs from shelters or that are owner-surrendered.
This is because the traits that make a great scat-sniffing dog, like boundless
energy, constant desire to play, and obsession with their ball or other toys, often
make these dogs incompatible pets. Additionally, many of the dogs have special
needs, aggressive tendencies, or health problems from trauma inflicted upon them
before they were adopted by the program. The dogs are trained, cared for by
105
professional animal trainers, and travel the world assisting in conservation
biology research. Dogs sometimes bond with a particular handler, providing them
with human companionship as well as a chance to expend their energy and play as
much as possible. And as a CK9 dog becomes too old or infirm to work in the
field, they are retired to good homes with the approval of the CK9 director Heath
Smith. This program is truly remarkable for the services it provides to scientific
research as well as the benefits to shelter dogs who would likely be euthanized
without their adoption by CK9. I have been truly honored to work with this
program, feel that it is a valuable and innovative use of resources, and sincerely
hope that the non-profit will survive the recession to work on future conservation
biology studies.
A drastically different set of disciplines that contributed to the design of
this study and will benefit from the findings are microbiology and biochemistry.
Bench microbiological techniques were required to complete this research, as was
a great deal of background research in biochemical and microbiological theory.
Research in these disciplines are sometimes considered a narrow field utilizing
primarily controlled laboratory experiments to gather information on specific
bacterial strains, genes, or processes of interest, typically geared toward human
health research. However, the growing field of environmental microbiology is
contributing to more in-situ (in place) studies that examine how bacterial
organisms and species interact with one another to create a process. The
development of more sensitive and deep PCR techniques has made genome
sequencing more affordable and practical. This technology has given rise to the
106
research of the “microbiome”, the assemblage of bacterial species present in a
given area. Studies on the microbiome of the sea surface, soils, and environments
such as caves, sediment, and intestinal flora of many species have been
conducted. These studies have revealed how interconnected the bacterial array in
an area are and that the assemblage, not the species, creates the ultimate function
and health of the environment. For example, a recent study on the relationship
between intestinal microbiota and obesity showed a direct link between bacterial
assemblage and weight gain. Separate groups of germfree mice were fed
uncultured microbiota from each member of four pair of human twins, one of
which of each pair was lean and one obese (Ridaura et al., 2013). The study
concluded that laboratory mice receiving fecal transplants from the bacterial
component of the obese co-twins’ microbiota gained significantly more weight
eating the same diet as the mice receiving the lean twin bacterial communities
(Ridaura et al., 2013). Increased understanding in the dynamics of bacterial
ecology paired with advances in technology make microbiological and
biochemical experiments increasingly applicable to real-world problems and
more-often relating to environmental processes. Research has used the
antimicrobial resistance patterns in fecal coliform bacteria to identify sources of
fecal contamination to water sources, for instance (Bernhard and Field, 2000;
Burnes, 2003). The research conducted on ARB occurrence and rates in SRKW
fecal was designed as a pilot project for the use of marine mammals in the Salish
Sea to serve as indicators of fecal and antimicrobial pollution, with the sample
location and wintering habits of the whale pods helping to identify “hotspots” for
107
acquisition of resistance genes, and in that aspect this study is very much aligned
to the field of applied environmental microbiology. By understanding the spread
and transmission of bacteria and resistance genes, microbiologists can work with
water resource professionals to trace contamination and learn more about the
extent of antimicrobial pollution and begin to create solutions for remediation.
The field of water resources connects to this project because of the
introduction of PCPPs and ARB into surface waters through WWTPs, leaky
septic tanks, agricultural land, and CSO. Reduction of anthropogenic compounds
and bacteria introduced to surface waters is directly tied to proper management
and regulation of these sources of pollution. Minimization of pharmaceuticals and
ARB in wastewater and runoff preserves water quality and decreases potential
problems that decrease sustainability of water resources, reducing need for large
scale engineering projects. Monitoring of ARB in water animals is one method
that engineers, hydrogeologists, and water scientists could observe the spread of
PCPP pollution through surface waters and thereby learn which systems are being
properly managed and how to improve resources.
