UNIVERSITY OF CALGARY
Blood on Filter Paper for Monitoring Caribou Health:
Efficacy, Community-Based Collection, and Disease Ecology in Circumpolar Herds
by
Patricia Sale Curry
A THESIS
SUBMITTED TO THE FACULTY OF GRADUATE STUDIES
IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
DEPARTMENT OF ECOSYSTEM & PUBLIC HEALTH
CALGARY, ALBERTA
September, 2012
© Patricia Sale Curry 2012
1
Abstract
Various pathogens of caribou and reindeer (Rangifer tarandus ssp.) have been detected
and isolated. Some are important from a human food-safety perspective, but relatively
little is known about the health impacts of these agents in Rangifer. This project sought to
contribute to the serological detection and understanding of caribou pathogens by
examining blood collection on filter paper (FP) as a tool for health monitoring. Filterpaper test performance was evaluated by comparing FP results to matched serum as the
relative standard. This was done in nine serological assays (enzyme-linked
immunosorbent assays and virus neutralization) for eight pathogens/groups: Brucella
spp., Neospora caninum, West Nile virus (WNV), bovine herpesvirus-1 (BHV-1),
parainfluenza virus type 3 (PI-3), bovine respiratory syncytial virus (BRSV), and bovine
viral diarrhea virus types I and II. Filter-paper performance in these tests was also
assessed under different storage and processing conditions. Performance was comparable
to serum (FP sensitivity and specificity ≥80%) within 2 months of collection and, in most
assays, after 1 year. Trials demonstrated that FP samples frozen directly upon collection
performed comparably to matched serum, and that FP samples failed in the fluorescence
polarization assay for Brucella. A serosurvey conducted on 550 animals from seven
migratory caribou herds of arctic North America and Greenland revealed relatively high
overall seroprevalence for Pestivirus (28%), BHV-1 (25%), and PI-3 (7%), and lowerthan-expected prevalence for Brucella and Toxoplasma gondii. No animals tested positive
for WNV or BRSV. In the two Greenland herds, only two animals were positive for one
pathogen (BHV-1). Filter-paper blood sampling was implemented in hunter-based
2
wildlife health-monitoring programs in arctic communities of Canada. Interviews with
harvesters revealed strong interest in wildlife disease, moderate interest in hunter
sampling, and hunter acceptance of the FP method. Findings also identified potential
cultural barriers to hunter sampling of wildlife, and challenges regarding program impact
and sustainability. This research validates a diagnostic field tool for multiple serological
assays in Rangifer, and adds new knowledge about pathogen exposure in caribou, and
about community engagement in research. It is relevant to other wildlife species, other
hunter-based monitoring programs, and other community-based initiatives in the North
and potentially worldwide.
3
Preface
Chapter 2 is a published scientific paper and Chapters 3 and 4 are journal articles in
submission. All three texts are formatted in the journal’s specified style requirements. As
lead author, Patricia Curry designed and conducted the research with guidance from her
committee, wrote these manuscripts, created all tables and figures, and integrated
comments and critique from coauthors. The complete citations for these articles are as
follows:
CURRY, P. S., B. T. ELKIN, M. CAMPBELL, K. NIELSEN, W. HUTCHINS, C.
RIBBLE, AND S. J. KUTZ. 2011. Filter-paper blood samples for ELISA detection of
Brucella antibodies in caribou. Journal of Wildlife Diseases 47 (1): 12-20.
CURRY, P. S., C. RIBBLE, W. C. SEARS, W. HUTCHINS, K. ORSEL, D. GODSON,
R. LINDSAY, A. DIBERNARDO, AND S. J. KUTZ. 2012. Filter-paper blood samples
for detecting antibodies to Neospora caninum, West Nile virus, and five bovine viruses in
Rangifer. Journal of Wildlife Diseases (in submission).
CURRY P. S., C. RIBBLE, W. C. SEARS, K. ORSEL, W. HUTCHINS, D. GODSON,
R. LINDSAY, A. DIBERNARDO, M. CAMPBELL, AND S. J. KUTZ. 2012. Filterpaper blood samples for wildlife serology: Evaluating storage and temperature challenges
of field collections. Journal of Wildlife Diseases (in submission).
4
Acknowledgements
Through the last weeks of writing this thesis at Stoney Lake in Ontario, I’ve been kept
company in the wee hours by wolves howling and loons calling, and, as always, by my
ever-steady arctic canine, Grizz (she’s known as “Dog” to some). The significance of
these presences and their ability to keeping me going, on track in the world, is a familiar
theme. The power of animals and the impacts of their connections with people are the
true essence of this research for me. I undertook this project in a way to acknowledge this
core importance in my life, and also to honour my own animistic beliefs. I wish to
recognize the caribou and reindeer who contributed to this research, the humans who
cared for some of them (Marianne Jorgensen especially), and all the people in the North
and elsewhere who revere caribou, other wildlife, and the animal world in general. I also
want to thank the many incredible and cherished four-legged companions who have
shared life with me in past and present. They have kept me grounded through this project
and other times, and have taught me so subtly and quietly about what is truly important
and how to live. I’ll continue to try to do better.
I have many, many humans to thank for their support, contributions, and friendship
throughout the past 5 years. First, deep gratitude goes to my supervisor, Susan Kutz, for
all that she has passed on to me and worked through with me during this project. This
experience has been full of challenges personal and professional, and valuable learning on
every level. In particular, I thank Susan for her enthusiasm, her endless big-picture
trajectories of Kutz Research Concepts and no-holds-barred plans, her guidance and faith
in me when my own faith sputtered at times, and her blatant and huge caring about the
5
Arctic, its animals, and its people. Thanks for making the efforts to honour our
differences, Susan. The fascinating travels, the hilarious times, the tough ones, a few tears
even; it has all been an unforgettable slice of life and much inspiration. Thank you.
My other supervisory committee members, Carl Ribble, Brett Elkin, and Wendy
Hutchins, have also been truly stellar. Each has contributed sage guidance and a steady
hand when needed, and doubtless Carl deserves some kind of medal for his mantra of
“putting a box around” things. These few but incisive words (typical of Carl) stated in
Committee Meeting #1 helped steer this project repeatedly throughout its duration, and in
my book they’re just good plain advice for life! Thank you so much Brett, Wendy, and
Carl for coming to the plate huge every single time I put out a call, and for mixing in your
expertise enthusiastically with the work I’ve been doing. I have been wildly fortunate to
have you three at my back. I wish every graduate student could be so lucky.
The Kutz Lab crew, hangers-on, and extended “lab diaspora” of amazing, wonderful
students, lab assistants, post docs, and other UofC friends have been beyond measure in
terms of their unending cheer, on-the-ground assistance, international dimension to
student life, and just generally their “good positive people” ways of being. I wish each
and every one all the wellbeing and success my heart and soul can muster. Each has
taught me something very valuable through this process, and that is everlasting. I just
wish I could attach a GPS device to each of these individuals and watch them go.
This was a massively collaborative project that involved a raft of agencies, diagnostic and
research laboratories, and wonderful, dedicated individuals. Each is acknowledged
6
chapter by chapter and I thank them all. Certain folks in the northern world, namely,
Debbie Jenkins, Grigor Hope, Alasdair Veitch, and George Koonoo, provided very
special support and assistance, and this project could not have unfolded as described in
this thesis without them. It is important to me to write their names large here. I also wish
to acknowledge all the northern communities with voices in this research, and specifically
the harvesters and others who generously gave their time to speak with me in interviews.
It was personally and professionally illuminating to sit with each of these individuals,
hear their words, and hopefully better understand human-animal connections in the North
and hunter-based wildlife monitoring there. I would also like to particularly mention the
Nasivvik Centre for Inuit Health and Changing Environments (Canadian Institutes of
Health Research), the Northern Scientific Training Program, and the Arctic Institute of
North America. The continuous funding support, some very special personal honours
received, and the general sense of belief and interest in this project that were extended
from these organizations over the years meant a tremendous amount to me. To all the
difference-makers in this project: I will never forget the displays of support—small and
large—that came my way. I am very grateful.
Within the UCVM faculty world, I am mindful of Karin Orsel, Susan Cork, and many
others who consistently voiced their interest in my research and who supported this work
and supported me, personally, with many kindnesses. These have meant much more than
they will ever know. I would also like to recognize Veterinary Medicine’s Librarian and
research resource expert, Lorraine Toews. Lorraine is one of the most dedicated people I
know at UCVM, and I thank her for teaching me some of her craft. These researching
7
skills are a gold mine for any student and will, no doubt, be part of my future career as
well.
It is hard to find ways to aptly express appreciation to all the others—my brothers Ian and
Peter, our friend Don, many precious family members and personal friends near and far—
who have added to this project. Their contributions were huge: basic belief in me; a tiny,
divine alpine cabin (the ideal mental and physical space) in which to write this thesis;
rejuvenating hikes, chats, and laughs on forest paths; special dinners; encouragement over
the phone, etc. I suspect that many of these individuals likely watched my return to
university at 45 years of age with silent trepidation (horror?) in relation to “career
trajectory.” Most will never know how much I’ve appreciated the laughs, the cooking (!),
and the tiniest of things they have added for me during this epic challenge. I thank them
all.
Finally, some words for three iconic people in my life. First, my mother, Elizabeth Curry,
a beauty in every way who stands by regardless of any new, potentially crazy turn I take
in life. Mum, thank you for listening to me and sticking with me during the ups and
downs of this experience, for extending your limitless positive outlook, generosity,
caring, and wise advice. You inspire me. You have no idea how much I appreciate you,
admire you. It is a lifelong endeavour to try to emulate you in my work and personal life.
I hope you feel good just knowing I have done this work (there is no need to try and read
it!).
8
Helen Stanbury, my grandmother and general life hero, is always on one of my virtual
shoulders and was there cheering me on throughout this PhD. I continue to envision
Gran’s example and she inspires so many aspects of work and day-to-day living, and
interactions with others. Most of all, Gran instilled in me wonder about nature and about
the small, exquisitely beautiful things in life. Her ever-positive approach to life helped me
through some crucial times in these past few years, and it is a treat to acknowledge Gran
here. My heart still misses her.
On my other shoulder during this entire research experience has been my father, Hugh
Curry. He is not here to read these pages either, but I thank him for all the good things he
modeled for me and that led me to complete this research (hopefully) well. Some of the
most outstanding of these were my father’s work ethic, his eclectic interest in the arts as
well as medicine and science, his decency and drive for personal best, and his high
ethical/moral standards. Dad, I am almost speechless/wordless. I wish so much that you
could have been here to witness this research process alongside me, vicariously visit the
Arctic with me, and discuss the iterations of this work as it evolved. I know you would
have loved to have been part of it. You were. I have always felt your presence during this,
Dad. It is partly for you (us), too.
9
Dedication
To the animals.
To all Northerners who are struggling.
To Mum, to Dad.
To the Seven Generations.
10
Table of Contents
Abstract ............................................................................................................................... 1
Preface................................................................................................................................. 3
Acknowledgements............................................................................................................. 4
Dedication ........................................................................................................................... 9
Table of Contents.............................................................................................................. 10
List of Tables .................................................................................................................... 14
List of Figures ................................................................................................................... 15
CHAPTER ONE  GENERAL INTRODUCTION.................................................. 16
Background ................................................................................................................... 17
Wildlife Disease: Context in the Year 2012............................................................ 17
Wildlife Disease Surveillance: Status and Considerations...................................... 19
Project Rationale and Problem Defined........................................................................ 22
Pathogens in Caribou: Detecting and Monitoring ................................................... 22
Caribou Health Monitoring and A Community-Based Diagnostic Method............ 24
Overview of Aims and Chapters................................................................................... 24
Chapters: A Brief Scan ............................................................................................ 27
CHAPTER TWO  FILTER-PAPER BLOOD SAMPLES FOR ELISA DETECTION OF BRUCELLA ANTIBODIES IN CARIBOU ......................... 29
Abstract ......................................................................................................................... 30
Introduction................................................................................................................... 31
Materials and Methods.................................................................................................. 33
Samples and Processing .......................................................................................... 33
Filter-Paper Elution ................................................................................................. 35
Immunoassays ......................................................................................................... 35
Statistical Analysis .................................................................................................. 37
Results........................................................................................................................... 37
Discussion..................................................................................................................... 38
Brucella Antibodies: Filter Paper vs. Serum ........................................................... 39
Variability Among Filter Papers From Individuals................................................. 40
Test Choice .............................................................................................................. 40
Benefits of Filter-Paper Sampling ........................................................................... 41
Acknowledgements....................................................................................................... 42
TABLES ....................................................................................................................... 44
FIGURES...................................................................................................................... 45
CHAPTER THREE  FILTER-PAPER BLOOD SAMPLES FOR DETECTING ANTIBODIES TO NEOSPORA CANINUM, WEST NILE VIRUS, AND FIVE BOVINE VIRUSES IN RANGIFER.................................. 50
Abstract ......................................................................................................................... 51
Introduction................................................................................................................... 52
11
Materials and Methods.................................................................................................. 53
Animals and Study Design ...................................................................................... 53
Sample Collection and Processing .......................................................................... 55
Testing ..................................................................................................................... 55
Competitive ELISAs......................................................................................... 56
Indirect ELISAs ................................................................................................ 58
Virus Neutralization Tests ................................................................................ 58
Analysis ................................................................................................................... 59
Test Performance .............................................................................................. 59
Other Evaluations ............................................................................................. 61
Results........................................................................................................................... 61
Discussion..................................................................................................................... 62
Filter-Paper Test Performance................................................................................. 64
Competitive ELISAs......................................................................................... 65
Indirect ELISAs ................................................................................................ 66
Virus Neutralization.......................................................................................... 67
Considerations with Multiple-Measures Data ......................................................... 68
Benefits and Application ......................................................................................... 68
Acknowledgements....................................................................................................... 70
TABLES ....................................................................................................................... 71
FIGURES...................................................................................................................... 76
CHAPTER FOUR  FILTER-PAPER BLOOD SAMPLES FOR WILDLIFE SEROLOGY: EVALUATING STORAGE AND TEMPERATURE CHALLENGES OF FIELD COLLECTIONS.................................................... 81
Abstract ......................................................................................................................... 82
Introduction................................................................................................................... 83
Materials and Methods.................................................................................................. 84
Design, Animals, and Antibody Status.................................................................... 84
Samples.................................................................................................................... 86
Processing and Treatments ...................................................................................... 86
Testing ..................................................................................................................... 88
Analysis ................................................................................................................... 89
Storage Time..................................................................................................... 89
Processing/Storage Conditions ......................................................................... 90
Results........................................................................................................................... 91
Part I – Storage Time............................................................................................... 91
Part II – Processing/Storage Conditions.................................................................. 92
Discussion..................................................................................................................... 92
Storage Time ........................................................................................................... 92
Processing/Storage Conditions................................................................................ 96
Other Implications and Recommendations ............................................................. 97
Acknowledgements....................................................................................................... 99
TABLES ..................................................................................................................... 100
FIGURES.................................................................................................................... 111
12
CHAPTER FIVE  SEROLOGICAL SURVEY FOR EIGHT PATHOGENS IN MIGRATORY CARIBOU HERDS OF NORTH AMERICA AND GREENLAND: SNAPSHOT DURING INTERNATIONAL POLAR YEAR 2007-2009.................................................................................................. 117
Abstract ....................................................................................................................... 118
Introduction................................................................................................................. 119
Material and Methods ................................................................................................. 125
Blood Samples....................................................................................................... 126
Testing ................................................................................................................... 127
Analysis ................................................................................................................. 130
Results......................................................................................................................... 131
Cross-Herd Findings: BHV-1................................................................................ 133
Three-Way Stratification ....................................................................................... 135
Cross-Herd Findings: PI-3..................................................................................... 136
Cross-Herd Findings: Pestivirus............................................................................ 136
Correlation ............................................................................................................. 137
Discussion................................................................................................................... 138
Overall Prevalence Findings.................................................................................. 138
Novel Biodiversity: Potential Research Directions ............................................... 144
Geographic Differences......................................................................................... 145
Sex, Age, Season, Pregnancy ................................................................................ 149
Herd / Ecological Factors ...................................................................................... 153
Surveillance Methodologies .................................................................................. 153
Synopsis and Conclusion....................................................................................... 155
Acknowledgements..................................................................................................... 155
TABLES ..................................................................................................................... 158
FIGURES.................................................................................................................... 167
CHAPTER SIX  HUNTER-BASED WILDLIFE SAMPLING AND FILTERPAPER BLOOD COLLECTION: PERCEPTIONS OF CARIBOU HARVESTERS FROM ACROSS NORTHERN CANADA............................ 168
Abstract ....................................................................................................................... 169
Introduction................................................................................................................. 170
Study Context: Hunter-Based Caribou Sampling Programs ................................. 173
Program in the North Baffin Region............................................................... 173
Program in the Sahtu Settlement Region........................................................ 175
Filter-Paper Blood Collection: A Tool for Hunter-Based Sampling..................... 178
Research Aim ........................................................................................................ 179
Methodology............................................................................................................... 180
Sampling: Participants and Recruitment ............................................................... 180
Researcher Information, Data Collection .............................................................. 181
Analysis ................................................................................................................. 184
Results......................................................................................................................... 185
1. Importance of Wildlife Disease......................................................................... 186
• Food: Safety, Taste, and Human Health ................................................... 186
13
• Science, Training, Disease: Ability to Detect and Monitor ...................... 188
• Humaneness and Disease Concerns for Wildlife at Large........................ 188
2. Importance of and Interest in Hunter-Based Wildlife Sampling....................... 189
• Food Safety, Human Health, and Science/Training.................................. 189
• Hunter-Based Sampling, Researchers, and Communities ........................ 190
• Other Issues and Pitfalls............................................................................ 191
3. Concerns about Filter-Paper Blood Collection.................................................. 193
Discussion................................................................................................................... 194
Emergent Themes .................................................................................................. 194
Basic Interest and Motivation................................................................................ 195
Results Reporting, Communication in General..................................................... 196
Engagement, Critiques, and Context ..................................................................... 201
Comparisons to Other Community-Based Initiatives............................................ 203
Research Fatigue and Capacity – Others’ Experience.................................... 205
Synopsis of Findings ............................................................................................. 207
Conducting the Study: Flexibility, Biases, Lessons Learned ................................ 208
Big-Picture Applications, Importance ................................................................... 211
Acknowledgements..................................................................................................... 211
TABLES ..................................................................................................................... 213
FIGURES.................................................................................................................... 218
CHAPTER SEVEN  WHERE DO WE GO FROM HERE? KNOWLEDGE GAINED AND FUTURE DIRECTIONS .............................. 225
Knitting It All Together .............................................................................................. 226
PHASE I: Efficacy of Rangifer Filter-Paper Samples for Serology ..................... 226
Future Directions ............................................................................................ 229
PHASE II: Serological Snapshot in Time: Eight Pathogens, Seven Migratory Herds ............................................................................................................. 231
Future Directions ............................................................................................ 234
PHASE III: Filter Papers and Community-Based Wildlife Monitoring ............... 235
Reflection and Learning.................................................................................. 235
Future Directions ............................................................................................ 237
Final Words................................................................................................................. 238
BIBLIOGRAPHY..................................................................................................... 241
APPENDIX 4A........................................................................................................... 280
APPENDIX 6A........................................................................................................... 283
APPENDIX 6B ........................................................................................................... 284
APPENDIX 6C ........................................................................................................... 292
APPENDIX 6D........................................................................................................... 300
APPENDIX 6E ........................................................................................................... 301
APPENDIX 6F ........................................................................................................... 305
14
List of Tables
Chpt 2. Table 1. Filter paper sample performance – Brucella assays............................... 44
Chpt 3. Table 1. Pathogen effects and study summary..................................................... 71
Chpt 3. Table 2. Filter-paper test performance ................................................................. 72
Chpt 3. Table 3. Laboratory and adjusted thresholds – PI-3, BRSV assays..................... 75
Chpt 4. Table 1. Study design......................................................................................... 100
Chpt 4. Table 2. Filter-paper performance vs. T1 serum ................................................ 102
Chpt 4. Table 3. Filter-paper performance vs. same-storage serum ............................... 105
Chpt 4. Table 4. Filter-paper performance – Processing/storage, WNV assay .............. 109
Chpt 4. Table 5. Recommendations for filter-paper use – Wildlife................................ 110
Chpt 5. Table 1. Pathogens and their effects .................................................................. 158
Chpt 5. Table 2. Herd parameters and collection dates .................................................. 159
Chpt 5. Table 3. Prevalence estimates by herd and overall ............................................ 160
Chpt 5. Table 4. Proportions of filter papers and serum tested....................................... 161
Chpt 5. Table 5. Results of overall one-way analyses .................................................... 163
Chpt 5. Table 6. Comparisons of risk across regions ..................................................... 164
Chpt 5. Table 7. Results of three-way stratification by region ....................................... 165
Chpt 6. Table 1. Interview locations............................................................................... 213
Chpt 6. Table 2. Background information – Interviewees.............................................. 214
Chpt 6. Table 3. Demographic characteristics – Interviewees........................................ 215
Chpt 6. Table 4. Interview questions relevant to the study............................................. 216
15
List of Figures
Chpt 2. Figure 1. Filter-paper blood collection................................................................. 45
Chpt 2. Figure 2. Filter-paper sets and drying racks......................................................... 46
Chpt 2. Figure 3. Filter paper vs. serum – Brucella cELISA............................................ 47
Chpt 2. Figure 4. Results for sample duplicates – Brucella cELISA ............................... 48
Chpt 2. Figure 5. Results for sample duplicates – Brucella iELISA ................................ 49
Chpt 3. Figure 1. Filter-paper set ...................................................................................... 76
Chpt 3. Figure 2. a-f. Filter paper vs. serum – ELISAs .................................................... 77
Chpt 3. Figure 3. a-e. Duplicate filter paper results – ELISAs......................................... 79
Chpt 4. Figure 1. a-f. Effects of storage time – ELISAs................................................. 111
Chpt 4. Figure 2. a, b. Effects of processing/storage regimes – PI-3, BRSV assays...... 115
Chpt 5. Figure 1. Herds surveyed ................................................................................... 167
Chpt 6. Figure 1. Example of wildlife harvesting cycles – Nunavut.............................. 218
Chpt 6. Figure 2. North Baffin Region, Nunavut ........................................................... 219
Chpt 6. Figure 3. Sahtu Settlement Region, Northwest Territories ................................ 220
Chpt 6. Figure 4. Hunter meeting and training – CHMP................................................ 221
Chpt 6. Figure 5. Hunter training and sample set – WHMP........................................... 222
Chpt 6. Figure 6. Schematic – Latent content analysis................................................... 223
16
CHAPTER ONE
GENERAL INTRODUCTION
17
Background
Wildlife Disease: Context in the Year 2012
Wildlife pathogens can be as enigmatic and compelling as their wild-animal hosts, and
there are many reasons for this. The past few decades have seen a significant ramping of
interest in the realm of wildlife disease, with curiosity about causal agents, etiologies, and
outcomes being fuelled and expanded by ecosystem events of global scale. Rapid and
accelerating species extinction, habitat depletion and fragmentation from industrial
development and land-use shifts, fluid and frequent human and animal transport globally,
rapid significant climate change, and various other anthropogenic perturbations are all
impacting ecosystems and contributing to unprecedented disease emergence (Daszak et
al., 2000, 2001; Aguirre and Tabor, 2008; Keesing et al., 2010). The complex factors that
trigger emerging infectious diseases (EIDs) have captured the attention of scientists in
many fields and have raised concerns for health organizations nationally and worldwide
(Kuiken et al., 2005; Merianos, 2007; Jones et al., 2008). Disease emergence and reemergence incurs significant socioeconomic burdens not only in the human-health realm,
but also through mass culling of domestic livestock and other linked events (Morens et
al., 2004; Jones et al., 2008). In the wild, some of the most acute concerns centre on
biodiversity loss, as research has shown that disease outbreaks can seriously threaten
already-endangered species (Smith et al., 2009) and may menace species with limited
genetic variation (Altizer et al., 2003), which can result from ecosystem perturbations.
