Characterization of the Fecal Microbiota of Commercial Mink (Neovison Vison)

Characterization of the Fecal Microbiota of Commercial Mink (Neovison vison)

Nicole Compo

A Thesis

presented to

The University of Guelph

In partial fulfilment of requirements

for the degree of

Doctor of Veterinary Science

Pathobiology

Guelph, Ontario, Canada

ABSTRACT

Characterization of the Fecal microbiota of Commercial Mink (Neovison vison)

Nicole Compo Advisor: University of Guelph, 2017 Dr. Patricia V. Turner

Mink, obligate carnivores, are the fifth largest livestock commodity group in Canada, with more than 3 million pelts produced annually and an industry valuation at almost $100 million (2015). To date, research in the field of microbiota; the collective, interacting genomes and the symbiotic microorganisms in the gastrointestinal tract, has focused primarily on herbivores and omnivores, with comparatively few studies carried out in carnivores. The goal of the current study was to characterize the fecal microbiota of farmed mink, Neovison vison. Fecal samples (n=366) were collected from adult females and weaned kits during two consecutive summers in 2014 and 2015 and adult females just before winter breeding in 2016. The V4 regions of the 16S rRNA genes were amplified, sequenced, and classified from extracted fecal bacterial DNA. The most predominant phyla were Firmicutes and Proteobacteria, with all other phyla together comprising <5% of the sequences identified. The relative abundance of Proteobacteria was found to be much higher than previous reports in mammals. Additionally, ranges in the relative abundances of both the predominant and many non-predominant taxa were found to be very large, suggesting that mink are tolerant to a broader community of organisms than has been seen in other mammals. Few differences were identified at each taxonomic level between adult females and weaned kits, though several genera within the families Lactobacillales and Clostridiales were identified to be differentially abundant in weaned kits (p<0.05). The years 2014 and 2016, the two periods that would have had the fewest number of animals in common, had the greatest number of taxa that were different. Farms were generally similar in the predominant phyla and genera from 2014 to 2015. The microbiota can likely be impacted by a range of dietary, environmental, and treatment factors, none of which are understood in mink. This body of work will contribute significantly to our understanding of how these factors impact the fecal microbiota of farmed mink, and the carnivore GI microbiota more generally, with the hope of ultimately utilizing this information to optimize the health, production, and welfare of mink.

DEDICATION

I dedicate this work to my parents, my anchors…my Mommy and DD, whose endless love, support, and laugh-until-I-cry moments have gotten me through the stormiest of waters.

iii ACKNOWLEDGEMENTS The completion of my program would undoubtedly not have been possible without a great number of people. From help and advice to laughter and encouragement, their roles were diverse and important in ways that I didn’t always realize. First, my advisory committee. To my advisor, Pat Turner, thank-you for your willingness to share your limitless knowledge with me, encouragement of my creative and independent thoughts, and always pushing me to be better and to do more. You’ve been an incredible resource and I am lucky to have been able to learn under you. To Scott Weese, your patience, kindness, and insight made navigating the world of microbiome and, ESPEPCIALLY, mothur, manageable. I may never have gotten through the analysis without you. Many thanks to David Pearl for all your help and guidance with the statistics for the mink post-mortem project. Finally, to Andrew Winterborn, who is both a mentor and a friend. Thank-you for always believing in me more than I believed in myself. You’ll never know how much it meant to me. Thank-you to the staffs of the Weese and Turner labs, especially Jutta and Joyce, you were always there when I wanted to poke my eye out with a pipette when my PCR didn’t work after days of pippetting. You’ve answered my many millions of questions with patience and enthusiasm, so thank-you for training and guiding me through this project. Next, my quad-mates, and greater scope room peeps, without whom this thesis likely would have been finished many months ago. Thank-you for the laughter, support, and, most importantly, friendship. I wish you all the best in wherever your endeavours take you. I hope we will find reason to come together again. To my one and only deskie, Chris, I appreciate someone to talk football with in a country where I’m surrounded by hockey fans. You’ve been a great friend and I look forward to our many monkey path consults. To my LLP, who knows who she is but would hate the recognition: thank-you for being you. The fur butts and I will miss you immensely, as I don’t see you ever taking a vacation, much less to an island of sunshine, beaches, and every meal outside. I’ll smuggle pickles for you any time. And, finally, my amazing family. I’ve got a team of parents whose emotional and, most importantly, financial support made pursuing my dream of being a laboratory animal veterinarian possible. When I came up with this crazy idea of moving to Canada, my mother said, “What!? Are you sure? You know we’re not going to come visit, right?” I definitely wasn’t sure then, but I am now. I have grown so much as both a person and veterinarian because of their support. To

iv the little bro, who will forever be the rational to my emotional, the common sense to my absent- mindedness, and the artist to my scientist, thank-you for always answering my random calls in the middle of the night when I just want to say hi and need a laugh. To Katlyn, who, in so many ways, is a mini-me. I so look forward to watching you conquer the world of medicine; you’re going to be amazing whatever you choose to do and wherever you choose to do it. And to my perfect nephew, Jude, I’m so happy I’ll soon be closer and get to see more of you growing up. They’re my best friends and favorite people to be with and without them, this journey would not have been possible. I love you guys so much. One final thanks to my financiers: the Canadian Mink Breeders Association and the Ontario Ministry of Agriculture, Food, and Rural Affairs. None of this would have been possible without them.

v DECLARATION OF WORK PERFORMED

All work presented in this thesis was completed by myself, with the following exceptions: • Collection of fecal samples was completed by Brian Tapscott • Initial receipt and packaging of samples was conducted by Dr. Jutta Hammermueller and Sabrina Stevens • Aliquoting was done by myself and Alison Thomas • 16S rRNA gene sequencing using the Illumina MiSeq was conducted by Jeffrey Gross of the University of Guelph’s Advanced Analysis Centre, Guelph, ON • Post-mortem examinations of mink kits included in Appendix A were conducted by Drs. Patricia V. Turner, Marina Brash, and Ms Amanda Storer • Microbiological evaluation of enteric tissue samples in Appendix A was completed by Dr. Durda Slavic

vi TABLE OF CONTENTS ABSTRACT ...... ii ACKNOWLEDGEMENTS ...... iv DECLARATION OF WORK PERFORMED ...... vi TABLE OF CONTENTS ...... vii LIST OF TABLES ...... ix LIST OF FIGURES ...... x LIST OF APPENDICES ...... xi ABBREVIATIONS ...... xii CHAPTER ONE: INTRODUCTION TO MINK AND THE GASTROINTESTINAL MICROBIOTA ...... Error! Bookmark not defined. 1.1 INTRODUCTION ...... 1 1.1.1 Production Cycle in Captivity ...... 1 1.1.2 Anatomy and Physiology of the Mink Gastrointestinal Tract ...... 2 1.1.3 Diet ...... 2 1.2 INFECTIOUS ENTERIC AGENTS OF FARMED MINK ...... 3 1.3 THE ENTERIC MICROBIOTA ...... 5 1.3.1 Methods of Characterizing the Enteric Microbiota ...... 6 1.3.2 Role of the Bacterial Gastrointestinal Microbiota in Health and Disease ...... 7 1.4 THESIS OBJECTIVES AND HYPOTHESES ...... 15 1.4.1 Hypotheses ...... 15 1.4.2 Study objectives ...... 15 CHAPTER TWO: THE FECAL BACTERIAL MICORBIOTA OF COMMERCIAL MINK (NEOVISON VISON): YEARLY, LIFE STAGE, AND SEASONAL COMPARISONS ...... 18 2.1 INTRODUCTION ...... 18 2.2 MATERIALS AND METHODS ...... 21 2.2.1 Farm recruitment and sample collection ...... 21 2.2.2 Sample processing ...... 21 2.2.3 DNA extraction, amplification, and sequencing of bacterial 16s rRNA gene ...... 22 2.2.4 Statistical analyses ...... 23 2.3 RESULTS ...... 24 2.3.1 Relative abundance ...... 24 2.3.2 LEfSe analyses ...... 25

vii 2.3.3 Alpha and beta diversity ...... 25 2.4 DISCUSSION ...... 26 CHAPTER THREE: GENERAL DISCUSSION ...... 47 3.1 The commercial mink fecal microbiota: What is normal? ...... 47 3.2 Optimization of production through manipulation of the microbiota ...... 49 3.3 Proteobacteria: Unhealthy bacteria in healthy animals ...... 50 3.4 Future Directions ...... 51 3.5 Summary and Conclusions ...... 52 REFERENCES ...... 54 APPENDICES ...... 70

viii LIST OF TABLES

Table 1.1 Metabolizable energy requirements for growing kits ...... 17

Table 2.12Number of pooled fecal samples collected from adult females and weaned kits collected from commercial mink farms in Ontario...... 32

Table 2.23Number of samples used in the final analysis of the fecal microbiota of commercial mink. .... 33

Table 2.34Relative abundance of predominant taxonomic classifications of bacteria from the feces of commercial mink ...... 34

Table 2.45Relative abundance and false discovery rate p-values, by year, at each taxonomic level ...... 35

Table 2.56Relative abundance and false discovery rate (FDR) p-values for significantly different taxa by life stage in mink fecal microbiota...... 37

Table 2.67Relative abundance and false discovery rate (FDR) p-values for significantly different taxa (p>0.05) by season in the fecal microbiota of mink ...... 38

Table 2.78Bacterial genera enriched, by year, from the feces of mink ...... 40

Table 2.89Differences (p-values) between groups compared with UniFrac, AMOVA, and Parsimony analysis in membership and structure of the fecal microbiota of mink ...... 41

Table A.1 Prevalence of gross post mortem lesions in 5,657 preweaned, found dead mink kits from 21 commercial mink farms in Ontario...... 90

Table A.2 Bacterial culture results from 45 found dead kits with suspected enteritis ...... 91

Table A.3 Producer-reported production characteristics and incidence of mortality from 2013 ...... 92

Table A.4 Producer-reported farm demographics from biosecurity survey ...... 93

Table A.5 Specific biosecurity protocols implemented by mink producers (n=11) ...... 94

Table A.6 Summary of responses for biosecurity-specific questions on survey ...... 95

Table A.7 Summary of responses for biosecurity-specific questions on survey with yes/no answers or multiple answers ...... 96

ix LIST OF FIGURES

Figure 2.1 Relative abundance of the 20 most predominant bacterial genera present in the feces of commercial mink ...... 42 Figure 2.2 LEfSe analysis indicating differentially enriched bacterial communities at the genus level by life stage ...... 43 Figure 2.3 LEfSe analysis indicating differentially enriched bacterial communities at the genus level by season...... 44 Figure 2.4 Alpha diversity parameters (a: Richness, b: evenness, c: diversity) observed in the feces of commercial mink...... 45 Figure 2.5 Principal Coordinate Analysis (PCoA) of community structure and membership of the fecal microbiota of commercial mink ...... 46 Figure A.6: Weight distribution of preweaned, found dead mink kits ...... 98 Figure A.7: Producer-reported common health problems ...... 99

Figure B.1 Relative abundance of 20 most predominant bacterial genera present in the feces of commercial mink ...... 100

Figure C.1: Dendrogram representing the community membership and structure of the fecal microbiota of commercial mink...... 101

x LIST OF APPENDICES

APPENDIX A: ON-FARM BIOSECURITY PRACTICES AND CAUSES OF PREWEANING MORTALITY IN CANADIAN COMMERCIAL MINK KITS ...... 71 A.1 INTRODUCTION ...... 72 A.2 MATERIALS AND METHODS ...... 74 A.2.1 Part I – Preweaned Kit Mortality Surveillance ...... 74 A.2.2 Part II - Survey of Husbandry and Management Practices ...... 76 A.2.3 Statistical analyses ...... 76 A.3 RESULTS ...... 77 A.3.1 Part I: Preweaned Kit Mortality Surveillance ...... 77 A.3.2 Part II: Survey of Husbandry and Management Practices ...... 78 A.4 DISCUSSION ...... 81 A.5 REFERENCES ...... 86 Table A.1: Prevalence of gross post mortem lesions in 5,657 preweaned, found dead mink kits from 21 commercial mink farms in Ontario...... 91 Table A.2: Bacterial culture results from 45 found dead kits with suspected enteritis ...... 92 Table A.3: Producer-reported production characteristics and incidence of mortality, 2013 .... 93 Table A.4: Producer-reported farm demographics from biosecurity survey ...... 94 Table A.5: Specific biosecurity protocols implemented by mink producers ...... 95 Table A.6: Summary of responses for biosecurity-specific questions on survey ...... 96 Table A.7: Summary of responses for biosecurity-specific questions on survey with yes/no answers or multiple answers...... 97 Figure A.1: Weight distribution of preweaned, found dead mink kits at the time of post mortem examination from 21 fur farms in Ontario ...... 99 Figure A.2: Producer-reported common health problems ...... 100 APPENDIX B: FECAL MICROBIOTA COMPOSITION OF COMMERCIAL MINK IN ONTARIO AT THE FARM LEVEL ...... 101 APPENDIX C: COMMUNITY MEMBERSHIP AND STRUCTURE OF THE FECAL MICROBIOTA OF COMMERCIAL MINK ...... 102

xi ABBREVIATIONS

ADV Aleutian disease virus bp Base pairs DNA Deoxyribonucleic acid FDR False discovery rate GF Germ free LDA Linear discriminant analysis MEC Mink enteric calicivirus MEV Mink enteritis virus OR Odd’s ratio OTU Operational taxonomic units PCR Polymerase chain reaction PCoA Principal coordinate analysis rRNA Ribosomal ribonucleic acid

xii CHAPTER ONE: INTRODUCTION TO MINK AND THE GASTROINTESTINAL MICROBIOTA

1.1 INTRODUCTION Mink comprise several species in two genera of semi-aquatic carnivores within the family Mustelidae, along with polecats, weasels, otters, ferrets, and badgers. There are two extant species: the American mink (Neovison vison) and the European mink (Mustela lutreola). American mink are an important animal commodity group in Canada with approximately 800,000 mink on 236 fur farms and more than 3 million pelts sold, valued at almost $100 million in 2015 (Statistics Canada, 2017).

1.1.1 Production Cycle in Captivity

Mink are triggered to breed by increasing day-length and produce one litter per year (Sundqvist et al., 1989; Hildebrandt, 2014). In captivity, one male is typically bred to multiple females in late February to early March. Kits are weaned at 6-8 weeks of age and the female is removed to another pen. In North America, producers vaccinate kits against Pseudomonas, Clostridium botulinum toxin, mink enteritis virus, and distemper virus in July-August of the production year, or when kits are about 10-12 weeks of age, using a multivalent vaccine (Hildebrandt, 2014). A common practice on mink farms is to feed animals for rapid growth and high pelt quality leading up to pelting in the late fall (when animals are euthanized and their furs are removed), then to calorie restrict replacement and breeding females through the late fall and early winter, prior to breeding (Boudreau et al., 2014). The principle behind this practice is to 1maximize pelt yield, but optimize reproductive success, because ideal body condition score is positively correlated with increased litter size and decreased kit mortality (Lagerkvist et al., 1994). However, recent research suggests that maintaining the females throughout the year at an ideal body score (rather than having marked weight fluctuations) results in larger live litters, compared to traditionally raised controls (Boudreau et al., 2014). This suggests that the practice of calorie restriction prior to the mating season, though common, may not be the most efficient or welfare-friendly practice for mink.

1 1.1.2 Anatomy and Physiology of the Mink Gastrointestinal Tract Mink are true carnivores and the gastrointestinal tract, as in other mustelids, is relatively simple and short. They lack a cecum and the distal aspect of the small intestine is demarcated from the large intestine only by a sudden change in the mucosal layer (Stevens and Hume, 2004). Mink are often compared anatomically to ferrets and cats due to similarities in diet, but the gastrointestinal tract most closely resembles bears and raccoons (Stevens and Hume, 2004). In nonruminant species, the total food transit time is determined by the total time spent in the large intestine (Hendriks et al., 2012). In mink, the short overall length of the gastrointestinal tract leads to a relatively fast total food transit time. Bleavins and Aulerich (1981) found that the average food passage time was 187 minutes in mink, while Sibbald et al, (1962) found an average of 142 minutes (range: 62-215 minutes). Research on the relationship between gastrointestinal transit time and the intestinal microbiota is limited. In one human study, increasing the colonic transit time by 186% via drugs that alter colonic motility led to a significant decrease in stool weight and total bacterial mass compared to pre-study values (Stephen et al., 1987), but whether this altered the relative composition of the microbiota was not evaluated. The significance of this finding from a nutritional standpoint relates to the amount of time bacteria have to assimilate nutrients, as a fast transit time alters bacterial production of short chain fatty acids and breakdown of amino acids, when compared to species with slower gastrointestinal transit time (Stephen et al., 1987).

1.1.3 Diet In the wild, mink eat a wide variety of prey, including small mammals, frogs, fish, crayfish, insects, worms, and birds, depending on what is available in the environment. In captivity, diets are frequently comprised of large portions (70-80%) of offal or food animal processing by-products unused for human consumption, such as fish filleting waste, cracked eggs, and various poultry and mammalian waste organs and tissues; the remainder being composed of cereal, vitamins, and minerals (Urlings et al., 1993; Hunter and Lemieux, 1996). Young mink begin to consume solid food at 3-5 weeks and nurse until 8-10 weeks, but are weaned as early as 6 weeks in farm settings (NFACC, 2013). Nutritional requirements of mink vary by age and stage of production. The guidelines for adult mink, based on several different studies and considering factors such as ambient

2 temperature and sex, is an average of 140 kcal metabolizable energy (ME)/kg body weight throughout the year (NRC, 1982). Formal studies have not been conducted on pregnant females to determine their energy requirements, but the generally accepted recommendation is to feed a comparable diet to that of early growing kits: around 200 kcal ME/kg BW daily (NRC, 1982). From the onset of consumption of animal-based protein, to weaning, to growth into adulthood, the energy needs of kits change dramatically and the general requirements over the course of the production cycle are summarized in Table 1.1. Because feed contributes most significantly to the cost of production on a farm and the overall welfare of animals, appropriate feed practices are of paramount importance to the mink industry.

1.2 INFECTIOUS ENTERIC AGENTS OF FARMED MINK While the scope of this review does not include an exhaustive review of single agent gastrointestinal pathogens in mink, a basic overview of those known to be present in healthy and diseased animals facilitates an understanding of the role of the microbiota in these animals. For example, Campylobacter jejuni has been shown experimentally to cause disease in healthy mink after oral inoculation with this bacterium (Hunter et al., 1986). However, C. jejuni is routinely cultured from the intestinal contents of mink with nursing disease (a condition seen in lactating females and characterized by dehydration, electrolyte imbalances, and death (Hildebrandt, 2014)) as well as healthy controls, all from farms in Southern Ontario (Schneider and Hunter, 1993). The authors concluded from this, as well as a lack of intestinal inflammation in culture- positive females, that Campylobacter spp. isolation is likely unrelated to disease. Similarly, a large-scale study of 500 mink from three Wisconsin ranches concluded that presence of C. jejuni in the gut was not necessarily associated with enteric disease in healthy mink (Bell and Manning, 1990). Together, these studies suggest that other contributing factors are likely to be important for determining the pathogenicity of C. jejuni and that development of disease is multifactorial. Similarly, multiple studies indicate that mink likely naturally harbour Staphylococcus delphini; cultured from the skin of 100% (118/118) mink in Denmark in one study (Guardabassi et al., 2012). The organism may also act as an opportunistic pathogen, as it has been associated with an outbreak of “sticky kit” disease (a syndrome of mucoid diarrhea and excess secretion by apocrine glands, giving affected mink kits the appearance of being dirty and unkempt (Sledge et al., 2010)), that resulted in mortality of >5000 kits on one Michigan mink farm (Sledge et al.

3 2010). Escherichia coli has also been isolated from both healthy and diarrheic mink (Vulfson et al., 2001), although phage and toxin typing was not investigated. Studies examining different single agents present in both healthy and diseased animals indicate that while certain bacteria may be present in disease, the actual cause of clinical signs is likely to be multifactorial. Additionally, this highlights the need for a more holistic view of the role of the microbiota as it relates to disease in farmed mink. As for enteric bacterial infections, certain enteric viruses have been identified in both healthy and diarrheic mink. Recently, a novel bocavirus was isolated from ~30% of mink fecal samples obtained from 30 healthy and diarrheic mink in China, of which only half of the mink had clinical signs of diarrhea (Yang et al., 2016). Mink enteritis virus (MEV), a parvovirus, is an important pathogen of mink, and much studied. There is rapid, transient shedding of the virus in feces, which can last for at least one year following recovery from disease (Uttenthal et al., 1990; Bouillant and Hanson, 1965). While recovered animals develop long-lived immunity to the virus, acute mortality rates in kits can reach 80% (Uttenthal et al., 1990; Bouillant and Hanson, 1965). Of particular interest given that there are few biosecurity measures in place on most fur farms (Compo et al., 2017), mink will exhibit signs of enteritis when inoculated with feline panleukopenia, another parvovirus, and develop immunity to MEV once recovered (Barker et al., 1982; Macpherson 1956). This indicates that mink are susceptible to feline parvovirus and highlights the need for control of companion cat and wildlife access to farms. Although enteric virus infections are most likely to manifest as gastrointestinal disease, certain systemic viral infections may indirectly affect the gastrointestinal microbiota through immune modulation or alteration of bacterial burden. For example, Aleutian disease virus (ADV), a parvovirus of significant economic concern to mink farmers, causes multi-organ plasmacytic infiltrates, in addition to chronic immunomodulatory effects (Hildebrandt, 2014; Martino and Martino, 1996). In found-dead mink kits, those from ADV-positive farms had higher bacterial burdens than those from ADV- negative farms, though whether the role of the organisms were as primary pathogens, or secondary invaders, was unclear (Martino and Martino, 1996). This indicates that, especially in kits, infection with ADV can increase susceptibility to secondary bacterial infection and likely influences the composition of the intestinal microbiota. Modern sequencing technology has also aided in the identification of viruses as possible etiologic agents in diseases where causes could not previously be elucidated. For example, a

4 novel orthoreovirus was identified by PCR in mink from China with diarrhea, which were negative by PCR for other classical mink viruses. Koch’s postulates were fulfilled since experimental inoculation of healthy mink induced diarrhea, confirming the role of the novel virus as the causative agent (Lian et al., 2013). Mink enteric calicivirus (MEC) has been detected by electron microscopy and RT-PCR from fecal samples of mink with diarrhea in the USA (Guo, 2001). Similarly, during a mink diarrhea outbreak on a farm in China, a novel circovirus was identified by use of degenerate PCR primers in 100% of afflicted mink and none of the healthy mink, following exclusion by PCR of all other classical endemic viruses (Lian et al., 2013). The virus appeared closely related to a human strain that had previously been isolated from two human children with necrotizing encephalopathy (Lian et al., 2013). Finally, a novel astrovirus was isolated, sequenced, and epidemiologically linked to pre-weaned mink kits with diarrhea in Northern Europe (Englund et al., 2002; Mittelholzer et al., 2003; Mittelholzer et al., 2003). Interestingly, a virus with ~80% sequence similarity to that same astrovirus was subsequently isolated from the brains of mink from Denmark exhibiting signs of “shaking syndrome”, indicating that the virus may have subtypes with differential affinity for tissues (Blomström et al., 2010). Intensive diagnostic efforts were undertaken to try to identify the cause, but it was not until random amplification and sequencing was undertaken years later that this astrovirus was identified as the etiologic agent by fulfilling Koch’s postulates (Blomström et al., 2010). Next generation sequencing is expected to be an invaluable tool in determining the etiopathogenesis of newly identified viruses, and will help to clarify whether virus infections are the result of interspecies transmission (important when considering the close proximity of farmed mink and humans) or virulent mutations of viruses normally present in the virome of mink. This is another reason why an understanding of what is “normal” is paramount to our understanding of disease in farmed mink.

1.3 THE ENTERIC MICROBIOTA The gastrointestinal microbiota is composed of the collective, interacting genomes of the symbiotic microorganisms in the gastrointestinal tract (Kinross et al., 2011). It has recently been shown to have a significant role not only in the pathogenesis of disease, but also in health and development (Velasquez-Manoff, 2015, Ericsson et al. 2015). This finding is not surprising as there are approximately 10x more bacterial cells in the mammalian body than there are host cells,

5 and as much as 45% of the host metagenome (all of the genetic material present in a host) is non- host genetic material (Haynes and Rohwer, 2011; Minot et al., 2011; Swanson et al., 2011). As a result, there is significant interest in research in this area.

1.3.1 Methods of Characterizing the Enteric Microbiota Historically, identification of bacteria relied primarily on direct culture. However, bacterial culture, while sensitive for the detection of certain bacteria, has several shortcomings that limit its sensitivity in other bacterial species (Timoney, 2004; Libby, 2010). Some studies indicate that only 10-20% of bacteria can be cultured by classic techniques (Pang et al., 2012; Suchodolski, 2011), while others put this estimate even lower at <1% (Štursa et al., 2009). Additionally, once cultured, bacteria must be correctly identified and classified based on available data of known species, further limiting the number of rare or previously unidentified species that can be characterized (Suchodolski, 2011). Because of these limitations, bacterial species that are readily cultured may be over-represented using this technique, whereas those that are difficult to culture may be under-represented. The surge in microbiota research stems from the development of techniques that do not rely on culture. The first DNA sequencing techniques were developed in the 1970s (Sanger and Coulson 1975) and were the gold-standard for DNA sequencing for nearly 30 years. These methods are laborious and expensive and have recently been supplanted by next-generation sequencing, which allows for the faster, less expensive sequencing of the microbiota (Suchodolski et al., 2009). PCR-based methods can detect bacterial species based on nucleic acids extracted directly from a sample. Using specific primers, the region of interest can then be amplified (Rogers et al., 2009). By using genes that are highly conserved among bacterial species, the amplified regions can be compared to databases and used to identify the microbes present in a sample. Next generation sequencing largely relies on the use of the 16S rRNA gene. Community fingerprinting, such as terminal restriction fragment length polymorphism (T-RFLP), denatured gradient gel electrophoresis (DGGE) and automated ribosomal intergenic spacer analysis (ARISA), uses PCR and relies on the heterogeneity of the length of the 16S gene, assuming that each variation in length represents a different group of bacteria (Fujimoto et al., 2003; Tiquia, 2010; Fisher and Triplett, 1999). This marker is both highly conserved among bacteria and

6 adequately divergent to allow for differentiation of different microorganisms through sequence differences that have resulted from random changes and thus represent evolution (Janda and Abbot, 2007). The 16S rRNA gene is composed of hypervariable regions interspersed with strongly conserved regions (Hartmann et al., 2010). It is the hypervariable regions that are the source of species-specific divergence, and, following PCR primer binding to the conserved regions, are amplified, sequenced, analyzed using computer algorithms, and compared to reference databases (Hartmann et al., 2010). Specific analysis systems allow not only for quick identification of microbes in a sample, but also will determine the relative abundance of organisms and comparisons between samples (Quail et al., 2012).