Finally, the fields of environmental and public health are critical to the
development and application of this research on ARB in the Salish Sea. These
fields are grouped together because there is truly no environmental health
problem that will not soon become a human health problem. The framework for
this integrated concept that the health of humans is connected to the health of
animals and the ecosystem has been coined the “One Health” movement (CDC,
2013b). The One Health movement relies on interdisciplinary, collaborative
108
approaches to address potential risks at the forefront of human, animal, and
environmental health (One Health Global, 2012). The One Health method
monitors and controls public health threats by working with physicians,
ecologists, and veterinarians to create an overall landscape-scale understanding of
how diseases and environmental contaminants move through the ecosystem, and
uses this information to reduce potential hazards in the interactions of the three
spheres. Animals are used as indicator organisms for zoonotic pathogens, such as
West Nile virus in birds, and serve as an early warning sign of human disease
threats (CDC, 2013b). This thesis research was based on the idea that the marine
mammal serves as an indicator organism for ARB pollution in surface waters. It
was the hope of this researcher that this study could connect marine mammal
veterinary health measurements to the environmental health risk of ARB in
surface waters, which has important consequences for human health if these ARB
enter our water supply or transmit genes to human or zoonotic pathogens. The
holistic approach to environmental and public health that has been adopted by the
One Health movement was a major motivation for the design of this project, and
the interdisciplinary and holistic measurement of contamination and disease in the
environment based on the interconnectedness of these three domains is important
for future studies.
Lastly, environmental policy becomes important to this study because of
the myriad ways antimicrobial drugs and PCPPs are infiltrating the environment,
and the number of opportunities that arise to prevent the pollution.
109
First, policies regulating discharge of WWTPs and private industry under
the Clean Water Act, known as National Pollutant Discharge Elimination System
(NPDES) permits, does not cover emerging pollutants like PCPPs. The Clean
Water Act required the monitoring of other water quality parameters, such as
fecal coliform, nitrogen, and total suspended solids, and made at least secondary
treatment practices mandatory, which in the Washington Department of Ecology’s
report was shown to decrease the concentrations of PCPP compounds in WWTPs
(Lubliner, 2010). Tertiary treatment methods decreased the concentrations even
further, making the requirement for tertiary treatment one possibility for reducing
the amount of pharmaceuticals discharged into surface waters. This would have to
be accompanied by monitoring for PCPPs in stormwater and non-point runoff as
well to be effective.
Another way to minimize prescription drugs entering the environment is
decreasing the amount of drugs that enter the wastewater system. This could be
accomplished by assuring that only compounds that have been partially degraded
by the human waste system are flushed. The promotion of drug take-back
programs would help people dispose of unused or expired medicines safely and
without environmental harm. However, there is no consistent drug take-back
program in the USA, and the Take Back Your Meds program in Washington
operates only intermittently. In Canada, pharmacies provide free drug disposal to
customers, and the services are relatively well publicized. Implementing and
advertising permanent drug take-back programs could prevent the flushing of
unwanted drugs down the toilet and prevent them from ever entering the WWTP,
110
decreasing the concentration of drugs not eliminated by secondary treatment
systems.
Finally and most importantly, the spread of ARB and introduction of
PCPPs to the environment can be prevented through consumer education.
Physicians prescribe antimicrobial drugs in the absence of a bacterial infection
because patients often demand drugs for viruses and colds which will not respond
to these drugs in an effort to feel better. This practice is thought to contribute to a
rise in ARB in community infections and increasing unmetabolized PCPPs in
wastewater. The engagement of patients and doctors working together to treat
health holistically rather than looking for quick fixes to illnesses or larger health
problems would be a step in the right direction to decreasing ARB gene spread.
Consumer education about products that contain antimicrobials would also be
beneficial. Antimicrobials used subtheraputically in animal agriculture, products
like toothpastes, hand soaps, and deodorants, and antimicrobial hand gels and
cleaning products are ubiquitous in the supermarket and in our homes. Using
these products can help prevent the spread of infections when used prudently, but
the majority of these products are used constantly, defeating the purpose of
including drugs in the product. Aside from this, products containing triclosan and
other unregulated antimicrobials are not required to be labelled as containing
drugs, and make their way into products that consumers may not suspect contain
these compounds. Consumer education and more transparent labelling would help
ameliorate the overuse of antimicrobials in our homes, slowing community-based
ARB infections and decreasing PCPPs entering the environment.