Within the past decade, public-health alarms have sounded globally in response to several
EIDs that are zoonoses (defined here as agents transmitted to people from animals),
18
including outbreaks of severe acute respiratory syndrome (SARS) and highly pathogenic
avian influenza. Concern about wildlife pathogens has also been heightened by scientific
evidence that approximately 60% of EID events are zoonotic and most of these (>70%)
originate from wild species (Jones et al., 2008).
Today, health and social costs, public fears, and science interests related to wildlife
pathogens and their links with human and ecosystem health are placing wildlife disease
under the microscope more than ever before. One positive effect of the “meeting of the
sciences” on EIDs has been inspiration: different thinking about and understanding of
wildlife pathogens and their effects within conservation medicine, which is a new
ecosystem-and-public-health concept and approach (Aguirre et al., 2002; Daszak et al.,
2004). Another overall effect has been a wave of urgency to close the knowledge and
regulatory gaps that crises such as SARS, avian influenza, and others have exposed.
Regarding surveillance, member countries of the World Organization for Animal Health
(OIE) report a list of pathogens that affect international trade, and there are also some
international early-warning and monitoring systems that focus mainly on animal
pathogens with economic or public health relevance (Kuiken et al., 2005). However, the
quality of animal surveillance varies significantly among countries, and there is a stark
lack of (or inferior levels of) disease surveillance for wildlife compared to that for
humans and domestic livestock (Daszak et al., 2000; Kuiken et al., 2005).
19
Wildlife Disease Surveillance: Status and Considerations
Some worldwide organizations and individual countries have systems in place for
monitoring pathogens and outbreaks in free-ranging wildlife, and examples include the
Internet-based Program for Monitoring Emerging Diseases [Pro-Med] and Canada’s
Wildlife Disease Strategy (CNWDS, 2004). At broad scale, however, authors have
identified politics, funding, and infrastructure as important barriers to global wildlife
disease surveillance (Kuiken et al., 2005; Butler, 2006; Merianos, 2007). Nordic
countries were some of the first to establish wildlife disease surveillance programs in the
1930s and 1940s, and these schemes and several others like them are operating in other
parts of Europe and in North America today (Mörner et al., 2002). Most of these
programs are based on volunteer collections and carcass submissions, and they feature
varying degrees of national coordination (Mörner et al., 2002). In general, the existing
systems are relatively few and limited in scope, and they tend to focus on protecting
domestic-animal health and provide almost no coverage of developing nations, where key
risk factors for EIDs prevail (Daszak et al., 2000, 2007; Mörner et al., 2002; Butler, 2006;
Merianos, 2007; Jones et al., 2008). Another fundamental issue is that pathogen
surveillance systems for humans, domestic animals, and wildlife are not functionally
linked; therefore, cases can fall through various cracks in regulatory coverage (Kuiken et
al., 2005).
Given the risks that EIDs pose to ecosystem and public health, and considering the
anthropogenic environmental degradation that is triggering disease emergence in wildlife,
there is concern that disease surveillance in animals—particularly wildlife—is
20
insufficient (Daszak et al., 2001; Kuiken et al. 2005; Merianos, 2007). At global
ecohealth levels, scientists and health experts are calling for “One Health”
interdisciplinary cross-sector approaches to wildlife disease surveillance, concerted action
by world stakeholder organizations, funding for infrastructure and diagnostic networks,
and inclusion of critical local expertise (Daszak et al. 2000, 2004, 2007; Mörner et al.,
2002; Kuiken et al. 2005; Butler, 2006; Merianos, 2007). Within science disciplines
focused on wildlife health, ecology, management, and conservation, there is recognition
that much remains to be learned about wildlife pathogens. Disease is one of many
ecological factors that affect wildlife, and it cannot be considered in isolation (Wobeser,
2006). Studying free-ranging populations has inherent challenges, and one of the most
basic knowledge gaps is detection and identification of infectious agents that are
circulating in host populations. Other gaps in our current understanding are pathogen
transmission dynamics, underlying drivers of disease emergence (including
anthropogenic factors and interactions of these with ecological factors in natural systems
[Stallknecht, 2007]), host-pathogen interactions, risks associated with species
translocations, and impacts of disease on individuals and populations (Williams and
Barker, 2001; Leighton, 2002; Altizer et al., 2003; Daszak et al., 2007; Smith et al.,
2009). Surveillance and monitoring programs are important means of detecting new and
emerging diseases, and are some of the first steps to understanding the health status of
wild animal populations (Mörner et al., 2002; Daszak et al., 2007). Information about
occurrence of diseases in wildlife is an essential part of local and global knowledge of
wild animal populations, and is needed in order to safeguard biodiversity and public
health (Mörner et al., 2002). Merianos (2007) stressed the need for funding support for
21
wildlife surveillance globally and identified optimized study design, sampling
methodology, and diagnostics as important areas of research.
Authors have discussed various elements of animal disease surveillance systems, as well
as relevant terminology. Thurmond (2003) defined surveillance as an active, formal,
systematic process aimed at early detection and with pre-specified intervention strategies,
and specified numerous considerations for developing infectious disease surveillance
systems for animals. Two key components this author stressed were i) a sampling scheme
that maximizes the probability of detecting an infected individual as early as possible; ii)
diagnostic tests that are highly likely to detect disease/exposure if it exists in the
population. The same article identified several performance milestones for these
programs: accuracy (sensitivity and specificity), precision (repeatability of the sampling
and testing), timeliness (functionality in real time), multiple utility (i.e., applicability in
numerous surveillance settings), and value (i.e. various potential benefits of detecting and
intervening). While these are constructive targets for a wildlife surveillance program, the
applications and examples Thurmond (2003) described are geared largely to livestock
settings and are not all attainable in wildlife studies.
In contrast, Stallknecht (2007) presented a somewhat more nuanced and practical
approach to wildlife disease surveillance and epidemiological studies. This author pointed
out that there is no one formula for pathogen surveillance programs because each
situation has different inherent goals, complexities, and circumstances. While recognizing
the need for increased programs and diagnostic infrastructure for wildlife disease
surveillance, Stallknecht noted some fundamental challenges that can be daunting: i) case
22
and data collection (obtaining a representative sample); ii) interpretation of results
(availability of reliable diagnostic assays for wildlife species); iii) validation of field
observations through experimental research. As well, Stallknecht (2007) discussed pros
and cons of different forms of surveillance, citing the potential for detection and reporting
problems with some passive systems (e.g., carcass disappearance in mortality
assessments [Wobeser and Wobeser, 1992]) and the potential for various biases in others
(e.g., spatial, temporal, and health-, age-, and sex-related biases in samples of hunterkilled animals). Among other guideposts for wildlife surveillance and research,
Stallknecht highlighted the need to archive samples for future studies, to maintain quality
control through supportive diagnostic tools and validated results, to take care in
interpreting findings (and that such restraint often reveals insights into complex systems),
to explore non-traditional approaches, and to not be intimidated by the often complex
challenges of wildlife surveillance endeavours. All of these elements are salient to the
doctoral research presented in this dissertation.
Project Rationale and Problem Defined
Pathogens in Caribou: Detecting and Monitoring
Caribou (Rangifer tarandus ssp.) are a keystone species in arctic ecosystems and are
highly valued across the North because of their integral worth as wild animals and their
spiritual, cultural, and nutritional importance (Receveur and Kuhnlein, 1998; Byers,
1999; Gordon, 2003; Festa-Bianchet et al., 2011). Many people in the Arctic depend on
23
caribou as a traditional and affordable source of food (Receveur and Kuhnlein, 1998;
Batal et al., 2005; Lambden et al., 2007). Food safety is a concern because some zoonotic
diseases (for example, brucellosis, which causes chronic joint and reproductive system
disease in humans) have been documented in barrenground caribou herds (Tessaro and
Forbes, 1986; Gunn et al., 1991; Ferguson, 1997). In addition to nutrition and safety
aspects, in 2012 food security is considered a crisis issue for Canadian Aboriginal people
owing to widespread poverty, possible effects of environmental change on wildlife foodsource sustainability, and other factors (Chan et al., 2006; Furgal and Seguin, 2006;
Lambden et al., 2007; Power, 2008; Wesche et al., 2011).
Population estimates from aerial surveys of caribou herds indicate that numerous
migratory caribou herds experienced severe (>75%) declines after peaking in the 1980s
and 1990s (Vors and Boyce, 2009; CARMA, 2012). Although caribou populations are
known to cycle over decades and some herd declines may be stabilizing (CARMA,
2012), the specific causes of these events are unknown and they raise conservation
concerns. There is question whether caribou populations can rebound in the face of new
anthropogenic disturbances, including effects of industrial development, the rapidity and
extent of climate change, other environmental changes in the North, and possible
interactions of these with harvest pressures (Anisomov et al., 2007; Festa-Bianchet et al.,
2011). Disease may be among the complex factors that contribute to wildlife population
declines and cycles (Hudson et al., 1998; Daszak et al., 2000, 2001), and to caribou
population cycles specifically (Albon et al., 2002). However, relatively little is known
about the health impacts of pathogens that circulate in caribou herds.
24
Caribou Health Monitoring and A Community-Based Diagnostic Method
In addition to the worldwide call for increased surveillance and deeper understanding of
the occurrence of and contributors to disease in wildlife, concerns about caribou
conservation and food safety indicate a need to assess disease and pathogen exposure in
arctic caribou herds. However, sample collection from free-ranging wildlife—particularly
in remote arctic settings—can be challenging. Reliable, validated sampling techniques are
needed (Stallknecht, 2007). One solution for serological detection of pathogen exposure
or infection is to establish a method that is practical for subsistence harvesters to use
during hunts. For broad acceptance, such a technique must be compatible with harvesters’
existing hunting methodologies and sensitive to cultural beliefs and values regarding
wildlife and harvesting/animal-use practices (Byers, 1999; Kofinas et al., 2003). If
successfully implemented, this type of diagnostic tool could increase the temporal and
spatial reach of a caribou disease surveillance program. Hunters are typically on the land
in all seasons and access host ranges in ways that are not logistically or financially
feasible for scientists.
Overview of Aims and Chapters
The main objectives of this doctoral research were as follows:
•
To evaluate a sampling method and diagnostic tool for caribou, filter-paper (FP)
blood sampling by subsistence hunters, that is novel and non-invasive in this
25
application and that could be used to detect exposure to pathogens (some
zoonotic) of known or potential relevance to Rangifer;
•
To apply this blood-collection technique in a coordinated serological survey of
arctic migratory caribou herds during International Polar Year 2007-2009 and,
thus, gain insights into the presence, patterns, and ecology of disease/pathogens in
these populations;
•
To implement the FP method in arctic communities of Canada, and to assess the
perceptions of subsistence harvesters of caribou regarding fundamental aspects of
hunter-based sampling and the FP method.
Filter-paper blood sampling was introduced to human medicine in the 1960s for neonatal
screening (Guthrie and Susi, 1963). The continued use of this technique in contemporary
medicine attests to its utility and reliability, and FP applications have expanded to a long
list of blood analytes that are tested in various types of research, including veterinary
studies and human epidemiological studies in remote settings (Mei et al., 2001; McDade
et al., 2007). Although various advantages of FP collection make this method well suited
for wildlife field collections, validation is generally lacking. Collection of FP samples by
northern subsistence hunters as part of a wildlife health-monitoring program (i.e., a
system somewhat less systematic/formal than surveillance and one that does not
necessarily involve an established intervention plan) has not been formally reported or
evaluated to date.
The aims of this research address several of Stallknecht’s (2007) main guidelines for
optimizing wildlife surveillance and research. They also seek to establish baseline
26
information about pathogen exposure in migratory caribou herds that could be valuable in
the relatively near future. More extensive environmental change and habitat
fragmentation may be approaching for arctic caribou as the rapidity of climate change
and the spectre of an ice-free Northwest Passage threaten to bring even more extensive
mining and oil/gas exploration to the North. Such perturbations could potentially impact
disease in caribou and other arctic species (Daszak et al., 2001; Kutz et al., 2005).
Specifically, marine traffic and port construction and activity in the Northwest Passage
could facilitate greater infrastructure on or near traditional calving grounds or caribou
migratory routes. However, these and other aspects of caribou space use and movement
are difficult to document for a number of reasons: i) lack of updated caribou population
data for some regions where there are sites targeted for development; ii) apparent low or
declining numbers of caribou in some of these areas currently, iii) the cost and extensive
logistics of conducting caribou movement and census surveys (Jenkins and Goorts, 2011;
CARMA, 2012; Giroux et al., 2012). Increased or year-round marine traffic and
associated land-use changes in the Arctic could also increase the risk of translocating
invasive pathogens from the south, and such invaders could negatively affect the marine
environment and terrestrial ecosystems that include caribou (Kutz et al., 2005; Tryland et
al., 2012). Another potentially valuable aspect of detecting and monitoring pathogens of
migratory caribou is that, currently, the Arctic features relatively simple biological
systems that may be good models for teasing out the ways in which climate change can
affect disease ecology (Kutz et al., 2009). Further, considering the evolutionary origins of
caribou (COSEWIC, 2011) and the likelihood that pathogens circulate within herds over
the long term, pathogen-exposure findings in ancestrally related population groupings
27
might facilitate predictions of pathogen groups circulating in important caribou
populations that are difficult or impossible to sample. The endangered subspecies Peary
caribou (R. tarandus pearyi), which reside in the High Arctic islands, is one example.