1.3.2 Role of the Bacterial Gastrointestinal Microbiota in Health and Disease Development of the microbiota begins with a complex microbe-mucosa signaling pathway involving microbial-associated molecular patterns (MAMPs) and host pattern recognition receptors (PRRs), ultimately leading to tolerance by the host’s immune system to certain bacteria (Sharma et al., 2010). These core, commensal bacteria provide competition for pathogenic bacteria, preventing the persistence of these pathogens (Alverdy et al., 2008). In addition to direct prevention of proliferation of pathogenic bacteria, commensal bacteria have long been known to have a significant role in digestion, contribution of nutrients, and promotion of intestinal enzyme activity (Savage, 1986). More recently, bacteria have been shown to contribute in a variety of ways to normal host development. There have been numerous studies on the role of indigenous intestinal bacteria to the normal development of the host immune system. Germ-free mice have fewer, smaller Peyer’s patches and fewer macrophages and lymphocytes within the intestinal lamina propria (Thompson and Trexler, 1971). Atarashi et al, (2011) and Umesaki et al., (1993) demonstrated that colonization of germ-free mice with flora from specific pathogen free (SPF) mice significantly increased the number of T lymphocytes in the lamina propria of the colon and the small intestines to levels comparable with conventional mice. Beyond the gastrointestinal tract, lymph nodes of germ-free mice have smaller and less cellular medullary cords comprised of small lymphocytes compared to conventional mice, in which medullary cords are larger and have greater cellular diversity (Bauer et al., 1963).

7 Stappenbeck et al (2002), conducted a series of experiments to characterize the contributions of the gastrointestinal microbiota to angiogenesis in the developing intestinal tract. The study compared germ-free (GF)/Paneth cell negative transgenic mice to GF/Paneth positive mice, Bacteroides thetaiotaomicron positive/Paneth negative mice and Bacteroides thetaiotaomicron positive/Paneth positive mice. In the absence of Paneth cells, regardless of the presence of B. thetaiotaomicron, the bacteria alone were unable to exert a significant effect on villus vasculature density. However, the presence of both Paneth cells and B. thetaiotaomicron, showed an additive effect with a greater than two-fold increase in vasculature density when compared to GF/Paneth cell positive mice. This indicates not only the importance of the intestinal microbiota on vascular development, but also the presence of the appropriate host cells on which they exert their influence, and speaks to the coevolution of the gastrointestinal microbiota with the host.

1.3.2.1 Establishment of the Microbiota The gut is sterile at birth in most species (Breitbart et al., 2008; Combes et al., 2013; Songjinda et al., 2005). In humans and other animals, the gut is rapidly colonized and then undergoes large fluctuations in the population over the first year of life (Breitbart et al., 2008; Songjinda et al., 2005; Palmer et al., 2007). For example, members of certain phyla, such as Bacteroidetes, become relatively stable once present in the feces of humans, while others, such as the genera Prevotella, Acinetobacter, Desulfovibrio, Veillonella and Clostridium are more transient, even disappearing and reappearing repeatedly during the first year of life. Leading up to weaning, the fecal microbiota of human individuals, regardless of method of delivery or parental microbiota composition, converges to become more similar to that of the adult microbiota (Palmer et al., 2007). These observations lead to the conclusion that during infancy, the gastrointestinal microbiota is colonized more opportunistically by bacteria that the infant encounters, such as from the mother’s vagina or feces, and the environment, until which time the bacteria that have co-evolved with the human gut can effectively establish their symbiotic relationship. The dominant phyla of intestinal bacteria are relatively consistent across species of mammals, with Bacteroidetes and Firmicutes representing the majority of sequences (Swanson et al., 2011; Pflughoeft and Versalovic, 2012; Pang et al., 2012; Muegge et al., 2011). However,

8 when examined to a species and strain level, there are significant differences and minimal overlap between animal species (Suchodolski, 2011), with the greatest variation attributed to inter-individual differences rather than temporal variability within an individual (Palmer et al., 2007). Additionally, the functional maturation of the human gut during the first three years of life is remarkably similar, regardless of distant geographic location (Yatsunenko et al., 2012). Despite this, when viewed across time, an individual’s microbiota has pronounced variability (Caporaso et al., 2010). Together, these suggest that once the gastrointestinal microbiota is established within an individual, its inhabitants may change, but its core functionality is likely persistent, a finding supported by the identification of a core microbiome in twins (Turnbaugh et al., 2009).

1.3.2.2 Composition of the Gastrointestinal Microbiota in Non-Mink Animal Species The microbial mass of the stomach of cats, another carnivore, is low. The stomach averages 103 bacteria/mL (NRCNA, 2006), and is composed primarily of Helicobacter species, regardless of technique used, similar to humans (Neiger et al., 1998; Bik et al., 2005). Additionally, the carnivore stomach is typically much more acidic, averaging a pH of 2.5 (NRCNA, 2006), compared to a pH of 6-7 in herbivores (Van Soest, 1994). This adaptation allows the consumption of rotting meat by carnivores without acquisition of disease, as stomach acidity is one of the first lines of defense of the host against pathogens. It also impacts the overall bacterial composition of the proximal gut. Because of the invasive methods needed to obtain samples, comparatively few studies focus on the small intestinal bacterial microbiota. In dogs, the small intestine has been shown to have a microbial population that is subject to large fluctuations in both number and prevalence of microbes present (Mentula et al., 2005). Depending on the section sampled, the microbial population of both cats and dogs is markedly different between sections of small intestine (Suchodolski et al., 2008; Suchodolski, et, al, 2009; Ritchie et al., 2008). This is in contrast to the microbial population of the colon and feces, which are much more stable over time (Mentula et al., 2005; Suchodolski et al., 2008). Thus, the fecal microbiota is not representative of the upper gastrointestinal tract. Moreover, multiple studies in humans have shown that the composition of the fecal microbiota is significantly different than that of even the lower gastrointestinal tract, such as the rectum and cecum (Momozawa et al., 2011; Marteau et al.,

9 2001). This is in contrast to studies in both pigs and horses, in which the bacterial microbiota of the feces were found to accurately reflect the bacterial microbiota of the distal gastrointestinal tract (Zhang, et al., 2013; Schoster, et al., 2013). So, though studying the fecal microbiota as an indicator of what is occurring in the upper, or even lower, gastrointestinal tract may be of limited value, its study can shed light on the impact on the microbial population over time and in response to alterations in environment, diet, disease state, or use of medication. The number of studies on the fecal microbiota has rapidly increased in the past several years. Studies in other domesticated species have largely examined companion and production animals. In dogs, the relative abundance of dominant phyla vary between studies, but generally include Firmicutes (14-95%), Fusobacteria (<1-7%), Bacteroidetes (<1-70%), and Proteobacteria (<1-15%) (Handl et al., 2011; Middlebos et al., 2010; Swanson et al., 2011; Garcia-Mazcorro et al., 2011). The differences between studies may be related to differences in sequencing technique or primers used (e.g., 454 pyrosequencing vs Illumina), bias in sampling, reporting method (e.g., mean vs median) or to the innate differences in composition between unrelated populations. For these reasons, direct comparisons between studies require critical analysis of the methods and techniques utilized. In healthy cats, a more closely related carnivore to mink than dogs, there are also similar dominant phyla, but with different overall proportions: Firmicutes (13-92%), Actinobacteria (4- 7%), Bacteroidetes (<2-68%), and Proteobacteria (<1-14%) (Tun et al., 2012; Garcia-Mazcorro et al., 2011; Ritchie et al., 2008). As mentioned previously, the differences between studies likely represents differences in methodology, reporting or inherent differences in populations. Similarly, studies in other carnivores, including the polar bear, Tasmanian devil, leopard cat, Eurasian otter, and raccoon dog show a predominance of Firmicutes (Cheng et al., 2015; Glad et al., 2010; An et al., 2016). Only one study exists on a small sample of ferrets, a mustelid related to mink, which used bacterial culture to characterize the fecal microbiota (Nizza et al., 2014). This study reported data in frequency of isolation from samples as a percentage, with Clostridium acetobutylicum (a Firmicute), the only organism yielded in >50% of samples. Other species included other members of phylum Firmicutes, as well as members of the phyla Actinobacteria and Proteobacteria (Nizza et al., 2014). These results fit with the high percentage from the phylum

10 Firmicutes in other carnivore species discussed previously, regardless of the use of culture or culture-independent techniques.

1.3.2.3 Host Differences and the Influence on the Bacterial Microbiota There have been a number of studies in numerous species exploring the ways in which host differences influence the composition of the microbiota. For example, the relative abundance of bacteria have been found to be different in obese vs lean mice, and to be heavily influenced by mouse host genotype (Ericsson et al., 2015; Zhao et al., 2013), as well as geographic differences in diet in humans (Pflughoeft and Versalovic, 2012). Furthermore, while Ericsson et al. (2015) did not find significant differences in the microbiota between male and female mice, Zhao et al (2013), and Markle et al (2013), did demonstrate sex as a factor in the microbiota composition in chicken and humans, respectively. So, while the overall abundance of phyla is relatively stable within an individual, the slight differences in bacterial species composition attributed to host genotype, phenotype, and environment, may play a significant role in health and disease. The confounding role this may play in research has yet to fully be determined.

1.3.2.4 Relationship between the Bacterial Microbiota and the Host Immune System There is ample evidence of the intimate relationship between the host immune system and the intestinal microbiota. Germ-free mice have smaller, less cellular spleens and mesenteric lymph nodes, and GF mice and rabbits show impaired development of the gut-associated lymphoid tissue (GALT), a primary mode of defense against pathogens in the gut (Pollard and Sharon, 1970; Rhee et al., 2004). Additionally, removal of all bacteria of the gut of mice through broad-spectrum antibiotic administration eliminated MyD88-dependent production of cytokines, even in the presence of severe intestinal injury, resulting in severe morbidity and mortality compared to those mice in which only certain bacteria were eliminated with a narrow-spectrum antibiotic (Rakoff-Nahoum et al., 2004). In the presence of pathogenic bacteria (vs commensal, non-pathogenic bacteria), GF mice had higher organ bacterial burdens and decreased movement of T lymphocytes into the peritoneal cavity (Inagaki et al., 1996), supporting the role of gut

11 commensal bacteria not only in development, but in homeostasis and activation of the immune system in disease states as well. Several studies have explored the possible connection between mode of birth and risk of development of autoimmune disease later in life. Negele et al, (2004) reported an increased risk in humans of allergic sensitization in the first few years of life when born via cesarean section. Debley et al (2005) found an increased risk of asthma later in life in babies born premature via cesarean section, with no difference between full term vaginally-delivered and cesarean- delivered babies. Additionally, although Laubereau et al., (2004) did not find an increased risk for atopic dermatitis in the first 12 months of life to babies born via cesarean section, they did find an increased risk of nutritional allergies. Nutritional allergies, have, in turn, been shown to have a positive association with later development of asthma and more severe atopic dermatitis, when present. Clearly, there is more at play than simply mode of delivery and a factor in this may be which flora a baby is initially exposed to (i.e., vaginal or skin). Supportive of this, Dominguez-Bello et al., (2010) found intestinal microbiota differences between babies born vaginally compared to those born via cesarean section. Of note is that none of these studies explicitly factored in increased likelihood of antibiotic use following cesarean section vs vaginal births (Tanaka et al., 2009), the use of which can heavily impact the composition of the gastrointestinal microbiota. Causative relationships between the microbiota and the host are more difficult to establish due to the myriad factors leading to any given phenotype or disease state. Mice deficient in toll-like receptor 5, important in microbial recognition, develop signs of an obese metabolic syndrome (Vijay-Kumar et al., 2010). Germ-free mice with the same receptor deficiency do not develop metabolic syndrome; however, after oral inoculation with cecal contents from their non-germ-free counterparts, a similar metabolic syndrome developed. This suggests that the genetic condition is necessary, but not sufficient, to cause disease and that the gut microbiota may more heavily influence the disease phenotype than does the host genotype (Pflughoeft and Versalovic, 2012; Vijay-Kumar et al., 2010). Germ-free mice are nearly completely lacking in Th17 cell populations, important in defense against bacterial and fungal infections. Addition of a specific member of the normal microbiota, segmented filamentous bacteria (SFB), to the gastrointestinal tract was sufficient to induce Th17 cells in the intestinal lamina propria of germ-free mice (Gaboriau-Routhiau et al.,

12 2009; Fritz et al., 2013). In a separate study, SFB were shown to be vital to induction of intestinal Th17 cells. Mice lacking these bacteria have increased susceptibility to Citrobacter rodentium, infection, potentially due to decreased expression of genes important to intestinal inflammation and antimicrobial defense (Ivanov et al., 2009). Additionally, a microbial product, polysaccharide A (PSA), has been shown to be a key element in lymphoid organogenesis and prevention of intestinal inflammation (Mazmanian et al., 2005; Mazmanian et al., 2008).

1.3.2.5 Alterations to the Bacterial Microbiota in Disease Beyond their role as symbionts and contributors to host health, a correlative relationship exists between alterations in gastrointestinal microbiota, particularly changes in the relative abundances of dominant phyla, and several disease processes. Swabs collected from the lower esophagus of healthy adult humans yielded predominantly Streptococcus species, while those with evidence of esophagitis or Barrett’s esophagus had greater bacterial diversity and included more anaerobic species (Pflughoeft and Versalovic, 2012; Yang et al., 2009). Chronic diarrhea in humans due to Clostridium difficile has been associated with antibiotic use, leading to an overall decrease in diversity of the intestinal microbiota, shifting, in some cases, to as much as 100% of sequences in the phylum Firmicutes (Chang et al., 2008). This finding has led to the use of fecal microbial transplantation in humans as a means to overcome this treatment-resistant condition (Duplessis et al., 2012). Whether it is the presence of certain bacteria that leads to the pathological condition, or the pathological condition that creates an environment conducive to the altered bacterial state, is still being elucidated, and will likely prove to be different for each disease. In addition to changes in the overall diversity of phyla and species present leading to pathologic conditions, the opposite has also been shown to occur, with disease states resulting in an altered microbiota. For example, in humans, adults with type II diabetes have a significant reduction in the relative abundance of a class of Firmicutes (Clostridia) possibly owing to a differential ability to metabolize and utilize carbohydrates in response to endocrine dysfunction of the host (Pflughoeft and Versalovic, 2012). The applications of microbiota research to treat pathological conditions are numerous. D’Argenio et al., (2013), showed that nutritional therapy could be used to normalize the intestinal microbiota in patients affected with Crohn’s disease, decreasing inflammation and

13 clinical signs. Fecal transplantation and bacteriotherapy (treatment of a disease with bacteria) have been well documented as a treatment for refractory Clostridium difficile colitis to restore the normal, non-pathogenic microbiota (Duplessis et al., 2012; Kelly et al., 2012; Brandt and Reddy, 2011). Additionally, microbiota research can be used in the diagnosis of disease (Liu et al., 2012) and has made substantial contributions to our understanding of numerous aspects of development (Claus et al., 2008; Atarashi et al., 2011; Furusawa et al., 2013). While most studies in humans are necessarily correlative (de Vos and de Vos, 2012), the use of animal models provides a means to explore cause-and-effect relationships, especially through the use of germ-free or germ-specific (gnotobiotic) models. As such, a deeper understanding of the microbiota of species used in research is necessary.

1.3.2.6 Potential Impact of the Microbiota in Production Animal Species While the host genome is not easily manipulated, the microbiota has been shown to change in response to changes in diet, use of medication, and in response to disease or introduction of pathogens (Ballou et al., 2016; Danzeisen et al., 2015; Mon et al., 2015). Studying the impact of these changes on the microbial community and establishing what constitutes a ‘healthy’ microbial population, has shown a significant potential for the use of probiotics in order to optimize production in a variety of species. In chickens, the use of a Lactobacillus probiotic was shown to be protective against Brachyspira pilosicoli, with Lactobacillus-treated birds having higher body weights, lower fecal moisture content, and improved egg weights (Mappley et al., 2013). In chicks, administration of an oral Salmonella vaccine at 1-day post-hatching had as much impact on the composition of the gastrointestinal microbiota at 28 days as did continuously fed probiotics, suggesting that manipulation of the microbiota may best be achieved in the immediate neonatal period (Ballou et al., 2016). Similarly, probiotic-fed veal calves showed an enhanced growth rate, average daily weight gain and feed efficiency and diminished mortality and coliform counts compared to controls, particularly in calves which were considered to be unhealthy (Timmerman et al., 2006). There have been several studies on the use of probiotics in pigs, a monogastric more closely related to mink than either chickens or ruminants. Enterococcus faecium-supplemented sows showed improved feed intake, with larger litters that had higher weights (Böhmer, 2006). Similarly, Di Giancamillo et al. (2008), showed that Pediococcus acidilactici-supplemented

14 piglets had a higher post-weaning average daily weight gain than untreated controls and, microscopically, increased villus height and crypt depth, which can be protective against enteric pathogens. In looking at the development of the immune system of Bacillus-supplemented sows and piglets, treated piglets had enhanced populations of intraepithelial and lamina proprial lymphocytes and less frequent occurrence of pathogenic Escherichia coli serogroups within the feces (Scharek et al., 2007). To conclude, these studies show that there likely is not one ‘best’ type of bacteria or microbial community composition, but that supplementation with a variety of ‘positive’ commensal bacteria may confer significant beneficial effects to overall health, immunity and production of individual animals. Following the study of the microbiota of howler monkeys, Amato et al (2013) concluded that, although animals had distinct gastrointestinal microbial communities, healthy animals tended to be much more similar than did unhealthy animals, which exhibited much less similarity in the composition of their microbiota. If the same is true in mink, this study could provide a benchmark for what a ‘healthy’ fecal microbiota looks like, which could, in turn, be used to optimize production through, for example, the formulation of mink- specific probiotics.

1.4 THESIS OBJECTIVES AND HYPOTHESES 1.4.1 Hypotheses 1) The dominant phyla in the feces of commercially farmed mink (both at weaning and in adult females) will be Firmicutes, followed by Proteobacteria and Bacteroidetes.

2) Weaned kits will show a higher proportion of lactic acid bacteria and pathogenic bacteria in their feces compared to adult females.

3) The summers of 2014 and 2015 will be relatively similar where farm practices are consistent from year-to-year (i.e., the same diet and vaccine use).

4) The restricted diet provided to adult females in winter will alter the relative abundances of the predominant phyla, classes, orders, families and genera, but will not have an impact on the overall predominant groups.

15 1.4.2 Study Objectives The objectives of the current study are: 1) To sequence and characterize the fecal bacterial microbiota from mink fur farms across Ontario.

2) To compare and contrast the mink fecal microbiota between: i. Adult females and weaned kits for determination of how the life stage of a carnivore influences the composition of the fecal microbiota. ii. Summers of 2014 and 2015 to study how the Clostridium botulinum toxoid vaccine shortage, as occurred in North America in 2014, influenced the overall composition of the fecal microbiota, and determine the stability of the fecal microbiota in a carnivore over time. iii. Adult females in summer and winter, to assess the effect of summer fattening vs winter caloric restriction on the composition of the fecal microbiota.

Table 1.1 Metabolizable energy requirements for growing kits (adapted from the NRC,1982)

17 CHAPTER TWO: THE FECAL BACTERIAL MICORBIOTA OF COMMERCIAL MINK (NEOVISON VISON): YEARLY, LIFE STAGE, AND SEASONAL COMPARISONS

The development of culture-independent sequencing technology has greatly accelerated knowledge about the role that gastrointestinal microbiota plays in host health, development, and disease. Our objective was to characterize the fecal microbiota of farmed mink (Neovison vison) to contribute to increased understanding of the carnivore microbiota and to gain insight into the various management factors that may affect its composition. Pooled fecal samples were collected from adult females and weaned kits in the summers of 2014 and 2015, and from females in the winter of 2016, a time when females undergo marked calorie restriction. Bacterial DNA was extracted and the V4 region of the 16S rRNA gene was amplified. Approximately 22 million sequences were identified following quality control filtering. The phyla Firmicutes and Proteobacteria account for >95% of the sequences identified. Comparisons were made by life stage, season and year; the greatest number of differences in the relative abundance at each taxonomic level were noted between samples collected in 2014 versus 2016, and the fewest between adult females and weaned kits. Significantly more operational taxonomic units (OTUs) were found in 2014 than 2015 or 2016 (p<0.05), 2014 was richer than 2016 (p=0.013), and was more even but less diverse than 2015 (p<0.01). There were significant differences in community membership and structure by year and season (all p-values <0.001). The predominant phyla and genera at the farm level were similar from year to year. Together, these indicate that environment and time are important factors in the stability of gastrointestinal microbiota, once the host reaches maturity. This study will contribute significantly to our understanding of the microbial dynamics of carnivores and will provide a baseline for future work aimed at optimizing the health, welfare, and production of farmed mink.

2.1 INTRODUCTION The intestinal microbiome is the collective, interacting genomes and symbiotic microorganisms in the gastrointestinal (GI) tract (Kinross et al., 2011). In recent years, the integral role of the GI microbiota in disease, health, and development has been well established in a variety of species, including humans, mice, dogs, and pigs, (Velasquez-Manoff, 2015; Ericsson et al., 2015; Alverdy and Chang, 2008; Thompson and Trexler, 1971; Atarashi et al., 2011; Umesaki et al., 1993; Stappenbeck et al., 2002). To date, most studies have focused on

18 herbivores and omnivores and few studies have been conducted in carnivores. Moreover, with the exception of a few studies in domestic cats, prior studies conducted in carnivores have been based on a limited number of samples (i.e. ten or fewer) (Cheng et al., 2015; Glad et al., 2010; An et al., 2016). Mink (Neovison vison) are an important commercial animal commodity group in Canada, and have a short and simple GI tract, resulting in a relatively fast total food transit time (Bleavis and Aulerich, 1981; Sibbald et al., 1962). The significance of this from a nutritional standpoint relates to the amount of time bacteria have to assimilate nutrients, as a faster transit time reduces bacterial production of short chain fatty acids and breakdown of amino acids. In humans, a faster GI transit time leads to a significantly lower bacterial mass compared to baseline, but whether this has an impact on the overall composition of the GI or fecal microbiota is unclear (Hendriks et al., 2012). The ferret, a common companion and research animal, is a mustelid closely related to mink. In a small sample of ferrets, using bacterial culture to characterize the fecal microbiota, only one organism, Clostridium acetobutylicum (phylum Firmicutes), was yielded in greater than 50% of samples cultured. Other species identified included other members of the phylum Firmicutes, as well as members of Actinobacteria and Proteobacteria, findings similar to other carnivores, regardless of the use of bacterial culture or culture-independent techniques (Nizza et al., 2014). Also using bacterial culture on fecal swabs from farmed mink in Denmark, but with the aim of identifying specific bacteria, Vulfson et al (2001) isolated Escherichia coli, enterococci, and lactic acid bacteria from 50%, 90%, and 90% of samples, respectively. However, culture-dependent techniques are unable to provide a detailed or even necessarily accurate reflection of the complex fecal microbiota. This is reflected in the discrepancy identified by Bahl et al (2017) between the organisms identified from the intestinal mucosa of mink by bacterial culture and those identified by 16S rRNA gene sequencing. In this study, Clostridia were underrepresented by culture, but made up a large proportion of the sequences identified by next generation sequencing. This provides evidence that bacteria that are difficult to culture by classical techniques have likely been underrepresented in culture-based studies and, thus, their importance to the GI microbiota has potentially been undermined. It is unclear if the GI microbiota of carnivores is as extensive, functional, or influential on the nutritional status and overall productivity of the host as in non-carnivores. The large

19 intestinal microbiota of the cat is active, allowing utilization of nutrients, and has been shown to be affected by diet (Barry et al., 2011). However, this is unlike what has been seen in gnotobiotic ferrets, which had minimal morphological or metabolic changes with a selected, rather than conventional, GI microbiota (Bell and Manning, 1990). Moreover, mink have substantially fewer colonic bacteria (2-4 orders of magnitude) than most other mammals, which is likely attributable to their fast GI transit time (Williams et al., 1998). As such, there is a need to more fully understand the role of the GI microbiota in the carnivore. The knowledge of the GI microbiota of other farmed species has been used to optimize various production parameters. Chicks have been shown to be resistant to Brachyspira pathology when orally inoculated by Lactobacillus reuteri (Mappley et al., 2013). Both multispecies and chick-specific probiotics increased the overall productivity of broiler chickens, with the chicken- specific probiotic having a more enhanced effect (Timmerman et al., 2006). Similarly, Pediococcus acidilactici-supplemented piglets showed higher body weights, post-weaning daily weight gain, and number of proliferating enterocytes compared to untreated controls (Di Giancamillo et al., 2008). Investigations into how the microbiota of farmed mink could be optimized to improve health and production parameters have yet to be carried out and by understanding what is present in healthy mink, a baseline can be created for future interventional studies. The objectives of this study were to characterize the fecal microbiota of commercial mink, and with the resulting data, to compare the microbiota of adult females and weaned kits in summer, adult females between summer and winter, between three consecutive years, and to determine if the fecal microbiota is similar at the farm-level from year-to-year. Based on results from other mammalian species, we hypothesized that the predominant phyla would be Firmicutes and Proteobacteria, that adult females and weaned kits would have similar fecal microbiota, particularly within a farm, and that significant differences would be between summer and winter results, because of differences in diet and nutritional levels of overwintered females kept for breeding. Additionally, we hypothesized that the fecal microbiota would be relatively stable from year-to-year at the farm level.