111
Changing the way we conceptualize medicinal treatment is also an
important consideration. Antimicrobials are rapidly becoming obsolete, and drug
companies are no longer working on developing new drugs because the immense
cost is not worth the development of a drug that will only be effective for a short
time (Shnayerson and Plotkin, 2002). Human medicine is in danger of entering a
new era where antimicrobials no longer exist, which would drastically change
treatment of diseases and infections. While engineering, water resources, policy,
and educational advances can be made to slow or prevent this change, the posing
of this problem is perhaps indicative of an overall need for a new approach to
health care. Instead of treating symptoms, perhaps physical health could be
improved through preventative measures of diet, exercise, and mental state.
Wellness has long been considered a science by the majority of Americans, who
lead the developing world in health problems such as obesity and diabetes and
rank 37th overall in international health (International Diabetes Foundation, 2013;
WHO, 2013). If holistic health practices and concentration on prevention rather
than treatment was adopted by a larger majority of the population, the need for
antimicrobials and PCPPs could be decreased, eliminating the problem of these
contaminants in the environment altogether. This would require a paradigm shift
in the individual’s responsibility for and understanding of personal health, a large
transition for the majority of the nation, and until then, people will look to
medications to ease pain and cease illness, making PCPP and ARB pollution a
problem in human, environmental, and animal health.
112
This thesis incorporates the One Health interdisciplinary concept of
interconnectedness of human, animal, and environmental health in an attempt to
draw conclusions on inputs of antimicrobial compounds and ARB to the Salish
Sea. Although the limited sample size and unrefined geographic scope in this
study made relationships between human pollution and environmental and animal
health hard to draw, the findings of ARB in scat of the Southern Resident Killer
Whales suggest that monitoring the health of this species could help draw
conclusions about the health of all the creatures and the waters of the Salish Sea.
This study presents the first findings of antimicrobial resistant bacteria in
the scat of Orcinus orca, specifically the SRKW of the Salish Sea. While the
findings warrant more scientific investigation, this study was unable to pinpoint
any environmental or organism traits that cause trends in resistance. Further study
to improve upon this research must include the following components to increase
understanding of this phenomenon.
Subdivision of study areas: This study only used a coarse-grained spatial
analysis to analyze environmental variables as risk factors. Without using
extensive GIS mapping to further subdivide the study area, the six sites linked to
the environmental variables is only a rough estimate of correlation. Using more
advanced GIS, the population and number of septic tanks, as well as other
information like watershed outflow volume and number of WWTP outfalls, could
be directly spatially related to each sample site rather than related by assigned
area, making the spatial assessment of colony growth variability more complete.
113
Genetic identification of whales for cultured samples: The data from 2013
yielded more information on the diversity and prevalence of resistance. However,
without the identification of animal age, sex, and pod, the individual animal risk
assessment was not possible for these samples. Reviewing the field log showed
that each day a cultured sample was collected, sampling was occurring
simultaneously on all 3 pods, so there was no way to narrow down the identity of
each sample. With the genetic information on the whales, both environmental and
organism risk assessments can be conducted, to gain a better understanding of
relative contribution of each of these components to colonization with ARB.
Less time between sample collection and genetic analysis: The results
from the DNA concentration testing indicate that the nucleic acids used as a
template for PCR and genetic identification degraded. This is supported by the
large number of resistant colonies growing on fresh agar and the small number of
samples that tested positive for resistant genes. For future studies, immediate
culturing followed by colony isolation, identification, and genetic assessment of
resistance genes would improve this study’s usefulness.
114
REFERENCES
1. Al-Bahry SN, Mahmoud IY, Al-Zadjali, M, Elshafie A, Al-Harthy A, AlAlawi W. 2010. Antimicrobial resistant bacteria as bio-indicator of
polluted effluent in the green turtles, Chelonia mydas, in Oman.
Marine Environmental Research doi:
10.1016/j.marenvres.2010.12.005
2. Alekshun MN, Levy S. 2007. Molecular mechanisms of antibacterial
multidrug resistance. Cell. 128:1037-1050.
3. Austin DJ, Kristinsson KG, Anderson RM.1999. The relationship between
the volume of antimicrobial consumption in human communities and
the frequency of resistance. Proc Natl Acad Sci USA. 96: 1152–1156.
4. Barbosa TM, Levy SB. 2000. The impact of antibiotic use on resistance
development and persistence. Drug Resistance Updates. 3: 303–311.
5. Bartholow M. 2012. Top 200 drugs of 2011. Pharmacy Times.com Web
site. July 2012.
http://www.pharmacytimes.com/publications/issue/2012/July2012/Top
-200-Drugs-of-2011. Accessed May 1, 2013.