Chapters: A Brief Scan
Chapters 2 and 3 of this thesis describe experimental studies that assessed the
performance of FP samples from Rangifer in assays for antibodies to eight distinct
pathogens or pathogen groups: Brucella spp., Neospora caninum, West Nile virus, and
five bovine viruses (bovine herpesvirus type 1, parainfluenza virus type 3, bovine
respiratory syncytial virus, and bovine viral diarrhea virus types I and II).
Chapter 4 documents another experimental study that examined the effects of different
storage times and processing/storage conditions (freezing vs. room temperature) on
results from Rangifer FP samples in these same antibody tests.
Moving to application of the technique, Chapter 5 describes a wide-scale collaborative
observational study of caribou herd serology with epidemiological analysis and
interpretation that touches on some aspects of disease ecology. This serosurvey for nine
pathogens/pathogen groups (the eight listed above plus Toxoplasma gondii) was
conducted on serum and FP samples that were collected from seven migratory caribou
herds in parts of the circumpolar world during International Polar Year 2007-2009. Most
of these populations were sampled in multiple seasons and multiple years, and all
collections were coordinated and executed by members and affiliates of the CircumArctic
Rangifer Monitoring and Assessment Network (CARMA, 2012).
28
Chapter 6 documents an evaluation of some key elements of two hunter-based wildlife
monitoring programs that have been implemented in Canada’s North during the past
decade. These initiatives seek to foster teamwork and expertise-sharing approaches
between hunters and scientists, and both the implemented programs feature FP blood
collection. The chapter explores aspects of the human dimension of community-based
wildlife surveillance programs. Face-to-face interviews were conducted with Aboriginal
caribou harvesters in communities across the Canadian Arctic, and qualitative and
quantitative analyses focused on three fundamentals: i) harvesters’ views on the
importance of wildlife disease, ii) their levels of interest in hunter-based sampling, and
iii) their acceptance of the FP method.
Taken together, this dissertation tells a narrative of sorts, one that begins with the efficacy
of FP sampling, then describes application of this tool and discoveries from a serosurvey
of migratory caribou herds, and then finally examines how caribou harvesters view FP
sampling and hunter-based wildlife sampling in general. This body of work contributes
new knowledge and insights on wildlife diagnostics, caribou disease ecology, and
community-based monitoring of caribou health. At a broader scale, it is applicable for
blood-sampling methodology and FP serology in woodland caribou (R. tarandus ssp.
tarandus) and other free-ranging wildlife species. The work could potentially inform
community-based wildlife-health-monitoring initiatives in the Canadian North and
elsewhere, as well as research and other programs that focus on community engagement.
29
CHAPTER TWO
FILTER-PAPER BLOOD SAMPLES FOR ELISA DETECTION OF
BRUCELLA ANTIBODIES IN CARIBOU
Authors:
Curry, P.S., B.T. Elkin, M. Campbell, K. Nielsen, W. Hutchins, C. Ribble, and S.J. Kutz
© Wildlife Disease Association 2011. Published article reprinted with permission.
30
Abstract
We evaluated blood collected on Nobuto filter-paper (FP) strips for use in detecting
Brucella spp. antibodies in caribou. Whole blood (for serum) and blood-saturated FP
strips were obtained from 185 killed arctic caribou (Rangifer tarandus groenlandicus).
Sample pairs (serum and FP eluates) were simultaneously tested in duplicate using
competitive enzyme-linked immunosorbent assay (cELISA) and indirect ELISA
(iELISA) for Brucella spp. Prior work based on isolation of Brucella spp. revealed
sensitivity (SE) and specificity (SP) of 100% and 99%, respectively, for both these serum
assays in caribou. Infection status of the animals in the current study was unknown but
recent sampling had revealed clinical brucellosis and >40% Brucella antibody prevalence
in the herd. To assess the performance of FP relative to serum in these assays, serum was
used as the putative gold standard. On both assays, the findings for duplicate runs (A and
B) were similar. For cELISA run A, the FP Brucella prevalence (47%) was lower than
serum prevalence (52%), with SE 89% (95% confidence interval [CI]: 82–95%) and SP
99% (97–100%). For iELISA run A, serum and FP Brucella prevalence rates were
identical (43%), and the SE and SP of FP testing were 100% and 99% (97–100%),
respectively. The findings suggest better FP test performance with iELISA than with
cELISA; however, iELISA does not distinguish cross-reacting antibodies induced by
Brucella vaccination or exposure to certain other Gram-negative pathogens. Results for
duplicate FP eluates (prepared using separate FP strips from each animal) were strongly
correlated for both protocols (r=0.996 and 0.999 for cELISA and iELISA, respectively),
indicating minimal variability among FP samples from any individual caribou. Dried
31
caribou FP blood samples stored for 2 months at room temperature are comparable with
serum for use in Brucella spp. cELISA and iELISA. Hunter-based FP sampling can
facilitate detection of disease exposure in remote regions and under adverse conditions,
and can expand wildlife disease surveillance across temporo-spatial scales.
Introduction
Wildlife disease researchers face a variety of difficulties collecting and deriving useful
information from biologic specimens (Kuiken et al., 2005). The sampler needs practical,
reliable tools that can be transported easily, perform well under field conditions, and
provide multifaceted information about health status. Collecting blood on filter paper
(FP) may be a simple way to address some of these problems. This method can be
performed by laypeople and circumvents many of the logistic and cost issues associated
with obtaining, processing, and shipping conventionally sampled blood (Mei et al., 2001;
McDade et al., 2007). Advantages include the elimination of tube breakage, reduced
processing time and labour in the field, and no requirement for special equipment, such as
centrifuges and freezers, that can be difficult to transport, operate, and maintain in the
field.
Filter-paper blood sampling is not new; this method has been widely used in various
forms in human medicine since the 1960s and its applications continue to expand (Mei et
al., 2001). The current and prospective human-related uses of FP for clinical chemistry
and in remote situations are diverse. Recent field-related human FP publications focus on
32
nucleic acid- and antibody-based detection of human immunodeficiency virus, and the
agents of malaria, dengue fever, and other infectious diseases (Lederman et al., 2007;
Balmaseda et al., 2008; Castro et al., 2008; Corran et al., 2008). This mode of blood
collection has parallel benefits for assessing animal health (e.g., Beard and Brugh, 1977;
Hopkins, et al., 1998; Thangavelu et al., 2000; Dubay et al., 2006; Yu et al., 2007) and its
advantages as a field tool suggest a range of applications for wildlife health analysis.
However, validation for detection of biochemical analytes and pathogen exposure in freeliving species is lacking. This research addresses this for an important infectious disease
of wild ungulates, brucellosis.
Brucellosis is a zoonosis caused by bacteria of the genus Brucella. This disease has a
multi-systemic pathogenesis, numerous clinical signs, and tends to affect the reproductive
system in particular (Romich, 2008). Severe losses through infertility/abortion and
reduced productivity make brucellosis one of the most serious diseases of livestock
(Romich, 2008). Brucella suis biovar 4 is the causative agent in caribou and reindeer
(both Rangifer tarandus subspecies) and the clinical signs and lesions are similar to those
seen in cattle and other domestic species, but are often more severe (Forbes, 1991).
Rangiferine brucellosis occurs across northern Canada and in Alaska and Russia, and
poses a potential human health risk if proper precautions are not taken with carcass
handling and cooking (Forbes, 1991; Bradley et al., 2005). Certain favoured caribou
“country foods” are eaten raw or undercooked. Natural Brucella infections have been
documented in muskoxen, moose, captive Rocky Mountain bighorn sheep, and predator
species, and researchers have demonstrated experimental transmission from naturally
33
infected reindeer to cattle (Gates et al., 1984; Forbes and Tessaro, 1993; Honour and
Hickling, 1993; Kreeger et al., 2004). There remain major gaps in scientific knowledge
about the epidemiology of Brucella in Rangifer, including transmission patterns and
effects at individual and population levels (Forbes, 1991; Forbes and Tessaro, 2003).
Extensive surveillance for brucellosis in these animals has been limited by the high cost
and logistic difficulties of sampling in the North. A broader sampling strategy is needed
to help provide better understanding of the patterns and ecology of this disease in caribou.
Our aim was to assess the efficacy of blood collected on FP relative to serum derived
from clotted blood for Brucella serologic testing in caribou. In contrast to conventional
blood sampling of caribou, FP sampling could facilitate widespread hunter-based
surveillance of Brucella exposure in circumpolar herds.
Materials and Methods
Samples and Processing
In March 2008, paired serum and FP blood samples were obtained from each of 185
barrenground caribou killed for a territorial government scientific study near the
community of Coral Harbour (64°11’24” N, 83°21’36” W) on Southampton Island,
Nunavut, Canada (Nunavut Wildlife Research Permit WL 000892). All sampling was
done outside in the extreme cold (temperature range –39.2 °C to –18.3 °C). Each sample
pair (blood tubes and blood-soaked FP strips) was collected as soon after death as
possible and kept inside a shelter and above freezing until end-of-day processing 2–12 hr
34
later. Serum was obtained by collecting jugular or femoral venous blood into a glass
Vacutainer® tube without coagulant (Becton-Dickinson, Mississauga, ON, Canada).
Filter-paper samples were collected from the same source by saturating the full length of
all FP strips with blood and shaking off any excess (Fig. 1). For each animal, 15 Nobuto
blood filter strips (Toyo Roshi Kaisha, Ltd., Tokyo, Japan; distributor Advantec MFS
Inc., Dublin, CA, USA) mounted in sets of five strips on a handmade lightweight
cardboard “handle” (Fig. 2a) were collected. According to manufacturer specifications,
the blood-absorbing section of each strip holds 100 µl of whole blood (approximately 40
µl of serum depending on hematocrit). Each animal’s FP sets were kept in an
antimicrobial-lined #10 envelope (Quality Park, St. Paul, MN, USA) inside a zip-closure
plastic bag. Care was taken to avoid touching FP strips during preparation of FP sets for
sampling and throughout collection, processing, and storage. For processing, blood tubes
were centrifuged (15 min at 3,500 revolutions per minute) and aliquots of serum were
stored at –20 °C until analysis. Collected FP sets were air-dried in racks at room
temperature (15–22 °C) overnight (Fig. 2b) and returned to a dry antimicrobial envelope.
Multiple envelopes (up to 25) were placed together in a large zip-closure plastic bag with
eight to 10 desiccant packs (Humidity Sponge™, VWR International LLC, Mississauga,
ON, Canada) and stored at room temperature. Desiccant was checked regularly and
replaced as required based on the manufacturer’s color indicator insert. Tissue specimens
were not obtained for culture, and Brucella infection status was unknown for the animals
in this study. Recent annual sampling of the herd had revealed clinical evidence of
disease and >40% antibody prevalence (M. Campbell, unpublished data). As the aim was
to evaluate FP results relative to serum findings, serum was used as the gold-standard
35
indicator of Brucella exposure.
Filter-Paper Elution
After 2 months of storage, two eluates (A and B) were prepared from each animal’s FP
samples. A stock solution was made consisting of Dulbecco’s phosphate-buffered saline
with CaCl2 and MgCl2 (D-PBS 13, Gibco® Invitrogen™, Burlington, ON, Canada) and
an antibiotic mixture (penicillin–streptomycin liquid, Invitrogen). The final penicillin and
streptomycin concentrations in the stock solution were 100 U/ml and 100 µg/ml,
respectively. For each eluate, clean (flamed and cooled) small scissors were used to cut
the absorbent portions (Fig. 2a) of two FP strips into five or six pieces directly into a 1.5ml microcentrifuge tube (MCT-200-C tubes, Axygen Scientific, Union City, CA, USA).
Eight hundred microliters of stock solution were added (as per the Nobuto FP
manufacturer’s instructions of 400 µl per strip) and the tube was finger-flicked to ensure
all fragments were in full contact with the fluid. Tubes were stored at 4 °C for 16 hr, and
all (dark red) fluid was pipetted from each into a new, labeled 1.5-ml microcentrifuge
tube. Eluates were spun very briefly (15 sec) to draw all fluid to the tube bottom, and then
stored at 220 °C until testing. Each resultant two-strip eluate was 400–440 µl and
estimated to be 1:10 serum concentration, according to the FP manufacturer’s
specifications.
Immunoassays
All samples were tested at the Brucellosis Centre of Expertise (BCE) in Ottawa, Canada
(Canadian Food Inspection Agency, Government of Canada). The BCE uses enzyme-
36
linked immunosorbent assay (ELISA) protocols and a fluorescence polarization assay
(FPA) that were developed for diagnosing brucellosis in Cervidae. Gall et al. (2001)
evaluated these tests in R. tarandus ssp. (reindeer and woodland caribou) and other
cervids, and some of the reindeer sera tested were from animals that were culture-positive
for B. suis biovar 4. The data from that study support the use of competitive ELISA
(cELISA) and FPA for diagnosing brucellosis in caribou, and the authors identified FPA
as the diagnostic test of choice for this purpose. However, multiple attempts to run FPA
with our caribou FP eluates failed. Thus, the two assays we used for our study were the
cELISA and indirect ELISA (iELISA) assessed by Gall et al. (2001); protocols were
detailed by Nielsen et al. (1994, 1996). Both assays are based on antigen of Brucella
abortus, which is the main cause of brucellosis in cattle and known to infect bison and
elk. Gall et al. (2001) tested caribou serum samples using the Brucella cELISA with a
threshold value of 16% inhibition and observed sensitivity (SE) 100% (n=102) and
specificity (SP) 99% (n=308). The corresponding n values, SE, and SP for iELISA
(threshold 11% positivity) were identical to those observed with cELISA. Currently, the
BCE uses higher thresholds for both these tests (30% inhibition for cELISA and 20%
positivity for iELISA). Raising these thresholds had minimal effect on SE and SP (K.
Nielsen, unpublished data) and the BCE has adopted these higher thresholds in an effort
to set universal values for brucellosis testing across animal species. Note that the Brucella
iELISA does not distinguish antibodies induced by Brucella spp. exposure from
antibodies elicited by Brucella vaccination or certain infectious agents, such as Yersinia
enterolitica (Nielsen, 1990).
37
Nobuto FP eluates are estimated to be 1:10 serum concentration; therefore, initial dilution
steps in the ELISA protocols were adjusted such that the assays would yield results for
eluates and 1:10 serum (one serum/FP pair per animal tested simultaneously in each of
two duplicate runs). The 185 serum/FP Southampton Island caribou sample pairs were
tested with cELISA in July 2008, and sera/ eluates were refrozen and kept at –20 °C until
iELISA was done in September 2008. Both immunoassays were carried out in duplicate
(i.e., run A: 1:10 serum vs. eluate A; run B: 1:10 serum vs. eluate B). Sera for runs A and
B in both assays were drawn from the same serum aliquot per animal. As noted, eluates A
and B were prepared from different FP samples from the same animal.
Statistical Analysis
To evaluate the performance of FP testing relative to serum (the putative gold standard),
data from each test run (run A and run B) were categorized and analyzed using Win
Episcope 2.0 (de Blas et al., 2005) to generate prevalence and SE and SP values. To
evaluate variability between FP tests and between serum tests (i.e., multiple testing of
samples collected from each animal), the correlation between duplicate results for each
sample type (serum A vs. serum B, eluate A vs. eluate B) was determined for cELISA
and for iELISA (Microsoft® Office Excel 2003, Microsoft Corp., Redmond, WA, USA).
Results
With cELISA, the measures of FP test efficacy for the duplicate runs (A and B) were
38
similar (Table 1). For run A, analysis of results using the laboratory’s established
threshold value (30% inhibition) revealed a 52% prevalence of Brucella antibody on the
basis of serum, compared with 47% prevalence on the basis of FP (SE 89% and SP 99%)
(Table 1, Fig. 3). With iELISA, the measures of FP test efficacy for the duplicate runs (A
and B) were also similar (Table 1). For run A, the serum and FP-based Brucella antibody
prevalence were identical (43%), and FP SE and SP were 100% and 99%, respectively
(Table 1). There was only one FP-serum results mismatch throughout the entire two runs
of 185 iELISA sample pairings (370 tests). The iELISA Brucella antibody prevalence
values were lower than observed with cELISA, and SE and SP were both higher than
observed with cELISA. The cELISA FP eluate A and eluate B results were strongly
correlated, as were the cELISA serum results from runs A and B (r=0.996 and 0.994,
respectively; Fig. 4). The iELISA eluate A and eluate B results were also strongly
correlated, whereas the serum results for run A vs. run B were slightly more variable
(r=0.999 and 0.974, respectively; Fig. 5).
Discussion
Previous investigators have assessed FP blood samples in ELISAs for brucellosis
diagnosis in humans and for B. abortus antibody detection in cattle (McLean and Hilbink,
1989; Takkouche et al., 1995). We examined FP blood testing as a means of identifying
Brucella spp. exposure in caribou (Rangifer spp. in general). Bacteriologic isolation
remains the only absolute method for establishing Brucella infection status (Gall and
39
Nielsen, 2004). This information was not available for the 185 caribou we investigated;
thus, the true brucellosis status of these animals was unknown. As mentioned, the
accuracy of caribou serum testing for brucellosis has been investigated previously (Gall
et al., 2001). Our intent was to assess the efficacy of FP relative to serum for detecting
anti-Brucella antibodies in Rangifer.
Brucella Antibodies: Filter Paper vs. Serum
For cELISA, the FP and serum results were in close agreement and we conclude that FP
cELISA is effective (comparable with serum cELISA) for detecting Brucella antibodies
(i.e., exposure) in caribou. Our data indicate that FP iELISA is also comparable with
serum iELISA for detecting Brucella exposure in caribou; however, there were some
interesting findings. The iELISA for Brucella is known to detect cross-reacting
antibodies, and on the basis of the literature one would anticipate higher prevalence with
iELISA than with cELISA (Nielsen, 1990; Gall et al., 2001; Gall and Nielsen, 2004;
Nielsen et al., 2004), yet we found the opposite. The BCE laboratory iELISA yielded
serum Brucella antibody prevalence 43%, which was 8% lower than the prevalence
detected with cELISA. Running the two separate ELISAs involved a freeze/thaw cycle
and there was an 8-wk interval between cELISA and iELISA testing. However, it is
unlikely that these factors would cause enough antibody degradation to explain this
extent of divergence. Brucella cELISA and iELISA are different test methods and
standard controls were run on each antigen-coated plate. The discrepancy between these
results may reflect interference with the monoclonal antibody used in the cELISA, which
would yield apparently higher levels of inhibition. This effect is occasionally observed
40
with sera from other animals, including cattle and pigs, and the cause is unknown (K.