20 2.2 MATERIALS AND METHODS

2.2.1 Farm recruitment and sample collection All known mink producers in Ontario were contacted through the Ontario Fur Breeders Association (OFBA) and the Ontario Ministry of Agriculture, Food, and Rural Affairs (OMAFRA) for potential enrollment. A total of 43 farms were initially enrolled, representing >93% of Ontario commercial mink farms in 2014. Over the three-year study period, several farms either pelted out, only participated in subsequent years or were new farms in 2015, such that a total of 49 different farms participated: 43 in 2014, 45 in 2015 and 42 in 2016. Participation was voluntary and producers were advised that they could withdraw from the study at any time. Fecal sample collection occurred between July and October of 2014 and 2015 and from January to February in 2016. Samples collected from underneath three cages of females or three cages of weaned kits were pooled and two pooled samples from females and two pooled samples from weaned kits were collected from each farm. Farmers were given detailed instructions on sample collection, including quantities, when, and from where within the fecal pile samples were to be collected and samples were primarily collected by a government/industry technical representative (BT). After collection and before freezing, pooled samples were mixed, such that the feces from underneath the three cages were homogenized into one sample. For winter 2016 sample collection, only one pooled female sample was collected from each farm. In 2014, only two pooled samples (1 female, 1 weaned kit) were collected from 3 farms. In 2015, only one pooled female sample was collected from two farms. The total number of samples collected each year is summarized in Table 2.1. Samples were collected and delivered to the University of Guelph, where they were stored in a -80°C freezer until processing.

2.2.2 Sample processing Samples were thawed for several hours at room temperature inside a biosafety cabinet and 0.2g of feces was aliquoted into 1 ml Eppendorf tubes and refrozen at -80°C pending DNA extraction.

21 2.2.3 DNA extraction, amplification, and sequencing of bacterial 16s rRNA gene DNA was extracted using the E.Z.N.A. Stool DNA Kit (Omega Bio-Tek, Inc, Doraville, Georgia, USA) according to the manufacturer’s instructions. Following extraction, spectrophotometry (NanoDrop, Roche, Mississauga, ON, Canada) was used to assess the quantity of nucleic acids present. The V4 region of the 16S rRNA gene was amplified using the PCR protocol described by Caporaso et al (2010) and the following primers: forward S-D-Bact-0564-a-S-15 (5′- AYTGGGYDTAAAGNG-3′) and reverse S-D-Bact-0785-b-A-18 (5′- TACNVGGGTATCTAATCC-3′). Both primers were designed with regions that overlap with the Illumina sequencing primers (Forward: TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG, Reverse: GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG), to allow for annealing to the Illumina universal index sequencing adaptors, plus the 8 bp identifier indices (Forward: AATGATACGG CGACCACCGAGATCTACAC-index-TCGTCGGCAGCGTC, Reverse: CAAGCAGAAGACGGCATACGAGAT- index-GTCTCGTGGGCTCGG). For each sample, a reaction mixture was prepared using 12.5µl KAPA Ready Mix, 9.0µl sterile water, 0.5µl each of the forward and reverse primers (10pM/µl), and 2.5µl of the extracted DNA (5ng/µl) and the PCR parameters were as follows: 1) 3 min at 94°C for denaturation, 2) 45 sec at 94°C for denaturation, 3) 60 sec at 53°C for denaturation, 4) 1.5 min at 72°C for elongation, and 5) 10 min at 72°C. Steps 2-4 were repeated for a total of 27 cycles. To ensure that bands of appropriate length (~254 bp) were present following the first PCR, electrophoresis of the PCR product was completed in a 2% agarose gel. PCR products were stored at 4°C until purification. Following purification of PCR products with Agencourt AMPure XP (Beckman Coulter Inc., Mississauga, ON, Canada), a second PCR was completed using 2.5µl of the purified product, 12.5µl KAPA Ready Mix, 9.0µl sterile water, and 1.0µl of the Illumina Forward and Reverse Index Primers (I501-I508 or S513, S15-S18, S20-S22 and I701-712 or N716, N718- N729, respectively). The PCR parameters were as follows: 1) 3 min at 94°C, 2) 45 sec at 94°C, 3) 60 sec at 50°C, 4) 1.5 min at 72°C, and 5) 10 min at 72°C. Steps 2-4 were repeated for a total of 8 cycles. The second PCR product was purified as above, with 40µl of AMPure XP and 35µl 10mM Tris pH 8.5 Buffer. The final purified products were evaluated for bands of appropriate length using 2% gel electrophoresis. When appropriate bands were not observed, the nucleic acid

22 was quantified using spectrophotometry and adjustments were made to the volumes of KAPA Read Mix, DNA product, or sterile water to correct for this. Samples were normalized to a final concentration of 12 nM. The library pool was submitted to the University of Guelph’s Advanced Analysis Centre and sequenced with an Illumina MiSeq (Illumina RTA v1.17.28, San Diego, California, USA) for 250 cycles from each end.

2.2.4 Statistical Analyses Microbiota analyses were completed using mothur software (v 1.38; https://www.mothur.org; Kozich et al., 2013). Paired end reads were aligned and sequences that contained ambiguous bases, were longer than 275 bp, or that contained homopolymer runs > 8 bp were removed. Paired-end reads were aligned to the SILVA 16S rRNA reference database (Quast et al., 2013) and any that were misaligned with the target region were removed. Chimeras and non-bacterial sequences (i.e., chloroplasts, mitochondria, Archaea and Eukaryotes) were removed. Sequences were identified using the RDP classifier (Wang et al., 2007) and binned into phylotypes using a closed OTU-picking approach. The relative abundance of the predominant phyla, classes, orders, families, and genera were calculated and stacked column graphs were created. A Wilcoxon rank-sum test was used to compare the relative abundance of different taxa between seasons (summer vs winter, adult females only, n=198) and life stages (adult females vs weaned kits, 2014 and 2015 only, n=332) within the JMP 12 Response Screening Platform (SAS Institute Inc., Cary, NC). A Wilcoxon rank-sum test for multiple comparisons was used to between years (n=366) and by farm between years. Only those farms for which all 8 samples were available (i.e., 2 adult females and 2 adult kits, for both 2014 and 2015) were used for farm-level comparisons (n=8 for each farm x 30 farms). Multiple comparison adjustments were conducted using the Benjamini & Hochberg’s False Discovery Rate (FDR), with an adjusted p< 0.05 accepted as statistically significant (R Foundation for Statistical Computing, Vienna). Subsampling was performed to normalize sequence numbers and sampling coverage was assessed by Good’s Coverage. Measures of alpha diversity, inverse Simpson’s, Chao1, and Shannon’s evenness indices were used to calculate diversity, richness, and evenness, respectively. Beta diversity was measured using the Jaccard and Yue and Clayton indices,

23 measures of community membership, and membership and structure, respectively, and used to created dendrograms using FigTree (v.1.4.3; http://tree.bio.ed.ac.uk). Unweighted UniFrac, analysis of molecular variance (AMOVA) and parsimony tests were used to compare seasons, life stage, and years on both the Jaccard and Yue and Clayton indices. Visual similarities and clustering of each of these groups was plotted with principal coordinate analyses (PCoA). Within mothur, linear discriminant analysis effect size (LEfSe) (Segata et al., 2011) was performed to identify bacterial taxa that were differentially abundant between groups.

2.3 RESULTS A total of 366 mink fecal samples were included in the final analyses; 198 from adult females and 168 from weaned kits (Table 2.2). Twenty-one samples were not included due to poor DNA yield after extraction or a repeated inability to obtain adequate sequence numbers. Following quality control filtering, a total of 22,238,782 sequences were yielded, with a range of 9,692 to 205,174 (median 57,953) sequences per sample. A subsample of 11,000 reads/sample was used to normalize sequence numbers across samples and was considered adequate, as evidenced by greater than 99% coverage for all samples and plateau of rarefaction curves (raw data accessible at: Compo, et al., 2017, doi:10.5683/SO/OEIP77). A total of 31 bacterial phyla were identified; however, only 4 comprised >0.5% of the total sequences identified with Firmicutes and Proteobacteria together comprising 95% of the total sequences. Greater than 1300 genera were identified; the 20 overall most predominant genera are depicted in Figure 2.1 by year and life stage and by farm in Appendix B. The predominant classifications are summarized in Table 2.3 at each taxonomic level.

2.3.1 Relative abundance At each taxonomic level, differences were identified in the relative abundance of fecal bacterial taxa by year, life stage, and season. All comparisons between years had more total taxa that were significantly different than seen in comparisons of life stage and season, with the greatest number of significantly different taxa identified between 2014 and 2016, and the fewest taxonomic differences between adult females and weaned kits. These differences are summarized in Tables 2.4-2.6.

24 A total of 30 farms had complete data for 2014 and 2015 and each of these farms was individually compared between 2014 and 2015. The 9 and 15 predominant phyla and genera, respectively, were compared by farm. Of the 30 farms, 11 (37%) and 6 (20%) had zero or only one phylum, respectively, which were significantly different between 2014 and 2015. When evaluating only those phyla for which the overall relative abundance was >1.0% (Firmicutes, Proteobacteria, and Bacteroidetes), 17 of 30 farms (57%) had no significant differences and 11 (37%) had only one phylum that was significantly different between 2014 and 2015. When only the two predominant phyla are considered (i.e., Firmicutes and Proteobacteria), together accounting for >95% of the total sequences, only 5 farms had one or both phyla that were significantly different between 2014 and 2015. Overall, 83% (223/270) of the comparisons made were similar between years, by farm, and by phylum. Similarly, 80% of the genera had similar relative abundance by farm from 2014 to 2015.

2.3.2 LEfSe analyses When comparing years, 109 OTUs were identified as differentially abundant (p<0.05). Those with an LDA score >4 are listed in Table 2.7. Comparing life stage (adult females vs weaned kits) and season (summer vs winter), 40 and 88 OTUs were differentially abundant, respectively (p<0.05); those with LDA scores >3 are depicted in Figures 2.2 and 2.3. All except one of the OTUs enriched in weaned kits were Firmicutes, of which only the families Lactobacillales and Clostridiales were represented.

2.3.3 Alpha and beta diversity There were no differences in alpha diversity measurements when comparing season (adult females only) and life stage (adult females vs weaned kits). Samples from 2014 had significantly more OTUs than in 2015 or 2016 (p=0.006 and 0.013, respectively), were more rich than 2016 (p=0.013), and the 2014 microbiota results were more even, but less diverse than in 2015 (p=0.002 and 0.001, respectively) (Fig 2.4). The composition of the fecal microbiota of commercial mink was significantly different in community membership, based on the classic Jaccard index, and structure, based on the Yue and Clayton index of dissimilarity, by year and season (all p-values <0.001). The community membership and structure of the fecal microbiota from adult females and weaned kits were

25 significantly different by AMOVA (p=0.008 and <0.001, respectively), but not with parsimony or UNIFRAC unweighted analyses (all p>0.1). When comparisons were made by life stage and year (e.g., females from 2014, kits from 2015), fecal microbiota from females and kits from 2015 samples were not significantly different in membership on both the AMOVA and parsimony tests and in structure on the parsimony test, but not on the AMOVA test. Similarly, females and kits from 2014 were similar in structure on both the AMOVA and parsimony tests and in membership on the parsimony test, but not the AMOVA test (Table 2.8). These differences in membership and structure by year, life stage and season are depicted visually by principal coordinate analyses (Fig 2.5) and dendrograms (Figure S2 in Supplementary Material, Fig 2.2).

2.4 DISCUSSION The predominant phylum of the feces of commercial mink overall, regardless of year, life stage, or season, was Firmicutes. Firmicutes is followed by a much larger proportion of Proteobacteria than has been previously reported in healthy dogs, cats, and other carnivores (34% vs <15%) (Handl et al., 2011; Middlebos et al., 2010; Swanson et al., 2011; Garcia- Mazcorro et al., 2011; Tun et al., 2012; Ritchie et al., 2008; Cheng et al., 2015; Glad et al., 2010; An et al., 2016). This is consistent with the high relative abundance of both Firmicutes and Proteobacteria found by Bahl et al., (2017), but unlike Zhao et al., (2017), who found a higher overall relative abundance of Proteobacteria than Firmicutes. As has been seen in other mammals when the microbiota is compared between adults and weanlings (Bian et al., 2016; Jacquay, 2017; Odamaki, et al., 2016; Schloss, et al., 2012), there were few significant differences identified between adult female mink and weaned kits. At the order level, organisms from Lactobacillales comprise more than one-third of the total bacterial sequences identified from the feces of mink, more than twice the next most abundant order, Xanthomonadales. Lactobacillales, to which the genus Lactobacillus belongs, consists primarily of the lactic acid bacteria (LAB), a group of aerotolerant anaerobic bacteria that produce lactic acid as the end-product of carbohydrate fermentation. This is an unexpected finding, as previous studies characterizing the fecal microbiota of cats and pigs found a much lower relative abundance and total count of LAB, respectively (Deusch et al., 2015; Su et al., 2008; Konstantinov et al., 2006). Furthermore, each of these studies saw a significant, age- related decrease in LAB. This is in contrast to Frese (2015), who found a significant increase in

26 the relative abundance of LAB from nursing to weaning. In our study, although weaned kits had a higher overall proportion of Lactobacillales than adult females, this difference was not significant. Interestingly, though, while the relative abundance of Lactobacillales was not significantly different between the groups, the genera Lactococcus and Lactobacillus, each LAB, were significantly enriched in the weaned kits, and the former had a significantly higher relative abundance in weaned kits than adult females. This suggests that while the composition of the microbiota in the weaned population that our samples came from had reached, or was close to reaching, maturity, there was persistence of certain genera that are typically associated with preweaning (Bian et al., 2016; Koenig et al., 2011; Yatsunenko et al., 2012). An interesting finding of the current study is the significantly higher proportion of the class Clostridia identified in the feces of mink in 2014 compared to 2015 or 2016 (17.1 vs 5.4 and 6.9, respectively). In 2014, there was a North American-wide shortage of the Clostridium botulinum toxoid vaccine, which is typically given to mink kits at approximately 10 weeks of age (Hildebrandt, 2014). It is unclear, however, the mechanism that would have caused this increase, as vaccination against the toxoid it unlikely to impact the bacterium itself. Although it is unknown if there was a corresponding increase in morbidity or mortality due to infection with Clostridium spp. in 2014, there was a clear difference in Clostridia within the gut of commercial mink during that year, particularly when one considers that the majority of mink farms in Ontario were included in this study. In the future, correlating changes in standard preventative care practices with both the clinical outcomes and the effects on the composition of the GI microbiome would offer valuable insight into the overall impact on the health status of these animals. An unexpectedly high proportion of organisms were identified from the genus Ignatzschineria. Several species from the genus have been found in conjunction with human maggot infestation and associated bacteremia (Barker et al., 2014; Le Brun et al., 2015). The genus was first identified from the gut of the larvae of an obligate parasitic fly, Wohlfahrtia magnifica (Tóth et al., 2001). Members of the genus have since been identified from the gut of adult flesh flies, Sarcophaga sp., and the Rocky Mountain wood tick, Dermacentor andersoni (Dergousoff and Chilton, 2011; Gupta et al., 2011). This suggests that this might represent contamination of fecal samples from insects. Further supporting this is the significant decrease in the overall relative abundance of this genus from summer to winter, when the cold temperatures

27 of Ontario all but eliminate most environmental insects. Moreover, Bahl et al (2017) found that the fecal microbiota of mink does not cluster with the microbiota of the host’s feed source, providing additional evidence that the high relative abundance of Ignatzschineria was the result of contamination of the feces (rather than the food source) with flies. However, we cannot definitively rule out that these organisms are true inhabitants of the mink gut, potentially from contaminated feed, as there are generally minimal biosecurity measures on mink farms and the risk of biocontamination of the feed is high (Compo et al., 2017; Appendix A). The impact of year appeared to be more common and profound than that of life stage and season. The microbiota can likely be impacted by a range of dietary, environmental, management and treatment factors, none of which are understood in mink. It is unsurprising that the largest number of differences were identified between 2014 and 2016, as these years would have had the fewest number (if any) of the same individual mink contributing to the samples collected (i.e., some of the kits from 2014 were the adult females from 2015 and some of the adult females from 2015 were the same females collected from 2016, but there would be almost no overlap of animals between 2014 and 2016). This is consistent with what has been seen in healthy adult humans, in which the fecal microbiota of an individual is more similar to itself over time, than to another’s fecal microbiota (Faith et al., 2013; Zoetendal et al., 1998; Costello et al., 2009; Rajilic-Stojanovic et al., 2012). Additionally, the dietary management of the mink industry relies heavily on what protein products are available, such as offal from the poultry, beef, pork, and seafood industries, expired deli meats or cheeses, or other meat products unfit for human consumption, which may not be consistent from year-to-year, or even from season to season. The differences we observed in composition of the fecal microbiota over the three years examined may be a reflection of differing diets or other management practices not examined here. At the farm-level, the predominant phyla and genera were found to be similar from 2014 to 2015. This is consistent with what has been seen in humans; individuals and family members (who have a shared environment) have a more similar fecal and oral microbiota than do unrelated individuals (Faith et al., 2013; Mosites et al., 2017; Song et al., 2013). The implications of this from a research resources standpoint, is that this suggests that there is more value in sampling from a larger number of environments (e.g. more farms) than a greater number of individuals from the same environment. It has additionally been shown that the skin microbiota, but not the fecal microbiota, of owners and their cohabitating pets were more similar

28 than noncohabitating pets (Song et al., 2013). We have previously determined that hygienic practices and biocontamination oversight are generally lacking on mink farms (Compo et al., 2017). It would be interesting to determine if there are any corresponding effects on the skin or fecal microbiota of mink farm workers, who are in constant contact with mink pelts and, potentially, mink feces, which is pervasive in the environment. Although there were few significant differences in the relative abundance of the predominant genera between adult female mink and weaned kits, a number of genera were identified as differentially abundant in weaned kits. The only genus of the phylum Proteobacteria that was enriched in weaned kits was Wohlfahrtiimonas, which is closely related to Ignatzschineria and similarly found in the gut of parasitic flies (Tóth et al., 2008). A reason for the feces of weaned kits being enriched with this genus is not immediately apparent, but may be related to preference of the host fly for the feces of the weaned animals, possibly due to consistency or content, neither of which were explored in this study. The gut of the mink is unique in that it is very short with an acidic stomach, averaging a pH of 2.5, presumably making them more tolerant of the effects of consuming bacteria associated with rotting meat (Bleavins and Aulerich, 1981; NRCNA, 2006). In the present study, a much higher proportion of Proteobacteria was identified than has previously been seen in mammals, including carnivores, regardless of year, life stage, or season. Bahl et al., (2017) found a similar high relative abundance of Proteobacteria, with Firmicutes still having the overall highest relative abundance. This is in contrast to Zhao et al., (2017) in which Proteobacteria was found to have the highest overall relative abundance and Firmicutes having the next highest relative abundance. However, the latter study used the V3 region of the 16S rRNA gene (vs the V4 region used presently), so the findings may not be directly comparable and may account for this difference. In humans, increases in the relative abundance of Proteobacteria have been associated with metabolic disorders, enteric inflammation, and dysbiosis (Frank et al., 2007; Manichanh et al., 2006; Tong et al., 2013; Shin et al., 2015). Additionally, metagenomic gene families with very high variance in abundance across hosts (“variable”) were almost entirely specific to the phylum Proteobacteria, whereas those invariable gene families (i.e., low variance across hosts) were generally specific to Firmicutes and Bacteroidetes, indicating that, in the human gut, the abundance of Proteobacteria may influence functional variability (Bradley and Pollard, 2017). All animals selected for this study were clinically healthy, yet showed a very high

29 relative abundance of Proteobacteria. One potential explanation for this is that mink, due to their carnivorous nature, are more resistant to what is generally considered an ‘unhealthy’ population of microbes in the gastrointestinal tract. Alternatively, the rapid transit time of the mink gastrointestinal tract lessens the ability of potentially pathogenic bacteria to take hold and cause clinical disease. Additional studies would be required to elucidate the reason for resistance to these potentially pathogenic bacteria in development of disease. By determining what bacteria are found in the feces of healthy mink, we have provided a baseline for the future study of how the normal composition is altered in disease states and, potentially, how the microbiota can be used to optimize production of farmed mink. The use of probiotics has been shown, for example, to increase overall productivity when given to growing broiler chickens and daily weight gain in weaning piglets (Timmerman et al., 2006; Di Giancamillo et al., 2008). As preweaning mortality represents a significant loss to mink farmers in Ontario, and the odds of being found dead decrease with increasing body weight during the preweaning period, promotion of weight gain through the use of probiotics could represent a means to increase production and animal welfare overall (Compo et al., 2017). A shortcoming of this study was related to how fecal samples were collected. Some samples may have been fresher than others, collected at different times of the day, or at different times in relation to husbandry duties. Thus, there may have been differences in exposure to the elements and potential for contamination. For future studies, uniformity in sample collection would aid in interpretation of results. Additionally, our collect-freeze-thaw-freeze-thaw procedure was used for convenience as samples were often aliquoted and processed weeks to months following collection. Though Song et al (2016) found that the fecal microbiome can shift over 8 weeks when subjected to freeze-thaw temperature fluctuations, it was found that the effect size was generally less than the variation between individuals. Regardless, the use of a preservative was found to reliably stabilize the microbiome and should be considered for future projects where temperature fluctuations or prolonged storage are to be expected (Song et al., 2016). An additional shortcoming of this study is its limited applicability to the microbiome in general. The unique findings of this study indicate that its representation of non-carnivores is likely limited. However, the findings between years, seasons, and life stages will contribute to our understanding of the dynamics of the microbiota more generally. Furthermore, it is currently

30 unknown if the fecal microbiota is representative of the microbiota of other sites in the gastrointestinal tract of the mink, such as the small or large intestines. For example, Pang et al (2013) determined that the microbiota of the cecum does not cluster with the fecal microbiota in mice. Similarly, distinct, though overlapping, microbial communities were identified by site along the gastrointestinal tract of dogs and cats, and, in the latter, the microbiota clustered by individual, rather than by body site (Suchodolski et al., 2008; Ritchie et al., 2008). Thus, though study of the fecal microbiota offers a noninvasive way to compare groups, it may not be representative of the gastrointestinal health of an animal overall and additional studies looking at other sites within the gastrointestinal tract of mink are warranted. To conclude, this study characterized the fecal microbiota of farmed mink and the influence of season, life stage, and year on its composition. In addition to expanding the knowledge of microbial dynamics in a carnivore, we have created a baseline for future studies exploring how the microbiota can be used to optimize the health, welfare, and production of farmed mink.

Table 2.12Number of pooled fecal samples collected from adult females and weaned kits collected from commercial mink farms in Ontario.

Adult Weaned Year & Females Kits Season (n=212) (n=175) 2014 83 83 summer 2015 90 92 summer 2016 not 39 winter collected

Table 2.23Number of samples used in the final analysis of the fecal microbiota of commercial mink.

Adult Weaned Year & Females Kits Season (n=198) (n=168) 2014 81 82 (summer) 2015 83 86 (summer) 2016 not 34 (winter) applicable

Table 2.34Relative abundance of predominant taxonomic classifications of bacteria from the feces of commercial mink (n=366).