6. BD Diagnostics. 2009. Manual of microbiological culture media, second
edition. Becton, Dickinson and Company. Sparks, MD, USA.
7. Bearden D. 2013. Personal communication, September 6, 2013.
8. Bernhard AE, Field KG.2000.Identification of nonpoint sources of fecal
pollution in coastal waters by using host-specific 16S ribosomal DNA
genetic markers from fecal anaerobes. Appl. Environ. Microbiol. 66:
1587–1594.
9. Blackburn JK, Mitchell MA, Blackburn CH, Curtis A, Thompson BA.
2010. Evidence of antimicrobial resistant bacteria in free-swimming
top level marine predatory fishes. J. Zoo Wildl. Med. 41: 7-16.
10. Bonomo RA. Editorial Commentary: New Delhi Metallo-β-Lactamase and
Multidrug Resistance: A Global SOS? Clin Infect Dis. 52: 485-487
11. Bossart GD. 2006. Case study: Marine mammals as sentinel species for
oceans and human health. Oceanography. 19: 134-137.
115
12. Braine T. 2011. Race against time to develop new antimicrobials. Bulletin
of the World Health Organization 2011; 89:88–89.
13. Buller NB. 2003. Bacteria from fish and other aquatic animals: a practical
identification manual.CABI, Wallingford.
14. Burnes BS. 2003. Antibiotic resistance analysis of fecal coliforms to
determine fecal pollution sources in a mixed-use watershed.
Environmental Monitoring and Assessment. 85: 87–98.
15. Buschmann AH, et al. 2002. Salmon aquaculture and antimicrobial
resistance in the marine environment. PLoS ONE 7, e42724.
16. Cabello FC. 2006. Heavy use of prophylactic antibiotics in aquaculture: a
growing problem for human and animal health and for the
environment. Environmental Microbiology, 8: 1137–1144.
17. Calva JJ, Sifuentes-Osornio J, Cerón C. 1996. Antimicrobial resistance in
fecal flora: Longitudinal community-based surveillance of children
from urban Mexico. Antimicrobial Agents and Chemotherapy. 40:
1699–1702.
18. Casewell M, Friis C, Marco E, McMullin P, Phillips I. 2003. The
European ban on growth-promoting antibiotics and emerging
consequences for human and animal health. J. Antimicrob. Chemother.
52: 159-161.
19. CDC. 2012. National Antimicrobial Resistance Monitoring System
(NARMS). CDCs Emerging Infections Program, About.
http://www.cdc.gov/narms/about_narms.htm
20. CDC. 2013. Diseases/Pathogens associated with antimicrobial resistance.
Centers for Disease Control and Prevention, National Center for
Emerging and Zoonotic Infectious Diseases. Accessed January 15,
2013.
21. CDC. 2013b. One health. Centers for Disease Control and Prevention,
National Center for Emerging and Zoonotic Infectious Diseases
(NCEZID). Accessed October 18, 2013.
116
22. Chadwick DJ, Goode J, organizers. 1997. Antibiotic resistance: Origins,
evolution, selection and spread. Ciba Foundation Symposium 207.
New York: John Wiley. Chap 1-3, 5, 8, 9.
23. Chee-Sanford JC, Aminov RI, Krapac IJ, Garrigues-Jeanjean N, Mackie
RI. 2001. Occurrence and diversity of tetracycline resistance genes in
lagoons and groundwater underlying two swine production facilities.
Applied and Environmental Microbiology. 67: 1494-502.
24. Corcoran J, Winter MJ, Tyler CR. 2010. Pharmaceuticals in the aquatic
environment: A critical review of the evidence for health effects in
fish. Critical Reviews in Toxicology. 40: 287- 304.
25. Corpet DE. 1988. Antibiotic resistance from food. New England Journal
of Medicine.318: 1206–1207.
26. Dortet L, Poirel L, Anguel N, Nordmann P. 2012. New Delhi metallo-βlactamase 4–producing Escherichia coli in Cameroon [letter]. Emerg
Infect Dis.
27. Dougherty JA, Swarzenki PW, Dinicola RS, Reinhard M. 2010.
Occurrence of herbicides and pharmaceutical and personal care
products in surface water and groundwater around Liberty Bay, Puget
Sound, WA. J. Environ. Qual. 39:1173-1180.
28. EB Medicine. 2005. Antibiotics in the ED: How to avoid the common not
wisely, but too well. Table 4. EB Medicine, 2013, Accessed December
9, 2013.