Nielsen, unpublished data). Our results suggest that if this effect occurs with Rangifer
specimens, it occurs to similar degrees in serum and FP samples.
Variability Among Filter Papers From Individuals
Analysis of variation between duplicate FP tests and between duplicate serum tests done
on the same animal revealed very strong correlations for both the cELISA and the
iELISA (Figs. 4,5). This encompassed data generated by four FP strips from each
individual (each eluate was prepared from two FP strips because of the volume required
for testing).
Test Choice
Although Gall et al. (2001) support the use of cELISA and FPA for serodiagnosis of
brucellosis in caribou, our attempts to use FPA with FP elutions failed. Excessive
background fluorescence was the suspected problem, and possible reasons for this
include 1) protein aggregation after drying on FP and subsequent elution; 2) release of
cellulose from the FP causing interference with light transmission; 3) emission of an
autofluorescing chemical from the paper matrix. Our experience with FPA underlines the
importance of evaluating individual assays before use in new applications.
We found that caribou FP samples worked very well in both the iELISA and cELISA for
Brucella, and we observed slightly better agreement between FP and serum results in the
iELISA. Although this appears to promote iELISA as the better FP screening test for
41
Brucella in caribou, this assay does not distinguish antibodies induced by Brucella
exposure from antibodies elicited by vaccination or Y. enterolitica (Nielsen, 1990).
Considering this, FP cELISA is likely the more informative Brucella assay for Rangifer
population surveillance, where disease and pathogen exposure levels and, in some cases,
vaccination status in a population may be unknown or only estimated.
Benefits of Filter-Paper Sampling
Filter papers are inexpensive sample media that are well suited for harsh environmental
conditions as they facilitate simple, rapid blood collection in the field and can be dried,
transported, and stored relatively economically. Filter-paper blood testing has been used
and studied in various veterinary and wildlife applications (Beard and Brugh, 1977;
Stallknecht and Davidson, 1992; Yamamoto et al., 1998; Sacks et al., 2002; Chomel et
al., 2004; Jordan et al., 2005; Dubay et al., 2006; Trudeau et al., 2007; Yu et al., 2007;
Duscher et al., 2009); however, the full potential of FP for wildlife disease testing and
monitoring has yet to be realized. One important feature of this tool is that it requires no
special training and can thus be used by hunters, biologists, and others. For many
Northerners, caribou are a local affordable food source and part of culture/tradition.
Filter-paper blood collection during hunting would increase sample size and expand
temporal and spatial scales of data collection. Such samples would otherwise be lost from
animals killed for subsistence. It is apt that those who rely on caribou should be key
agents in monitoring of populations. A team strategy with collection by hunters and
analysis by scientists engages communities in caribou health assessment. Additionally,
harvesters’ observational skills and traditional knowledge can inform scientific
42
perspectives on wildlife disease investigation (Brook et al., 2009). To our knowledge,
validated FP blood testing has not been applied in a community-based wildlife or
domestic animal health-monitoring context anywhere. This mode of disease surveillance
is the ultimate goal of our FP efficacy research.
Two common challenges in assessing diagnostic tests for wildlife are sample size and
availability of assays designed for wildlife species. We analyzed FP samples from 185
barrenground caribou using two serologic assays that were developed for cervids and
evaluated using caribou serum samples (Gall et al., 2001). Our FP-serum comparisons
and FP-FP (eluate) correlations present strong evidence that FP is an effective tool
(comparable with serum) for Brucella ELISA screening in Rangifer species. Confidence
in use of the FP method for detecting antibody to Brucella and other infectious agents in
Rangifer will help pave the way for researchers to acquire baseline data on circumpolar
caribou, and better understanding of disease in these populations. Future work will
explore the efficacy of caribou FP testing for Brucella after 1 and 2 years of storage. The
FP method will also be investigated in several other infectious disease contexts for
Rangifer, and with samples subjected to different collection-temperature and storage-time
regimes that mimic field conditions.
Acknowledgements
We recognize the lives of the animals that were taken from the Southampton Island
caribou herd as part of the broader investigation that facilitated this study. This research
43
could not have been done without the expertise of hunters Aaron Emiktowt, Chris Jones,
John Nakoolak, Mark Pootoolik, and Greg Ningeocheak, all from Coral Harbour,
Nunavut. Technical work and logistics by Linda Kelly (BCE) and Johnathon Pameolik
(Dept. of Environment, Government of Nunavut) were also instrumental. We are most
grateful for the support of the Government of Nunavut, which facilitated collection of
these valuable samples. Thanks to our funders: NSERC International Polar Year Funding
(Government of Canada), Nasivvik Centre for Inuit Health and Changing Environments
(Canadian Institutes of Health Research), Northern Scientific Training Program (Indian
and Northern Affairs, Government of Canada), University of Calgary Faculty of
Veterinary Medicine, and the CircumArctic Rangifer Monitoring and Assessment
Network (CARMA, www.carmanetwork.org).
44
TABLES
Chpt 2. Table 1. Filter paper sample performance – Brucella assays
Findings for filter-paper (FP) blood testing relative to serum in competitive and indirect
immunosorbent assays (cELISA and iELISA) for Brucella spp. in wild caribou (n=185).
Results for duplicate test runs (eluates A and B) are shown. The threshold for cELISA is
% inhibition (sample’s optical density [OD] relative to that of the buffer well [uninhibited
control]). The threshold for iELISA is % positivity (sample’s OD expressed as a
percentage of the OD of a positive control).
VARIABLE cELISA THRESHOLD: 30% for serum and FP iELISA THRESHOLD: 20% for serum and FP Eluate A Eluate B Eluate A Eluate B SPECIFICITY (%) (CI) 88.5 (82.2‐94.9) 98.9 (96.7‐100) 89.4 (83.1‐95.6) 98.9 (96.8‐100) 100 99.1 (97.2‐100) 100 100 SERUM PREVALENCE (%) (CI) 51.9 (44.7‐59.1) 50.8 (43.6‐58.0) 42.7 (35.6‐49.8) 42.7 (35.6‐49.8) FP PV (‐) Test (%) 46.5 (39.3‐53.7) 98.8 (96.6‐100) 88.9 45.9 (38.7‐53.1) 98.8 (96.5‐100) 90.0 43.2 (36.1‐50.4) 98.8 (96.3‐100) 100 42.7 (35.6‐49.8) 100 100 (CI) (82.7‐95.1) (84.1‐95.9) SENSITIVITY (%) (CI) FP PREVALENCE (%) (CI) FP PV (+) Test (%) (CI) Abbreviations: CI: 95% confidence interval; PV: predictive value 45
FIGURES
Chpt 2. Figure 1. Filter-paper blood collection
Collecting filter-paper blood samples from the jugular vein of a hunter-killed caribou
(photo: S. Kutz).
46
Chpt 2. Figure 2. Filter-paper sets and drying racks
(a) A set of seven Nobuto filter-paper (FP) strips mounted on light cardboard and with
the 3-cm absorbent portion (A) of each strip identified; (b) saturated FP sets in simple
drying racks made from hard foam material and duct tape.
47
Chpt 2. Figure 3. Filter paper vs. serum – Brucella cELISA
Results from the competitive enzyme-linked immunosorbent assay (cELISA) for
Brucella: The paired serum and filter-paper (FP) values for each animal (plotted
according to ascending serum values) in test run A (n=185). The discrete FP and discrete
serum values, respectively, are shown joined by lines to facilitate visual comparison of
the FP and serum results.
48
Chpt 2. Figure 4. Results for sample duplicates – Brucella cELISA
Correlation of Brucella competitive enzyme-linked immunosorbent assay (cELISA)
results (% inhibition) for the duplicate runs of caribou (a) serum and (b) filter-paper (FP)
eluates (run A vs. run B pairs; n=185).
49
Chpt 2. Figure 5. Results for sample duplicates – Brucella iELISA
Correlation of Brucella indirect enzyme-linked immunosorbent assay (iELISA) results
(% positivity) for the duplicate runs of caribou (a) serum and (b) filter-paper (FP) eluates
(run A vs. run B pairs; n=185).
50
CHAPTER THREE
FILTER-PAPER BLOOD SAMPLES FOR DETECTING
ANTIBODIES TO NEOSPORA CANINUM, WEST NILE VIRUS, AND FIVE BOVINE VIRUSES IN RANGIFER
Authors: Curry, P.S., C. Ribble, W.C. Sears, W. Hutchins, K. Orsel, D. Godson,
R. Lindsay, A. Dibernardo, and S.J. Kutz
Article in submission – Journal of Wildlife Diseases
51
Abstract
Filter-paper (FP) blood sampling of wildlife by hunters and others can extend the
temporal and geographic reach of disease surveillance programs, but it is important to
validate FP samples for each serological test. We assessed Nobuto FP blood strips for
detecting antibodies to seven pathogens in Rangifer species. Serum and blood-saturated
FP samples were obtained from captive reindeer (Rangifer tarandus) between April 2008
and August 2009, and were tested within 2 months of collection. Sample pairs (serum and
FP eluates) were assayed in duplicate using competitive enzyme-linked immunosorbent
assay (cELISA) for Neospora caninum and West Nile virus, indirect ELISA (iELISA) for
bovine herpesvirus type 1 (BHV-1), parainfluenza virus type 3 (PI-3), and bovine
respiratory syncytial virus (BRSV), and virus neutralization (VN) for bovine viral
diarrhea virus (BVDV) types I and II. The FP samples performed well in all assays when
compared to serum. Using laboratory threshold values, FP specificity estimates were
consistently high, ranging from 92% in the cELISAs for N. caninum and WNV to 98% in
the iELISAs for PI-3 and BRSV. Sensitivity was >85% for five of the seven tests (most
≥95%) but was lower (71% to 82%) for the PI-3 and BRSV iELISAs. Lowering the
threshold for FP results in these two ELISAs raised sensitivity to ≥87% and reduced
specificity only slightly (≥90% in three of the four total test runs), thus making FP
performance comparable to that in all other assays. Filter-paper duplicates from
individual animals were strongly correlated (r≥0.96 for all ELISAs). Although limited
sample sizes affected estimate precision, FP proved effective on both ELISA platforms.
Stronger performance with cELISA point to this test as superior for detecting antibodies
52
with FP samples from Rangifer. These samples also functioned well in BVDV VN but
initial sample dilutions in the protocol reduced FP test sensitivity, and cell toxicity may
be an issue with FP samples. Overall, the results show that, when testing within 2 months
of collection, FP samples from Rangifer perform comparably to serum in all seven
antibody assays, and multiple FP samples from individual animals vary minimally. The
finding that FP samples from Rangifer can be used in serological screening for an
additional seven pathogens expands the potential utility of this sampling method.
Introduction
Assessing wildlife health in the field can be challenging, and practical sampling tools
such as blood collection on filter paper (FP) offer attractive alternatives to traditional
methods. The relative ease of collecting, storing, and transporting FP blood samples, their
durability compared to traditional blood tubes, their light weight and compactness, and
lack of need for special field-processing equipment are all significant advantages of this
technique for wildlife work (see Chapter 2). In addition to its practicalities and
convenience, FP blood collection by subsistence hunters or other trained laypersons can
enhance spatial and temporal scales of sampling (Brook et al., 2009; see Chapter 2).
Authors are calling for wildlife disease surveillance to be expanded, refined, and better
integrated with public health systems, and FP blood sampling could contribute to these
advancements (Glaser, 2004; Kuiken et al., 2005; Jones et al., 2008; Brook et al. 2009).
Filter-paper blood testing was introduced to human medicine in the 1960s and its use has
53
expanded from neonatal screening to human epidemiological studies, with more than 100
assays available today (Mei et al., 2001; Clague and Thomas, 2002; McDade et al.,
2007). This method has also been applied in diverse veterinary and wildlife settings, and
studies by Dubay et al. (2006), Trudeau et al. (2007), Yu et al. (2007), and Kalayou et al.
(2011) are recent examples; however, the efficacy of each new FP application in freeranging species needs to be rigorously assessed. A previous investigation by our research
group demonstrated that FP blood samples are effective for enzyme-linked
immunosorbent assay (ELISA) detection of Brucella antibodies in caribou (see Chapter
2). In this paper, we expand the scope of FP evaluation to antibody assays for seven
additional pathogens with known or possible health ramifications for Rangifer species:
Neospora caninum, West Nile virus, bovine herpesvirus type 1 (BHV-1), parainfluenza
virus type 3 (PI-3), bovine respiratory syncytial virus (BRSV), and bovine viral diarrhea
virus types I and II (BVDV-I and -II) (Table 1). Our aim was to assess whether FP and
serum results are comparable on antibody assays for these seven agents, and to assess
variability between multiple FP samples from individuals.
Materials and Methods
Animals and Study Design
Blood samples were collected from live captive reindeer between April 2008 and August
2009 in accordance with the guidelines of the Canadian Council on Animal Care. All
samples were tested within 2 months of collection. Test performance of the FP samples
54
was determined by comparing to antibody results for matched sera. Testing was done in
blinded fashion at animal diagnostic laboratories, and FP-serum pairs were tested on the
same day and with identical control samples on FP and serum test plates. Presence of
pathogen or infection was not confirmed by culture or any other means.
Table 1 summarizes the pathogens, sources of reindeer used, pathogen/antigen exposure
information, numbers of sample pairs collected (serum and FP from each animal), and
collection dates. The reindeer were from a commercially ranched herd in central Alberta,
Canada (approx. 52°20' N, 112°41' W) and a small research herd at the University of
Calgary in southern Alberta (51°5' N, 114°5' W). In order to assess the efficacy of FP for
detecting antibodies to a given pathogen, there must be known seropositive animals in the
group that is tested. In the commercial herd, previous diagnostic serology had confirmed
that some animals were antibody-positive for one or more of N. caninum, WNV, PI-3,
and BRSV (P. Curry and S. Kutz, unpublished data). Significant WNV activity had been
documented in the province of Alberta just prior to the study period (PHAC, 2007). In the
research herd, we administered a primary label dose of the killed-virus cattle vaccine
Triangle® 4 + BVDV-II (Wyeth Animal Health, Guelph, Ontario, Canada) and a booster
dose 4 weeks later to induce antibody production for five bovine pathogens (BHV-1, PI3, BRSV, BVDV-I, BVDV-II). In order to examine a range of antibody levels in the
serum and FP samples, the vaccinated group was sampled at three time points: baseline
(pre-vaccination) and 4 and 8 weeks after the booster dose.
55
Sample Collection and Processing
From each animal, we collected venous blood into 10-mL Vacutainer® tubes (BectonDickinson, Mississauga, ON, Canada) without anticoagulant. We also drew a 20-mL
syringe of venous blood and immediately (prior to clotting) applied this to Nobuto blood
filter strips (Toyo Roshi Kaisha, Ltd., Tokyo, Japan; Advantec MFS Inc., Dublin, CA,
USA distributor) that were mounted in sets on small cardboard “handles.” The absorbent
portion of each strip (Fig. 1) was completely saturated with blood. According to
manufacturer’s specifications, the absorbent portion holds 100 µl of whole blood
(approximately 40 µl of serum depending on hematocrit). The tube and FP blood samples
were processed as described previously (see Chapter 2), and sera were stored at -20 °C
and FP samples were stored at room temperature (15–22 °C) with desiccant packs
(Humidity Sponge™, VWR International LLC, Mississauga, ON, Canada).
Testing
Antibody testing was carried out at animal health laboratories in Canada using
competitive ELISA (cELISA, for N. caninum and WNV), indirect ELISA (iELISA, for
BHV-1, PI-3, BRSV), and virus neutralization (VN, for BVDV-I and BVDV-II). Elution
of the FP samples was performed as detailed previously (see Chapter 2) and the eluate
was a dark-red fluid estimated to contain 1:10 dilution of serum (FP specifications: Toyo
Roshi Kaisha, Ltd., Tokyo, Japan). Depending on the eluate volume that the lab required
for each assay, one FP strip (N. caninum and WNV tests) or multiple FP strips (all other
tests) were eluted per assay. For six of the seven pathogens (all but WNV), two identical
56
FP eluates (for duplicate testing) were prepared for each animal and these were frozen at
–20 °C and shipped to the diagnostic laboratory within days.
For WNV testing, two dry FP samples per animal (one strip per assay) were shipped in
separate envelopes so that the receiving laboratory could prepare duplicate FP eluates on
site. This was necessary because the WNV assay procedure employs a blocking buffer
(phosphate-buffered saline containing 5% skim milk) as diluent for reagents and,
therefore, the FP eluates needed to be made with this buffer instead of phosphatebuffered saline. These eluates were prepared at the time of WNV testing and were not
frozen prior to assay.
Laboratories received one tube of frozen serum per animal along with the FP eluates (or
dry FP strips as noted for WNV testing) and assays were performed as soon as possible
upon receipt. All samples were tested within 2 months of collection. Given that eluates
contained an estimated 1:10 dilution of serum, assay protocols were adjusted so that
results for serum and FP samples were comparable. Serum-FP pairs from each animal
were assayed simultaneously (same day) and in duplicate (Run A pairs and Run B pairs).
The only exception was with WNV testing, in which serum was run only once,
simultaneous with the Run A and Run B FP samples.
Competitive ELISAs:
Neospora caninum antibody testing was done at the Animal Health Centre (Abbotsford,
BC, Canada) using a commercial kit (Neospora caninum Antibody Test Kit, cELISA;
VMRD Inc., Pullman, WA, USA). Although validated for cattle only, the test format
57
allows samples from other species to be tested (VMRD Inc., Pullman, Washington,
USA). The kit procedure involves no initial dilution of serum; thus, to generate directly
comparable results for serum and FP samples, sera were diluted 1:10 as the first step, and
eluates were tested undiluted. Results were expressed as % inhibition based on samples’
optical densities (OD), and were calculated as 100 – ([sample OD ÷ mean OD of negative
control serum] × 100). The kit threshold value for cattle (≥30% inhibition) was used to
identify antibody-positive animals.
The WNV antibody testing was done in the Zoonotic Diseases and Special Pathogens
section of the Public Health Agency of Canada (Winnipeg, MB, Canada) using a cELISA
developed by Blitvich et al. (2003) that allows detection of WNV antibodies in samples
from domestic mammals. To ensure that test data for serum and FP samples were directly
comparable, serum was diluted 1:10 as per the assay protocol, whereas eluates were
tested undiluted. An animal’s antibody status was determined based on two
simultaneously run tests, each involving a different monoclonal antibody: 3.112G-NS-1
(hereafter referred to as mAb1) or C03301-Env (mAb2). Results were expressed as %
inhibition, calculated as 100 – ([OD of blanked sample ÷ mean OD of blanked control
serum] × 100). In the protocol, results of ≥30% inhibition are positive and three WNV
antibody status categories are generated based on the two test findings: positive (both
tests positive), negative (both tests negative), equivocal (one test positive, the other
negative). For test performance analysis, and to maximize data from the limited sample
available for this assay (n=26), we converted each animal’s results to binomial outcomes
58
as conservatively as possible with respect to sensitivity: only samples with ≥30%
inhibition on both tests were identified as WNV antibody-positive.