Classification Level Taxon Relative Abundance (%) Phylum Firmicutes 60.5 (>1%, among 31 Proteobacteria 34.4 identified) Bacteroidetes 1.7 Bacilli 43.7 Gammaproteobacteria 30.9 Clostridia 15.2 Class (>1%, among 83 Betaproteobacteria 1.7 identified) Alphaproteobacteria 1.5 Flavobacteria 1.4 Erysipelotrichia 1.2 Lactobacillales 33.7 Xanthomonadales 15.2 Clostridiales 15.1 Bacillales 9.3 Order (>1%, among Enterobacteriales 6.4 148 identified) Pseudomonadales 5.4 Aeromonadales 2.0 Burkholderiales 1.6 Flavobacteriales 1.4 Erysipelotrichales 1.2 Xanthomonadaceae 15.2 Lactobacillaceae 12.8 Enterococcaceae 7.1 Enterobacteriaceae 6.4 Peptostreptococcaceae 4.8 Carnobacteriaceae 4.7 Planococcaceae 4.3 Family (>2%, among Clostridiales_Incertae_Sedis_XI 4.0 336 identified) Streptococcaceae 3.6 Moraxellaceae 2.9 Pseudomonadaceae 2.5 Staphylococcaceae 2.4 Incertae_Sedis_XI 2.3 Clostridiaceae_1 2.2 Aerococcaceae 2.2 Aeromonadaceae 2.0 Ignatzschineria 14.0 Lactobacillus 8.0 Enterococcus 5.4 Escherichia_Shigella 4.3 Genus (>2%, among Atopotipes 3.6 1,255 identified) Tissierella 2.8 Anaerosphaera 2.1 Lactococcus 2.1 Peptostreptococcus 2.1 Clostridium_cluster XI 2.0

Table 2.45Relative abundance and False Discovery Rate p-values, by year, for which at least 1 significant difference was identified between years from the fecal microbiota of mink, at each taxonomic level (n=366). Taxonomic 2014- 2014- 2015- Level 2014 2015 2016 2015 2016 2016 (overall Taxon Median% Median% Median% FDR FDR FDR cutoff) (Min-Max) (Min-Max) (Min-Max) p-value p-value p-value 62.1 54.5 65.4 Firmicutes 0.1053 0.0746 0.0040 (11.9 – 97.2) (8.9 – 98.8) (11.5 – 99.0) Phylum 33.1 31.4 25.6 Proteobacteria 0.6556 0.0033 0.0037 (>0.1%) (1.3 – 87.6) (0.5 – 89.9) (0.6 – 88.2) 0.2 0.5 5.3 Bacteroidetes 0.0100 0.0003 0.0100 (<0.1 – 7.8) (<0.1 – 47.7) (<0.1 – 22.8) 37.8 46.6 51.2 Bacilli 0.0070 0.0003 0.0384 (6.2 – 91.4) (6.0 – 97.2) (10.8 – 96.1) 27.5 25.5 22.8 Gammaproteobacteria 0.7535 0.0070 0.0384 (1.2 – 87.6) (0.5 – 88.4) (0.5 – 88.1) 17.1 5.4 6.9 Clostridia 0.0004 0.0028 0.0420 Class (0.9 – 72.6) (0.1 – 50.6) (0.1 – 55.6) (>0.1%) 0.1 0.3 4.1 Flavobacteria 0.0004 0.0003 0.0154 (<0.1 – 6.9) (<0.1 – 43.2) (<0.1 – 16.6) 0.8 0.4 0.6 Erysipelotrichia 0.0004 0.0725 0.1531 (<0.1 – 9.5) (<0.1 – 6.8) (<0.1 – 6.2) <0.1 <0.1 <0.1 Deltaproteobacteria 0.7535 0.0012 0.0397 (<0.1 – 1.0) (<0.1 – 1.8) (<0.1 – <0.1) 14.1 4.8 2.0 Xanthomonadales 0.0004 0.0003 0.0008 (0.2 – 79.6) (<0.1 – 84.2) (<0.1 – 5.3) 17.0 5.2 6.9 Clostridiales 0.0004 0.0043 0.0465 (0.9 – 72.6) (0.1 – 50.5) (0.1 – 55.2) 4.6 7.7 11.0 Bacillales 0.0004 0.0003 0.0092 (0.1 – 36.5) (0.5 – 87.7) (2.6 – 93.1) Order 0.3 0.4 3.7 Aeromonadales 0.5745 0.0060 0.0465 (>0.1%) (<0.1 – 18.2) (<0.1 – 43.9) (<0.1 – 20.5) 0.1 0.3 4.1 Flavobacteriales 0.0004 0.0003 0.0127 (<0.1 – 6.9) (<0.1 – 43.2) (<0.1 – 16.6) 0.8 0.4 0.6 Erysipelotrichales 0.0004 0.0993 0.1845 (<0.1 – 9.5) (<0.1 – 6.8) (<0.1 – 6.2) <0.1 <0.1 <0.1 Rhodospirillales 0.6418 0.0060 0.0730 (<0.1 – 1.9) (<0.1 – 3.0) (<0.1 – 0.1) 14.1 4.8 2.0 Xanthomonadaceae 0.0003 0.0003 0.0006 (0.2 – 79.6) (0.2 – 84.2) (<0.1 – 5.3) 4.4 7.6 9.2 Enterococcaceae 0.0003 0.0003 0.2725 (0.2 – 22.2) (0.4 – 25.1) (0.3 – 57.9) 3.9 1.3 2.5 Peptostreptococcaceae 0.0003 0.2662 0.0083 (<0.1 – 41.3) (<0.1 – 28.9) (0.1 – 27.8) 4.3 1.4 2.8 Carnobacteriaceae 0.0003 0.0042 0.2037 (<0.1 – 35.7) (<0.1 – 27.1) (0.1 – 29.5) 1.0 2.1 5.2 Planococcaceae 0.0003 0.0003 0.0006 (<0.1 – 20.7) (0.1 – 37.5) (0.9 – 85.6) 3.5 0.4 0.4 Clostridiales_Incertae_Sedis_XI 0.0003 0.0003 0.5566 (<0.1 – 37.1) (<0.1 – 18.4) (<0.1 – 8.3) Family 1.4 1.5 3.1 (>0.2%) Streptococcaceae 0.7949 0.0242 0.0337 (<0.1 – 43.5) (<0.1 – 48.0) (<0.1 – 31.1) 1.3 0.8 0.7 Moraxellaceae 0.0525 0.0443 0.4361 (<0.1 – 21.6) (<0.1 – 73.2) (<0.1 – 19.0) 0.9 1.0 1.5 Pseudomonadaceae 0.3435 0.0947 0.0084 (<0.1 – 54.7) (<0.1 – 11.9) (0.1 – 43.4) 0.8 0.6 2.0 Staphylococcaceae 0.7190 0.0315 0.1051 (<0.1 – 19.9) (<0.1 – 85.6) (<0.1 – 12.2) 1.0 0.1 1.0 Incertae_Sedis_XI 0.0003 0.0003 0.0337 (<0.1 – 39.2) (<0.1 – 16.1) (<0.1 – 0.2) 0.6 0.6 2.6 Clostridiaceae_1 0.3623 0.0157 0.0035 (<0.1 – 33.8) (<0.1 – 20.4) (<0.1 – 25.4) Aerococcaceae 1.1 0.4 0.1 0.0003 0.0003 0.0006

35 (<0.1 – 25.2) (<0.1 – 35.0) (<0.1 – 0.4) 0.3 0.4 3.7 Aeromonadaceae 0.3623 0.0031 0.0337 (<0.1 – 18.2) (<0.1 – 43.) (<0.1 – 20.5) 0.1 0.3 4.1 Flavobacteriaceae 0.0003 0.0003 0.0096 (<0.1 – 6.8) (<0.1 – 43.1) (<0.1 – 16.6) 0.8 0.4 0.6 Erysipelotrichaceae 0.0003 0.0765 0.1963 (<0.1 – 9.5) (<0.1 – 6.8) (<0.1 – 6.2) 0.2 0.1 0.7 Bacillaceae_2 0.2224 0.0244 0.0006 (<0.1 – 12.9) (<0.1 – 13.1) (<0.1 – 2.0) <0.1 <0.1 <0.1 Alteromonadaceae 0.0252 0.0383 0.0096 (<0.1 – 5.6) (<0.1 – 3.8) (<0.1 – <0.1) 15.4 7.2 1.9 Ignatzschineria 0.0002 0.0002 0.0006 (0.3 – 79.6) (0 – 74.4) (<0.1 – 4.9) 2.0 4.8 7.3 Enterococcus 0.0002 0.0002 0.0879 (<0.1 – 16.7) (0.2 – 21.9) (0.3 – 53.3) 0.8 1.3 8.2 Escherichia_shigella 0.0031 0.0060 0.3546 (0 – 44.3) (<0.1 – 77.4) (0.9 – 20.5) 4.2 0.6 2.0 Atopostipes 0.0002 0.0601 0.0409 (<0.1 – 35.1) (0 – 19.0) (0 – 29.5) 2.7 0.3 0.3 Tissierella 0.0002 0.0002 0.5969 (0 – 22.9) (0 – 14.9) (0 – 8.3) 0.4 <0.1 <0.1 Anaerosphaera 0.0002 0.0002 0.0152 (0 – 38.5) (0 – 16.0) (0 – 0.2) 0.5 0.7 0.5 Lactococcus 0.7105 0.1070 0.0415 (<0.1 – 40.5) (<0.1 – 11.1) (0 – 4.1) 0.6 0.1 0.1 Peptostreptococcus 0.0002 0.0002 0.6906 (<0.1 – 23.5) (0 – 6.4) (0 – 1.1) 0.9 0.4 1.7 Clostridium_XI 0.0002 0.4327 0.0037 (0.1 – 23.6) (0 – 19.7) (0 – 37.0) Genus 0.8 0.6 3.6 Oceanisphaera 0.6212 0.0060 0.0409 (>1.1%) (<0.1 – 18.2) (0 – 44.0) (<0.1 – 20.4) 0.3 0.6 0.6 Psychrobacter 0.0002 0.0172 0.6457 (0 – 13.7) (0 – 39.4) (0 – 18.9) 0.5 0.2 2.6 Streptococcus 0.1364 0.0002 0.0006 (<0.1 – 10.7) (0 – 24.7) (0 – 30.7) 0.4 0.3 0.8 Clostridium_sensu_stricto 0.0276 0.6416 0.0741 (<0.1 – 13.4) (0 – 7.5) (0 – 23.6) 0.5 0.5 1.3 Paenalcaligene 0.0109 0.6821 0.3546 (<0.1 – 11.6) (0 – 27.5) (0 – 3.8) 0.7 0.1 <0.1 Facklamia 0.0002 0.0002 0.0006 (0 – 18.4) (0 – 30.7) (0 – 0.3) 0.4 0.3 0.4 Pseudomonas 0.1525 0.1565 0.0489 (<0.1 – 54.0) (0 – 1.9) (<0.1 – 11.2) 0.3 0.1 <0.1 Wohlfahrtiimona 0.0006 0.0002 0.0006 (<0.1 – 16.5) (0 – 20.0) (0 – 0.4) 0.1 0.1 <0.1 Acinetobacter 0.0002 0.0002 0.0019 (0 – 8.1) (0 – 25.3) (0 – 2.5) 0.8 0.3 0.5 Erysipelothrix 0.0002 0.0021 0.5969 (<0.1 – 9.3) (0 – 6.8) (0 – 3.2)

Table 2.56Relative abundance and false discovery rate (FDR) p-values for significantly different taxa by life stage in mink fecal microbiota (n=332; 2014 and 2015 only). Adult Female Weaned Kit Taxonomic Level FDR Taxon Median % Median % (overall cutoff) p-value (Min-Max) (Min-Max) 35.1 29.5 Proteobacteria 0.0440 Phylum (1.1 – 87.6) (0.5 – 89.9) (>0.1%) <0.1 <0.1 Fusobacteria 0.0440 (<0.1 – 4.3) (<0.1 – 14.1) 29.7 22.4 Gammaproteobacteria 0.0301 (1.1 – 87.6) (0.5 – 88.4) Class 0.6 0.2 Betaproteobacteria 0.0182 (>0.1%) (<0.1 – 30.1) (<0.1 – 15.2) <0.1 <0.1 Fusobacteria 0.0411 (<0.1 – 4.3) (<0.1 – 14.1) 11.8 6.9 Xanthomonadales 0.0331 (<0.1 – 79.6) (<0.1 – 84.2) 2.9 2.0 Pseudomonadales 0.0063 (<0.1 – 56.3) (<0.1 – 73.4) Order 0.8 0.2 Aeromonadales 0.0061 (>0.1%) (<0.1 – 43.9) (<0.1 – 29.7) 0.5 0.2 Burkholderiales 0.0061 (<0.1 – 30.1) (<0.1 – 15.1) <0.1 <0.1 Fusobacteriales 0.0061 (<0.1 – 4.3) (<0.1 – 14.1) 11.8 6.9 Xanthomonadaceae 0.0360 (0 – 79.6) (<0.1 – 84.2) 4.7 6.7 Enterococcaceae 0.0360 (0.2 – 25.1) (0.3 – 23.4) 3.4 2.2 Carnobacteriaceae 0.0273 (<0.1 – 35.7) (<0.1 – 33.9) 1.3 1.6 Streptococcaceae 0.0473 (<0.1 – 43.5) (<0.1 – 48.0) 1.3 0.8 Moraxellaceae 0.0473 (<0.1 – 39.4) (<0.1 – 73.2) Family 1.0 0.4 Staphylococcaceae 0.0062 (>0.2%) (<0.1 – 85.6) (<0.1 – 34.0) 0.1 0.3 Incertae_Sedis_XI 0.0400 (<0.1 – 39.2) (<0.1 – 37.1) 1.0 0.5 Aerococcaceae 0.0360 (<0.1 – 35.0) (<0.1 – 17.9) 0.8 0.2 Aeromonadaceae 0.0016 (<0.1 – 43.9) (<0.1 –29.7) 0.5 0.1 Alcaligenaceae 0.0016 (<0.1 – 3<0.1) (<0.1 – 15.1) 0.2 0.1 Bacillaceae_2 0.0093 (<0.1 – 13.1) (<0.1 – 11.8) 11.4 6.2 Ignatzschineria 0.0234 (0 – 79.6) (0.2 – 79.0) 3.0 5.1 Enterococcus 0.0267 (<0.1 – 21.9) (0.1 – 19.8) 1.9 1.0 Atopostipes 0.0267 (0 – 35.1) (0 – 33.4) 0.1 0.2 Anaerosphaera 0.0267 (0 – 38.5) (0 – 27.8) Genus 0.6 1.0 Lactococcus 0.0234 (>1.1%) (<0.1 – 40.5) (0 – 44.1) 0.7 0.1 Oceanisphaera 0.0015 (0 – 44.0) (0 – 3<0.1) 0.5 0.3 Psychrobacter 0.0240 (0 – 39.4) (0 – 73.2) 0.5 0.1 Paenalcaligene 0.0020 (0 – 27.5) (0 – 14.8) 0.4 0.1 Jeotgalicoccus 0.0015 (0 – 53.0) (0 – 11.6)

Table 2.67Relative abundance and false discovery rate (FDR) p-values for significantly different taxa (p>0.05) by season in the fecal microbiota of mink (n=198; adult females only). Summer Winter Taxonomic Level FDR Taxon Median % Median % (overall cutoff) p-value (Min-Max) (Min-Max) 55.2 65.4 Firmicutes 0.0034 (10.9 – 98.8) (11.5 – 99.0) 35.1 25.6 Proteobacteria 0.0003 (1.1 – 87.6) (0.6 – 88.2) 0.3 5.3 Bacteroidetes 0.0028 (<0.1 – 47.7) (<0.1 – 22.8) Phylum <0.1 <0.1 Acidobacteria 0.0143 (>0.1%) (<0.1 – 3.2) (<0.1 – <0.1) <0.1 0.1 Fusobacteria 0.0205 (<0.1 – 4.3) (<0.1 – 0.8) <0.1 <0.1 Verrucomicrobia 0.0003 (<0.1 – 2.0) (<0.1 – <0.1) <0.1 <0.1 Tenericutes 0.0050 (<0.1 – 0.9) (<0.1 – 5.6) 40.4 51.21 Bacilli 0.0001 (6.4 – 97.2) (10.8 – 96.1) 29.7 22.8 Gammaproteobacteria 0.0012 (1.1 – 87.6) (0.5 – 88.1) Class 0.1 4.1 Flavobacteria 0.0005 (>0.1%) (<0.1 – 43.2) (<0.1 –16 6.) <0.1 0.1 Fusobacteria 0.0289 (<0.1 – 4.3) (<0.1 – 0.8) <0.1 <0.1 Deltaproteobacteria 0.0114 (<0.1 – 1.7) (<0.1 – <0.1) 11.8 2.0 Xanthomonadales 0.0004 (<0.1 – 79.6) (<0.1 – 5.3) Order 5.9 11.0 Bacillales 0.0004 (>0.1%) (0.1 – 87.7) (2.6 – 93.1) 0.1 4.1 Flavobacteriales 0.0004 (<0.1 – 43.2) (<0.1 – 16.6) 11.8 2.0 Xanthomonadaceae 0.0005 (<0.1 – 79.6) (<0.1 – 5.3) 4.7 9.2 Enterococcaceae 0.0023 (0.2 – 25.1) (0.3 – 57.9) 1.3 5.2 Planococcaceae 0.0005 (<0.1 – 33.5) (0.9 – 85.6) 1.2 0.4 Clostridiales_Incertae_Sedis_XI 0.0039 (<0.1 – 34.5) (<0.1 – 8.3) Family 1.3 3.1 Streptococcaceae 0.0074 (>0.2%) (<0.1 – 43.5) (<0.1 – 31.1) 0.1 1.0 Incertae_Sedis_XI 0.0012 (<0.1 – 39.2) (<0.1 – 0.2) 0.5 2.6 Clostridiaceae_1 0.0019 (<0.1 – 25.5) (<0.1 – 25.4) 1.0 0.1 Aerococcaceae 0.0005 (<0.1 – 35.0) (<0.1 – 0.4) 0.1 4.1 Flavobacteriaceae 0.0005 (<0.1 – 43.1) (<0.1 – 16.6) 11.4 1.9 Ignatzschineria 0.0003 (0 – 79.6) (<0.1 – 4.9) 3.1 7.3 Enterococcus 0.0003 (<0.1 – 21.9) (0.3 – 53.3) 1.1 8.2 Genus Escherichia_shigella 0.0496 (0 – 77.4) (0.9 – 20.5) (>1.1%) 0.9 0.3 Tissierella 0.0029 (0 – 22.9) (0 – 8.3) 0.1 <0.1 Anaerosphaera 0.0006 (0 – 38.5) (0 – 0.2) Oceanisphaera 0.7 3.6 0.0003

38 (0 – 44.0) (<0.1 – 20.4) 0.5 1.3 Paenalcaligene 0.0003 (0 – 27.5) (0 – 3.8) 0.3 0.4 Pseudomonas 0.0003 (0 – 54.0) (<0.1 – 11.2) 0.2 <0.1 Wohlfahrtiimona 0.0003 (0 – 20.0) (0 – 0.4) 0.3 <0.1 Acinetobacter 0.0300 (0 – 25.3) (0 – 2.5) 0.5 0.5 Erysipelothrix 0.0003 (0 – 9.3) (0 – 3.2)

Table 2.78Bacterial genera enriched, by year, from the feces of adult female mink and weaned kits (n=366; p<0.05; Linear Discriminate Analysis (LDA) scores >4)

2014 2015 2016 Unclassified Flavobacteriaceae Peptostreptococcus Escherichia_Shigella Facklamia Ignatzschineria Streptococcus Lactobacillus Unclassified Enterococcus Tissierella Lactobacillaceae Unclassified Atopostipes Enterococcaceae

Table 2.89Differences (p-values) between groups compared with UniFrac, AMOVA, and Parsimony analysis in membership and structure of the fecal microbiota of mink UNIFRAC AMOVA Parsimony Unweight The Yue and Clayton Index Female – Kits 0.188 <0.001 0.168 Summer – Winter <0.001 <0.001 0.001 2014 – 2015 <0.001 0.001 2014 – 2016 <0.001 0.001 2015 – 2016 <0.001 0.001 14– 15– 16 <0.001 <0.001 F14 – F15 <0.001 0.001 F14 – F16 <0.001 0.001 F15 – F16 <0.001 0.001 F14 – K14 0.097 0.824 F15 – K14 <0.001 0.001 F16 – K14 <0.001 0.001 F14 – K15 <0.001 0.001 F15 – K15 0.029 0.238 F16 – K15 <0.001 0.001 K14 – K15 <0.001 0.001 F14– F15– F16– K14– K15 <0.001 <0.001 By farm <0.001 The Classic Jaccard index Female – Kits 0.547 0.008 0.430 Summer – Winter <0.001 <0.001 0.001 2014 – 2015 <0.001 0.001 2014 – 2016 <0.001 0.001 2015 – 2015 <0.001 0.001 14 – 15 – 16 <0.001 <0.001 F14 – F15 <0.001 0.001 F14 – F16 <0.001 0.001 F15 – F16 <0.001 0.001 F14 – K14 0.031 0.643 F15 – K14 <0.001 0.001 F16 – K14 <0.001 0.001 F14 – K15 <0.001 0.001 F15 – K15 0.181 0.334 F16 – K15 <0.001 0.001 K14 – K15 <0.001 0.001 F14-F15-F16-K14-K15 <0.001 <0.001 By farm <0.001 AMOVA, analysis of molecular variance

Figure 2.1 Relative abundance of the 20 most predominant bacterial genera present in the feces of commercial mink (n=366). F=adult females, K=weaned kits, followed by the year (summers 2014 and 2015, winter 2016) from which the sample was taken.

LDA Score

Figure 2.2 Linear discriminant analysis effect size (LEfSe) indicating differentially enriched bacterial communities (with LDA score >3) at the genus level by life stage from samples from the summers of 2014 and 2015 (n=332). Blue=adult females (n=164), red=weaned kits (n=168).

LDA Score

Figure 2.3 Linear discriminant analysis effect size (LEfSe) indicating differentially enriched fecal bacterial communities (with LDA score >3) at the genus level of adult females by season (n=198). Blue=winter (n=34), red=summer 164).

Figure 2.4 Alpha diversity parameters (a: richness, b: evenness, c: diversity) observed in the feces of commercial mink by year (n=366), season (n=198), and life stage (n=332).

Figure 2.5 Principal coordinate analysis (PCoA) of community structure based on the Yue-Clayton Index (a) and membership based on the Jaccard Index (b) of the fecal microbiota of commercial mink, by year (n=366), life stage (n=332), and season (n=198).

46 CHAPTER THREE: GENERAL DISCUSSION

The current study was the first of its kind to explore the fecal microbiota of mink in Ontario to better understand how management practices influences its composition. A number of discoveries were made, both novel and consistent with previous mammalian studies. This study demonstrated that both the most predominant and some rare taxa have large ranges in relative abundance, and suggests that the carnivore fecal microbiota may be less relevant to the overall health of the animal than it is in other species. Additionally, this body of work will provide an important basis for the future optimization of the health, production, and welfare of farmed mink.

3.1 The commercial mink fecal microbiota: What is normal? Consistent with a recent study comparing the fecal microbiota of fasted and unfasted commercial mink, we found significant intersample variability in both the predominant and non- predominant taxa (Bahl et al., 2017). The range of relative abundance of Firmicutes, the most predominant phylum regardless of year, life stage, or season, was 90 (9% to 99%). At the genus level, despite identifying more than 1300 individual genera, the most predominant genus, Ignatzschineria, had a range of 80 (0% to 80%). Even several genera with an overall relative abundance of less than 0.5%, such as Jeotgalicoccus and Pseudomonas, have ranges greater than 50 (0% to 53% and 0% to 54%, respectively). Visually, this difference is evident when comparing stacked bar graphs of the relative abundance of the predominant phyla of the fecal microbiota of the rabbit, an herbivore that relies heavily on commensal bacteria to maintain gastrointestinal health, to mink, at the sample level (Appendix B; Kylie et al., 2016). In addition to the wide ranges seen here and reported by Bahl et al. (2017), a previous study in gnotobiotic ferrets, a mustelid closely related to mink, saw minimal morphologic or metabolic changes with a selected, rather than conventional fecal microbiota (Manning and Bell, 1990). Similarly, ferrets are generally considered resistant to the enteric side effects of oral antimicrobial administration that can be quite severe, and even deadly, in certain species, such as rabbits and horses (Hoefer et al., 2012). Together, these findings suggest that mink may be less reliant on their resident GI microbiota than other species, such as rabbits, horses, mice, and humans, or, possibly, that the fecal microbiota is less relevant than the microbiota of other parts of the gastrointestinal tract (Michelland et al., 2010; Schoster et al., 2013; Pang et al., 2012;

47 Eckburg et al., 2005). Moreover, mink appear to be relatively resilient to alterations to their gastrointestinal microbiota, as well as tolerant to a great range in the relative abundances of both the predominant and non-predominant taxa. The presence of bacterial DNA in a sample, or even of the live organism itself, does not necessarily indicate an infection, with consequent effect on an animal’s health. Differences in the composition of the fecal microbiota have been identified in several species of carnivores. Diarrheic cats had significantly increased levels of the phyla Proteobacteria and Firmicutes, with corresponding increases in the classes Gamma- and Beta-proteobacteria and Bacilli, compared to healthy cats, which was accompanied by changes in bacterial functional gene categories (Suchodolski et al., 2015). Smura et al. (2016) identified increased levels of Firmicutes, but not Proteobacteria, in diarrheic, compared to healthy, arctic foxes. Interestingly, there were no significant differences identified at any bacterial taxonomic level between healthy and diarrheic mink (Smura et al., 2016), which suggests that the cause of diarrhea in those cases was not bacterial in origin. This is consistent with our finding that 35% of suspected enteritis cases in found dead mink kits were negative on bacterial culture of their gastrointestinal contents (Compo et al, 2017; Appendix A). Rather, enteric viral pathogens, such as mink enteritis virus (MEV) and epizootic catarrhal gastroenteritis virus of mink (a coronavirus), which are known to cause significant morbidity and mortality in mink, may be present and contribute to disease (Gorham et al., 1990). A large proportion of mink kits from the previous study that were culture positive had growth of bacteria that might be considered as commensals, further suggesting that the cause of enteritis in mink kits is likely multifactorial (Compo et al., 2017; Appendix A). That said, only a small proportion of gut samples were cultured in the previous study, so it is difficult to make broad conclusions from these limited results. Our findings will contribute significantly to our understanding of the complicated interplay of the microorganisms found in the carnivore gut, and, through additional comparative and interventional studies, help elucidate those that may be contributing to gastrointestinal disease in mink. A shortcoming of this study was the lack of oversight as to how samples were collected, which may have altered our results. Instructions were provided to farmers for how to package and store samples, but freshness was not indicated, nor where from within the ‘pile’ a sample was collected. Furthermore, farms have different husbandry and management procedures (Compo et al., 2017; Appendix A), so potential for contamination or exposure to the elements

48 likely differed by farm. Additionally, it is unknown how our collect-freeze-thaw-freeze-thaw procedure might have altered our findings. The fecal microbiome has been shown to shift over an 8-week period when exposed to temperature fluctuations (as is seen with multiple freezing and thawing episodes), though this effect was generally less than the variation seen between individual samples (Song et al., 2016). Similarly, prolonged storage (14 days) at refrigeration temperature (4°C) or at room temperature had minimal effect on the composition, membership, and structure of the fecal microbiota in dogs and cats (Weese and Jalali, 2014; Gavriel, 2017). The use of preservatives, such as 95% ethanol, FTA cards, or OMNIgene Gut, reliably stabilized the microbiome and should be considered for future projects where temperature fluctuations or prolonged storage are to be expected (Song et al., 2016).

3.2 Optimization of production through manipulation of the microbiota Research on the fecal microbiota of carnivores is limited and it was our goal not only to contribute to this knowledge base, but to establish normal parameters for the fecal microbiota of healthy mink to determine if this information can be used to optimize the health, welfare, and production of commercial mink. A variety of strategies have proven successful for different health and production parameters in chickens, veal calf, and pigs, including continuous administration of probiotics, administration of probiotics to the mother, and perinatal vaccination (Mappley et al., 2013; Ballou et al., 2016; Timmerman et al., 2006; Di Giancamillo et al., 2008; Scharek et al., 2007). Each of these studies looked at production parameters (e.g., daily weight gain) during the neonatal or juvenile period, so it is unclear if this resulted in an overall increase in production value over time. However, morbidity and mortality represent a significant loss to production of farmed mink, and preweaning mortality is estimated at 20% to 25% for the Ontario industry (Compo et al, 2017; Appendix A). The odds of being found dead through the preweaning period has been shown to decrease significantly with increased weight, so even small gains in weight, promoted through probiotic use during this time, as has been shown in other species (Mappley et al., 2013; Timmerman et al., 2005; Di Giancamillo et al., 2008; Scharek et al., 2007), may contribute to overall decreased preweaning mortality and, in this way, increase production overall (Compo et al, 2017; Appendix A). Despite the high intersample variability in fecal microbiota, we demonstrated that the fecal microbiota of adult female mink and weaned kits are similar. Mink from up to 49 Ontario

49 farms were represented, suggesting that this similarity persists regardless of differences in farm environment and individual animal history. This is consistent with what is seen in humans, pigs, horses, and mice, where the mature, adult microbiota is reached at or around weaning, regardless of previous individual differences (Palmer et al., 2007; Bian et al., 2016; Jacquay, 2017; Odamaki et al., 2016; Schloss et al., 2012). From this, it could be inferred that attempts at manipulation of the fecal microbiota would be fruitless; however, this does not account for interruptions to the gastrointestinal microbiota caused by infectious disease, which could potentially be staved off with promotion of a ‘healthy’ gastrointestinal flora, as has been seen in chicks given probiotics, then challenged with Campylobacter jejuni, an opportunistic enteric pathogen (Morishita et al., 1997). In addition to the potential for optimizing production through decreased mortality by promotion of weight gain and resistance to infectious disease, another target for optimization of mink production is through identification of taxa associated with increased feed conversion efficiency. After mortality, feed represents a large cost to farmers. It has been established in both beef and dairy cattle that certain bacterial species, metabolic pathways, and microbiome genes were not only enriched in animals with higher feed efficiency, but that they also accurately predicted the animal’s feed efficiency phenotype (Shabat et al., 2016; Li and Guan, 2017). Recent work in pigs has suggested that certain taxa, even if present at very low relative abundance, may contribute to increased feed efficiency (McCormack et al., 2017). Similarly, swine enterotypes, or identifiable microbial ecosystem clusters, were positively associated with body weight and average daily weight gain in piglets (Arumugam et al., 2011; Ramayo-Caldas et al., 2016). However, it still remains to be determined if such relationships in pigs, and, more broadly, animals that are seemingly less dependent on their gastrointestinal microbiota to maintain homeostasis, are causal. While the study of metabolomics was outside the scope of this work, future studies exploring the functional differences in metabolic pathways between animals that are determined to have low or high conversion of feed may provide valuable insight into how the microbiota can be used to optimize the production of farmed mink.