29. Edge TA, Hill S. 2005. Occurrence of antibiotic resistance in Escherichia
coli from surface waters and fecal pollution sources near Hamilton,
Ontario. Canadian Journal of Microbiology. 51: 501-505.
30. EPA (Environmental Protection Agency). 1977. Summary of the Clean
Water Act. Washington, DC. Environmental Protection Agency.
31. ESRI. 2010. USA Population Density.
32. ESRI. 2013. Canada: Population Density.
33. Fraser DA, Gaydos JK, Karlsen E, Rylko MS. 2006. Collaborative
science, policy, development, and program implementation in the
117
transboundary Georgia Basin/Puget Sound ecosystem. Environmental
Monitoring Assessments 113: 49-69.
34. Ford JKB, Ellis GM, Balcomb K. 2000. Killer whales: The natural history
and genealogy of Orcinus orca in British Columbia and Washington.
UBC Press, Vancouver.
35. Gao L, Shi Y, Li W, Liu J, Cai Y. 2012. Occurrence, distribution and
bioaccumulation of antimicrobials in the Haihe River in China. J
Environ Monit. 14:1248-55.
36. Gaydos JK, Balcomb KC, Osborne RW, Dierauf L.2004. Evaluating
potential infectious disease threats for southern resident killer whales,
Orcinus orca: a model for endangered species. Biological Conseration.
117:253-262.
37. Gaydos JK, Dierauf L, Kirby G, Brosnan D, Davis GE. 2009. Top ten
principles for designing healthy coastal ecosystems like the Salish Sea.
EcoHealth 5: 460-471.
38. Gilliver MA, Bennett M, Begon M, Hazel SM, Hart CA. 1999. Antibiotic
resistance found in wild rodents. Nature. 401: 233–234.
39. Glad T, Bernhardsen T, Nielsen KM, Brusetti L, Andersen M, Aars J,
Sundset A. 2010. Bacterial diversity in faeces from polar bear (Ursus
maritimus) in Arctic Svalbard. BMC Microbiology 10:10.
40. Grudén M. 2013. Antimicrobial resistance associated with bacteria in
irrigation water - a case study from Southern Sweden. Swedish
Institute of Agricultural Sciences. Alnarp: SLU, Plant Breeding and
Biotechnology. Second cycle, A2E.
41. Harris G. 2009. Administration Seeks to Restrict Antimicrobials in
Livestock. The New York Times, published July 13, 2009.
42. Harrison JW, Svec TA. 1998. The beginning of the end of the
antimicrobial era? Part II. Proposed solutions to antimicrobial abuse.
Quintessence International 29: 223–9.
43. Harwood VJ, Whitlock J, Withington V. 2000. Classification of
antimicrobial resistance patterns of indicator bacteria by discriminant
118
analysis: use in predicting the source of fecal contamination in
subtropical waters. Appl. Environ. Microbiol. 66:3698–3704.
44. Hogg S. 2005. Essential Microbiology. John Wiley and Sons, Ltd. West
Sussex.
45. Hurst CJ, Crawford RL. (eds). 2002. Manual of environmental
microbiology, second edition. Amer Society for Microbiology,
Washington, D.C
46. Interagency Task Force on Antimicrobial Resistance. 2011. 2011 progress
towards implementation of: A public health action plan to combat
antimicrobial resistance.
47. International Diabetes Federation. IDF Diabetes Atlas, 6th edn. Brussels,
Belgium: International Diabetes Federation, 2013.
http://www.idf.org/diabetesatlas
48. Johnson A, Carey B, Golding S. 2004. Results of a Screening Analysis for
Pharmaceuticals in Wastewater Treatment Plant Effluents, Wells and
Creeks in the Sequim-Dungeness Area.
49. Johnson SP, Nolan S, Gulland FMD. 1998. Antimicrobial susceptibility of
bacteria isolated from pinnipeds stranded in central and northern
California. J. Zoo Wildl. Med. 29: 288–294.
50. Keil R, Salemme K, Forrest B, Neibauer J, Logsdon M. 2011. Differential
presence of anthropogenic compunds dissolved in the marine waters of
Puget Sound, WA and Barkley Sound, BC. Marine Pollution Bulletin
62:2404-2411.
51. Khachatourians GG. 1998. Agricultural use of antibiotics and the
evolution and transfer of antibiotic-resistant bacteria. Canadian
Medical Association Journal. 129:1129-1136.