Indirect ELISAs:
Tests for antibodies to BHV-1, PI-3, and BRSV were performed at Prairie Diagnostic
Services (Saskatoon, SK, Canada) using a protocol developed for cattle (Durham and
Sillars, 1986; Durham and Hassard, 1990) that had been adapted for cervids. The adapted
test employs a protein G-enzyme conjugate instead of an anti-bovine-conjugate. Protein
G is not species-specific and binds a variety of immunoglobulin G molecules.
Preliminary observations made at the laboratory with a limited number of samples
indicated that the function of protein G in this test with cervid samples was similar to that
with cattle samples (D. Godson, unpublished data). To ensure that test data for serum and
FP samples were directly comparable, serum was diluted 1:50 initially as per the
protocol, whereas FP eluates were diluted 1:5 (equivalent to 1:50 serum dilution). Results
were expressed in ELISA units (EU) and calculated as (OD blanked sample ÷ OD
blanked positive control [arbitrarily assigned a value of 100]) × 100 (Durham and
Hassard, 1990). The thresholds used were the laboratory thresholds for cattle, which were
based on comparisons with virus neutralization (Durham and Hassard, 1990). For all
three iELISAs, a result ≥14 EU indicated antibody-positive status.
Virus Neutralization Tests:
Antibody testing for BVDV-I and BVDV-II was done at Prairie Diagnostic Services
using VN assays described for cattle (Waldner and Campbell, 2005). There are no
59
species-specific reagents involved in these assays and results are expressed as titers (i.e.,
the dilution at which antibodies in the sample no longer block BVDV infection of the
susceptible cell line). In brief, the protocol for serum involved threefold dilutions that
started at 1:3 (a serum dilution of 1:6 when the virus inoculum was added). Each test was
run in duplicate. If there was disagreement between the duplicates, for example a 1:6
result for one well versus 1:18 (one dilution step higher) for the duplicate well, the
halfway point between them (1:12 in the example) was the final result recorded. Filterpaper eluates were assayed identically to serum, except that eluates (estimated 1:10
serum) were added undiluted to virus inoculum in the first well, thus making 1:20 the
lowest titer detectable for the FP samples. For cattle serum, results 1:6 or higher on these
BVDV VN assays indicate antibody-positive status (D. Godson, unpublished data).
Applying these same thresholds, we identified reindeer with serum titers ≥6 and FP titers
≥20 (lowest detectable for FP samples) as positive.
Analysis
Test Performance:
For all seven assays in the study, FP test performance was analyzed using EpiTools
calculators (Sergeant, 2009) to determine sensitivity (SE), specificity (SE), serum
prevalence, FP prevalence, and positive and negative predictive values for FP.
“Comparable to serum” was defined as SE ≥80% and SP ≥80% based on test proficiency
criteria of the United States Clinical Laboratory Improvement Advisory Committee
(Astles, 2010). Clopper-Pearson confidence intervals (CI), which are exact and
conservative (i.e., 95% CI indicating at least 95% confidence) were calculated. For five
60
of the pathogens investigated (BHV-1, PI-3, BRSV, BVDV-I, BVDV-II), FP-serum
sample pairs were obtained from a group of 12 vaccinated animals at three time points
(Table 1). These multiple measures per animal were used as independent data in test
performance analysis for several reasons: i) the purpose of the experiment was strictly to
compare serum and FP for detection of antibodies; ii) testing a range of antibody levels is
more robust than testing at a single post-vaccination time point; iii) the three blood
collections were separated by intervals of 1 month minimum; iv) it was difficult to
acquire large numbers of animals for these experiments (only a small number of reindeer
were available for vaccination; very few commercial herds were accessible). The
limitations of the FP test performance estimates and CIs for these five pathogens are
recognized and are addressed in the discussion.
On both BVDV VN assays, some test-plate wells containing the first dilution of serum
(1:6) and the first two FP eluate dilutions (i.e., up to 1:60) were unreadable. The
laboratory attributed this distinct loss of susceptible cells to sample toxicity or another
cause unrelated to virus cytopathogenicity. The laboratory-recorded results in these
“toxic” cases were ranges that made it impossible to categorize the sample as positive or
negative (e.g., ≤6 for serum or ≤20, ≤40 or ≤60 for FP). Thus, any sample pair with a
toxic result for serum or FP (or both) was excluded from analysis (see n values in Table
2).
61
Other Evaluations:
The five ELISA assays generated continuous data (% inhibition or EU) and these results
were plotted to allow visual comparison of the serum and FP data for each individual
sample pair tested (PASW Statistics 18.0, SPSS Inc., Chicago, IL, USA).
We analyzed variability between duplicate FP strips collected from individual reindeer by
correlation and calculated 95% predictive intervals. Multiple-measures data from the 12
vaccinated animals were excluded from this analysis because it is statistically impossible
to calculate 95% predictive intervals for such data. Thus, we correlated FP duplicates for
the N. caninum, WNV, PI-3, and BRSV tests (all ELISAs), but did not do so for the
BVDV-I or –II tests (VN assays).
Results
Filter-paper test performance results relative to serum in the seven antibody assays are
shown in Table 2 and Figure 2. Sensitivity estimates exceeded 85% for five of the seven
tests (most SE ≥95%) but were lower (range, 71% to 82%) for the PI-3 and BRSV
iELISA runs. Specificity estimates were consistently high, ranging from 92% for the N.
caninum and WNV cELISAs to 98% for the PI-3 and BRSV iELISAs.
On the BHV-I, PI-3, and BRSV iELISAs, the FP results tended to be lower than the
matched serum results (Fig. 2d-f). In the cELISA for WNV with mAb2, FP results tended
to be higher than serum results (Fig. 2c). No consistent pattern was discernable when FP
62
and serum findings were compared in the cELISAs for N. caninum and for WNV with
mAb1, respectively (Fig. 2a,b).
Correlations between FP duplicates were strong on all the ELISAs (Fig. 3a-e), with the
lowest r value being 0.96 (WNV mAb1 test). The 95% predictive intervals were widest
for FP duplicates in the WNV mAb1 cELISA. The 95% predictive intervals for the other
four assays were comparable, with N. caninum cELISA intervals the narrowest of all.
Discussion
We assessed the performance of FP samples relative to serum for detecting antibodies to
seven pathogens of known or potential importance in Rangifer species (Table 1). These
agents may pose threats to reproduction, recruitment, or food safety, and may be
introduced to naïve Rangifer populations by host range shifts or invasions of domestic or
wild species.
Neospora caninum is an apicomplexan parasite that causes abortion storms, neonatal
mortality, and is transmitted transplacentally in cattle (Dubey, 2003). Clinical neosporosis
has been suspected, though not confirmed, in Rangifer (Kutz et al., 2012a). Overt disease
and vertical transmission have been reported in other cervids, and antibodies to N.
caninum have been detected in free-ranging caribou and deer populations (Woods et al.,
1994; Dubey et al., 1996; Dubey, 2003; Dubey and Thulliez, 2005; Dubey et al., 2009;
Stieve et al., 2010). In this paper, we report high seroprevalence of N. caninum (73%
63
[33/45], Table 2) in captive reindeer.
West Nile virus is a zoonotic flavivirus that amplifies in avian hosts and is transmitted by
Culex spp. mosquitoes (Romich, 2008). Neither clinical disease nor antibodies to WNV
have been documented in free-ranging Rangifer; however, severe neurological disease
and death from WNV has been reported in naturally infected captive reindeer in the
Midwestern United States (Palmer et al., 2004a). Northward shifts of host and vector
ranges with climate change could potentially increase the risk of WNV infection for
caribou and reindeer. In this paper, we report high seroprevalence of WNV (54% [14/26],
Table 2) in a captive reindeer herd in Alberta, Canada.
Bovine herpesvirus-1, PI-3, BRSV, and BVDV-I and -II are all contributors to the bovine
respiratory disease complex. Bovine herpesvirus-1 and BVDV also impact other organ
systems, with effects ranging from abortion and weak calves to diarrhea and lethal
disease (Radostits et al., 2007). These five viruses are important for Rangifer species in
that i) current and future agricultural development may bring cattle pathogens into
contact with caribou and reindeer, and ii) bovine virus assays may detect antibodies that
are induced by related but (in some cases) undescribed viruses that are circulating in
Rangifer populations. The effects of these bovine viruses or their relatives in caribou and
reindeer are not fully understood; however, antibodies to BHV-1, PI-3, and BVDV have
been detected in and related viruses have been isolated from Rangifer species (Elazhary
et al., 1979, 1981; Zarnke, 1983; Ek-Kommonen et al., 1986; Rockborn et al., 1990;
Rehbinder et al., 1992; Stuen et al., 1993; Avalos-Ramirez et al., 2001; Jordan et al.,
2003; Tessaro et al., 2005; Das Neves et al., 2010; Johnson et al., 2010).
64
Filter-Paper Test Performance
The Nobuto FP samples from reindeer performed well relative to serum in all seven
antibody assays (see further discussion of PI-3 and BRSV iELISAs below), and we
observed minimal variability between duplicate FP strips from individual animals
(r>0.96 for ELISAs; Fig. 3). The same was found in a previous study on caribou FP
samples for detection of Brucella antibodies (see Chapter 2).
Several points about the animals’ immunoreactivity, the test protocols, and our study
design are noteworthy for assessing antibody detection. None of the vaccinated reindeer
tested seropositive on any of the five bovine assays at baseline (i.e., the day of
vaccination), which suggests that the reactivity later detected in those animals was
induced by the vaccine antigens. The antibodies detected in the non-vaccinated reindeer
(the animals tested for N. caninum, WNV, and some of those tested for PI-3 and BRSV)
were induced by natural exposure to agents that we did not specifically isolate or identify.
Although cervid- or Rangifer-specific pathogens related to those in this study circulate in
populations, few Rangifer-specific antibody assays have been developed (Tessaro et al.,
2005; Avalos-Ramirez et al., 2001; Das Neves et al., 2010). In our study, we used tests
that were initially developed for pathogens of bovines or other domestic mammals, and
for use in these host species. The optimal laboratory threshold values for reindeer serum
in these assays are not known, so we applied the thresholds that the test developers had
established for serum of domestic mammals (the WNV assay) or for cattle specifically
(all other assays in the study). We recognize that cattle thresholds may not be the most
accurate for indirect antibody assays performed with serum from a different species, and
65
that good-quality reindeer or other cervid immunoglobulin standards are needed for
comparison studies and validations. However, our iELISA protocols used protein G
conjugate, thus minimizing effects related to cross-reactivity or to binding affinity of
secondary antibody. Moreover and most salient, our sole purpose was to assess the
performance of Nobuto FP samples relative to serum in the seven antibody tests. To
ensure this was done in the most robust manner, for each respective assay we took several
measures that allowed results for matched FP-serum samples to be directly compared:
We tested FP eluates and sera against the same controls, used the same kit or reagent lot
number if applicable or possible, tested sample pairs simultaneously, and tested pairs in
duplicate (WNV serum testing the only exception).
Competitive ELISAs:
We found that Nobuto FP samples from reindeer performed comparably to serum in the
N. caninum and WNV cELISAs (SE 98% and SP 92% for both; Table 2). The SE of FP
in the N. caninum kit assay may be reduced (normally neat serum is tested, whereas FP
eluates are 1:10 serum dilution); however, the kit manufacturer states that only 5% of
bovine sera fall within ± 5% of the threshold value (VMRD Inc., Pullman, WA, USA)
and our results for reindeer were consistent with this (Fig. 2a). Other investigations of
matched serum-FP samples from animals and humans in cELISAs have also shown
strong serum-FP concordance (Afshar et al., 1987; Chitambar and Chadha, 2000).
Specific to cervids, our group previously compared 185 caribou serum-FP pairs in a
Brucella cELISA and demonstrated FP efficacy on this platform (SE 89% and SP 99%;
see Chapter 2). Other researchers validated FP disc blood samples from domestic Sika
66
deer (Cervus nippon) in blocking and iELISAs for hepatitis E virus and successfully
tested hunter-collected FP samples from wild Sika deer (Yu et al., 2007). Authors have
also noted superior performance of FP blood samples in cELISA compared to other
antibody tests, referring to specific benefits such as reduced cross-reactivity (in caribou)
and possible earlier detection of antibody response (in cattle and sheep) compared to FP
iELISA (see Chapter 2; Afshar et al., 1987).
Indirect ELISAs:
Studies in humans and various animals, including some cervids, have shown that blood
samples collected on various forms of FP work well in iELISAs for hepatitis A and E,
rubella, and Aujeszky’s disease viruses, as well as bluetongue virus (BTV) and epizootic
hemorrhagic disease virus (EHDV) (Banks, 1985; Afshar et al., 1987; Helfand et al.,
2001; Yu et al., 2007). Testing of 185 caribou Nobuto FP samples relative to serum in a
Brucella iELISA yielded estimates of SE 100% and SP 99% (see Chapter 2). In the
present study, we found that, in general, Nobuto FP samples from reindeer performed
well (relative to serum) in the BHV-1, PI-3, and BRSV iELISAs using the laboratory
thresholds (which were established for cattle serum but which had appeared to be
comparable for cervids in the laboratory’s preliminary testing). Relative SP was high for
all three iELISAs (SP ≥96%) and SE was also high for BHV-1, but was somewhat lower
(range, 71% to 82%) on the PI-3 and BRSV tests (Table 2, Fig. 2 e, f). Incremental
analysis was beyond the scope of this study, but reducing the FP threshold value from 14
EU to 10 EU for these two assays increased SE considerably (SE ≥87% on all test runs)
while retaining acceptable SP (SP ≥90% on three of the four runs) (Table 3). This
67
adjustment brought FP performance in these two tests into accord with all other assays in
the study.
Virus Neutralization:
There are fewer reports on FP use in VN, but evaluation of two deer species’ serum-FP
pairs on VN assays for BTV and EHDV showed FP samples to be reliable (Stallknecht
and Davidson, 1992; Dubay et al., 2006). We found that reindeer Nobuto FP samples
performed comparably to serum in the BVDV-I and -II assays (SE ≥91% for all but one
run, SP 93%; Table 2) and we note that SE was potentially compromised by the higher
minimal detectable titer with FP samples (1:20) than with serum (1:6).
We encountered toxicity problems with reindeer serum and FP on VN, which made it
necessary to exclude those results from the analysis. For the BVDV-I analysis, four
sample pairs were excluded from both runs (i.e., eight sample pairs excluded in total;
excluded pairs were from the same four animals in each of the two runs) and only the FP
sample was toxic in six of these eight pairs. For the BVDV-II analysis, six sample pairs
were excluded from Run A and 10 sample pairs were excluded from Run B (i.e., 16 pairs
excluded in total). Only the FP sample was toxic in 15 of these 16 sample pairs. Most of
the toxic samples (serum and FP) were obtained at one specific collection time point and
had been stored for 2 months prior to testing, whereas the others had been stored for only
1 month. Although not statistically tested, the numbers suggest that FP eluates may be
more likely to fail (unreadable results) on VN than serum. In general, this may reflect
effects of free hemoglobin or other red blood cell intracellular contents in FP eluates;
68
however, there may also have been an issue peculiar to this one collection time point
(e.g., contamination, field conditions, incomplete drying of FP strips before storage) that
negatively affected the FP samples primarily. It may also be that longer storage affects
the integrity of FP samples for VN testing. To our knowledge, toxicity with FP blood
samples on VN has not been reported previously.
Considerations with Multiple-Measures Data
The need to vaccinate some reindeer in order to validate FP samples for detection of
antibodies to some of the pathogens (BHV-1, PI-3, BRSV, BVDV-I, BVDV-II) presented
analytical challenges. Three collections of serum-FP from these animals at serial time
points (Table 1) allowed us to examine samples with a range of antibody levels during
immune response post-vaccination; however, this created a situation of matched pairs
(serum and FP from each animal) combined with potential random effects (individual
animal effects, potential dependence between time points of collection). We are unaware
of statistical testing that can account for random effects with matched-pairs data. Given
that these observations were used as independent data in the test performance analysis,
the resultant point estimates may be somewhat high and the calculated CIs somewhat
narrow. Note that the estimates and CIs for PI-3 and BRSV are relatively more precise as
these are based on considerably larger sample sizes that included animals in naturally
exposed herds that were tested only once.
Benefits and Application
Wildlife disease surveillance requires sensitive assays that can be used to screen
69
populations for disease or for pathogen exposure. Our findings indicate that, when tested
within 2 months of collection, Rangifer Nobuto FP samples are comparable to serum in
all seven of the antibody assays examined, and there is minimal variation among FP
samples from an individual animal. Though limited sample sizes may have affected the
precision of our estimates, the ELISA test performance values (using adjusted FP
threshold for PI-3 and BRSV) were in line with a larger study of caribou FP samples in
ELISAs for Brucella (see Chapter 2). Our work reveals that FP samples from Rangifer
are effective in five additional ELISA protocols, and the higher and more consistent SE
and SP in the cELISAs (N. caninum and WNV) suggest that this platform is preferable to
iELISA when using FP samples. The same preference was noted in the Brucella antibody
detection study (see Chapter 2). Filter-paper samples from Rangifer also function well in
BVDV VN assays, but inherent features of this protocol constrain FP test performance to
some degree. As well, there is an apparent issue with toxicity of FP samples leading to
results that cannot be interpreted, and this bears further investigation.
Caribou and reindeer have integral value in northern ecosystems and are culturally and
economically important throughout the Arctic (Ulvevadet and Klokov, 2004). Robust
methods for monitoring Rangifer population health are needed to improve disease
surveillance, inform management, and identify food safety risks. Filter-paper blood
sampling is a practical, effective tool for wildlife health assessment, and FP collection by
hunters and other laypersons in the field can extend the temporal and geographic scope of
surveillance programs. This technique appears to have real potential for widespread
application; however, the level of validation described herein is necessary to be confident
70
that pathogen exposure levels indicated by FP samples truly reflect serum findings. The
novel finding that Rangifer FP testing is effective for detecting antibodies to seven
additional pathogens expands the utility of this method.
Acknowledgements
We are mindful and appreciative of the reindeer who were the sample sources for this
study. Marianne Jorgensen’s meticulous care of the U of Calgary reindeer herd was a
major contribution to this research, and we also acknowledge Barbara Smith and Dr.