3.3 Proteobacteria: Unhealthy bacteria in healthy animals Proteobacteria is a phylum of facultative anaerobes found ubiquitously in the environment, but with overall low relative abundance in the gastrointestinal tract of healthy

50 mammals (Shin et al., 2015). In humans and other species, a high relative abundance of Proteobacteria in the gut is considered indicative of a dysbiotic state, with elevated levels seen in neonates, and in conditions associated with metabolic disorders, malnutrition, and gastrointestinal inflammation and neoplasia (Jakobsson et al., 2014; Fei and Zhao, 2013; Subramanian et al., 2014; Morgan et al., 2012). Additionally, in the human gastrointestinal tract, the abundance of Proteobacteria may influence functional variability, as those genes that were variable across subjects were primarily from Proteobacteria, whereas those that were invariable were primarily from Firmicutes and Bacteroidetes (Bradley and Pollard, 2017). However, the presence of Proteobacteria in high relative abundance alone does not indicate disease, as evidenced in the current study and by the transient presence in healthy humans of Proteobacteria up to a relative abundance of 45% (Caporaso et al., 2010). Rather, it would appear that Proteobacteria act more opportunistically, blooming to a higher relative abundance with changes in the host environment or disease state. Because they have an exceptionally short gastrointestinal tract, mink have a gastrointestinal environment that may be more transient than is seen in other mammals, and the very high overall relative abundance of Proteobacteria, regardless of year, life stage, or season, may indicate an overall unstable microbial community.

3.4 Future Directions This study examined the fecal microbiota, but it is currently unknown if the bacteria found in the feces of mink is truly representative of other sites in the gastrointestinal tract of mink. In hindgut fermenting species, such as the rabbit and horse that rely heavily on their symbiotic bacteria to maintain homeostasis, the fecal microbiota has been found to reflect the composition of the distal gastrointestinal microbiota (though not the upper gastrointestinal tract) (Michelland et al., 2010; Schoster et al., 2013). However, in mice, the fecal microbiota does not cluster with the cecal microbiota (Pang et al., 2013). Similarly, dogs and cats, which more closely reflect the gastrointestinal physiology of mink than mice, have distinct, though overlapping, microbial communities at each level of the gastrointestinal tract (Suchodolski et al., 2008; Ritchie et al., 2008). Additionally, in dogs, cats, and humans, the microbiota of an individual animal tends to cluster with itself, rather than with the gastrointestinal site from which the sample came (Suchodolski et al., 2008; Ritchie et al., 2008; Eckburg et al., 2005). Thus, though study of the fecal microbiota offers a noninvasive way to compare groups, it may not be

51 representative of the gastrointestinal health of an animal overall and additional studies looking at other sites within the gastrointestinal tract of mink are warranted to determine if these findings would prove true in mink. As for other species, the gastrointestinal microbiota of mink is likely impacted by a range of dietary, environmental, management, and treatment factors, none of which is well understood. By gaining an understanding of the composition of the gastrointestinal microbiota of healthy farmed mink, we have set the stage for exploration of how the composition may change in response to disease states. Additional interventional studies could then be used to associate differences in the composition of the fecal microbiota with causation to optimize the health and production of farmed mink. In the current study, we identified taxa that were present in both high and very low relative abundance; however, the presence of an organism alone does not necessarily equate to having an actual clinical effect on an animal. As discussed previously, Proteobacteria are typically associated with poor gastrointestinal health, but all animals in this study were apparently healthy, despite some animals having a very high relative fecal abundance of this phylum. Research into the differences between groups of animals (i.e., between healthy and diseased animals or high- and low-producing animals) using metabolomics, which looks more specifically at the metabolic products present in a sample, would offer additional insight into these findings and might help to better understand what the differences in fecal microbial composition actually mean for the health and production success of an animal. Finally, this study explored only the bacterial component of the fecal microbiota. There are a number of additional microorganisms, such as viruses and fungi, which contribute to the microbiota and there is a need a to understand the collective interactions between all of these components in order to more fully understand the role of the microbiota in both health and disease of the host. Such work has already begun in our lab, which is in the process of characterizing the virome and associated bacteriophages within the same pooled fecal samples studied in the current project.

3.5 Summary and Conclusions 1) Firimicutes and Proteobacteria are the predominant bacterial phyla present in the feces of mink, regardless of year, life stage, or season.

2) Yearly factors have a more profound effect on the fecal microbiota of farmed mink than do life stage or season.

3) Farmed mink, regardless of year, life stage, or season, have a higher relative abundance of Proteobacteria than has been reported for other mammals, including other carnivores.

4) The composition of the fecal microbiota of farmed mink cluster by year and season, but not life stage.

5) There are few significant differences in the composition of the fecal microbiota between adult female mink and weaned kits.

6) At the farm level, the composition of the fecal microbiota of the predominant phyla and genera were not generally significantly different between 2014 and 2015, indicating that, when resources are limited, more information would be gained from sampling a greater number of environments (i.e., farms) than obtaining a greater number of samples from the same environment.

7) A large number of bacterial taxa were identified from the feces of farmed mink, the vast majority of which were present at very low relative abundance, and large ranges were identified in the relative abundance of many taxa.

8) There were significantly fewer Ignatzschineria spp. identified in fecal samples of adult female mink in winter than in summer, suggesting that this bacterium is likely an environmental contaminant.

53 REFERENCES

Alverdy, J.C. and Chang, E.B. (2008). The re-emerging role of the intestinal microflora in critical illness and inflammation: why the gut hypothesis of sepsis syndrome will not go away. J. Leukoc. Biol. 83(3), 461-466.

Amato, K.R., Yeoman, C.J., Kent, A., Righini, N., and Carbonero, F. (2013). Habitat degradation impacts black howler monkey (Alouatta nigra) gastrointestinal microbiomes. ISME J. 7, 1344-53. An C., Okamoto, Y., Xu, S., Eo, K.Y., Kimura, J., Yamamoto, N. (2016). Comparison of fecal microbiota of three captive carnivore species inhabiting Korea. K Vet Med Sci. 79(3), 542-546.

Arumugam, M., Raes, J., Pelletier, E., Paslier, D.L., Yamada, T., Mende, D.R., et al. (2011). Enterotypes of the human gut microbiome. Nature. 473, 174-180.

Atarashi, K., Tanoue, T., Shima, T., Imaoka, A., Kuwahara, T., Momose, Y., et al. (2011). Induction of colonic regulatory T cells by indigenous Clostridium species. Science. 331(6015), 337-341.

Bahl, M.I., Hammer, A.S., Clausen, T., Jakobsen, A., Skov, S., and Andresen, L. (2017). The gastrointestinal tract of farmed mink (Neovison vison) maintains a diverse mucosa- associated microbiota following a 3-day fasting period. MicrobiologyOpen. 6, e434.

Ballou, A.L., Ali, R.A., Mendoza, M.A., Ellis, J.C., Hassan, H.M., Croom, W.J., et al. (2016). Development of the chick microbiome: How early exposure influences future microbial diversity. Front Vet Sci. 3, 2. doi: 10.3389/fvets.2016.00002

Barker, H.S., Snyder, J.W., Hicks, A.B., Yanoviak, S.B., Southern, P., Dhakal, B.K., et al. (2014). First case reports of Ignatzschineria (Schineria) indica associated with myiasis. J. Clin. Microbiol. 52(12), 4432-4434. doi: 10.1128/JCM.02183-14.

Barker, I.K., Povey, R.C., and Voigt, D.R. (1983). Response of mink, skunk, red fox, and raccoon inoculation with mink virus enteritis, feline panleukopenia and canine parvovirus and prevalence of antibody to parvovirus in wild carnivores in Ontario. Can. J. Comp. Med. 47(2), 88.

Barry, K.A., Wojcicki, B.J., Bauer, L.L., Middelbos, I.S., Vester Boler, B.M., Swanson, K.S., et al. (2011). Adaptation of healthy adult cats to select dietary fibers in vivo affects gas and short-chain fatty acid production from fiber fermentation in vitro. J. Anim. Sci. 89(10), 3163-9. doi: 10.2527/jas.2010-3445.

Bauer, H., Horowitz, R.E., Levenson, S.M., and Popper, H. (1963). The response of the lymphatic tissue to the microbial flora. Studies on germfree mice. Am. J. Pathol. 42(4), 471.

54 Bell, J.A. and Manning, D.D. (1990) Prevalence of Campylobacter jejuni in ranch mink at pelting: Cultural, serological and histological evidence of infection. Can. Vet. J. 31, 67- 370.

Bian, G., Ma, S., Zhu, Z., Su, Y., Zoetendal, E.G., Mackie, R., et al. (2016). Age, introduction of solid feed and weaning are more important determinants of gut bacterial succession in piglets than breed and nursing mother as revealed by a reciprocal cross-fostering model. Environ Microbiol. 18, 1566-1577.

Bik, E.M., Echburg, P.B., Gill, S.R., Nelson, K.E., Purdom, E.A., Francois, F., et al. (2005). Molecular analysis of the bacterial microbiota in the human stomach. Prod. Natl. Acad. Sci. USA.103(3), 732-737.

Bleavins, M.R. and Aulerich, R.J. (1981). Feed consumption and food passage time in mink (Mustela vison) and European ferrets (Mustela putorius furo). Lab Anim. 31(3), 268-269.

Blomström, A.L., Widén, F., Hammer, A.S., Belák, S., and Berg, M. (2010). Detection of a novel astrovirus in brain tissue of mink suffering from shaking mink syndrome by use of viral metagenomics. J. Clin. Mirobiol. 48(12), 4392-4396

Böhmer, B.M., Kramer, W., and Roth-Maier, D.A. (2006). Dietary probiotic supplementation and resulting effects on performance, health status, and microbial characteristics of primiparous sows. J. Anim. Physiol. Anim. Nutr. 90, 309–15.

Boudreau, L., Benkel, B., Astatkie, T., Rouvinen-Watt, K. (2014). Ideal body condition improves reproductive performance and influences genetic health in female mink. Anim. Reprod. Sci. 145(1), 86-98.

Bouillant, A., Lee, V.H., and Hanson, R.P. (1965). Epizootiology on mink enteritis: Musca domestica L as a possible vector of the virus. Can. J. Comp., Med. Vet. Sci. 29, 148-152.

Bradley, P.H. and Pollard, K.S. (2017). Proteobacteria explain significant functional variability in the human gut microbiome. Microbiome. 5, 36. doi: 10.1186/s40168-017-0244-z.

Brandt, L.J., and Reddy, S.S. (2011). Fecal microbiota transplantation for recurrent Clostridium difficile infection. J. Clin. Gastroenterol. 45, S159-S167.

Breitbart, M., Haynes, M., Kelley, S., Angly, F., Edwards, R.A., Felts, B., et al. (2008). Viral diversity and dynamics in an infant gut. Res. Microbiol. 159(5), 367-373.

Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., et al. (2010). QIIME allows analysis of high-throughput community sequencing data. Nat. Methods. 7(5), 335-336. doi:10.1038/nmeth.f.303.

55 Chang, J.Y., Antonopoulos, D.A., Kalra, A., Tonelli, A., Khalife, W.T., Schmidt, T.M., et al. (2008). Decreased diversity of the fecal microbiome in recurrent Clostridium difficile— associated diarrhea. J. Infect. Dis. 197(3), 435-438.

Cheng, Y., Fox, S., Pemberton, D., Hogg, C., Papenfuss, A.T., and Belov, K. (2015). The Tasmanian devil microbiome—implications for conservation and management. Microbiome. 2015. doi: 10.1186/s40168-015-0143-0

Claus, S.P., Tsang, T.M., Wang, Y., Cloarec, O., Skordi, E., Martin, F.P., et al. (2008). Systemic multicompartmental effects of the gut microbiome on mouse metabolic phenotypes. Mol. Syst. Biol. 4, 219.

Combes, S., Fortun-Lamothe, L., Cauquil, L., and Gidenne, T. (2013). Engineering the rabbit digestive ecosystem to improve digestive health and efficacy. Animal. 7(9), 1429-1439.

Compo, N.R., Pearl, D.L., Tapscott, B., Storer, A., Hammermueller, J., Brash, M., et al (2017). On-farm biosecurity practices and causes of preweaning mortality in Canadian commercial mink kits. Acta Vet Scanda. (submitted)

Compo, N.R., Turner, P.V., and Weese, J.S. (2017). Characterization of the fecal microbiota of commercial mink. Scholars Portal Dataverse, V1.0, doi:10.5683/SO/OEIP77.

Costello E.K., Lauber, C.L., Hamady, M., Fierer, N., Gordon, J.I., and Knight, R. (2009). Bacterial community variation in human body habitats across space and time. Science. 326, 1694–1697. de Vos, W.M. and de Vos, E.A. (2012). Role of the intestinal microbiome in health and disease: from correlation to causation. Nutr Res. 70(suppl 1): S45-S56.

D’Argenio, V., Precone, V., Casaburi, G., Miele, E., Martinelli, M., Staiano, A., et al. (2013). An altered gut microbiome profile in a child affected by Crohn's disease normalized after nutritional therapy. Am. J. Gastroenterol. 108(5), 851.

Danzeisen, J.L., Clayton, J.B., Huang, H., Knights, D., McComb, B., Hayer, S.S., et al. (2015). Temporal Relationships exist between cecum, ileum, and litter bacterial microbiomes in a commercial turkey flock, and subtherapeutic penicillin treatment impacts ileum bacterial community establishment. Front Vet Sci. 2(56). doi: 10.3389/fvets.2015.00056

Debley, J.S., Smith, J.M., Redding, G.J., and Critchlow, C.W. (2005). Childhood asthma hospitalization risk after cesarean delivery in former term and premature infants. Ann Allergy Asthma Immunol. 94(2), 228-233.

Dergousoff, S.J. and Chilton, N.B. (2011). Novel genotypes of Anaplasma bovis, "Candidatus Midichloria" sp. and Ignatzschineria sp. in the Rocky Mountain wood tick, Dermacentor andersoni. Vet. Microbiol. 150(1-2), 100-6. doi: 10.1016/j.vetmic.2011.01.018.

56 Deusch, O., O’Flynn, C., Colyer, A., Swanson, K.S., Allaway, D., and Morris, P. (2015). A longitudinal study of the feline faecal microbiome identifies changes into early adulthood irrespective of sexual development. PLoSONE. 10(12), e0144881. doi: 10.1371/journal.pone.0144881.

Di Giancamillo, A., Vitari, F., Savoini, G., Bontempo, V., Bersani, C., Dell’Orto, V., et al. (2008). Effects of orally administered probiotic Pediococcus acidilactici on the small and large intestine of weaning piglets. A qualitative and quantitative micro-anatomical study. Histol. Histopathol. 23, 651–54.

Dominguez-Bello, M.G., Costello, E.K., Contreras, M., Magris, M., Hidalgo, G., Fierer, N., et al. (2010). Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Prod. Natl. Acad. Sci. USA. 107(26), 11971- 11975.

Duplessis, C.A., You, D., Johnson, M., and Speziale, A. (2012). Efficacious outcome employing fecal bacteriotherapy in severe Crohn’s colitis complicated by refractory Clostridium difficile infection. J. Infect. 40(4), 469-472.

Eckburg, P.B., Bik, E.M., Bernstein, C.N., Purdom, E., Dethlefsen, L., Sargent, M., et al. (2005). Diversity of the human intestinal microbial flora. Science. 308(5728), 1635-1638.

Englund, L., Chriél, M., Dietz, H.H., and Hedlund, K.O. (2002). Astrovirus epidemiologically linked to pre-weaning diarrhea in mink. Vet. Microbiol. 85(1), 1-11.

Ericsson, A.C., Davis, J.W., Spollen, W., Bivens, N., Givan, S., Hagan, C.E., et al. (2015). Effects of vendor and genetic background on the composition of the fecal microbiota of inbred mice. PloSONE. doi: https://doi.org/10.1371/journal.pone.0116704

Faith, J.J., Guruge, J.L., Charbonneau, M., Subramanian, S., Seedorf, H., Goodman, A.L., et al. (2013). The long-term stability of the human gut microbiota. Science. 341(6141), 1237439. doi: 10.1126/science.1237439.

Fei, N. and Zhao, L. (2013). An opportunistic pathogen isolated from the gut of an obese human causes obesity in germfree mice. ISME J. 7, 880-884.

Fisher, M.M. and Triplett, E.W. (1999). An automated approach for ribosomal intergenic spacer analysis of microbial diversity and its application to freshwater bacterial communities. Appl. Environ. Microbiol. 65(10), 4630-4636.

Frese, S.A., Parker, K., Calvert, C.C., and Mills, D.A. (2015). Diet shapes the gut microbiome of pigs during nursing and weaning. Microbiome. 3(1), 28.

Frank, D.N., St. Amand, A.L., Feldman, R.A., Boedeker, E.C., Harpaz, N., and Pace, N.R. (2007). Molecular-phylogenetic characterization of microbial community imbalances in

57 human inflammatory bowel diseases. Proc. Natl. Acad. Sci. USA. 104(34), 13780–13785. http://doi.org/10.1073/pnas.0706625104.

Fritz, J.V., Desai, M.S., Shah, P., Schneider, J.G., and Wilmes, P. (2013). From meta-omics to causality: experimental models for human microbiome research. Microbiome. 1(1), 14.

Fujimoto, C., Maeda, H., Kokeguchi, A., Takashiba, S., Nishimura, F., Arai, H., et al. (2003). Application of denaturing gradient gel electrophoresis to the analysis of communities of subgingival plaque. J Periodontal Res. 38(4), 440-5.

Furusawa, Y., Obata, Y., Fukuda, S., Endo, T.A., Nakato, G., Takahashi, D., et al. (2013). Commensal microbe-derived butyrate induces the differentiation of colonic regulatory T cells. Nature. 504(7480), 446-450.

Gaboriau-Routhiau, V., Rakotobe, S., Lecuyer, E., Mulder, I., Lan, A., Bridonneau, C., et al. (2009). The key role of segmented filamentous bacteria in the coordinated maturation of gut helper T cell responses. Immunity. 31(4), 677-689.

Garcia-Mazcorro, J.F., Lanerie, D.J., Dowd, S.E., Paddock, C.G., Grutzner, N., Steiner, J.M., et al. (2011). Effect of multispecies symbiotic formulation on fecal bacterial microbiota of healthy cats and dogs as evaluated by pyrosequencing. FEMS Microbiol Ecol. 78(3), 542- 554.

Gavriel, M.T. The Effect of Prolonged Ambient Temperature Exposure on the Feline Fecal Microbiota. American College of Veterinary Internal Medicine Forum 2017. 8 June 2017, Gaylord National Resort and Convention Center, National Harbor, MD.

Glad, T., Bernhardsen, P., Nielsen, K.M., Brusetti, L., Andersen, M., Aars, J., et al. (2010). Bacterial diversity in faeces from polarbear (Ursus maritimus) in Arctic Svalbord. BMC Microb. doi: 10.1186/1471-2180-10-10

Gorham, J.R., Evermann, J.F., Ward, A., Pearson, R., Shen, D., Hartsough, G.R., et al. (1990). Detection of coronavirus-like particles from mink with epizootic catarrhal gastroenteritis. Can. J. Vet. Res. 54(3), 383.

Guardabassi, L., Schmidt, K.R., Petersen, T.S., Espinosa-Gongora, C., Moodley, A., Agersø, Y., et al. (2012). Mustelidae are natural hosts of Staphylococcus delphini group A. Vet Microbiol. 159, 351-353.

Guo, M., Evermann, J.F., and Saif, L.J. Detection and molecular characterization of cultivable caliciviruses form clinically normal mink and enteric caliciviruses associated with diarrhea in mink. Arch. Virol. 146(2), 470-493.

Gupta, A.K., Dharne, M.S., Rangrez, A.Y., Verma, P., Ghate, H.V., Rohde, M., et al. (2011). Ignatzschineria indica sp. nov. and Ignatzschineria ureiclastica sp. nov., isolated from

58 adult flesh flies (Diptera: Sarcophagidae). Int. J. Syst Evol. Microbiol. 61, 1360-1369. doi: 10.1099/ijs.0.018622-0.

Handl, S., Dowd, S.E., Garcia-Mazcorro, J.F., Steiner, J.M., and Suchodolski, J.S. (2011). Massive parallel 16s rRNA gene pyrosequencing reveals highly diverse fecal bacterial and fungal communities in healthy dogs and cats. FEMS Microbol Ecol. 76(2), 301-310.

Hartmann, M., Howes, C.G., Abarenkov, K., Mohn, W.W., and Nilsson, R.H. (2010). V- Xtractor: an open-source, high-throughput software tool to identify and extract hypervariable regions of small subunit (16S/18S) ribosomal RNA gene sequences. J. Microbiol. Methods. 83(2), 250-253.

Haynes, M. and Rohwer, F. “The human virome.” In: Nelson, K.E., editor. Metagenomics of the human body. New York: Springer (2011). p. 63-77.

Hendriks, W.H., van Baal, J., and Bosch, G. (2012). Ileal and faecal protein digestibility measurement in humans and other non-ruminants–a comparative species view. Brit J Nutri. 108(S2), S247-S257.

Hildebrandt, H. “Management of mink.” In: Merck Veterinary Manual. Available at: http://www.merckvetmanual.com/exotic-and-laboratory-animals/mink/management-of- mink. Accessed: April 2015.

Hoefer, H.L., Fox, J.G., and Bell, J.A. (2012). “Ferrets: Gastrointestinal Diseases.” In: Queensberry, K.E. and Carpenter, J.W., editors. Ferrets, Rabbits, and Rodents: Clinical Medicine and Surgery. St. Louis: Elsevier. pp 27-45.

Hunter, D.B., Prescott, J.F., Hoover, D.M., Hlywka, G. and Kerr, J.A. (1986). Campylobacter colitis in ranch mink in Ontario. Can J Vet Res. 50, 47-53.

Hunter, D.B. and Lemieux, N. Mink…biology, health and disease. Guelph, ON: Graphic and Print Services, 1996.

Inagaki, H., Suzuki, T., Nomoto, K., Yoshikai, Y. (1996). Increased susceptibility to primary infection with Listeria monocytogenes in germfree mice may be due to lack of accumulation of L-selectin+ CD44+ T cells in sites of inflammation. Infect Immun. 64(8), 3280-3287.

Ivanov, I.I., Atarashi, K., Manel, N., Brodie, E.L., Shima, T., Karaoz, U., et al. (2009). Induction of intestinal Th17 cells by segmented filamentous bacteria. Cell. 139(3), 485-498.

Jacquay, E. (2014). Colonization and maturation of the foal fecal microbiota from birth through weaning and the effect of weaning method. [dissertation/masters’s thesis]. [Manhattan (KS)]: Kansas State University.

59 Jakobsson, H.E., Abrahamsson, T.R., Jenmalm, M.C., Harris, K., Quince, C., Jernberg, C., et al. (2014). Decreased gut microbiota diversity, delayed Bacteroidetes colonisation and reduced Th1 responses in infants delivered by Caesarean section. Gut. 63, 559-566.

Janda, J.M. and Abbott, S.L. (2007). 16S rRNA gene sequencing for bacterial identification in the diagnostic laboratory: pluses, perils, and pitfalls. J Clin Microbiol. 45(9), 2761-2764.

Kelly, C.R., de Leon, L., and Jasutkar, N. (2012). Fecal microbiota transplantation for relapsing Clostridium difficile infection in 26 patients: methodology and results. J Clin Gastroenterol. 46(2), 145-149.

Kinross, J.M., Darzi, A.W., and Nicholson, J.K. (2011). Gut microbiome-host interactions in health and disease. Genome Med. 3(3), 14.

Koenig, J.E., Spor, A., Scalfone, N., Fricker, A.D., Stombaugh, J., Knight, R., et al. (2011). Succession of microbial consortia in the developing infant gut microbiome. Proc. Natl. Acad. Sci. U.S.A. 108, 4578–4585. doi: 10.1073/pnas.1000081107.

Konstantinov, S.R., Awati, A.A., Williams, B.A., Miller, B.G., Jones, P., Stokes, C.R., et al. (2006). Post-natal development of the porcine microbiota composition and activities. Environ. Microbiol. 8, 1191–1199. doi: 10.1111/j.1462-2920.2006.01009.x

Kozich, J.J., Westcott, S.L., Baxter, N.T., Highlander, S.K., and Schloss, P.D. (2013). Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Appl. Environ. Microbiol. 79(17), 5112-5120. doi:10.1128/aem.01043-13. Accessed Jan-May 2017.

Kylie, J. (2016). An investigation into the fecal microbiota of domestic rabbits (Oryctolagus cuniculus) and factors influencing its composition. [dissertation/doctor’s thesis]. [Guelph (ON)]: University of Guelph

Lagerkvist, G., Johansson, K., Lundeheim, N. (1994). Selection for litter size, body weight, and pelt quality in mink (Mustela vison): correlated responses. J Anim Sci. 72, 1126-1126.

Laubereau, B., Filipiak-Pittroff, B., Von Berg, A., Grübl, A., Reinhardt, D., Wichmann, H.E., et al. (2004). Caesarean section and gastrointestinal symptoms, atopic dermatitis, and sensitisation during the first year of life. Arch Dis Child. 89(11), 993-997.