52. King LJ, Anderson LR, Blackmore CG, Blackwell MJ, Lautner EA,
Marcus LC, Meyer TE, Monath TP, Nave JE, Ohle J, Pappaioanou M,
Sobota J, Stokes WS, Davis RM, Glasser JH, Mahr RK. 2008.
Executive summary of the AVMA One Health Initiative Task Force
report. JAVMA Vet Med Today: Special Report. 233:259-261.
119
53. Klare I, Badstubner D, Konstabel C, Bohme G, Claus H, Witte W. 2009.
Decreased incidence of VanA-type vancomycin-resistant enterococci
isolated from poultry meat and from fecal samples of humans in the
community after discontinuation of avoparcin usage in animal
husbandry. Microb Drug Resist. 5: 45–52.
54. Klekota J, Brauner E, Schreiber SL. 2005. Identifying biologically active
compound classes using phenotypic screening data and sampling
statistics. J. Chem. Inf. Model. 45: 1824-1836.
55. Kohn MH, Wayne RK. 1997. Facts from feces revisited. Trends in
Ecology and Evolution, 12:6.
56. Krahn MM, Wade PR, Kalinowski ST, Dahlheim ME, Taylor BL, Hanson
MB, Ylitalo GM, Angliss RP, Stein JE, Waples RS. 2002. Status
review of Southern Resident Killer Whales (Orcinus orca) under the
Endangered Species Act. NOAA Technical Memeorandum NMFSNWFSC-54 National Oceanic and Atmospheric Administration
(NOAA) National Marine Fisheries Service, Seattle, WA. xii-134.
57. Krahn MM, Hanson MB, Baird RW, Boyer RH, Burrows DG, Emmons
CK, Ford JKB, Jones LL, Noren DP, Ross PS, Schorr GS, Collier TK.
2007. Persistent organic pollutants and stable isotopes in biopsy
samples (2004/2006) from Southern Resident killer whales, Marine
Pollution Bulletin 54:1903-1911.
58. Kümmerer K. 2004. Resistance in the environment. J. Antimicrob.
Chemother. 54: 311-320.
59. Kümmerer K. 2008. Pharmaceuticals in the environment: Sources, fate,
effects and risks. 3rd edn. Springer, Berlin. K. Kümmerer.
60. Kümmerer K. 2009. Antibiotics in the aquatic environment- A reviewPart II. Chemosphere, 75: 435-441.
61. Le Bris H, Pouliquen H. 2004. Experimental study on the bioaccumulation
of oxytetracycline and oxolinic acid by the blue mussel (Mytilus
edulis). An evaluation of its ability to bio-monitor antimicrobials in the
marine environment. Marine Pollution Bulletin, 48: 434-440.
120
62. Levin-Edens E, Soge OO, No D, Stiffarm A, Meschke JS, Roberts MC.
2012. Methicillin-resistant Staphylococcus aureus from northwest
marine and fresh water recreational beaches. FEMS Microbiol
Ecol.79: 412–420.
63. Levy SB. 1992. The antibiotic paradox. New York: Plenum Press.
64. Littnan CL, Stewart BS, Yochem PK, Braun R. 2007. Survey for selected
pathogens and evaluation of disease risk factors for endangered monk
seals in the main Hawaiian Islands. EC Health. 3: 232–244.
65. Livermore DM. 2003. Bacterial Resistance: Origins, Epidemiology, and
Impact
Clin Infect Dis. 36 (Supplement 1): S11-S23 doi:10.1086/344654
66. Lockwood SK, Chovan JL, Gaydos JK. 2006. Aerobic bacterial isolations
from harbor seals (Phoca vitulina) stranded in Washington: 1992–
2003. J. Zoo Wildl. Med. 37: 281–291.Lorian, V. 1996.
Antimicrobials In laboratory medicine, 4th edition. Lippincott Williams
and Wilkins, Baltimore, MD, USA.
67. Lubliner B, Redding M, Ragsdale D. 2010. Pharmaceuticals and personal
care products in municipal wastewater and their removal by nutrient
treatment technologies. Washington State Department of Ecology,
Olympia, WA.
68. McQuaig S, Griffith J, Harwood VJ. 2012. Association of fecal indicator
bacteria with human viruses and microbial source tracking markers at
coastal beaches impacted by nonpoint source pollution. Appl Environ
Microbiol. 78: 6423–6432.