Greg Muench in this regard. The samples could not have been obtained without the
support of the entire Kutz Lab team 2007-2009. We owe a great debt of thanks to this
group, whose dedication and cheerful assistance was invaluable and maximized safety
and respect for the animals involved. Lori Hassard, Laura Bond, and Roberta Yemen did
critical technical work and the contributions of John Verhoeven were also sincerely
appreciated. This research was supported by numerous funding sources: International
Polar Year Funding from the NSERC Special Research Opportunity Program,
Environment Canada/Natural Resources Canada, Nasivvik Centre for Inuit Health and
Changing Environments (Canadian Institutes of Health Research), Alberta Innovates
Technology Futures, Northern Scientific Training Program (Indian and Northern Affairs,
Government of Canada), Arctic Institute of North America, and the University of Calgary
Faculty of Veterinary Medicine.
71
TABLES
Chpt 3. Table 1. Pathogen effects and study summary
Summary of pathogen effects, antigen exposure status at time of the study, and study
design information for the animal groups sampled. Competitive enzyme-linked
immunoassays (cELISA), indirect ELISAs (iELISA), or virus neutralization assays (VN)
detected antibodies to seven pathogens (abbreviations defined in main text). Paired
samples (serum and filter-paper blood strips) were obtained from each animal at
collection.
PATHOGEN EEFECTS IN Rangifer sp. N. caninum Potential reproductive impact West Nile virus Neurological disease, death BHV‐1 Oral lesions, enteritis, reproductive impacts PI‐3 HERD: ANTIGEN EXPOSURE: Comm. (C) Vaccinal (V) Research (R) Natural (N) C N R TEST a
SAMPLE COLLECTION PAIRS DATES TESTED (n) cELISA 39 Apr 2008 Unknown cELISA 6 Aug 2008 C N cELISA 26 Aug 2008 R V iELISA 36 May/Jul/Aug 2009 C R N V iELISA iELISA 26 c
36 Aug 2008 May/Jul/Aug 2009 Aug 2008 b
c
Unknown BRSV Unknown C N iELISA R V iELISA 26 c
36 May/Jul/Aug 2009 BVDV‐I Unknown R V VN 36 c
May/Jul/Aug 2009 BVDV‐II Unknown R V VN 36 c
May/Jul/Aug 2009 Abbreviations: Comm.: Commercial a
Seropositives detected during prior herd health testing No vaccinal exposure prior to collection c
12 animals were sampled at three time points: Pre‐vaccination and 4 and 8 weeks after booster (which was 4 weeks after initial injection) b
72
Chpt 3. Table 2. Filter-paper test performance Filter-paper (FP) test performance for the antibody assays in the study with 95% ClopperPearson exact confidence intervals (CI) shown. Sample pairs were tested in duplicate
(Run A and Run B) and FP performance estimates were calculated by comparing to
serum results. In cases where the initial point estimate derived was 100%, the median
unbiased estimate (denoted with *) was calculated and reported instead. Competitive
enzyme-linked immunoassays (cELISA), indirect ELISAs (iELISA), or virus
neutralization assays (VN) detected antibodies to seven pathogens (abbreviations defined
in main text).
73
Table 2.
74
Table 2. (cont’d)
75
Chpt 3. Table 3. Laboratory and adjusted thresholds – PI-3, BRSV assays
Relative sensitivity (SE) and specificity (SP) of reindeer filter-paper testing for the PI-3
and BRSV indirect ELISA (iELISA) runs. Results using the laboratory threshold (14 EU
for serum and FP) are compared with results using an adjusted threshold for FP (10 EU).
In cases where the initial point estimate derived was 100%, the median unbiased estimate
(denoted with *) was calculated and reported instead. Clopper-Pearson exact 95%
confidence intervals (CI) are shown.
AGENT and TEST SAMPLE PAIRS ANALYZED (n) RUN LABORATORY SE % (CI) PI‐3 iELISA 62 A 62 B 80.0 (61.4‐92.3) 77.4 (58.9‐90.4) ADJUSTED SE % (CI) LABORATORY SP % (CI) ADJUSTED SP % (CI) 96.7 (82.8‐99.9) 87.1 (70.2‐96.4) 96.9 (83.8‐99.9) 97.8* (90.8‐100) 90.6 (75.0‐98.0) 90.3 (84.7‐99.9) 92.9 (76.5‐99.1) 89.3 (71.8‐97.7) 97.1 (84.7‐99.9) 98.0* 91.6‐100 79.4 (62.1‐91.3) 97.1 (84.7‐99.9) BRSV iELISA 62 A 62 B 82.1 (63.1‐93.9) 71.4 (51.3‐86.8) 76
FIGURES
Chpt 3. Figure 1. Filter-paper set
A set of Nobuto filter-paper strips with the absorbent portion (A) of a strip identified.
77
Chpt 3. Figure 2. a-f. Filter paper vs. serum – ELISAs
Comparison of serum and filter-paper (FP) eluate results for the five pathogens
investigated with competitive or indirect enzyme-linked immunosorbent assay (cELISA,
iELISA). For each agent/test (abbreviations defined in main text), the run with the lowest
sensitivity is displayed (see Table 2). The WNV assay was a two-step protocol with
separate monoclonal antibodies (mAb1 and mAb2). To plot the graphs, the pairs of data
(serum and FP results) were first sorted according to ascending serum values. Each
number on the x-axis has a pair of data (the animal’s serum result and its matching FP
result) plotted vertically above it. Dotted lines denote laboratory threshold values.
78
Figure 2.
79
Chpt 3. Figure 3. a-e. Duplicate filter paper results – ELISAs
Plots of duplicate filter-paper (FP) results (Run A and B) for the competitive and indirect
enzyme-linked immunosorbent assays (cELISAs, iELISAs) with each graph showing line
of best fit, 95% predictive intervals, and r value. Some data points (black circles) in the
graphs overlap. The samples for the PI-3 and BRSV analyses (n=38) were from 26
commercial reindeer and 12 vaccinated research animals (the sample set collected 8
weeks post-vaccination). The WNV cELISA was a two-test protocol with two distinct
monoclonal antibodies (mAb1 and mAb2).
80
Figure 3.
81
CHAPTER FOUR
FILTER-PAPER BLOOD SAMPLES
FOR WILDLIFE SEROLOGY: EVALUATING STORAGE AND TEMPERATURE CHALLENGES OF FIELD COLLECTIONS
Authors: Curry P.S., C. Ribble, W.C. Sears, K. Orsel, W. Hutchins, D. Godson,
R. Lindsay, A. Dibernardo, M. Campbell, and S.J. Kutz
Article in submission – Journal of Wildlife Diseases
82
Abstract
Filter-paper (FP) blood sampling can facilitate wildlife research and expand disease
surveillance. Nobuto FP samples from caribou and reindeer (Rangifer tarandus ssp.) have
been shown to perform comparably to serum in competitive enzyme-linked
immunosorbent assays (cELISAs) for Brucella spp., Neospora caninum, and West Nile
virus, and in indirect ELISAs (iELISAs) for Brucella spp., bovine herpesvirus-1 (BHV1), parainfluenza virus type 3 (PI-3), and bovine respiratory syncytial virus (BRSV).
Similar sensitivity (SE) and specificity (SP) were also observed in virus neutralization
(VN) for bovine viral diarrhea virus (BVDV) types I and II, though toxicity of FP eluates
may be an issue in this platform. To assess the robustness of FP samples for wildlife
serology, we evaluated the performance of FP samples from Rangifer in these nine assays
after simulating potential challenges of northern field collections: extended storage and a
small trial of processing/storage regimes involving freezing and/or drying. Sample pairs
(serum and FP) were collected from Rangifer populations between 2007 and 2010, and
were tested in duplicate. Filter-paper performance was calculated based on serum results,
and SE and SP ≥80% were taken to indicate comparability. Filter-paper performance was
determined after 2 months of storage dry at room temperature, and after two longer
periods. After 1 year (compared to frozen serum stored for the same period), SE was
≥88% for all but two assays (68% BHV-1; 75% PI-3) and SP remained >90%. Thus,
Nobuto FP samples from Rangifer performed comparably to serum in seven of the
antibody assays after 1 year of dry storage at room temperature. The trial of
processing/storage conditions suggested no detrimental effect of freezing FP samples at
83
collection as opposed to drying them for storage.
Introduction
Wildlife sampling tends to be challenging, and the simplicity and practicality of blood
collection on filter paper (FP) make this an appealing alternative to conventional
collection. Filter papers are lightweight, non-fragile, and compact for sampling kits, and
FP samples are easy to collect, carry, and ship, and require no processing equipment in
the field. As well, FP blood collection by subsistence hunters and other laypeople could
increase the temporal and spatial coverage of disease-surveillance programs (Brook et al.,
2009; Curry, 2010; see Chapter 2).
Human FP blood testing was introduced in the 1960s and its uses continue to expand
(Guthrie and Susi, 1963; Mei et al., 2001; McDade et al., 2007). For wildlife, relatively
few FP validation studies have been done and the literature is narrower in scope. Recent
publications on the use of FP blood samples from animals describe serological and
toxicological assays (Sacks et al., 2002; Trudeau et al., 2007; Yu et al., 2007; Dusek et
al., 2011; Kalayou et al., 2011). Specific to cervids, FP samples were reported to be
effective in antibody assays for hepatitis E virus in Sika deer (Cervus nippon) (Yu et al.,
2007), for bluetongue virus and epizootic hemorrhagic disease virus in mule deer and
white-tailed deer (Odocoileus virginianus and O. hemionus, respectively) (Stallknecht
and Davidson, 1992; Dubay et al., 2006), and for Brucella spp., Neospora caninum, West
84
Nile virus (WNV), and five bovine viruses in caribou and reindeer (Rangifer tarandus
ssp.) (see Chapters 2 and 3).
However, even the best diagnostic test is limited by sample quality, and field conditions
and logistics are key determinants of this. For high-latitude fauna such as caribou and
reindeer, potential threats to sample integrity include severe cold, lack of controlled
transport conditions from remote sampling locations to processing sites, frequent need for
long-term storage before shipping for analysis, and temperature fluxes and in-transit
delays during shipping. Wildlife professionals grapple with issues of sample-storage
space and costs, and often have to decide how to collect, process, and store samples in
non-ideal situations. To inform decisions about FP use for wildlife serology, we tested
the robustness of Rangifer FP samples in nine antibody assays after exposure to storage
and collection challenges that typify remote, cold-climate field settings.
Materials and Methods
Design, Animals, and Antibody Status
The two-part study spanned July 2007 through December 2010. Part I probed the effects
of storage time on FP test results, and Part II was a small trial that primarily assessed the
effects of processing/storage conditions (Table 1). The assays were competitive and
indirect enzyme-linked immunosorbent assays (cELISA and iELISA, respectively) for
Brucella spp.; cELISAs for N. caninum and WNV; iELISAs for bovine herpesvirus type
85
1 (BHV-1), parainfluenza virus type 3 (PI-3), and bovine respiratory syncytial virus
(BRSV); virus neutralization assays (VN) for bovine viral diarrhea virus types I and II
(BVDV-I and -II) (see more details below).
Blood sampling of live animals was done in compliance with the guidelines of the
Canadian Council on Animal Care. The samples for Brucella spp. serology came from
free-ranging hunter-killed caribou near Coral Harbour (64°11' N, 83°21' W) on
Southampton Island, Nunavut, Canada (Nunavut Wildlife Research Permit WL 000892)
(see Chapter 2). Those for N. caninum testing were collected from a commercial reindeer
herd in central Alberta, Canada (approx. 52°20' N, 112°41' W) (see Chapter 3). All other
samples were from a previously unvaccinated small research herd of reindeer at the
University of Calgary in Calgary, Alberta (51°5' N, 114°5' W). For the five bovine
viruses (BHV-1, PI-3, BRSV, BVDV-II and –II), we injected a primary label dose of the
killed-virus vaccine Triangle® 4 + BVDV-II (Wyeth Animal Health, Guelph, ON,
Canada) and a booster dose 4 weeks later (Part I of study, n=12 animals) or 3 weeks later
(Part II, n=3 animals). To examine a range of antibody levels, these animals were
sampled at baseline (pre-vaccination), at or after the booster dose, and several weeks after
booster (Table 1). For WNV, we injected a primary label dose of the killed-virus equine
vaccine West Nile-Innovator (Fort Dodge, Wyeth Animal Health, Fort Dodge, IA, USA)
followed by a booster 4 weeks later. Only adults (10 of the 14 reindeer sampled for WNV
testing) were vaccinated and blood samples were collected 6 months later.
86
Samples
Nobuto FP strips (Toyo Roshi Kaisha, Ltd., Tokyo, Japan; distributor Advantec MFS
Inc., Dublin, CA, USA) were used for blood sampling. At each collection time, we
obtained matched serum and FP (sample pairs) from each animal. Detailed descriptions
of these methods in free-ranging and captive Rangifer have been described (see Chapters
2 and 3). Briefly, for the hunter-killed caribou, a jugular or femoral vein was severed to
provide fresh whole blood for 10-mL Vacutainer® tubes without anticoagulant (BectonDickinson, Mississauga, ON, Canada), and for saturating the FP strips. For the live
reindeer, after venipuncture collection into 10-mL Vacutainer® tubes without
anticoagulant, we drew a separate 12- or 20-mL syringe of blood and immediately
saturated the strips. We collected multiple sets of Nobuto FP samples mounted on
cardboard “handles” (see Chapter 3 Fig. 1) and these were immediately placed in labeled
#10 envelopes (one per animal).
Processing and Treatments
Blood tubes were centrifuged 15 minutes at 3,500 g and serum aliquots were stored at –
20 °C until analysis. Filter-paper samples for each pathogen assay were subjected to one
or more of four possible processing/storage regimes or “treatments” (Table 1):
•
Dry-Dry (DD): FP samples were collected at ambient temperature and the FP
envelope was placed in a Ziploc® Brand (SC Johnson, WI, USA) plastic bag (i.e.,
maintained above freezing in same conditions as tube samples until processing).
Upon return to the lab 2 to 8 hours later, the samples were dried in racks on the bench
87
overnight, then stored dry at room temperature (RT; 15-22 °C) until analysis;
•
Dry-Freeze (DF): As for DD, except stored dry at RT for only 2 weeks, then stored at
–20 °C until analysis;
•
Freeze-Freeze (FF): FP samples were collected as above and the Ziploc® bag with
FP envelope was immediately placed in an ice cooler. Upon return to the lab 2 to 8
hours later, the samples were transferred to a –20 °C freezer until analysis;
•
Freeze-Dry (FD): As for FF, except stored at –20 °C for only 2 weeks, then
thawed/dried overnight and stored dry at RT until analysis.
For dry storage, FP envelopes were grouped and placed in Ziploc® plastic bags with
desiccant (Humidity Sponge™, VWR International LLC, Mississauga, ON, Canada). The
bags were kept in multipurpose, non-airtight, clear polypropylene storage boxes
(approximately 60-L size) with snap lids and desiccant was checked regularly (2 week
intervals for month 1, approximately 12-week intervals thereafter) and replaced as
needed. For –20 °C storage, no desiccant was used and each FP envelope was kept in an
individual Ziploc® bag in a laboratory-grade –20 °C freezer unit without auto-defrost.
In Part I of the study (storage time), all FP samples were dried immediately and stored
dry at RT (DD treatment). Serum-FP pairs were assayed at one initial time point (T1 , ≤2
months of storage) and at two longer storage times (Table 1). For the 24-month Brucella
testing, it was only possible to assay a subset of the 184 total serum-FP pairs (n=88
which, with controls, filled one 96-well test plate). To ensure the range of anti-Brucella
reactivity was represented in the subset, we categorized the 2-month serum results for the
Brucella cELISA as low, medium, or high reactivity (<20%, 20-40%, and ≥80%
88
inhibition, respectively) and used a random-number generator to identify 34, 21, and 33
samples, respectively (i.e., as close to equal numbers for each reactivity level as were
available) from these groups.
In Part II of the study (trial of processing/storage conditions), all samples were tested 12
months after collection.
Testing
The cELISAs, iELISAs, and VNs performed and the respective threshold values used are
listed in Table 1. Specifics for Nobuto FP elution methods, shipping, laboratories, assay
protocols, and test thresholds are detailed (with original references cited) in Chapters 2
and 3. Filter-paper samples were eluted according to manufacturer instructions (400 uL
buffer per FP strip), yielding eluates approximately equivalent to 1:10 serum (Toyo Roshi
Kaisha, Ltd., Tokyo, Japan). The FP samples stored at –20 °C were dried overnight
before elution. Laboratories carried out batches of FP and serum tests simultaneously
(same day) and in blinded fashion, and samples were assayed as soon as possible upon
receipt. Technicians adjusted the initial protocol steps as needed to account for the 1:10
dilution factor with FP samples and, thus, make serum and FP samples directly
comparable (see Chapters 2 and 3). Serum-FP pairs from each animal were assayed
simultaneously (same day) and in duplicate runs using separate FP samples for Run A
and Run B. The WNV assay was a two-step testing process, with each step involving a
monoclonal antibody that targeted a different protein of WNV. At the time of testing, the
monoclonals in use were 3.112G-NS-1 and 7H2-Env, which targeted nonstructural and
89
envelope proteins, respectively. We followed the same conservative method and rationale
previously described for establishing positive or negative status in this two-test assay (see
Chapter 3).
Analysis
Storage Time:
We analyzed FP test performance by i) comparing FP samples at each stage of storage to
the “ideal” (T1, or shortest-storage) serum results; ii) comparing the paired FP-serum
results after each storage period. EpiTools calculators (Sergeant, 2009) were used to
determine FP test sensitivity (SE), specificity (SP), serum prevalence, FP prevalence, and
positive and negative predictive values for FP. Clopper-Pearson confidence intervals
were calculated for estimates. Comparability to serum was defined as SE ≥80% and SP
≥80% (Astles, 2010; see Chapter 3).
Each of the five ELISAs generated continuous data (% inhibition, % positivity, or ELISA
Units [EU]). These were plotted to show FP results at successive storage times relative to
T1 serum and thresholds. For each assay, the Run A serum results at T1 were plotted with
FP results for each stage of storage (the lowest-sensitivity FP run identified during testperformance analysis).
It was necessary to exclude some of the BVDV VN results from test performance
analysis. On both these assays, the threshold values were titer ≥1:6 for serum and titer
≥1:20 for FP (detailed explanation in Chapter 3). In these tests, some of the wells
90
containing the first dilution of serum (1:6) and the first two FP eluate dilutions (i.e., titer
steps up to 1:60) were unreadable because of loss of susceptible cells resulting from
sample toxicity (as opposed to pathogenic virus effects). The laboratory results for these
toxic cases were listed as ranges as opposed to individual titers (e.g., ≤1:6 for serum or
≤1:20, ≤1:40 or ≤1:60 for FP), and were, therefore, impossible to categorize as positive or
negative based on the threshold values specified above. Consequently, any sample pair
with a toxic result for serum or FP (or both) was excluded.