Le Brun, C., Gombert, M., Robert, S., Mercier, E., and Lanotte, P. (2015). Association of necrotizing wounds colonized by maggots with Ignatzschineria–associated septicemia. Emerg. Infect. Dis. 21(10), 1881-1883. doi:10.3201/eid2110.150748.

Li, F. and Guan, L.L. (2017). Metatranscriptomic profiling reveals linkages between the active rumen microbiome and feed efficiency in beef cattle. Appl. Environ. Microbio. 83(9). doi: 10.1128/AEM.00061-17.

60 Lian, H., Liu, Y., Zhang, S., Zhang, F., Hu, R. (2013). Novel orthoreovirus from Mink, China, 2011. Emerg Infect Dis. 19(12), 1985-1988.

Libby, S.J., Halsey, T.A., Altier, C., Potter, J., and Gyles, C.L. “Salmonella.” In: Gyles, C.L., Prescott, J.F., Songer, J.G., and Thoen, C.O., editors. Pathogenesis of Bacterial Infections in Animals. 4th ed. Ames, Iowa: Wiley-Blackwell. 2010. p.143-145.

Liu, B., Faller, L.L., Klitgord, N., Mazumdar, V., Ghodsi, M., Sommer, D.D., et al. (2012). Deep sequencing of the oral microbiome reveals signatures of periodontal disease. PloSONE 7. doi:e37919.

Macpherson, L.W. (1956). Feline enteritis virus—Its transmission to mink under natural and expiermental conditions. Can. J. Comp. Med. 20(6), 197.

Manichanh, C., Rigottier-Gois, L., Bonnaud, E., Gloux, K., Pelletier, E., Frangeul, L., et al. (2006). Reduced diversity of faecal microbiota in Crohn’s disease revealed by a metagenomic approach. Gut. 55(2), 205–211.

Manning, D.D. and Bell, J.A. (1990). Derivation of gnotobiotic ferrets: perinatal diet and hand- rearing requirements. Lab. Anim. Sci. 40(1), 51-5.

Mappley, L.J., Tchórzewska, M.A., Nunez, A., Woodward, M.J., Bramley, P.M., et al. (2013). Oral treatment of chickens with Lactobacillus reuteri LM1 reduces Brachyspira pilosicoli-induced pathology. J. Med. Microbiol. 62, 287–96.

Markle, J.G.M., Frank, D.N., Mortin-Toth, S., Robertson, C.E., Feazel, L.M., Rolle-Kampczyk, U., et al. (2013). Sex differences in the gut microbiome drive hormone-dependent regulation of autoimmunity. Science. 339(6123), 1084-1088.

Marteau, P., Pochart, P., Dore, J., Bera-Maillet, C., Bernalier, A., and Corthier, G. (2001). Comparative study of bacterial groups within the human cecal and fecal microbiota. Appl Environ Microbiol. 67(10), 4939-4942.

Martino, P.E. and Martino, J.J. (1996) Prevalence of microorganisms in dead mink kits from Aleutian-disease-infected and non-infected farms. Vet Res. 27(6), 607-12.

Mazmanian, S.K., Liu, C.H., Tzianabos, A.O., and Kasper, D.L. (2005). An immunomodulatory molecule of symbiotic bacteria directs maturation of the host immune system. Cell. 122(1), 107-118.

Mazmanian, S.K., Round, J.L., and Kasper, D.L. (2008). A microbial symbiosis factor prevents intestinal inflammatory disease. Nature. 453(7195), 620-625.

McCormack, U.M., Curiao, T., Buzoianu, S.G., Prieto, M.L., Ryan, T., Varley, P., et al. (2017). Exploring a possible link between the intestinal microbiota and feed efficiency in pigs. Appl. Environ. Microbiol. doi: 10.1128/AEM.00380-17.

Mentula, S., Harmoinen, J., Heikkila, M., Westermarck, E., Rautio, M., Huovinen, P., et al. (2005). Comparison between cultured small-intestinal and fecal microbiota in beagle dogs. Appl Environ Microbiol. 71(8), 4169-4175.

Michelland, R.J., Combes, S., Monteils, V., Cauquil, L., Gidenne, T. and Fortun-Lamothe, L. (2010). Molecular analysis of the bacterial community in digestive tract of rabbit. Anaerobe. 16(2), pp.61-65.

Middelbos, I.S., Vester Boler, B.M., Qu, A., White, B.A., Swanson, K.S., and Fahey, G.C. (2010). Phylogenetic characterization of fecal microbial communities of dogs fed diets with or without dietary fiber using 454 pyrosequencing. PLoSONE. 5(3), e9768. doi: dx.doi.org/10.1371/journal.pone.0009768.

Minot, S., Sinha, R., Chen, J., Li, H., Keilbaugh, S.A., Wu, G.D., et al. (2011). The human gut virome: inter-individual variation and dynamic response to diet. Gen Res. 21(10), 1616- 1625.

Mittelholzer, C., Englund, C., Hedlund, K.O, Dietz, H.H., Svensson, L. (2003). Detection and sequence analysis of Danish and Swedish strains of mink astrovirus. J. Clin. Microbiol. 41(11), 5192-5194.

Mittelholzer, C., Hedlund, K.O. Englund, L., Dietz, H.H., and Svensson, L. (2003). Molecular characterization of a novel astrovirus associated with disease in mink. J. Gen. Virol. 84(11), 3087-3094.

Momozawa, Y., Deffontaine, V., Louis, E., and Medrano, J.F. (2011). Characterization of bacteria in biopsies of colon and stools by high throughput sequencing of the V2 region of bacterial 16s rRNA gene in human. PLoSONE. doi: http://dx.doi.org/10.1371/journal.pone.0016952.

Mon, K.K.Z., Saelao, P., Halstead, M.M., Chanthavixay, G., Chang, H-C., et al. (2015) Salmonella enterica serovars enteritidis infection alters the indigenous microbiota diversity in young layer chicks. Front Vet Sci. 2, 61. doi: 10.3389/fvets.2015.00061.

Morgan, X.C., Tickle, T.L., Sokol, H., Gevers, D., Devaney, K.L., Ward, D.V., et al. (2012). Dysfunction of the intestinal microbiome in inflammatory bowel disease and treatment. Genome Biol. 13: R79.

Morishita, T.Y., Aye, P.P., Harr, B.S., Cobb, C.W. and Clifford, J.R. (1997). Evaluation of an avian-specific probiotic to reduce the colonization and shedding of Campylobacter jejuni in broilers. Avian Dis. pp. 850-855.

Mosites, E., Sammons, M., Otiang, E., Eng, A., Noecker, C., Manor, O., et al. (2017). Microbiome sharing between children, livestock and household surfaces in western Kenya. PLoS ONE. 12(2), e0171017. http://doi.org/10.1371/journal.pone.0171017.

Muegge, B.D., Kuczynski, J., Knights, D., Clemente, J.C., González, A., Fontana, L., et al. (2011). Diet drives convergence in gut microbiome functions across mammalian phylogeny and within humans. Science. 332(6032), 970-974.

National Farm Animal Care Council. “Code of Practice for the care and Handling of Farmed Mink.” Canada, 2013, electronic book. Available at: https://www.nfacc.ca/pdfs/codes/mink_code_of_practice.pdf

National Research Council. (1982). Nutrient Requirements of Mink and Foxes, Second Revised Edition, 1982. Washington, DC: The National Academies Press. doi:https://doi.org/10.17226/1114.

National Research Council. 2006. Nutrient Requirements of Dogs and Cats. Washington, DC: The National Academies Press. doi:https://doi.org/10.17226/10668. pgs 5-8.

Negele, K., Heinrich, J., Borte, M., Berg, A., Schaaf, B., Lehmann, I., et al. (2004). Mode of delivery and development of atopic disease during the first 2 years of life. Pediatr Allergy Immunol. 5(1), 48-54.

Neiger, J., Dietrich, C., Burnens, A., Waldvogel, A., Corthesy-Theulaz, I., Halter, F., et al. (1998). Detection and prevalence of Helicobacter infection in pet cats. J. Clin. Microbiol. 36(3), 634-637.

Nizza, S., Rando, F., Fiorito, F., Pagnini, U., Iovane, G., De Martino, L. (2014). Fecal microbiota and antibiotic resistance in ferrets (Mustela putorius furo) from two captive breeding facilities in Italy. Res Vet Sci. 96(3), 426-428.

Odamaki, T., Kato, K., Sugahara, H., Hashikura, N., Takahashi, S., Xiao, J., et al. (2016). Age- related changed in gut microbiota composition from newborn to centenarian: a cross- sectional study. BMC Microbiol. 16(90). doi: 10.1186/s12866-016-0708-5.

Palmer, C., Bik, E.M., DiGiulio, D.B., Relman, D.A., and Brown, P.O. (2007). Development of the human infant intestinal microbiota. PLoS Biol 5. 7, e177.

Pang, W., Vogensen, F.K., Nielsen, D.S., and Hansen, A.K. (2012). Faecal and caecal microbiota profiles of mice do not cluster in the same way. Lab Animal. 46(3), 231-236.

Pflughoeft, K.J. and Versalovic, J. (2012). Human microbiome in health and disease. Annu. Rev. Pathol. Mech. 7, 99-122.

Pollard, M. and Sharon, N. (1970). Responses of the Peyer's patches in germ-free mice to antigenic stimulation. Infect Immun. 2(1), 96-100.

63 Quail, M.A., Smith, M., Coupland, P., Otto, T.D., Harris, S.R., Connor, T.R., et al (2012). A tale of three next generation sequencing platforms: comparison of Ion Torrent, Pacific Biosciences and Illumina MiSeq sequencers. BMC Genomics. 13(1), 341.

Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., et al. (2013). The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucl. Acids Res. 41(D1), D590-D596. doi: 10.1093/nar/gks1219.

Rajilic-Stojanovic, M., Heilig, H.G.H.J., Tims, T., Zoetendal, E.G., and de Vos, W.M. (2012). Long-term monitoring of the human intestinal microbiota composition. Environ. Microbiol. doi: 10.1111/1462-2920.12023.

Ramayo-Caldas, Y., Mach, N., Lepage, P., Levenez, F., Denis, C., Lemonnier, G., et al. (2016). Phylogenetic network analysis applied to pig gut microbiota identified an ecosystem structure linked with growth traits. ISME J. 10, 2973-2977.

Rakoff-Nahoum, S., Paglino, J., Eslami-Varzaneh, F., Edberg, S., and Medzhitov, R. (2004). Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis. Cell. 118(2), 229-241.

Rhee, K., Sethupathi, P., Driks, A., Lanning, D.K., and Knight, K.L. (2004). Role of commensal bacteria in development of gut-associated lymphoid tissues and preimmune antibody repertoire. J. Immunol. 172(2), 1118-1124.

Ritchie, L.E., Steiner, J.M., and Suchodolski, J.S. (2008). Assessment of microbial diversity along the feline intestinal tract using 16s rRNA gene analysis. FEMS Microbiol Ecol. 66(3), 590-8.

Rogers, G.B., Carroll, M.P., and Bruce, K.D. (2009). Studying bacterial infections through culture-independent approaches. J. Med. Microbiol. 58, 1401-1418.

Sanger, F. and Coulson, A.R. (1975). A rapid method for determining sequences in DNA by primed synthesis with DNA polymerase. J. Mol. Biol. 94(3), 441-448.

Savage, D.C. (1986). Gastrointestinal microflora in mammalian nutrition. Annu Rev Nutri. 6(1), 155-178.

Scharek, L., Altherr, B.J., Tölke, C., and Schmidt, M.F.G. (2007). Influence of the probiotic Bacillus cereusvar toyoi on the intestinal immunity of piglets. Vet Immunol Immunopathol. 120, 136–47.

Schloss, P.D., Schubert, A.M., Zackular, J.P., Iverson, K.D., Young, V.B., Petrosino, J.F. (2012). Stabilization of the murine gut microbiome following weaning. Gut Microbes. 3(4), 383- 393.

64 Schoster, A., Arroyo, L.G., Staempfli, H.R. and Weese, J.S. (2013). Comparison of microbial populations in the small intestine, large intestine and feces of healthy horses using terminal restriction fragment length polymorphism. BMC research notes. 6(1), p.91.

Schneider, R.R. and Hunter, D.B. (1993). Mortality in mink kits from birth to weaning. Canadian Vet J. 34(3), 159.

Segata N., Izard, J., Waldron, L., Gevers, D., Miropolsky, L., Garrett, W.S., et al. (2011). Metagenomic biomarker discovery and explanation. Genome Biol. 12, R60.

Shabat, S.K.B., Sasson, G., Doron-Faigenboim, A., Durman, T., Yaacoby, S., Miller, M.E.B., et al. (2016). Specific microbiome-dependent mechanisms underlie the energy harvest efficiency of ruminants. ISME J. 10, 2958-2972.

Sharma, R., Young, C., and Neu, J. (2010). Molecular modulation of intestinal epithelial barrier: contribution of microbiota. J. Biomed. Biotechnol. 2010: 305879. doi: http://dx.doi.org/10.1155/2010/305879.

Shin, N.R., Whon, T.W., and Bae, J.W. (2015). Proteobacteria: microbial signature of dysbiosis in gut microbiota. Trends Biotechnol. 33(9), 496–503.

Sibbald, T.R., Sinclair, D.G., Evans, E.V., and Smith, D.L.T. (1962). The rate of passage of feed through the digestive tract of the mink. Can. J. Biochem. Physiol. 40(10), 1391-1394.

Sledge, D.G., Danieu, P.K., Bolin, C.A., Bolin, S.R., Lim, A., Anderson, B.C., et al. (2010). Outbreak of neonatal diarrhea in farmed mink kits (Mustella vison) associated with enterotoxigenic Staphylococcus delphini. Vet Pathol. 47(4), 751-757.

Smura, T., Aaltonen, K., Virtanen, J., Moisander-Jylha, A.M., Nordgren, H., Peura, J., et al. (2016). “Fecal microbiota of healthy and diarrheic farmed arctic foxes (Vulpes lagopus) and American mink (Neovison vison)—a case-control study.” In: Mäki-Tanila, A., Valaja, J., Mononen, J., Sironen, T., and Vapalahti, O., editors. Proceedings of the XIth International Scientific Congress in Fur Animal Production. Helsinki, Finland. pp 17-21.

Song, S.J., Lauber, C., Costello, E.K., Lozupone, C.A., Humphrey, G., Berg-Lyons, D., et al. (2013). Cohabiting family members share microbiota with one another and with their dogs. eLife. 2, e00458. http://doi.org/10.7554/eLife.00458.

Song S.J., Amir, A., Metcalf, J.L., Amato, K.R., Xu, Z.Z., Humphrey, G., et al. (2016). Preservation methods differ in fecal microbiome stability, affecting suitability for field studies. mSystems. 1(2), e00021-16. doi: 10.1128/mSystems.00021-16.

Songjinda, P., Nakayama, J., Kuroki, Y., Tanaka, S., Fukuda, S., Kiyohara, C., et al. (2005). Molecular monitoring of the developmental bacterial community in the gastrointestinal tract of Japanese infants. Biosci Biotech Bioch. 69(3), 638-641.

65 Stappenbeck, T.A., Hooper, L.V., and Gordon, J.I. (2002). Developmental regulation of intestinal angiogenesis by indigenous microbes via Paneth cells. P. Natl. Acad. Sci. 99(24), 15451-15455.

Statistics Canada. Table 003-0013. “Number and value of pelts produced, annual.” CANSIM (database). http://www.statcan.gc.ca/pub/11-627-m/11-627-m2017011-eng.htm Accessed: June 1, 2017.

Stephen, A.M, Wiggins, H.S., and Cummings, J.H. (1987). Effect of changing transit time on colonic microbial metabolism in man. Gut. 28(5):601-9.

Stevens, C and Hume, E. Comparative Physiology of the Vertebrate Digestive System. Cambridge University Press: New York, 2004, pgs 58-60.

Štursa, P., Uhlík, O., Kurzawová, V., Koubek, J., Ionescu, M., Strohalm, M., et al. (2009). Approaches for diversity analysis of cultivable and non-cultivable bacteria in real soil. Plant Soil Environ. 55(9), 389-396.

Su, Y., Yao, W., Perez-Gutierrez, O.N., Smidt, H., Zhu, W.Y. (2008). Changes in abundance of Lactobacillus spp. and Streptococcus suis in the stomach, jejunum and ileum of piglets after weaning. FEMS Microbiol. Ecol. 66 (3), 546-555. doi: 10.1111/j.1574- 6941.2008.00529.x

Subramanian, S., Huq, S., Yatsunenko, T., Haque, R., Mahfuz, M., Alam, M.A., et al. (2014) Persistent gut microbiota immaturity in malnourished Bangladeshi children. Nature. 510(7505), 417-21. doi: 10.1038/nature13421.

Suchodolski, J.S. (2011). Intestinal microbiota of dogs and cats: a bigger world than we thought. Vet Clin N Am-Small. 41(2), 261-272.

Suchodolski, J.S., Camacho, J., and Steiner, J.M. (2008). Analysis of bacterial diversity in the canine duodenum, jejunum, ileum, and colon by comparative 16s rRNA gene analysis. FEMS Microbiol Ecol. 66(3), 567-78.

Suchodolski, J.S., Dowd, S.E., Westermarck, E., Steiner, J.M., Wolcott, R.D., Spillmann, T., et al. (2009). The effect of the macrolide antibiotic tylosin on microbial diversity in the canine small intestine as demonstrated by massive parallel 16S rRNA gene sequencing. BMC Microbiol. 1, 210.

Suchodolski, J.S., Foster, M.L., Sohail, M.U., Leutenegger, C., Queen, E.V., Steiner, J.M. et al. (2015). The fecal microbiome in cats with diarrhea. PloSONE. 10(5), p. e0127378.

Sundqvist, C., Amador, A.G., Bartke, A. (1989). Reproduction and fertility in the mink (Mustela vison). J. Reprod. Fertil. 85(2), 413-441.

66 Swanson, K.S., Dowd, S.E., Suchodolski, J.S., Middelbos, I.S., Vester, B.M., Barry, K.A., et al. (2011) Phylogenetic and gene-centric metagenomics of the canine intestinal microbiome reveals similarities with humans and mice. ISME J. 5, 639-649.

Tanaka, S., Kobayashi, K., Songjinda, P., Tateyama, A., Tsubouchi, M., Kiyohara, C., et al. (2009). Influence of antibiotic exposure in the early postnatal period on the development of intestinal microbiota. FEMS Immunol. Med. Microbiol. 56(1), 80-87.

Thompson, G.R. and Trexler, P.C. (1971). Gastrointestinal structure and function in germ-free or gnotobiotic animals. Gut. 12(3), 230-235.

Timmerman, H.M., Veldman, A., van den Elsen, E., Rombouts, F.M., and Beynen, A.C. (2006). Mortality and growth performance of broilers given drinking water supplemented with chicken-specific probiotics. Poult. Sci. 85(8), 1383-8.

Timoney, J.F. Steptococcus. In: Gyles, C.L., Prescott, J.F., Songer, J.G. and Thoen, C.O, editors. Pathogenesis of Bacterial Infections in Animals. 4th ed. Ames, Iowa: Wiley-Blackwell, 2008. pg 23.

Tiquia, S.M. (2010). Using terminal restriction fragment length polymorphim (T-RFLP) analysis to assess microbial community structure in compost systems. Method Mol Biol. 599, 89- 102.

Tong, M., Li, X., Wegener Parfrey, L., Roth, B., Ippoliti, A., Wei, B., et al. (2013). A modular organization of the human intestinal mucosal microbiota and its association with inflammatory bowel disease. PLoSONE. 8(11), e80702. doi: http://doi.org/10.1371/journal.pone.0080702.

Tóth, E.1., Kovács, G., Schumann, P., Kovács, A.L., Steiner, U., Halbritter, A., et al. (2001). Schineria larvae gen. nov., sp. nov., isolated from the 1st and 2nd larval stages of Wohlfahrtia magnifica (Diptera: Sarcophagidae). Int. J. Syst. Evol. Microbiol. 51(Pt 2), 401-7.

Tóth, E.M., Schumann, P., Borsodi, A.K., Kéki, Z., Kovács, A.L., Márialigeti, K. (2008). Wohlfahrtiimonas chitiniclastica gen. nov., sp. nov., a new gammaproteobacterium isolated from Wohlfahrtia magnifica (Diptera: Sarcophagidae). Int. J. Syst. Evol. Microbiol. 58(Pt 4), 976-81. doi: 10.1099/ijs.0.65324-0.

Turnbaugh, P.J., Hamady, M., Yatsuneko, T., Cantarel, B.L, Duncan, A., Ley, R.E. (2009). A core gut microbiome in obese and lean twin. Nature. 457(7228), 480-484.

Tun, H.M., Brar, M.S., Khin, N., Jun, L., Hui, R.K., Dowd, S.E., et al. (2012). Gene-centric metagenomics analysis of feline intestinal microbiome using 454 junior pyrosequencing. J. Microbiol. Methods. 88(3), 369-376.

67 Umesaki, Y., Setoyama, H., Matsumoto, S., and Okada, Y. (1993). Expansion of alpha beta T- cell receptor-bearing intestinal intraepithelial lymphocytes after microbial colonization in germ-free mice and its independence from thymus. Immunology. 79(1), 32.

Urlings, H.A., Jonge, G.D., Bijker, P.G., and Van Logtestijn, J.G. (1993). The feeding of raw, fermented poultry byproducts: using mink as a model. J. Anim. Sci. 71(9), 2427-2431.

Uttenthal, A., Larsen, S., Lund, E., Bloom, M.E., Storgard, E., and Alexandersen, S. (1990). Analysis of experimental mink enteritis virus infection in mink: In situ hybridization, serology, and histopathology. J. Virol. 64(6), 2768-2779.

Van Soest, P.J. (1994). Nutritional Ecology of the Ruminant. Cornell University Press: Ithaca. pgs 4256-264

Velasquez-Manoff, M. (2015). Gut microbiome: The peacekeepers. Nature. S3-S11.

Vijay-Kumar, M., Aitken, J.D., Carvalho, F.A., Cullender, T.C., Mwangi, S., Srinivasan, S., et al. (2010). Metabolic syndrome and altered gut microbiota in mice lacking Toll-like receptor 5. Science. 328(5975), 228-231.

Vulfson, L., Pedersen, K., Chrie, M., Frydendahl, K., Andersen, T.H., Madsen, and M., Dietz, H.H. (2001). Serogroups and antimicrobial susceptibility among Escherichia coli isolated from farmed mink (Mustela vison Schreiber) in Denmark. Vet. Microbiol. 79, 143-153.

Wang, Q., Garrity, G.M., Tiedje, J.M., and Cole, J.R. (2007). Naïve Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ. Microbiol. 73(16), 5261-7.

Weese, J.S. and Jalali, M. (2014). Evaluation of the impact of refrigeration on next generation sequencing-based assessment of the canine and feline fecal microbiota. BMC Vet. Res. 10, 230.

Williams, C., Elnif, J., and Buddington, R.K. (1998). The gastrointestinal bacteria of mink (Mustela vison): influence of age and diet. Acta Vet. Scand. 39(4), 473-82.

Yang, S., Wang, Y., Li, W., Fan, Z., Jiang, L., et al. (2016). A novel bocavirus from domestic mink, China. Virus Genes. 52(6), 887-890.

Yang, L., Lu, X., Nossa, C.W., Francois, F., Peek, R.M., Pei, Z. (2009). Inflammation and intestinal metaplasia of the distal esophagus are associated with alterations in the microbiome. Gastroenterol. 137(2), 588-597.

Yatsuneko, T., Rey, F.E., Manary, M.J., Trehan, I., Dominguez-Bello, M.G., Contreras, M., et al. (2012). Human gut microbiome viewed across age and geography. Nature. 486(7402), 222-7. doi: 10.1038/nature11053.

68 Zhang, Q., Widmer, G., and Tzipori, S. (2013). A pig model of the human gastrointestinal tract. Gut Microbes. 4(3), 193–200. http://doi.org/10.4161/gmic.23867

Zhao, L., Wang, G., Siegel, P., He, C., Wang, H., Zhao, W., et al. (2013). Quantitative genetic background of the host influences gut microbiomes in chickens. Sci Rep. 3, 1163. doi: 10.1038/srep01163

Zhao, H., Sun, W., Wang, Z., Zhang, T., Fan, Z., Gu, Haijun., et al. (2017). Mink (Mustela vison) gut microbial communities from Northeast China and its internal relationship with gender and food additives. Curr Microbiol. doi: 10.1007/s00284-017-1301-3.

Zoetendal, E.G., Akkermans, A.D., and De Vos, W.M. (1998). Temperature gradient gel electrophoresis analysis of 16S rRNA from human fecal samples reveals stable and host- specific communities of active bacteria. Appl. Environ. Microbiol. 64, 3854–3859.

69 APPENDIX A: ON-FARM BIOSECURITY PRACTICES AND CAUSES OF PREWEANING MORTALITY IN CANADIAN COMMERCIAL MINK KITS

This Appendix chapter has been submitted for publication as:

Nicole Compo, David L. Pearl, Brian Tapscott, Amanda Storer, Jutta Hammermueller, Marina Brash, Patricia V. Turner. (2017). On-farm Biosecurity Practices and Causes of Preweaning Mortality in Canadian Commercial Mink Kits. Acta Veterinaria Scandanavia (Submitted)

Background: Mink are an important animal commodity group in Canada and excessive kit mortality represents a significant loss to production. National biosecurity standards have been developed for Canadian mink farms, but it is unclear how well these standards have been implemented as there are no studies correlating management practices of mink producers with causes of death in mink kits. To that end, we surveyed Ontario mink producers on their biosecurity and management practices and conducted almost 5,660 post mortem examinations on found-dead, preweaned kits to characterize mink farm biosecurity practices and causes of death in preweaned kits. Results: We found that very few biosecurity and management practices were uniformly used by producers, despite good awareness of appropriate practices. Use of personal protective equipment was implemented by fewer than 50% of respondents, while control of mink shed access, disinfection of feed containers after use, and use of a rodent control program were the only practices implemented by greater than 70% of respondents. Only 18% of producers reported regular use of antimicrobials in feed or water, although 91% stated they used antimicrobials for treatment of suspected bacterial diseases on a regular basis. On post mortem examination, no gross abnormalities were noted in 71% of the kits, 45% were thought to be stillborn or aborted, 27% had some form of abnormal fluid distribution in the body, and 2% had a congenital lesion. A subset of 69 gastrointestinal tract samples was submitted for bacterial culture, of which 45 samples yielded sufficient growth. Most interesting was the identification of Salmonella enterica serovar Heidelberg in 11% of samples. Conclusions: The results of this study will provide a benchmark for Canadian mink producers and their veterinarians, defining the areas to which greater attention should be given to ensure more rigorous biosecurity practices are in place. Ultimately, these improvements will promote increased mink production and animal well-being.