69. Miller RV, Gammon K, Day MJ. 2009. Antimicrobial resistance among
bacteria isolated from seawater and penguin fecal samples collected
near Palmer Station, Antarctica. Canadian Journal of Microbiology.
55:37.
70. Moore WEC, Cato EP, Holdeman LV. 1969. Anaerobic bacteria of the
gastrointestinal flora and their occurrence in clinical infections. J.
Infect. Dis. 119:641-649.
121
71. Naylor RL, Goldburg RJ, Primavera JH, N Kautsky, Beveridge MCM,
Clay J. 2000. Effect of aquaculture on world fish supplies. Nature 405:
1017–1024.
72. Nilsen E, Rosenbauer R, Furlong E, Burkhardt M, Werner S, Greaser L,
Noriega M. USGS. 2007. Pharmaceuticals and personal care products
detected in streambed sediments of the lower Columbia River and
selected tributaries.
73. Okeke IN, Edelman R. 2001. Dissemination of antimicrobial resistant
bacteria across geographic borders. Clinical Infectious Diseases
33:364-9.
74. One Health Global. 2012. What is One Health? One Health Global
Network Webportal, Accessed 7 October 2013.
75. Palumbi S, Martin A, Romano S, McMillian WO, Stice L, Grabowski G.
2002. The simple fool’s guide to PCR. Department of Zoology and
Kewalo Marine Laboratory, University of Hawaii.
76. Peak N, Knapp CW, Yang RK, Hanfelt MM, Smith MS, Aga DS, Graham
DW. 2007. Abundance of six tetracycline resistance genes in
wastewater lagoons at cattle feedlots with different antibiotic use
strategies. Environmental Microbiology 9: 143-151.
77. Puget Sound Partnership. 2011. Puget Sound Science Update, April 2011
version. Accessed from http://www.psp.wa.gov/. Puget Sound
Partnership. Tacoma, Washington.
78. Qiagen, 2013. DNA protocols and applications. Protocol and Application
Guidelines. Available online: http://www.qiagen.com/knowledge-andsupport/spotlight/protocols-and-applications-guide/dna/
79. Rabinowitz P, Conti L. 2013. Links Among Human Health, Animal
Health, and
Ecosystem Health. Annual Review of Public Health. 34:189-204.
80. Ridaura VK, Faith JJ, Rey FE, Cheng J, Duncan AE, Kau AL, et al. 2013.
Gut microbiota from twins discordant for obesity modulate
122
metabolism in mice. Science. 341: 1241214
[DOI:10.1126/science.1241214]
81. Roberts MC, Soge OO, Giardino MA, Mazengia E, Ma G, Meschke JS.
2009. Vancomycin‐resistant Enterococcus spp. in marine
environments from the West Coast of the USA. Journal of Applied
Microbiology 107:300-307.
82. Roberts RR, Ahmad HB, Scott RD, Foster SD, Abbasi F, Schabowski S,
Kampe LM, Ciavarella GG, Supino M, Naples J, Cordell R, Levy SB,
Weinstein RA. 2009. Hospital and societal costs of antimicrobialresistant infections in a Chicago teaching hospital: Implications for
antimicrobial stewardship. Clin. Infect Dis. 49:1175-84.
83. Ross PS, Ellis GM, Ikonomou MG, Barrett-Lennard LG, Addison RF.
2000. High PCB concentrations in free-ranging pacific killer whales
(Orcinus orca): Effects of age, sex and dietary preference. Marine
Pollution Bulletin 40: 504-515.
84. Schaefer AM, Reif JS, Goldstein JD, McCulloch SD, Mazzoil M, Fair PA,
Bossart GD. 2011. Risk factors for Colonization of Antimicrobial
Resistant E. Coli in Atlantic Bottlenose Dolphins (Tursiops truncatus)
in the Indian River Lagoon, FL. J Environ Public Health.
85. Schroeder P, Raverty S, Zebek E, Cameron D, Eshghi A, Bain D, Wood
R, Rhodes L, Hanson B. 2009. Investigation into the microbial culture
and molecular screening of exhaled breaths of endangered Southern
Resident Killer Whales (SRKW) and pathogen screening of the sea
surface microlayer (SML) in Puget Sound. Proceedings of the 2009
Puget Sound Georgia Basin Ecosystem Conference.