Processing/Storage Conditions:
Performance in the WNV cELISA was analyzed as above. Though the n=3 animals in the
trial of four FP treatments precluded statistical testing, we graphed each animal’s findings
for serum and each FP treatment in the BHV-1 and PI-3 iELISAs. We also converted
these data to categorical results (antibody-positive or -negative according to threshold)
and assessed for agreement in two ways: i) comparing serum vs. FP results for each
treatment (n=9 per serum vs. FP comparison); ii) comparing FP vs. FP for all possible
pairings of the four treatments at each antibody level (i.e., “low” or pre-vaccination
collection, “medium” 3 weeks later [booster], and “high” 6 weeks later) (n=3 per FP vs.
FP comparison).
91
Results
Part I – Storage Time
The number of toxic results in the VNs reduced sample sizes such that no BVDV data
could be analyzed relative to T1 serum (Table 2). The same issue limited the staged
analysis for these assays to two storage periods (1 and 12 months) instead of three (Table
3).
The means of the test performance estimates for Runs A and B revealed that FP samples
performed comparably to T1 serum in all eight assays that were analyzed for storage time
(ranges: SE 89% to 98%, SP 91% to 99%; Tables 2, 3). One test-performance estimate
(50% SP for BRSV at T1) was excluded from the analysis as an outlier. When FP results
at 12 or 17 months storage were compared to T1 serum, SE was ≥81% (comparable to
serum) in the Brucella, N. caninum, and BRSV assays, but was lower in the BHV-1 and
PI-3 assays (62% and 54%, respectively) (Table 2). In contrast, when the FP results at 12
or 17 months storage were compared to their same-storage-time serum results, SE was
≥88% for Brucella, N. caninum, BRSV, and both BVDV assays, and was 68% and 75%
for the BHV-1 and PI-3 tests, respectively (Table 3; additional performance statistics in
Appendix 4A). After 12 or 17 months of storage, the average drop in SE relative to T1
serum was approximately 15% (range, 3-36%), whereas the corresponding drop in SE
relative to same-storage serum was 7% (range, 0-23%).
92
The slopes of the iELISA results for BHV-1, PI-3, and BRSV (i.e., samples from
vaccinated animals) were shallower crossing the threshold than those for the Brucella and
N. caninum assays (i.e., samples from naturally exposed animals) (Fig. 1).
Part II – Processing/Storage Conditions
The WNV cELISA performance estimates for FP samples after 12 months in DD and FF
conditions were the same (means SE 80% and SP 90%) (Table 4). For the other assays in
the FP treatments trial, there were no obvious differences among the four treatments (Fig.
2).
The categorical (positive/negative) test results (data not shown) for the PI-3 iELISA
results agreed in all treatment comparisons. For the BHV-1 iELISA, all comparisons
except those that involved FD treatment had complete agreement.
Discussion
Storage Time
Overall, the most important findings for the FP samples subjected to dry/RT storage were
i) the SE of FP testing diminished somewhat with longer storage whereas SP remained
high; ii) the decline in SE was less pronounced when FP results were compared to results
from sera that had been stored at –20 °C for the same period (as opposed to T1); iii) FP
performance at 6 months tended to be better than that observed at 12 months or longer;
93
however, even after 12 months or more of dry/RT storage, the SE of FP relative to samestorage serum was acceptable (≥85%) in six of the eight antibody assays evaluated for
storage time. Specifically, our results for comparison to same-storage serum suggest that,
even after ≥1 year of dry/RT storage (and in some cases after as long as 2 years), Nobuto
FP samples from Rangifer can still perform comparably to serum in the cELISA and
iELISA for Brucella, the cELISA for N. caninum, the iELISA for BRSV, and the VNs
for BVDV-I and -II. For these six assays, SE of FP remained ≥88% at this stage, and in
most cases SE and SP were both >90% (Table 3). As expected, in comparison to longer
storage, FP performance was generally, but not always, better after only 6 months of
storage. The poorest FP performance was in the BHV-1 assay, with an average estimated
SE of 77% at 6 months (Table 3).
Filter-paper products have different specifications for biomedical use, and drying whole
blood in FP matrices stabilizes most analytes (McDade et al., 2007). Nobuto FP strips are
made of high-purity cellulose, are highly absorbent and of uniform quality, and are
recommended for several veterinary serology applications (Toyo Roshi Kaisha, Ltd.,
Tokyo, Japan). Studies of the effects of storage on FP samples differ considerably in
design and some results are challenging to compare (Beard and Brugh 1977; Brugh and
Beard 1980; Banks, 1985; Ruangturakit et al., 1994; McDade et al., 2000; Mubarak et al.,
2004; Corran et al., 2008); however, evidence from animals and humans indicates that
antibodies in whole-blood FP samples generally withstand storage and temperature
challenges better than some other analytes (McDade et al., 2007; Trudeau et al. 2007;
Elbin et al., 2011). A drop in test SE after prolonged sample storage could reflect a
94
characteristic of the assay, a technical or reagent issue, or loss of analyte stability. Some
degree of inter-assay variation is predictable with any antibody assay that is repeated after
extended intervals, and reports note approximately ≥6% such variation for FP testing in
ELISAs and other platforms (Clague and Thomas, 2002; McDade et al., 2007). Although
we observed this in our study (Tables 2 and 3), the cumulative decreases in SE with FP
testing suggest that antibodies degraded within the Nobuto FP matrix over time. In line
with previous research on various serum analytes (Woodrum and York, 1998; Cray et al.,
2009), for most assays we also detected slight drops in antibody prevalence when we retested serum after extended storage at –20 °C (Appendix 4A). Thus, there is potential for
some loss of SE not only with FP samples in serological tests, but also with serum that is
banked at –20 °C for extended periods.
The findings indicated that vaccination and sampling close to time of seroconversion may
have affected our ELISA results (Table 1). We found that FP test performance was
generally closer to that of serum in the ELISAs on samples from herds with natural
immunity (Brucella and N. caninum testing) than in the ELISAs on samples from
vaccinates (BHV-1, PI-3, and BRSV testing) (Fig. 1, Table 3). The vaccinates had
shallower slopes of antibody test results and more values near the threshold (Fig. 1 d,e,f).
A large proportion of results near the threshold can impair FP test performance. Also, the
samples from animals that had recently seroconverted likely contained significant
primary-response immunoglobulin (Ig) M, whereas one would expect IgG to predominate
in animals with non-recent natural exposure. Studies of FP samples from humans in
ELISAs for dengue virus and human immunodeficiency virus have revealed loss of SE in
95
samples from patients who are newly seroconverted, and more rapid degradation of IgM
in samples from patients with primary infections (Behets et al., 1992; Ruangturakit et al.,
1994). Thus, it is possible that the performance of FP samples from herds naturally
exposed to Rangifer-specific relatives of BHV-1, PI-3, or BRSV might have better
concordance with serum antibody results than we have estimated for these three ELISAs
(and the same may hold true for the BVDV VNs as well). This is particularly relevant for
FP performance in the BHV-1 and PI-3 iELISAs because SE was lower (i.e., below the
80% defining value for “comparable to serum”) in these tests than in the other assays at 6
and 12 months of dry/RT storage.
We were unable to access Rangifer herds that were antibody-positive for all the
pathogens, thus, small sample size contributed to the width and overlapping of
confidence intervals (Tables 2 and 3, Appendix 4A). Assessing test performance in
duplicate test runs is more robust than doing so in single runs. Excluding one outlier SP
result for BRSV, the duplicate SE and SP results for all the assays provide evidence that
our estimates are reasonable indicators of FP test performance. For the five bovinepathogen assays, we obtained FP-serum sample pairs from 12 vaccinated reindeer at three
time points to assess FP comparability to serum over a range of antibody levels (Table 1).
We used these multiple measures as independent data in test performance analysis
because i) our objective was solely to compare serum and FP for detection of antibodies;
ii) testing a range of antibody levels is more robust than testing at a single time point
post-vaccination; iii) collections were separated by ≥1 month; iv) only a small number of
reindeer were available for vaccination. The results from these animals are multiple-
96
measures data, and we are unaware of statistical testing that can account for random
effects with matched-pairs binomial data. Since these observations were used as
independent data in performance analysis, the point estimates for the five bovine-virus
assays may be somewhat high and the confidence limits somewhat narrow.
Processing/Storage Conditions
Our WNV test results provide no evidence of a difference between dry/RT and –20 °C
storage conditions (DD and FF treatments) for FP samples from Rangifer (Table 4).
Comparison to previously published FP performance results in this cELISA for WNV
indicates that the drop in FP SE for this assay after 1 year of dry/RT storage may be
comparable to declines we observed in the other eight assays over this period of storage
(Part I); however, our sample size for WNV testing was small.
The small FP treatments trial revealed no obvious differences among the four
processing/storage treatments after 12 months (Fig. 2). Lack of statistical analysis
precludes strong conclusions; however, as opposed to drying FP samples initially (which
the Nobuto FP manufacturer recommends), the results suggest that freezing the saturated
FP strips directly upon collection and keeping them frozen (often necessary when
collecting in the North in winter) does not impair detection of antibodies to BHV-1 or PI3. Our sample size was limited, yet this finding is noteworthy because, in comparison to
studies of tropical-climate effects on FP samples, little has been reported on the effects of
freezing fresh FP samples prior to drying (Behets et al., 1992). Our small trial also
suggests that switching from frozen state (at collection) to dry storage might negatively
97
affect FP sample performance in serology. However, in general, further investigation of
freezing/drying effects on FP samples from wildlife is needed. Results from studies with
larger sample sizes could provide stronger evidence for decision making in the field.
Other Implications and Recommendations
For serological applications, the Nobuto FP manufacturer recommends avoiding sunlight,
ultraviolet light, water, extreme temperatures, and moisture, and to store “in a clean
indoor space” (Toyo Roshi Kaisha, Ltd., Tokyo, Japan). More specifics have been
published in blood-spot guidelines and reviews (NCCLS, 2003; McDade et al., 2007),
studies from the tropics and elsewhere (Punnarugsa and Mungmee, 1991; Behets et al.,
1992; Ruangturakit et al., 1994; McDade et al., 2000; Mubarak et al. 2004; Ferraz et al.,
2008; Corran et al., 2008; Mei et al., 2011; Chase et al., 2012), and a few animal reports
(Brugh and Beard, 1977; Beard and Brugh 1980; Banks, 1985; Trudeau et al., 2007).
However, the vast majority of sources focus on human blood spots that are dried and
analyzed from small FP discs, and we found only one source that assessed effects of
freezing FP blood samples directly upon collection (Behets et al., 1992).
Table 5 integrates our findings in a summary of key considerations for collecting, storing,
and shipping Nobuto FP samples for wildlife serology. The most important step in
ensuring FP sample integrity—and one that should be communicated explicitly to hunters
and other lay-person samplers—is to make every effort to collect a clean blood sample
and avoid bacterial contamination, which can degrade antibodies. It is also crucial that
Nobuto FP collectors be trained to dip/saturate the entire length (the absorbent portion) of
98
the FP strip, as elution requires a complete FP strip and a partial sample can negate
testing altogether. For archiving of Nobuto FP samples, one alternative to the suggestions
in Table 5 might be to elute and freeze the eluate, which might reduce antibody
degradation. However, FP eluate composition differs from that of serum and, to our
knowledge, there is no published evidence for doing this. In fact, freeze-thaw studies
suggest that keeping samples in the FP matrix may be somewhat protective if there is any
risk of thawing (McDade et al., 2000, 2004). Shipping decisions depend on the individual
situation and the transport of frozen FP samples needs careful consideration. If such FP
samples will be tested shortly after arrival at the laboratory and if there are risks of delays
and temperature fluctuations during transit, then it may be best to thaw/dry immediately
prior to shipping and ship dry with desiccant. Formaldehyde damages FP samples, thus,
exposure to this chemical or its fumes should be avoided (Toyo Roshi Kaisha, Ltd.,
Tokyo, Japan). The higher toxicity rates we have encountered with FP samples than with
serum in VN deserve investigation but, given our experience with this to date in two
studies, and given the relatively lower SE of FP samples in VN (compared to other test
platforms where there is no initial dilution step), we recommend using wildlife FP
samples in ELISA platforms rather than VN if possible. Filter-paper samples need to be
validated for each test in each species.
The FP method offers particular advantages for wildlife field collections and has lower
per-unit storage-space and storage-cost demands than serum. Our estimates of test
performance for FP samples from Rangifer in these antibody assays after different
storage times and under different processing/storage regimes can help guide users. We
99
hope that our evidence, experience, and multi-sourced recommendations will empower
decision makers to tailor the best Nobuto FP practices to their field and laboratory
constraints.
Acknowledgements
We acknowledge the contributions of all the animals involved in this research, and we
especially thank Marianne Jorgensen and Dr. Greg Muench for their care of the U of
Calgary reindeer herd. Caribou hunters Aaron Emiktowt, Chris Jones, John Nakoolak,
Mark Pootoolik, and Greg Ningeocheak from Coral Harbour, Nunavut, as well as Brett
Elkin, Jane Harms, and the entire Kutz Lab team 2007-2009 provided invaluable support
for sample collection. We are indebted to Dr. Klaus Nielsen’s laboratory at the
Brucellosis Centre of Expertise (Canadian Food Inspection Agency). Sincere appreciation
to Linda Kelly, Mélanie Sabourin, Lori Hassard, Laura Bond, Roberta Yemen, and
Johnathon Pameolik for their technical and logistical expertise. We acknowledge the
Government of Nunavut and our funders: International Polar Year Funding from the
NSERC Special Research Opportunity Program, Environment Canada/Natural Resources
Canada, Nasivvik Centre for Inuit Health and Changing Environments (Canadian
Institutes of Health Research), Alberta Innovates Technology Futures, Northern Scientific
Training Program (Indian and Northern Affairs, Government of Canada), Arctic Institute
of North America, the Circum-Arctic Rangifer Monitoring and Assessment Network
(CARMA, www.carmanetwork.org), and the University of Calgary Faculty of Veterinary
Medicine.
100
TABLES
Chpt 4. Table 1. Study design
Study design, showing numbers of sample pairs (serum and filter paper [FP] per animal)
tested and other details. Competitive enzyme-linked immunoassays (cELISA), indirect
ELISAs (iELISA), or virus neutralization assays (VN) detected antibodies to eight
pathogens or pathogen groups (pathogen abbreviations defined in main text). The FP
treatments were i) Dry-Dry (DD): collected at ambient temperature (above freezing),
dried overnight, stored dry at room temperature (RT) until analysis; ii) Dry-Freeze (DF):
collected at ambient temperature above freezing, dried overnight, stored dry at RT for 2
weeks, then stored at –20 °C until analysis; iii) Freeze-Dry (FD): placed in ice cooler at
collection, stored at –20 °C for 2 weeks, then thawed/dried overnight and stored dry at
RT until analysis; iv) Freeze-Freeze (FF): placed in ice cooler at collection, stored at –20
°C until analysis.
101
Table 1.
STUDY SECTION and PATHOGENS COLLECTION DATES SAMPLE PAIRS (n) FILTER PAPER TREATMENTS SAMPLE STORAGE TIMES ASSAY (threshold value) PART I: STORAGE TIME Brucella spp. Mar 2008 184 a
DD 2 , 12, 24 mo. Neospora caninum Apr 2008 Aug 2008 39 6 DD DD 1 , 6, 17 mo. b
1 , 6, 17 mo. DD DD DD b
1 , 6, 12 mo. b
2 , 6, 12 mo. b
1 , 6, 12 mo. b
1 , 6, 12 mo. BHV‐1, PI‐3, and BRSV Summer 2009 • May • July • Aug BVDV‐I and BVDV‐II Summer 2009 • May • July • Aug c
36 DD DD DD c
36 b
b
b
2 , 6, 12 mo. b
1 , 6, 12 mo. cELISA, (30% inhibition) iELISA (20% positivity) cELISA (30% inhibition) iELISA d
(14 or 10 EU) VN (titer 1:6 serum, titer 1:20 FP) PART II: PROCESSING/STORAGE CONDITIONS e
BHV‐1 and PI‐3 July, Aug 2007 • Vacc. day • 3 wks later • 6 wks later 9 West Nile virus Nov 2009 14 a DD, DF, FD, FF DD, DF, FD, FF DD, DF, FD, FF 12 mo. 12 mo. 12 mo. iELISA e
(14 or 10 EU) DD, FF 12 mo. cELISA (30% inhibition) A randomly selected subset (n=88) of the original 184 samples was assayed after 24 mo. storage. Data analyzed in FP validation studies (see Chapters 2 and 3). b c Twelve animals were sampled at three time points: Pre‐vaccination and 4 and 8 weeks after booster (which was given 4 weeks after initial injection). d In the iELISAs, the serum threshold value was 14 ELISA units (EU) and the FP thresholds were 14 EU for BHV‐1, 10 EU for PI‐3, 10 EU for BRSV (see Chapter 3).
e Three animals were sampled at three time points: Baseline (day of vaccination), 3 weeks later (time of booster injection), and 6 weeks (3 weeks post‐booster). 102
Chpt 4. Table 2. Filter-paper performance vs. T1 serum
Performance of filter-paper (FP) blood samples relative to the shortest-storage (T1)
serum. Sensitivity (SE), specificity (SP), and confidence intervals (CI) are shown in
duplicate (Run A and B) at T1 and after two longer storage times. All FP samples were
dried overnight after collection and stored dry at room temperature.
103
Table 2.
104
Table 2. (cont’d)
105
Chpt 4. Table 3. Filter-paper performance vs. same-storage serum
Performance of filter-paper (FP) blood samples relative to serum results for each
respective storage period. Sensitivity (SE), specificity (SP), and confidence intervals (CI)
are shown in duplicate (Run A and B). All FP samples were dried overnight after
collection and stored dry at room temperature.
106
Table 3.
107
Table 3. (cont’d)
108
Table 3. (cont’d)
109
Chpt 4. Table 4. Filter-paper performance – Processing/storage, WNV assay
Performance of filter-paper (FP) blood samples in the West Nile virus assay after two
different treatments: Dried immediately and stored dry at room temperature (DD) or
frozen immediately and stored at –20 °C (FF) until analysis. The FP-serum pairs were
compared after 12 months’ storage, and sensitivity (SE), specificity (SP), and confidence
intervals (CI) are shown in duplicate (Runs A and B). Results for a larger DD sample set
that was stored for 1 month and tested at the same laboratory are presented for
comparison.