70 A.1 INTRODUCTION Mink are purpose bred for their pelts and are an important animal commodity group in Canada. In 2014, there were approximately 784,000 breeder mink on 237 mink farms across Canada, with more than 3.2 million pelts sold and valued at almost $98 million [1]. While most costs to producers are relatively fixed, mink mortality can be highly variable and it can represent a significant loss to productivity. Published data are lacking in this area; however, empirical evidence suggest that early loss of preweaned kits likely represents the largest area of overall mortality on mink farms. Causes of preweaning mortalitiy in carnivores in general are poorly characterized. Newborn ferrets have been shown to be uniquely susceptible to fatal infection with influenza virus compared to their adult counterparts, in which the disease is mild and transient [2], and infection of the pregnant ferret with vesicular stomatitis virus leads to fetal resorption, abortion or neonatal death, with no clinical signs in the dam [3]. In a retrospective analysis of kittens from private homes and rescue shelters submitted for histopathologic analysis in the UK, the most common cause of death in neonatal and preweaning cats was infectious disease, of which more than half were of a viral etiology, including feline herpesvirus, calicivirus, and parvovirus [4]. However, in a study on specific pathogen-free cats, the most frequent cause of stillbirth was undetermined, followed by maternal neglect [5]. Historical preweaning mortality rates on mink farms range from 20-25% in Canada and Argentina, respectively [6,7], somewhat higher than that noted for other food animal commodity groups, such as swine at 15% [8], but similar to meat rabbits [9,10]. For the most part, post mortem examinations are not conducted routinely on found dead mink kits and a cause of death is not identified. Even in studies specifically evaluating mortality in mink kits, gross lesions were not present in up to 78% of animals dying within 4 days of birth [6]. Two previous studies identified systemic infection as the most common cause of death in unweaned mink kits >4 and 7 days of age, respectively, although specific diseases were not further characterized [6,7]. Other gross pathology findings in these studies included evidence of starvation, hypothermia, dystocia, anasarca, and congenital defects [6]. The Canadian Food Inspection Agency has indicated that the most serious infectious diseases that mink producers face in Canada are Aleutian disease, mink enteritis virus, distemper, and hemorrhagic pneumonia due to Pseudomonas aeruginosa [11]. Potential sources of

71 infectious agents vary by disease. For example, antibodies to Aleutian disease have been identified in skunk, wild mink, badgers, and small rodents [12-14], while foxes were the source of a major distemper outbreak on Danish mink farms between July 2012 and February 2013 [15,16]. Poorly managed manure and carcasses can contribute to the persistance of pathogens, such as mink enteritis virus, a parvovirus, which remains infectious in the environment for at least nine months [17]. Newly purchased mink stock, even when apparently healthy, can also harbor and shed mink enteritis virus in the feces for at least one year [17]. Feed and water sources have also been implicated, for example, an outbreak of swine influenza in Danish mink was determined to arise from feed containing infected swine lung tissue [18]. Additional sources of infectious agents include humans, companion animals, vehicles, and farm equipment [11]. Biosecurity resources are available to producers to help limit spread of disease between farms and wildlife, and to minimize on-farm losses. There are two guides produced for Canadian mink farmers, the Code of Practice for the Care and Handling of Farmed Mink [19], and the CFIA National Farm-Level Mink Biosecurity Standard [11]. The CFIA defines on-farm biosecurity as, “a set of well-organized, well-planned procedures that are applied at the farm level,” with the intent to, “reduce exposure of mink to infectious disease-causing agents, including their introduction, spread within the farmed mink population and release from the farm,” [11]. This document introduces standards based on well-known principles of isolation, sanitation, traffic control, herd health management, and maintenance of the biosecurity program once it is established. The standard provides basic and voluntary recommendations, which establish target outcomes on access management, animal health management, and operational management [11]. Although the document is readily available, it is unclear to what extent the standards are implemented at the farm level in Canada. To better characterize causes of production losses on Canadian mink farms as well as implementation of new biosecurity guidelines, the goals of this study were two-fold: 1) to determine the causes of death in found dead preweaned mink kits, and 2) to characterize uptake of on-farm biosecurity and management practices on mink farms.

72 A.2 MATERIALS AND METHODS

A.2.1 Part I – Preweaned Kit Mortality Surveillance

A.2.1.1 Farm recruitment and sample collection

Between April 1 and 30, 2013, all mink producers in Ontario were contacted through the Ontario Fur Breeders Association and the Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA) for enrollment and a total of 21 of 44 producers were enrolled. Participating producers were provided with supplies, including instructions for cadaver collection and storage, freezer bags, plastic totes, and permanent markers. Producers were instructed to collect dead kits daily into dated, labeled freezer bags, and to store frozen at -200C until pick up. Frozen, dead kits were collected at three and six weeks post-whelping from participating farms. Depending on the number of dead kits collected, a proportion of or all dead kits were collected. Farm and kit information were linked using anonymized identification codes for each farm. Dead mink kits between 0 and 45 days of age were collected and returned to the University of Guelph and stored at -200C until processed.

A.2.1.2 Sample processing Bags of cadavers from each farm were thawed at 40C, cadavers were weighed, and gross post mortem examinations were conducted in the Dept of Pathobiology, University of Guelph on a total of 5,657 kits, and abnormalities were recorded. Samples of sections of gastrointestinal tract from 69 kits with suspected enteritis, based on gross examinations, and sections of skin from three kits with dermatitis, were submitted for aerobic microbiologic culture to the Animal Health Laboratory, University of Guelph, following standard procedures.

A.2.1.3 Preweaning mortality survey Participating mink farms were contacted by OMAFRA by email or telephone and asked to submit production figures for the 2013 whelping season. Producers were asked to include the total number of kits born on the farm, how or when this count was performed (i.e., actual counts versus estimates and on what day the count occurred post-whelping), how many kits were weaned, and at what age kits were weaned. This information was used to calculate the self- reported incidence of preweaning mortality from all causes per participating farm.

A.2.1.4 Environmental data To determine whether there were any significant weather variations during the whelping season, daily environmental parameters were collected during the whelping season from Environment Canada (https://weather.gc.ca/canada_e.html) for the four counties encompassing the study farms: minimum and maximum daily temperatures, total precipitation, and highest relative humidity.

A.2.1.5 Statistical analyses Statistical analyses were conducted using Stata (StataSE 14 , College Station, TX). For post mortem results, models concerning the following outcomes were fitted using multilevel logistic regression: breath taken (yes/no), evidence of food in gastrointestinal tract (yes/no), presence of any congenital lesion (yes/no), presence of any abnormality (yes/no; including any congenital lesions, abnormal fluid distribution, signs of trauma, and/or evidence of dystocia), and producer-reported incidence of preweaning mortality. These models were used to examine statistical associations between the above dependent variable and the following independent variables: total kits born, average age weaned (in days), and when kit counts were taken (reported in days post-whelping). Random intercepts were included for farm and shed, except for mortality, for which only data on farm of origin were available. For postmortem body weight (in grams), the same associations were examined, except multi-level linear regression models were fitted. The assumption of linearity between independent and dependent variables (in the log odds scale for logistic regression models) was examined using lowess curves (i.e., locally weighted regression). If the assumption was violated, a quadratic term was included, if appropriate, or the independent variable was modeled as a dichotomous variable (i.e. above and below a median cut-off). For all outcomes, intercept-only multi-level logistic and linear models were fitted to estimate variance components and to calculate the variance partition coefficient at the farm, shed and kit level. To estimate the variance at the kit level for our multi-level logistic regression models, the latent variable technique was used [20]. Significance level for all analyses was set at α=0.05.

74 A.2.2 Part II - Survey of Husbandry and Management Practices In April, 2014, a detailed survey regarding farm practices was sent to 26 Ontario mink producers of which 11 responded. Participants were informed prior to enrollment of the survey’s purpose and that any identifying information would be confidential. Five of the survey respondents had participated in the preweaning mortality project and six respondents had not. The survey included questions on owner demographics (e.g., years in operation, number of employees, etc.); proximity of farm/mink to other livestock; water and feed source treatment; animal health, quarantine and carcass disposal; biosecurity and hygiene practices; and, manure disposal and pest control (see Supplementary Table 1). Some questions had open-ended answers, while others required selection from a fixed list of choices. In all cases, participants had an option to supply their own answers if they did not feel the options provided adequately described their farm’s practices. Additionally, participants were informed that they could refuse to answer any question and still remain in the study. Once completed, participants were instructed to return the survey via email or fax. Finally, participants were informed that they could withdraw their survey information at any time by contacting the study coordinator. The project was reviewed and approved by the University of Guelph Research Ethics Board (14MR013).

A.2.3 Statistical analyses Statistical analyses were conducted using Stata. For descriptive statistics concerning answers to survey questions which could be answered as yes/no proportions and their exact 95% confidence intervals are reported. For descriptive statistics concerning answers to survey questions which had continuous answers, median, range, mean, standard deviation and their exact 95% confidence intervals are reported. Using exact univariable logistic regression, associations between the use of each biosecurity and management practice measure were examined and the following independent variables evaluated (dichotomized as above/below the median): number of years the producer had been in operation, the number of adult females on farm, and the number of years females were kept for breeding. The significance level for all analyses was set at α=0.05.

75 A.3 RESULTS A.3.1 Part I: Preweaned Kit Mortality Surveillance A total of 5,657 post mortem examinations were completed on found dead preweaned mink kits, and of these, 11 kits were excluded from the final analysis due to incomplete data. The mean weight of the mink kits was 17.9g (range=2-1123g, SD=41.8g, 95%CI 16.8-19.0g) (Figure A.1) and 55.4% were estimated to have breathed prior to being found dead, based on lung flotation. Overall, 66% of kits born dead were <11g in weight, 32% were between 11-18.5g and just 2% were >18.5g. As weight increased, kits were more likely to have taken a breath (OR=1.03, p<0.001. Of kits born alive, 11% had food in their gastrointestinal tract at the time of death, representing 6.0% of all animals examined. No gross abnormalities were present in 70.6% of animals (Table A.1). The percent variance for the outcomes ‘breathed,’ ‘food in gastrointestinal tract,’ ‘congenital lesion,’ ‘any abnormality,’ and ‘weight’ explained at the kit level ranged from 78- 86%. Much less variance was explained at the farm level (11-20%) and almost no variance explained at the shed level (0-5%). This indicates that most of the variance in these outcomes is explained at the kit level rather than the farm or shed levels. Approximately 1% of kits examined had abnormalities consistent with dystocia during birth. Signs of trauma were noted in 1.3% of animals, although in some cases it was not possible to discern if the injuries were sustained ante or post mortem. Four abnormal fluid distribution patterns were commonly observed, including hydrothorax, hydroperitoneum, subcutaneous edema, and anasarca, a severe form of generalized edema. When considered together, 26.5% of kits had some form of abnormal fluid distribution (95% CI 25.4 – 27.7%) (Table 1). Congenital lesions, specifically, cleft palate, umbilical hernia, or other (i.e., low incidence conditions such as heart defects) and were seen in 0.92%, 0.87%, and 0.23% of animals, respectively (Table 1). When considered together, congenital abnormalities were present in 2% of all found dead preweaned kits (95% CI 1.7 – 2.4%). In suspected enteritis cases (n=69), gastrointestinal tract samples were taken and submitted for bacterial culture. The proportion of samples yielding significant (3+ to 4+) growth of 1, 2, 3 or 4 bacterial species was 23%, 20%, 16%, and 1%, respectively and 24 samples had minimal or no growth and were excluded from the final analysis. Ten bacterial species were isolated from the 45 samples that yielded sufficient growth (Table 2): five species from

76 Enterococcus, two species from Staphylococcus, and one species each from Escherichia, Salmonella and Streptococcus, which represent a variety of gastrointestinal tract commensal and potentially pathogenic or zoonotic organisms. Salmonella spp. was isolated from four different farms, and one farm yielded two separate positive samples. One skin sample from the three suspected dermatitis cases yielded 4+ growth of Staphylococcus delphini. Fifteen of 21 farms participating in the mortality surveillance study provided data on the total number of kits born, the day this number was counted (i.e., # of days post-whelping), total number of kits weaned, and the average day of weaning (Table 3). There was a statistically significant quadratic relationship between kit body weight at the time of post mortem examination and the day on which kits were counted (main effect: coefficient=-2.75, 95% CI - 9.0-3.5, p=0.391; quadratic effect: coefficient=0.33, 95% CI 0.01 – 0.65, p=0.042). Initially, body weight decreased as the day counted increased, but after 5 days, body weight increased with increasing day counted. Additionally, and not surprisingly, the odds of a kit having food in its gastrointestinal tract at post mortem examination was significantly increased if the day kits were counted exceeded 3 days post-whelping (OR=3.86, 95% CI=1.05-14.0, p=0.042). The odds of detecting a kit having any congenital lesion was significantly increased with each day increase following birth (OR=1.07, 95% CI=1.01-1.12, p=0.013). No other associations tested were statistically significant. No conclusions could be drawn relating kit mortality to general weather patterns and there were no environmental extremes in overall temperature, precipitation, and relative humidity patterns documented during the preweaning period for any of the counties monitored (data not shown).

A.3.2 Part II: Survey of Husbandry and Management Practices Of 26 biosecurity surveys distributed to Ontario mink farms, 11 were completed and returned (42% completion rate, or a 25% response rate for all mink farms in Ontario).

A.3.2.1 General farm information Farm demographics are summarized in Table 4. Farms had generally been in operation for many years, indicating that the mink industry may not be a growth industry, as no new farms are being established. In terms of the number of breeding females kept on farm, farms were not

77 generally very large. Employees working on the farm, including family members, were categorized as full time (FT), part time summer (PTS) and part time pelting (PTP), with farms having more PTP employees than FT or PTS (Table 4). There was a wide range in number of pelts produced (range=2,800-36,000) and total kits born (range=3,200-39,000) across Ontario mink farms in 2013 (Table 4). The average number of years that breeding females were kept was 2.7 (range=2-4, median=3), while for breeding males, it was 1.5 years (range=0-3, median=1). All producers reported being within 3 km of other agricultural species, including other mink farms, chickens, dairy and beef cattle, swine, and turkeys. Only four producers (4 of 11; 36%; 95% CI=11-69%) owned other production species, including broiler chickens, beef cattle, and horses. Most producers (9 of 11; 81%; 95% CI=48-98%) visited other farms, although not often (less than monthly), and most producers also reported having visitors from other farms, which included mink, beef and dairy cattle, and swine operations. All but one farm reported receiving farm water from wells (including deep, drilled, and shallow, dug wells) (95% CI 58.7 – 99.7%), with one farm using exclusively municipal water and one farm using both wells and municipal water. Six of 11 (55%; 95% CI 23.4 – 83.3%) farms tested the water for bacteria, with 50% testing on an annual basis, and 50% testing less than annually. Four of 11 farms (36%; 95% CI 10.9 – 69.2%) treated the water on-farm by a variety of methods, including dechlorination, use of an injection pump, hydrogen peroxide or descaling. Six of 11 (55%; 95% CI=23.4-83.3) farms had a perimeter fence surrounding mink sheds and 4 of 6 reported that the fence was at least 6 feet high. Eight of 11 (72%; 95% CI=39-94) farms used wet feed made on site, while the remaining 3 of 11 (27%) purchased wet feed, and no producers reported feeding pellets. Only 1 of 11 (9%) producers reported having cement floors in mink sheds, while the remaining 10 of 11 (91%) had packed soil.

A.3.2.2 Animal health responses One producer reported increased mortality in preweaned kits, one producer reported increased mortality in weaned kits, and one producer reported increased mortality in both preweaned and weaned kits in 2013 compared to previous years. Estimates of overall kit mortality ranged from 4.5% to10%, with one producer reporting only that it was more than 50% higher than the previous year. Only one farm reported treating kits with antimicrobials. The

78 most common of diseases (specifically related to mortality) in post-weaned kits were heat stress, foot pad disease, hemorrhagic pneumonia, diarrhea, and dermatitis (‘sticky kits’). Only one producer reported increased mortality in adult females for 2013. The most common diseases in adult females (specifically related to mortality) were foot pad disease, hemorrhagic pneumonia (likely bacterial), and mastitis. Three farms provided an estimate of female mortality, which ranged from 0.5% to 1.5%. Two farms indicated that enrofloxacin was given to adult females; however, information on frequency of use was not provided. A variety of health problems for both kits and adult females were reported to be seen commonly by producers (Figure 2). Ninety-one percent of producers (10 of 11; 95% CI=58.7- 99.7) responded that they used antimicrobial to treat disease of suspected bacterial origin. Various formulations were reported, which most frequently fell into the penicillin class, used either alone or as a combination product (9 of 11; 82%; 95% CI 48.2 – 97.7%). Other classes of antimicrobials used alone or as a combination product, included macrolides, sulfonamides tetracyclines, and fluoroquinolones, each with 4 of 11 (36%; 95% CI 10.9 – 69.2%) and polypeptides (3 of 11; 27%; 95% CI 6.0 – 61.0). Two producers reported using an antimicrobial combination product that included a penicillin, tetracycline, and sulfonamide. Only 2 of 11 producers (18%; 95% CI= 2.3-52.8%) reported using antimicrobials in the feed or water regularly.

A.3.2.3 Biosecurity responses There were no significant differences in responses to the survey between those farms categorized as ‘small’ (<3400 females) or ‘large’ (≥3400 females). As the number of years a female was kept increased, the odds of kits having any abnormality identified at post mortem exam significantly decreased (OR=0.42; 95% CI=0.19-0.92; p=0.029). There was no effect of farm size or number of years in operation on biosecurity practices. More than half of respondents did not feel that the mink industry had biosecurity standards in place and there were very few biosecurity and pre-emptive health management practices used uniformly by producers, with the exception of vaccinations. Three of 11 respondents indicated that no special biosecurity practices were in place on their farm. The use of personal protective equipment, policies on basic hand-washing and dedicated clothing were each implemented by fewer than 50% of respondents, while control of shed access, disinfection of

79 feed containers after use and use of a rodent control program were the only practices implemented by greater than 70% of respondents (Tables 5 and 6). Most producers (9 of 11; 81%; 95% CI=48.2 – 97.7%) used several methods for manure disposal. The most common methods for manure disposal (temporary vs permanent were not specified) were field spreading and an outdoor pile, with 8 of 11 (72%; 95% CI=39.0-94.0%) producers implementing each of these methods. Less commonly used methods included a temporary storage shed (1 of 11; 9%; 95% CI 0.2 – 41.3%), composting (3 of 11; 27%; 95% CI 6.9 – 61.0%), and hauling away from the farm (5 of 11; 45%; 95% CI 16.7 – 76.6%).

A.4 DISCUSSION Similar to previous studies [6,7], the cause of death in the majority of mink kits examined grossly in this study was not determined. As kit weight increased, the odds of being born dead decreased and more than 50% of kits that were born dead were <11g, suggesting that low kit birth weight may contribute to increased risk of early mortality. In 2013, the incidence of producer-reported mink kit preweaning mortality ranged from 3-15% and environmental conditions were within those expected, and likely not contributing significantly to the incidence of preweaned kit mortality. Gross findings are not expected for a range of other infectious and noninfectious causes of preweaning mortality in kits, including various virus infections (kit or pregnant female), chilling, poor maternal care, lack of milk, and failure to thrive syndrome. A shortcoming of this study was the lack of statistical power. Had more producers participated in both parts of the study, additional associations between biosecurity practices and post mortem findings may have been identified. Mink producers assisted with the development of national mink biosecurity guidelines which have been widely available to producers for close to a year at the time the survey was conducted [11,19]. Despite this, only 45% of respondents in this study felt that the industry had adequate biosecurity standards. This was also reflected in the overall lack of consistency of implementation of recommended biosecurity practices by the producers surveyed in the second part of this study. To that end, the results of this study will provide a benchmark for the Canadian mink industry to assess areas in which greater attention should be given to enhance biosecurity practices. The industry may also use this as an opportunity to provide additional training and educational programs for producers.

80 Although certain enteric viral pathogens, such as mink enteritis virus (MEV) and epizootic catarrhal gastroenteritis of mink (a coronavirus) [21,22], are known to cause significant morbidity and mortality in mink, the role of specific bacterial agents as a primary cause of preweaned kit enteropathy is not as well-established [23]. Samples from 45 of 69 kits with suspected enteritis were culture positive for at least 1 bacterial species. Enterococcus spp. were most common, with E. faecalis isolated from 43% of samples and E. faecium from nearly 20%. These agents are both common, commensal gastrointestinal tract bacteria of most mammals, including mink, and may not have been the primary cause of enteritis [24-26]. Enterococcus gallinarum was cultured in 3 of 69 samples and likely represents another commensal organism, as it has not previously been associated with disease in mink. Both E. avium and E. hirae are common gastrointestinal bacteria of birds, including poultry [27,28]. Given the frequent use of poultry offal for mink feed this finding is not unexpected. Some strains of E. coli are thought to be a significant cause of morbidity and mortality in farmed mink and this bacterium has been implicated as a cause of hemorrhagic pneumonia, enteritis, mastitis, and septicemia [21,29,20]. However, E. coli has also been shown to have a high prevalence in apparently healthy adult female mink and their kits and further genotyping and toxin isolation is needed to determine the significance of isolated strains [26,31]. Additionally, when prevalence of E. coli serotype was investigated, there was no difference between healthy kits and those affected by sticky kit disease [31]. Of concern was identification of Salmonella enterica serovar Heidelberg from 7% (5 of 69) of cultured samples, an agent of significant public health conern. This bacterium has been associated with food-borne disease outbreaks in Canada and the U.S.A. [32,33]. Salmonellosis is usually food-borne, with poultry and pork products representing the most common sources of contamination [34]. Salmonella spp. infection has been linked to abortion and stillbirths in mink, dogs and cats [35-38]. Furthermore, although less common, an asymptomatic carrier state can also occur in humans and animals [39]. Less than 50% of producers in this study reported that employees washed their hands after handling mink and only one producer had hand-washing policies in place. Thus, inadvertant zoonotic transmission from contaminated fecal matter could occur when poor hygiene practices are present, emphasizing again the importance of biosecurity practices on-farm.

81 Staphylococcus delphini, identified in 15% of gastrointestinal tract samples from kits with suspected enteritis and one skin sample, has previously been associated with an outbreak of ‘sticky kit’ syndrome with high mortality in the U.S.A [40]. This syndrome presents as neonatal diarrhea characterized by mucoid feces and wet and sticky fur of affected kits [40,41]. Both bacterial and viral agents have been isolated from affected kits in other studies, including astrovirus, coronavirus, mink enteric calicivirus, S. delphini, and Salmonella spp. [35,40-43] and the precise etiology of ‘sticky kit’ syndrome is unknown. Although enteritis was not a common finding in the present study, outbreaks of neonatal diarrhea can be a significant cause of morbidity and mortality on mink farms and the etiology is likely multifactorial. Lesions were present in 29% of kits examined in this study and the most common abnormality noted was excessive fluid accumulation, including hydrothorax, hydroperitoneum, subcutaneous fluid, and anasarca. Abnormal fluid distribution is a nonspecific clinical sign, which can arise from increased hydrostatic pressure, decreased oncotic pressure, increased vascular permeability or obstruction of fluid clearance. In dogs inoculated with minute virus of canines (a parvovirus), some dams had stillbirths or whelped pups with anasarca [44]. Further, causes of fading pups, pup abortion and pup stillbirth in dogs can include infection with canine herpes virus 1a [45-47] and canine parvovirus [44,48]. Similarly, in cats, abortion, stillbirth and fading kittens have been linked to infection with feline panleukopenia virus (a parvovirus) or feline infectious peritonitis virus [49-55]. Infection with Aleutian mink disease virus, a parvovirus of significant concern in mink, has been shown to decrease conception rate, litter size and weight, and increase neonatal mortality, even in kits born to clinically healthy females [56]. Mink-specific viruses are less well characterized compared with those of domestic dogs and cats and their role in kit stillbirth, abortion, and abnormal development are largely unknown. Moreover, mink have been shown to be susceptible to infection with feline panleukopenia virus [57] and canine distemper virus [15,58], which reinforces that companion animals should be excluded from mink sheds. Findings from this study suggest that more work is needed to characterize viral infections of mink and their impact on kit mortality. A significant noninfectious factor in mink kit survival is the mothering ability of the female, particularly nest building, kit retrieval, and nursing [59]. Mink kits are dependent on the female for warmth and nutrition, as they are altricial at birth with an undeveloped thermoregulatory system and minimal fat stores [59-62]. Kits are especially susceptible to

82 hypothermia and lose heat quickly during the first few days of life when exposed to cold [60,61]. Females with higher kit survival rates spend significantly more time exhibiting kit-directed behavior [59]. Further, compared to kits from females provided with either straw as a nesting substrate, a plastic or artificial nest, or both, kits from females provided with only a nest box and wood shavings (the current standard on most Canadian mink farms) had significantly lower body weights one week after birth and higher mortality rates compared to the other groups [63]. Females provided with nesting material were also quicker to retrieve kits that had been removed from the nest [63]. This suggests that provision of supplemental nesting materials and selection of females for mothering abilities, a heritable trait in many mammals, could significantly decrease kit mortality and contribute to overall greater production yields and improved kit welfare. In addition to human activity and companion animals as a source of pathogen introduction, wildlife pose a significant risk to mink health [12-16]. National mink biosecurity guidelines recommend the use of effective security fences, self-closing, lockable gates, and enclosed sheds to minimize wildlife access to mink [11,19]. Just over half of respondents reported having fences around their mink sheds/farm, although information on the types of sheds used was not collected. Reasons for not having such measures were not collected, but cost is a possible explanation. The logic for implementing biosecurity standards, specifically in mink production, is evident in the near eradication of Aleutian disease virus in Denmark. Voluntary implementation of testing, quarantine, and limited movement of mink decreased the number of positive farms from 100% in 1976 to 15% in 1996; further reduced to <5% in 2001 following implementation of additional government-mandated biosecurity measures [13]. Even if the goal is not eradication of a specific disease from a geographical area, limiting the spread of potential pathogens and preventing outbreaks of infectious disease should represent a viable goal for the Canadian mink farming industry. Despite mortality of preweaned kits being a significant cause of loss to producers, few studies have evaluated specific causes of death. In this study, we sought to estimate associations between management practices of farmers with causes of preweaned kit mortality, as well as characterizing the current state of biosecurity practices of the Canadian mink farming industry. Enhancing on-farm biosecurity practices as per national industry recommendations will assist

83 with reducing infectious and contagious causes of mortality in mink kits, likely resulting in increased productivity and animal well-being.