86. Senta I, Terzic S, Ahel M. 2012. Occurrence and fate of dissolved and
particulate antimicrobials in municipal wastewater treatment. Water
Research doi: 10.1016/j.watres.2012.10.041.
87. Shnayerson M, Plotkin MJ. 2002. The killers within: The deadly rise of
drug resistant bacteria. New York: Little Brown and Company.
123
88. Smith DW. 1999. Decreased antimicrobial resistance after changes in
antibiotic use. Pharmacotherapy 19: 129S–132S.
89. Smith TJ. 2012. Drugs and the environment: A call for change in drug
dispensing practices. Drug Intox Detox 1:1.
90. Soge OO, Meschke JS, No D, Roberts MC. 2009. Characterization of
methicillin-resistant Staphylococcus aureus and methicillin-resistant
coagulase-negative Staphylococcus spp. isolated from US West Coast
public marine beaches. Journal of Antimicrobial Chemotherapy 64:
1148-1155.
91. Stanley S. 2010. Puget Sound watershed characterization - Introduction to
the water flow assessment for Puget Sound. Washington State
Department of Ecology. Publication number 10-06-014.
92. Stoddard RA, Atwill ER, Gulland FMD, Miller MA, Dabritz HA, Paradies
DM, Worcester KR, et al. 2008. Risk Factors for infection with
pathogenic and antimicrobial-resistant fecal bacteria in northern
elephant seals in California. Public Health Reports123: 360–370.
93. Taberlet P, Wasites LP, Luikart G. 1999. Noninvasive genetic samplingLook before you leap. Trends in Ecology and Evolution 14:8.
94. Take Back Your Meds. 2011. Pharmaceuticals in the environment.
95. Taylor BL, Baird R, Barlow J, Dawson SM, Ford J, Mead JG,
Notarbartolo di Sciara G, Wade P, Pitman RL. 2013. Orcinus orca. In:
IUCN 2013. IUCN Red List of Threatened Species. Version 2013.2.
<www.iucnredlist.org>. Downloaded on 08 September 2013.
96. Taylor LH, Latham SM, Woolhouse ME. 2001. Risk factors for human
disease emergence. Philos Trans R Soc Lond B Biol Sci.356:983–989.
97. Todar K. 2011. Basic mechanisms of antimicrobial action and resistance.
Todar’s Online Textbook of Bacteriology.
98. Torrey EF, RH Yolken. 2005. Beasts of the earth. New Brunswick, NJ:
Rutgers University Press.
99. Union for Concerned Scientists. 2012. Prescription for trouble.
124
100. van den Bogaard AE, Bruinsma N, Stobberingh EE. 2000. The effect of
banning avoparcin on VRE carriage in the Netherlands. J Antimicrob
Chemother. 46: 146–147.
101. van den Bogaard, A. E., Willems, R., London, N., Top, J., and E.E.
Stobberingh. 2002. Antibiotic resistance of faecal enterococci in
poultry, poultry farmers and poultry slaughterers. The Journal of
Antimicrobial Chemotherapy 49: 497-505.
102. Wasser SK, Keim JL, Taper ML, Lele SR. 2011. The influences of wolf
predation, habitat loss, and human activity on caribou and moose in
the Alberta oil sands. Frontiers in Ecology and the Environment.
doi:10.1890/100071
103. Wasser SK, Davenport B, Ramage ER, Hunt KE, Parker M, Clarke M,
Stenhouse G. 2004. Scat detection dogs in wildlife research and
management: application to grizzly and black bears in the Yellowhead
Ecosystem, Alberta, Canada. Can. J. Zool. 82: 475–492.
104. World Health Organization (WHO). 2013. World health statistics 2013.
World Health Organization. Geneva, Switzerland.
105. Wiseman C, Matthews R, Vandersypen J. 2003. San Juan County
monitoring project final report. San Juan County Conservation
District.
106. Wong S. 2002. Ocean sentinels: Marine mammals and antimicrobial
resistance. September 27–30, 2002, San Diego, California. Proc. 42nd
Interscience Conf. Antimicrob. Agents Chemother.
107. Yang H, Byelashov OA, Geornaras I, Goodridge LD, Nightingale KK,
Belk KE, Smith GC, Sofos JN. 2010. Center for Meat Safety and
Quality, Food Safety Cluster, Department of Animal Sciences,
Colorado State University, 1171 Campus Delivery, Fort Collins, CO
80523-1171, USA.
125
Potter_SMES2013.pdf