84 A.5 REFERENCES

[1] Statistics Canada. Number and value of pelts produced, Table 003-0014. 2016. http://www5.statcan.gc.ca/cansim/a26?lang=eng&retrLang=eng&id=0030014&pattern=Number+and+val ue+of+pelts+produced&tabMode=dataTable&srchLan=-1&p1=1&p2=-1 (accessed 10 Apr 2017).

[2] Collie MH, Rushton DI, Sweet C, Smith H. Studies of influenza virus infection in newborn ferrets. J Med Microbiol. 1980;13:561-571.

[3] Suffin SC, Muck KB, Porter DD. Vesicular stomatitis virus causes abortion and neonatal death in ferrets. J Clin Microbiol. 1977;6(4):437-438.

[4] Cave TA, Thompson H, Reid SWJ, Hodgson DR, Addie DD. Kitten mortality in the United Kingdom: a retrospective analysis of 274 histopathological examinations (1986-2000). Vet Rec. 2002;151:497-501.

[5] Young C. Preweaning mortality in specific pathogen-free kittens. J Small Anim Pract. 1973;14:391- 397.

[6] Schneider RR, Hunter BD. Mortality in mink kits from birth to weaning. Can Vet J. 1993;34:159-163.

[7] Martino PE, Villar JA. A survey on perinatal mortality in young mink. Vet Res Commun. 1990;14:199-205.

[8] U.S. Dept. of Agriculture. Swine 2012 Part I: Baseline Reference of Swine Health and Management in the United States, 2012. 2015. https://www.aphis.usda.gov/animal_health/nahms/swine/downloads/swine2012/Swine2012_dr_PartI.pdf (accessed 10 Apr 2017).

[9] Kylie J, Brash M, Whiteman A, Tapscott B, Slavic D, Weese JS, Turner PV. Biosecurity practices and causes of enteritis on Ontario meat rabbit farms. Can Vet J. 2017;(in press).

[10] Whitney JC, Blackmore DK, Townsend GH, Parkin RJ, Hugh-Jones ME, Crossman PJ, Graham- Marr T, Rowland AC, Festing MFW, Krzysiak D. Rabbit mortality survey. Lab Anim. 1976;10:203-207.

[11] Canadian Food Inspection Agency. National Farm-Level Mink Biosecurity Standard. Canadian Food Inspection Agency, Guidance Document Repository. 2013. http://www.inspection.gc.ca/animals/terrestrial-animals/biosecurity/standards-and- principles/mink/eng/1376667870636/1376667871636 (accessed 10 Apr 2017).

[12] Nituch LA, Bowman J, Wilson PJ, Schulte-Hostedde AI. Aleutian mink disease virus in striped skunks (Memphitis memphitis): Evidence for cross-species spillover. J Wildl Dis. 2015;51(2):389-400.

[13] Themudo GE, Houe H, Agger JF, Østergaard J, Ersbøll AK. Identification of biosecurity measures and spatial variables as potential risk factors for Aleutian disease in Danish mink farms. Prev Vet Med. 2012;107:134-141.

[14] Knuuttila A, Aaltonen K, Virtala A-MK, Henttonen H, Isomursu M, Leimann A, Maran T, Saarma U, Timonen P, Vapalahti O, Sironen T. Aleutian mink disease virus in free-ranging mustelids in Finland – a cross-sectional epidemiological and phylogenetic study. J Gen Virol. 2015;96:1423–1435.

85 [15] Trebbien R, Chriel M, Struve T, Hjulsager CK, Larsen G, Larsen LE. Wildlife reservoirs of canine distemper virus resulted in a major outbreak in Danish farmed mink (Neovison vison). PLoS One. 2014;9(1):e85598. doi: 10.1371/journal.pone.0085598.

[16] Gregers-Jensen L, Agger JF, Hammer ASV, Andresen L, Chrièl M, Hagberg E, Jensen MK, Hansen MS, Hjulsager CK, Struve T. Associations between biosecurity and outbreaks of canine distemper on Danish mink farms in 2012–2013. Acta Vet Scand. 2015: 57: 66. doi:10.1186/s13028-015-0159-2

[17] Bouillant, A., Hanson R.P., 1965. Epizootiology of mink enteritis: I. Stability of the virus in feces exposed to natural environmental factors. Can J Comp Med Vet Sci. 29, 125-128.

[18] Chriél M, Jensen TH, Hjulsager C, Larsen LE, Jørgensen PH, Le Févre Harslund J, Rangstrup- Christensen L, Petersen B, Hammer AS. Consequences of outbreaks of influenza A virus in farmed mink (Neovison vison) in Denmark in 2009 and 2010. Proc Xth Int Sci Congr Fur Anim Prod. 2012:186-189.

[19] National Farm Animal Care Council. Code of Practice for the Care and Handling of Farmed Mink. Rexdale, 2013. http://www.nfacc.ca/codes-of-practice/mink (accessed 10 Apr 2017).

[20] Pearl DL. Making the most of clustered data in laboratory animal research using multi-level models. ILAR J. 2015;55(3):486-492.

[21] Hildebrandt H. Bacterial Diseases of Mink. October 2014. http://www.merckvetmanual.com/mvm/exotic_and_laboratory_animals/mink/bacterial_diseases_of_mink .html (accessed 10 Apr 2017).

[22] Gorham JR, Evermann JF, Ward A, Pearson R, Shen D, Hartsough GR, Leathers C. Detection of coronavirua-like particles from mink with with catarrhal gastroenteritis. Can J Vet Res. 1990;54(3):383- 384.

[23] Wilson DJ, Baldwin TJ, Whitehouse CH, Hullinger G. Causes of mortality in farmed mink in the Intermountain West, North America. J Vet Diagn Invest. 2015;27(4):470-475.

[24] Liu H, Yang C, Jing Y, Li Z, Zhong W, Li G. Ability of lactic acid bacteria isolated from mink to remove cholesterol: in vitro and in vivo studies. Can J Microbiol. 2013;59(8):563-569.

[25] Fisher K, Phillips C. The ecology, epidemiology and virulence of Enterococcus. Microbiol. 2009;155:1749-1757.

[26] Vulfson L, Pendersen K, Chriel M, Andersen TH, Dietz HH. Assessment of the aerobic faecal microflora in mink (Mustela vison Schreiber) with emphasis on Escherichia coli and Staphylococcus intermedius. Vet Microbiol. 2003;93(3):235-245.

[27] Chadfield MS, Christensen JP, Juhl-Hansen J, Christensen H, Bisgaard M. Characterization of Enterococcus hirae outbreaks in broiler flocks demonstrating increased mortality because of septicemia and endocarditis and/or altered production parameters. Avian Dis. 2005;49(1):16-23.

[28] Morishita TY. Overview of enterococcosis in poultry. 2015. http://www.merckvetmanual.com/mvm/poultry/enterococcosis/overview_of_enterococcosis_in_poultry.ht ml (accessed 10 Apr 2017).

86 [29] Salomonsen CM, Boye M, Høiby N, Jensen TH, Hammer AS. Comparison of histological lesions in mink with acute hemorrhagic pneumonia associated with Pseudomonas aeruginosa or Escherichia coli. Can J Vet Res. 2013;77(3):199-204.

[30] Tibbetts RJ, White DG, Dyer NW, Giddings CW, Nolan L. Characterization of Escherichia coli isolates incriminated in colisepticaemia in mink. Vet Res Comm. 2003;27:341-357.

[31] Jørgensen M, Schetz F, Strandbygaard B. Eschericia coli and virus isolated from “sticky kits”. Acta Vet Scand. 1996;37(2):163-169.

[32] U.S. Centers for Disease Control. Multistate outbreak of multidrug-resistant Salmonella Heidelberg infections linked to foster farms brand chicken (Final Update). Centers for Disease Control, Salmonella 2013 Outbreaks. 2013. http://www.cdc.gov/salmonella/heidelberg-10-13/ (accessed 10 Apr 2017).

[33] Demczuk W, Soule G, Clark C, Ackerman HW, Easy R, Rasik K, Rodgers F, Ahmed R. Phage- based typing scheme for Salmonella Enterica serovar Heidelberg, a causative agent of food poisonings in Canada. J Clin Microbiol. 2003;41(9):4279–4284.

[34] Rosengren LB, Waldner CL, Reid-Smith RJ, Checkley SL, McFall ME, Rajíc A. Antimicrobial resistance of fecal Salmonella spp. isolated from all phases of pig production in 20 herds in Alberta and Saskatchewan. Can J Vet Res. 2008;72(2):151-159.

[35] Dietz HH, Chriél M, Andersen T, Jørgensen J, Torpdahl K, Pedersen H, Pedersen K. Outbreak of Salmonella Dublin-associated abortion in Danish fur farms. Can Vet J. 2006;47:1201-1205.

[36] Givens MD, Marley MSD. Infectious causes of embryonic and fetal mortality. Theriogenol. 2008;70:270-285.

[37] Morse EV, Duncan MA. Canine salmonellosis: prevalence, epizootiology, signs, and public health significance. J Am Vet Med Assoc. 1975;167(9):817-820.

[38] Reilly GA, Ballie NC, Morrow WT, McDowell SW, Ellis WA. Feline stillbirths associated with mixed Salmonella typhimurium and Leptospira infection. Vet Rec. 1994;135(25):608.

[39] Giannella RA Salmonella. Chapter 21: Salmonella, in: Baron S, (ed). Medical Microbiology. 4th ed. Galveston (TX): University of Texas Medical Branch at Galveston; 1996.

[40] Sledge DG, Danieu PK, Bolin CA, Bolin SR, Lim A, Anderson BC, Kiupel M. Outbreak of neonatal diarrhea in farmed mink kits (Mustella vison) associated with enterotoxigenic Staphylococcus delphini. Vet Pathol. 2010;47(4):751-7.

[41] Englund L, Chriél M, Hietz HH, Hedlund KO. Astrovirus epidemiologically linked to preweaning diarrheoa in mink. Vet Microbiol.2002;85:1-11.

[42] Have P, Moving V, Svansson V, Uttenthal A, Bloch B. Coronavirus infection in mink (Mustela vison). Serological evidence of infection with a coronavirus related to transmissible gastroenteritis virus and porcine epidemic diarrhea virus. Vet Microbiol. 1992;31:1-10.

87 [43] Guo M, Evermann JF, Saif LJ. Detection and molecular characterization of cultivable caliciviruses from clinically normal mink and enteric caliciviruses associated with diarrhea in mink. Arch Virol. 2001;146(3):479-93.

[44] Carmichael LE, Schlafer DH, Hashimoto A. Pathogenicity of minute virus of canines (MVC) for the canine fetus. Cornell Vet. 1991;81(2):151-171.

[45] Ronnse V, Verstegen J, Thiry E, Onclin K, Aeberle C, Brunet S. Canine herpesvirus-1 (CHV-1): clinical, serological and virological patterns in breeding colonies. Theriogenol. 2012;64:61-74.

[46] Anvik O. Clinical considerations of canine herpesvirus infection. Vet Med. 1991;86:394-403.

[47] Van Gucht S, Nauwynck SH, Pensaert M. Prevalence of canine herpesvirus in kennels and the possible association with fertility problems and neonatal death. Vlaams Dierg Tijds. 2001;70(3):204-211.

[48] Truyen U, Wolf G, Carmichael LE. The "other" parvovirus: first description of the minute virus of canines (Canine parvovirus Type 1) in Germany. Tierarztl Prax. 1996;24:503-511.

[49] Truyen U, Addie D, Belák S, Boucraut-Baralon C, Egberink H, Frymus T, Gruffydd-Jones TE, Hartman K, Hosie MJ, Lloret A, Lutz H, Marsilio F, Pennisi MG, Radford AD, Thiry E, Horzinek MC. Feline panleukopenia. ABCD guidelines on prevention and management. J Fel Med Surg. 2009;11(7):538-46.

[50] Kilham L, Margolis G, Colby ED. Congenital infections of cats and ferrets by feline panleukopenia virus manifested by cerebellar hypoplasia. Lab Invest. 1967;17:465-480.

[51] Krustritz MVR. Clinical management of pregnancy in cats. Theriogenol. 2006;66(1):145-150.

[52] Johnston SD, Raksil S. Fetal loss in the dog and cat. Vet Clin North Am. 1987;17:535-554.

[53] Sharp NJ, Davis BJ, Guy JS, Cullen JM, Steingold SF, Kornegay JN. Hydraencephaly and cerebellar hypoplasia in two kittens atrributed to intrauterine parvovirus infection. J Comp Pathol. 1999;121:39-53.

[54] Hok K. Development of clinical signs and occurrence of feline coronavirus antigen in naturally infected barrier reared cats and their offspring. Acta Vet Scand. 1993;34:345-356.

[55] Hok K. Morbidity, mortality and coronavirus antigen in previously coronavirus free kittens placed in two catteries with feline infectious peritonitis. Acta Vet Scand. 1993;34:203-210.

[56] Reichert M, Kostro K. Effect of persistent infection of mink with Aleutian mink disease virus on reproductive failure. Bull Vet Inst Pulawy. 2014;58:369-373.

[57] Macpherson LW. Feline enteritis virus - its transmission to mink under natural and experimental conditions. Can J Comp Med Vet Sci. 1956;20(6):197-202.

[58] Sutherland-Smith MR, Rideout BA, Mikolon AB, Appel MJ, Morris PJ, Shima AL, Janssen DJ. Vaccine-induced canine distemper in European mink, Mustela lutreola. J Zoo Wildl Med. 1997;28(3):312-8.

[59] Malmkvist J, Gase M, Damm BI. Parturient behaviour in farmed mink (Mustela vison) in relation to early kit mortality. Appl Anim Behav Sci. 2007;107:120-132.

[60] Rouvinen-Watt K, Harri M. Observations on thermoregulatory ontogeny of mink (Mustela vison). J Therm Biol. 2001;26:9-14.

[61] Harjunpaa S, Rouvinen-Watt K. The development of homeothermy in mink (Mustela vison). Comp Biochem Physiol, Part A. 2004;137:339-348.

[62] Tauson AH. Postnatal development in mink kits. Acta Agric Scand Sect A, Anim Sci. 1994;44:177- 184.

[63] Malmkvist J, Palme R. Periparturient nest building: Implications for parturition, kit survival, maternal stress and behaviour in farmed mink (Mustela vison). Appl Anim Behav Sci. 2008;114:270-283.

Table A.10 Prevalence of gross post mortem lesions in 5,657 preweaned, found dead mink kits from 21 commercial mink farms in Ontario. Overall Diagnosis 95% CI prevalence (%) Breath taken 55.4 54.1 - 56.7 Evidence of dystocia 1.0 0.7 - 1.3 Trauma 1.3 1.0 - 1.6 Food in gut 6.0 5.4- 6.6 Abnormal fluid 26.5 25.4 - 27.7 distribution Hydrothorax 21.5 20.4 - 22.5 Hydroperitoneum 16.1 15.2 - 17.1 Subcutaneous fluid 11.7 10.9 - 12.6 Anasarca 1.2 0.9 - 1.5 Congenital lesions 2.0 1.7 - 2.4 Cleft palate 0.9 0.7 - 1.2 Umbilical hernia 0.2 0.1 - 0.4 Other 0.9 0.6 - 1.1 Any abnormality 29.4 28.2 - 30.6

Table A.11 Bacterial culture results from 45 found dead kits with suspected enteritis

Bacteria cultured No. of samples (%; 95% CI)* Enterococcus faecalis 30 (67; 51-80) Escherichia coli 16 (36; 22-37) Enterococcus faecium 10 (22; 11-37)

Staphylococcus delphini 10 (22; 11-37) Enterococcus hirae 5 (11; 4-24) Salmonella enterica 5 (11; 4-24) serovar Heidelberg Enterococcus gallinarum 3 (7; 1-18) Enterococcus avium 3 (7; 1-18)

Streptococcus canis 1 (2; 0.1-12) Staphylococcus sciuri 1 (2; 0.1-12) *Percentages do not equal 100 as some samples had multiples species present. Only 45 samples yielded sufficient growth for inclusion.

Table A.12 Producer-reported production characteristics and incidence of mortality from 2013 (results from 15 of 21 participating farms). Parameter Median Range Mean SD 95%CI 2,900- 9,983.6- Total kits born 13,698 13,860 6,713.1 27,855 17,735.7 Day counted (post- 3 0-21 5.3 5.7 2.0-8.6 whelping) 2,700- Total kits weaned 11,250 11,995 6,181.7 8,425.6-15,564 26,000 Average age weaned 56 42-75 59.3 10.2 53.6-64.9 (days) Incidence of mortality 6.8 3.1-15.4 7.8 3.8 5.5-10.1 (%)

Table A.13 Producer-reported farm demographics from biosecurity survey (n=11)

Parameter Range Mean Median SD 95% CI Farm size (# breeding 700-6,900 3,066 3,400 2,022 1,708-4,425 females) # of years in operation 15-85 59.5 67 24.7 43.0-76.1 # of full-time employees 0-11 5.2 6 3.7 2.7-7.7 # of part-time summer 0-11 2.5 1 3.4 0.2-4.73 employees # of part-time employees at 0-20 5.6 3 5.6 1.3-8.9 pelting # of pelts produced (2013) 2,800-36,000 14,668 16,900 10,734 7,457-21,879 # of kits born (2013) 3,200-39,000 5,948 17,500 11,648 8,123-23,774

Table A.14 Specific biosecurity protocols implemented by mink producers (n=11)

No. of producers Specific biosecurity practice utilizing specific practice

Foot baths 2

Restricted access to shed 3 Restricted access to property 6 Personal Protective Equipment (PPE) 4 Controlled traffic (i.e. young to old, 1 clean to dirty) Separate work washrooms 3 Designated clothes for sheds 3 Hand washing policies 1 Ante rooms 0 Locks on shed doors 0

94 Table A.15 Summary of responses for biosecurity-specific questions on survey (n=11). Unless otherwise specified, questions pertain to the respondent’s specific on-farm practices. Exact Exact Survey Question Y(%) Survey Question Y(%) 95% Survey Question Answer 95% CI CI In your opinion, does 9/11 Restricted access to storage 3/11 0-1: 2-4: CMBA or OFBA discuss 48.2-97.7 6.0-61.0 # visitors to sheds per week (81.8) manure? (27.3) 10/11 1/11 biosecurity at meetings? In your opinion, does Employees trained to 5/11 7/11 Guests wear special No: Boots: Boot covers: Coveralls: industry have biosecurity 16.7-76.6 recognize disease or 30.8-89.1 (45.5) (63.6) clothing? 7/11 2/4 1/4 1/4 standards? sickness? 8/11 Feed containers washed and 8/11 Storage of mink carcasses None: Freezer: Solid container: Compost: Controlled access to sheds 39.0-94.0 39.0-94.0 (72.7) disinfected after use? (72.7) until disposal 2/11 3/8 2/8 2/8 5/11 Any integrated fly control 7/11 Change storage between Y: Controlled access to farm 16.7-76.6 30.8-89.1 (45.5) programs used? (63.6) summer & winter? 3/11 Companion animals have 9/11 How often dead animals Daily: 2-4x/wk: Biweekly: N/A: Employees wash hands 48.2-97.7 access to sheds? (81.8) collected? 6/11 2/11 1/11 1 0/11 Dedicated clothing for work 5/11 How are dead mink Composting: Incineration: Both: Before entering sheds 0-28.5* 16.7-76.6 (0) in sheds (45.5) disposed of? 9/11 1/11 1/11 0/11 5/11 New mink brought on to Y: After leaving sheds 0-28.5* Dedicated farm shoes 16.7-76.6 (0) (45.5) farm? 7/11 0/11 Farm clothing washed 7/11 No: 2 wks: 1 mo: Before handling mink 0-28.5* 30.8-89.1 If so, quarantine? (0) separately (63.6) 2/7 1/7 4/7 5/11 Rodent control program on 8/11 New stock separate from No: Separate shed: Same shed, After handling mink 2.3-51.8 39.0-94.0 (45.5) farm? (72.7) main herd? 2/7 3/7 different row: 2/7 11/11 Natural: Chimney: Visitors wash hands Birds have access to sheds? 71.5-1* Type of ventilation (100) 8/11 3/11 0/11 11/11 No: W/ water: W/ disinfectant: Before entering sheds 0-28.5* Wood used in pens? 71.5-1* Kit pens cleaned (0) (100) 3/11 7/11 1/11 1/11 4/11 No: W/ water: W/ disinfectant: After leaving sheds 16.7-76.6 Mortality log kept? 10.9-69.2 Female pens cleaned (9.1) (36.4) 4/11 5/11 2/11 0/11 Borrow males from other 0/11 No: W/ water: W/ disinfectant: Before handling mink 0-28.5* 0-28.5* Kit pens cleaned (0) farms? (0) 6/11 4/11 1/11 2/11 6/11 Frequency waterlines are Sporadically: Yearly: Rarely: Weekly: After handling mink 2.3-51.8 Farm policies on traffic? 23.4-83.3 (18.1) (54.5) cleaned 1/11 3/11 5/11 2/11 Distemper: MEV**: Pseudomonas: Botulism: Vaccinations given? 11/1 10/11 10/11 10/11 aCanada Mink Breeders Association bOntario Fur Breeders Association c One-sided, 97.5% confidence interval d Mink enteritis virus

95 Table A.16 Summary of responses for biosecurity-specific questions on survey with yes/no answers (a) or multiple answers (b) (n=11). Unless otherwise specified, questions pertain to the respondent’s specific on-farm practices. a)

Survey Question Yes (%) Exact 95% CI

In your opinion, does CMBAa or OFBAb discuss 9/11 (81.8) 48.2-97.7 biosecurity at meetings? In your opinion, does industry have biosecurity 5/11 (45.5) 16.7-76.6 standards?

Controlled access to sheds? 8/11 (72.7) 39.0-94.0

Controlled access to farm? 5/11 (45.5) 16.7-76.6

Employees wash hands

Before entering sheds 0/11 (0) 0-28.5 c

After leaving sheds 0/11 (0) 0-28.5 c

Before handling mink 0/11 (0) 0-28.5 c

After handling mink 5/11 (45.5) 2.3-51.8

Visitors wash hands

Before entering sheds 0/11 (0) 0-28.5 c

After leaving sheds 1/11 (9.1) 16.7-76.6

Before handling mink 0/11 (0) 0-28.5 c

After handling mink 2/11 (18.1) 2.3-51.8

Restricted access to storage manure? 3/11 (27.3) 6.0-61.0

Employees trained to recognize disease or 7/11 (63.6) 30.8-89.1 sickness?

Feed containers washed and disinfected after use? 8/11 (72.7) 39.0-94.0

Any integrated fly control programs used? 7/11 (63.6) 30.8-89.1

Companion animals have access to sheds? 9/11 (81.8) 48.2-97.7

Dedicated clothing for work in sheds? 5/11 (45.5) 16.7-76.6

Dedicated farm shoes? 5/11 (45.5) 16.7-76.6

Farm clothing washed separately? 7/11 (63.6) 30.8-89.1

Rodent control program on farm? 8/11 (72.7) 39.0-94.0

Birds have access to sheds? 11/11 (100) 71.5-1c

96 Wood used in pens? 11/11 (100) 71.5-1 c

Mortality log kept? 4/11 (36.4) 10.9-69.2

Borrow males from other farms? 0/11 (0) 0-28.5 c

Farm policies on traffic? 6/11 (54.5) 23.4-83.3

Mortality log kept? 4/11 (36.4) 10.9-69.2 b)

Survey Question Response

No. of visitors to sheds per week 0-1: 10/11 2-4: 1/11

Boot covers: Guests wear special clothing? No: 7/11 Boots: 2/4 Coveralls: 1/4 1/4 Solid container: Storage of mink carcasses until disposal None: 2/11 Freezer: 3/9 Compost: 3/9 3/9

Change storage between summer & winter? Y: 3/11

How often dead animals collected? Daily: 6/11 2-4x/wk: 2/11 Biweekly: 1/11 N/A:1

Composting: Incineration: How are dead mink disposed of? Both: 1/11 9/11 1/11

New mink brought on to farm? Y: 7/11

If so, quarantine? No: 2/7 2 weeks: 1/7 1 month: 4/7

Same shed, Separate shed: New stock separate from main herd? No: 2/7 different row: 3/7 2/7

Type of ventilation Natural: 8/11 Chimney: 3/11

W/ Kit pens cleaned No: 3/11 W/ water: 7/11 disinfectant: 1/11 W/ Female pens cleaned No: 4/11 W/ water: 5/11 disinfectant: 2/11 W/ Kit pens cleaned No: 6/11 W/ water: 4/11 disinfectant: 1/11 Sporadically: Frequency waterlines are cleaned Yearly: 3/11 Rarely: 5/11 Weekly 2/11 1/11 Distemper: Pseudomonas: Botulism: Vaccinations given? MEVd: 10/11 11/11 10/11 10/11 aCanada Mink Breeders Association bOntario Fur Breeders Association c One-sided, 97.5% confidence interval d Mink Enteritis Virus

97 Figure A.1: Weight distribution of preweaned, found dead mink kits at the time of post mortem examination from 21 fur farms in Ontario (n=5,659)

98 Figure A.2: Producer-reported common health problems as answered on biosecurity survey (n=11). Percentages represent the number of producers identifying the condition as a problem.

a) Adult females

b) Preweaned kits

99 APPENDIX B: FECAL MICROBIOTA COMPOSITION OF COMMERCIAL MINK IN ONTARIO AT THE FARM LEVEL

B. 1 Relative abundance of 20 most predominant bacterial genera present in the feces of commercial mink, by farm (n=49)

100 APPENDIX C: COMMUNITY MEMBERSHIP AND STRUCTURE OF THE FECAL MICROBIOTA OF COMMERCIAL MINK a)

Figure C. 1: Dendrogram representing the community membership (a) and structure (b) of the fecal microbiota of commercial mink, based on the Jaccard Index and Yue and Clayton Index, respectively (red=2014, green=2015, 2016=blue). 101