FILTER a
PAPER ESTIMATE TREATMENT DD FF SE % CI SP % CI SE % CI SP % CI FP / SERUM STORAGE TIME and RUN n sample pairs tested b
1 mo. (n=26) 95.2* (80.7‐100) 91.7 (61.5‐99.8) n/a 12 mo. A (n=14) 75.0 (34.9‐96.8) 89.1* (60.7‐100) 85.7 12 mo. B (n=14) 85.7 (42.1‐99.6) 90.6* (65.2‐100) 75.0 (42.1‐99.6) (34.9‐96.8) n/a 90.6* 89.1* (65.2‐100) (60.7‐100) a Clopper‐Pearson 95% exact CIs are given. In cases where the initial estimate derived was 100%, the median unbiased estimate (denoted with *) was calculated and reported instead. b Results from Chapter 3. The monoclonals used were 3.112G‐NS‐1 (as in the current study) and C03301‐ Env. 110
Chpt 4. Table 5. Recommendations for filter-paper use – Wildlife
Key recommendations for collection, processing, storage, and shipping of wildlife
Nobuto filter-paper (FP) blood samples for use in serological testing.
RECOMMENDATIONS and CAUTIONS
a
COLLECTION, HANDLING Avoid touching the absorbent (long) portion of FP strip; Dip/saturate entire length in a pool of clean whole blood (serosanguinous fluid in a body cavity may contain antibodies but is not equivalent to a whole‐blood sample); Avoid contacting hair, feces, organ contents, other tissues/fluids PROCESSING Option 1: Dry FP samples completely at ambient room temperature (15‐22 °C at least overnight; timing depends on humidity at location); Avoid direct sunlight, extreme heat >30 °C, humidity >30%, formaldehyde (or fumes) exposure Option 2: Freeze FP samples upon collection, keep frozen STORAGE <2 years Dry FP samples: Store in zip‐type plastic bags with (but not directly contacting) desiccant and humidity‐indicator card; check/refresh desiccant regularly; ideally, refrigerate (4 °C) for up to 2 years; alternatively, can store dry/room temperature for several months (exact duration is assay‐specific) Frozen FP samples: Store in individual zip plastic bags in –20 °C laboratory‐grade freezer (no auto‐defrost) STORAGE b
≥2 years Dry FP samples: –20 °C laboratory‐grade freezer (no auto‐defrost) Frozen FP samples: –20 °C laboratory‐grade freezer (no auto‐defrost) ANALYSIS Dry samples prior to eluting; Keep dry at room temperature while preparing eluates; Test FP eluates immediately or freeze at –20 °C (if necessary) until testing Dry FP samples: Ship in zip‐type plastic bags with desiccant SHIPPING Frozen FP samples: *Depends on individual circumstances/risks – Can ship frozen in zip plastic bags or thaw/dry prior to shipping and ship in zip bags with desiccant All FP samples: Ship separate from any formaldehyde‐preserved samples FP Eluates: If eluates must be shipped, keep frozen during transit a
Main sources: NCCLS, 2003; McDade et al., 2007; Toyo Roshi Kaisha, Ltd., Tokyo, Japan; authors’ experience with wildlife Nobuto FP samples and evidence from the current study b
–20 °C is specifically recommended by the Clinical and Laboratory Standards Institute (formerly NCCLS) for ≥2 years’ storage of dried blood spots for human neonatal screening (NCCLS, 2003). 111
FIGURES
Chpt 4. Figure 1. a-f. Effects of storage time – ELISAs
Effects of storage time on filter-paper (FP) test performance. All FP samples were dried
immediately after collection and stored dry. Graphs for the enzyme-linked
immunosorbent assays (cELISAs and iELISAs) show the results for Run A of T1
(shortest-storage) serum (open circles) and the corresponding FP results for three
different storage times. For each storage period, the FP run with the lowest sensitivity
(see Table 3) was plotted using a distinct symbol. To create the graphs, data were first
sorted according to ascending serum values. Each number on the x-axis represents an
individual animal and has a set of data symbols (the animal’s serum result and its three
FP results, one for each storage time) vertically above it. Dotted lines denote assay
threshold values for serum and FP (identical unless noted separately).
112
Figure 1.
113
Figure 1. (cont’d)
114
Figure 1. (cont’d)
115
Chpt 4. Figure 2. a, b. Effects of processing/storage regimes – PI-3, BRSV assays
Effects of different treatments (Dry-Dry [DD], Dry-Freeze [DF], Freeze-Dry [FD], or
Freeze-Freeze [DF]; details in Methods) on filter-paper (FP) test performance. Results are
from three vaccinated animals that were sampled on the day of vaccination and then 3
weeks (booster) and 6 weeks later. Indirect enzyme-linked immunosorbent assays
(iELISAs) for bovine herpesvirus type 1 (BHV-1) and parainfluenza virus type 3 (PI-3)
were performed on sample pairs 12 months after collection. Dotted lines denote assay
threshold values for serum and FP (identical unless noted separately) and the three
animals’ results are distinguished by different symbols. Analysis of category
(positive/negative) agreement identified FD treatment (*) as the only processing/storage
regime that might negatively affect FP results.
116
Figure 2.
117
CHAPTER FIVE
SEROLOGICAL SURVEY FOR EIGHT PATHOGENS
IN MIGRATORY CARIBOU HERDS
OF NORTH AMERICA AND GREENLAND:
SNAPSHOT DURING INTERNATIONAL POLAR YEAR 2007-2009
118
Abstract
During 2007-2010, members of an international network for monitoring the health of
migratory caribou and wild reindeer coordinated unprecedented, simultaneous collections
of blood samples and data from seven caribou populations, five in arctic North America
and two in Greenland. For most herds, collections took place in multiple seasons and over
multiple years. Whole blood and/or filter-paper blood samples were used to test for eight
pathogens or pathogen groups: Brucella spp., Neospora caninum, West Nile virus
(WNV), Toxoplasma gondii, bovine herpesvirus type 1 (BHV-1), parainfluenza virus
type 3 (PI-3), bovine respiratory syncytial virus (BRSV), and Pestivirus. The aim was to
evaluate prevalence of exposure to these agents and identify patterns or relationships with
selected demographic, geographic, physiologic, temporal, and environmental parameters.
Overall prevalence was highest for BHV-1, PI-3, and Pestivirus (25%, 7%, and 28%,
respectively). Prevalence for Brucella, N. caninum, and T. gondii was ≤2%, which was
lower than expected based on historical findings or relevant contemporary studies. No
animals tested positive for WNV or BRSV. Ecological factors (herd values for mean June
and July temperatures and mean population density on summer range) were not
correlated with seropositivity (“risk” of exposure) for any of the agents. For BHV-1, PI-3,
and Pestivirus, adults had greater risk of exposure than young caribou, and the findings
suggest that fall season might be associated with greater risk of exposure than other
seasons of collection. Overall results suggest higher prevalence of Pestivirus exposure in
pregnant females than in non-pregnant females, but sample sizes limited robust analysis
and the explanation for such a difference is not clear. Prevalence differences by region
119
(West vs. Quebec vs. Greenland) were in accord with genetic evidence for the ancestral
origins of the caribou that occur in these general zones. Recognizing the limitations of
sample size and non-random sampling, the serosurvey was substantial and
unprecedented, and the results provide possible baselines and shed light on potential
differences among herds and regions. Aspects of the surveillance/collection methodology
and complexities of the cycling of these infectious agents are discussed, and some
directions for future research are identified.
Introduction
Infectious diseases of caribou and reindeer (Rangifer tarandus ssp.) are important from
Rangifer-health and human-health perspectives. Serological surveys and other pathogen
studies involving Rangifer began to enter the literature around the 1970s. Most have
reported prevalence for various agents in reindeer and free-ranging caribou from Alaska
(Dieterich et al., 1981; Zarnke 1983, 2000; Stieve et al., 2010; Dubey and Thulliez,
2005), in caribou from other parts of arctic North America (Elazhary et al., 1979, 1981;
Ferguson, 1997; Farnell et al., 1999; Kutz et al., 2001), and in reindeer from Scandinavia
(Rehbinder et al., 1992; Stuen et al., 1993; Åsbakk et al., 1999; Lillehaug et al., 2003;
Kautto et al., 2012). Some authors have also examined specific pathogen isolates from
Rangifer and their pathogenesis and transmission (Ek-Kommonen et al., 1982, 1986;
Rockborn et al., 1990; Becher et al. 1999, 2003; Avalos-Ramirez et al., 2001; Dubey et
al., 2002; Das Neves et al., 2009a,b; Tryland et al., 2005, 2009; Evans et al., 2012). Very
120
recent research has focused on cervid herpesvirus 2 (CvHV2), an alphaherpesvirus that is
endemic in semi-domesticated Scandinavian reindeer and strongly suspected to be
circulating in Alaskan reindeer and caribou (Das Neves et al., 2010; Evans et al., 2012).
In general, there is much still to learn about the specific pathogens and strains that infect
Rangifer, and about the distribution, transmission dynamics, and health impacts of these
agents.
This study investigated eight pathogens/pathogen groups that are of known or potential
relevance to Rangifer (Table 1). The bacterium Brucella suis biovar 4 is the causal agent
for brucellosis in caribou and reindeer, and is a chronic multisystemic disease that most
commonly presents as orchitis and reproductive failure, joint disease, and other forms of
debilitation (Forbes, 1991). This agent is also a zoonotic concern in the North and causes
multisystemic disease in humans as well, with transmission via contact with or
consumption of raw or partially cooked meat from affected caribou (Tessaro and Forbes,
1986; Gunn et al., 1991).
Neospora caninum and Toxoplasma gondii are apicomplexan protozoa with similar life
cycles that feature transplacental transmission (Dubey, 2003; Dubey and Jones, 2008).
Canids are the only known definitive hosts of N. caninum, an agent that causes abortion
storms, neonatal mortality, and is transmitted transplacentally in cattle (Dubey, 2003;
Dubey and Thulliez, 2005). Clinical neosporosis has not been confirmed in Rangifer but
clinical disease and vertical transmission have been reported in other cervids (Woods et
al., 1994; Dubey et al., 1996). Antibodies to N. caninum have been detected in numerous
free-ranging canids and ungulates, and serological evidence suggests a sylvatic cycle and
121
vertical transmission in caribou (Lindsay et al., 1996; Buxton et al., 1997; Lindsay et al.,
2001; Dubey, 2003; Dubey and Thulliez, 2005; Dubey et al., 2009; Stieve et al., 2010).
Seropositivity for T. gondii has been reported in a variety of northern wildlife (Elmore et
al., 2012) and in Rangifer worldwide (Oksanen et al., 1997; Zarnke et al., 2000; Kutz et
al., 2001; Vikoren et al., 2004; Stieve et al., 2010). Transplacental transmission, abortion,
and experimentally induced fatal enteritis have been documented in reindeer (Oksanen et
al., 1996; Dubey et al., 2002). Lynx are suspected definitive hosts of T. gondii in arctic
North America (Kutz et al., 2001; Zarnke et al., 2001; Stieve et al., 2010); however, the
generally low numbers of felid hosts across the Arctic suggest other potential
transmission and maintenance dynamics. There is evidence for introduction of T. gondii
to the North via migratory waterfowl, and the agent may spread and persist in animal
populations through carnivory and vertical transmission (Prestrud et al., 2007; Elmore et
al., 2012). Toxoplasmosis is a zoonosis that is usually subclinical but can cause serious
disease and abortion in humans (Dubey and Jones, 2008). People become infected by
ingesting T. gondii tissue cysts or oocysts, and seropositivity and apparent overt disease
have been documented in communities of northern Canada (McDonald et al., 1990;
Elmore et al., 2012). Various risk factors for human exposure to T. gondii have been
identified (summarized in Elmore et al., 2012). The main ones linked with caribou are i)
consumption of raw or partially cooked traditional “country foods” from an infected
animal, and ii) accidental ingestion after hands become contaminated during
butchering/skinning, which is typically done at remote sites with no washing facilities
(McDonald et al., 1990; Forbes et al., 2009). Further, the results of a serosurvey of Inuit
122
in Québec, Canada suggest that humans may be exposed to T. gondii oocysts in surface
water (Messier et al., 2009), and this might also be a source of infection for wildlife.
West Nile virus (WNV) is a zoonotic arbovirus that amplifies in avian hosts and is
transmitted by various mosquitoes, principally Culex species (CDC, 2003). Clinical
infection with WNV has not been diagnosed in free-ranging Rangifer, but severe
neurological disease and death were observed in naturally infected captive reindeer in the
United States in 2002, with pathology similar to that observed in horses (Palmer et al.,
2004a). Northward shifts of vector ranges with climate change could potentially increase
the risk of WNV infection for caribou and reindeer.
Bovine herpesvirus type 1 (BHV-1, an alphaherpesvirus), parainfluenza virus type 3 and
bovine respiratory syncytial virus (PI-3 and BRSV, respectively; both paramyxoviruses),
and members of the genus Pestivirus (flaviviruses that include bovine viral diarrhea virus
[BVDV] and border disease virus [BDV]) all contribute to the bovine respiratory disease
complex. In domestic bovids, BHV-1 and BVDV can also impact other organ systems
and cause a range of effects from abortion and weak calves to diarrhea and severe or
lethal disease (Radostits et al., 2007). Parainfluenza viruses (PIVs) and respiratory
syncytial viruses (RSVs) cause respiratory disease in a wide variety of mammals
(Williams and Barker, 2001). Most PIV infections are subclinical but may lead to
interstitial pneumonia, whereas the interstitial pneumonia caused by RSVs tends to be
more severe (Williams and Barker, 2001). Bovine RSV is a major cattle pathogen
worldwide (30% to 70% prevalence, 60% to 80% morbidity, and up to 20% mortality in
some outbreaks) and shares similar epidemiology with human RSV (Valarcher and
123
Taylor, 2007). In bovids, clinically severe RSV disease is observed in calves particularly,
with peak incidence at 2-6 months of age (Valarcher and Taylor, 2007). The effects of the
bovine agents BHV-1, PI-3, BRSV (or viruses closely related to them) and pestiviruses in
wild ruminants are unknown or not fully understood. Research suggests that reindeer may
contract alphaherpesviruses without developing overt disease, and that some of these
agents produce lesions under certain stress-related conditions, and may contribute to
disease outbreaks and development of necrobacillosis (Rockborn et al., 1990). The
alphaherpesvirus CvHV2 was identified as the primary agent in an outbreak of
keratoconjunctivitis in Norwegian reindeer (Tryland et al., 2009). Experimental infection
of two reindeer calves with a cytopathic BVDV strain caused diarrhea and transient
coronitis and laminitis (Morton et al., 1990). Investigation of a Pestivirus isolate from
reindeer has revealed that this strain is phylogenetically closer to BDV than to BVDV,
and recent virus neutralization testing of Swedish reindeer sera showed higher titers
against BDV strains than BVDV Type I (Becher et al., 1999, 2003; Kautto et al., 2012).
The specific health impacts of these viral groups in caribou and reindeer remain largely
unclear, and testing for BHV-1, PI-3, BRSV, and Pestivirus is of interest for several
reasons: i) there is evidence that Rangifer worldwide are exposed to alphaherpesviruses,
paramyxoviruses, and pestiviruses (Elazhary et al. 1979, 1981; Dieterich et al., 1981;
Zarnke, 1983; Ek-Kommonen et al., 1986; Rockborn et al., 1990; Rehbinder et al., 1992;
Stuen et al., 1993; Tryland et al., 2005; Johnson et al., 2010; Kautto et al., 2012); ii) these
are major pathogens of domestic animals, and agriculture can bring domestic hosts’
pathogens of varying virulence into contact with Rangifer (Williams and Barker, 2001;
124
Tryland, 2012); iii) assays for these agents based on domestic-animal antigens may detect
antibodies induced by closely related viruses that are circulating in Rangifer populations.
In 2004, as part of a wildlife monitoring initiative of the Arctic Council, an international
organization of scientists, managers, and community leaders known as the Circum-Arctic
Rangifer Monitoring and Assessment Network (CARMA) was established through the
Conservation of Arctic Flora and Fauna Group (CAFF, 2010). With a mandate for
collaborative monitoring of circumpolar herds and funding from the International Polar
Year program 2005-2011, CARMA spearheaded a pan-arctic coordinated program of
standardized sampling and analysis of Rangifer health. One component of this was multiherd blood-sample collection that spanned 2007-2010, and this was the genesis for the
current study. The unprecedented scope and scale of this sampling permitted a
simultaneous serological survey for the above-noted pathogens in seven arctic caribou
populations: three herds of barrenground caribou (R. t. groenlandicus and R. t. granti) in
the Western Arctic, two herds of migratory woodland caribou (R. t. caribou) in
Québec/Labrador, and two herds of migratory caribou (R. t. groenlandicus) in southwest
Greenland (Roed, 2005; COSEWIC, 2011). The objectives of this study were to evaluate
overall and herd seroprevalence for these eight pathogens/pathogen groups, and to
identify patterns or relationships with selected demographic, geographic, physiologic,
temporal, and environmental parameters.
125
Material and Methods
Between 2007 and 2010, CARMA members and partners from government agencies and
universities sampled seven migratory caribou herds of the Arctic, five in North America
(Porcupine, Bluenose-West, Bathurst, Rivière-aux-Feuilles, Rivière-George) and two in
southwest Greenland (Akia-Maniitsoq and Kangerlussuaq-Sisimiut) (Fig. 1). Blood
samples were obtained in addition to demographic and morphometric data, tissue
specimens, and gross observations of pregnancy status (Kutz et al., 2012b). Estimated
herd population size at the time of the study ranged from approximately 18,000 animals
in the Bluenose-West herd to more than 430,000 in the Rivière-aux-Feuilles herd
(CARMA, 2012; Taillon et al., 2012). Caribou were shot, in most cases by hunters, with
high-powered rifles and were sampled as swiftly as possible after death. The sampling
was directed by scientists in all herds except for Bluenose-West. The scientist-driven
collections were done according the collecting agency/researcher’s purposes and, thus,
were not random samples of the populations. Sampling from the Bluenose-West
population was also non-random, as most of these caribou were sampled during
community subsistence hunts. Table 2 provides background information on the herds, the
seasons and years of collection, and some herd parameters of potential relevance to
disease transmission.
All CARMA data and tissue-sample analyses were entered into a master database.
Caribou of both sexes were sampled for the study, but in some populations primarily
females or males were sampled. Caribou of different ages were sampled from all herds.
Age classes were assigned according to tooth age analysis where available (calf <1 year;
126
subadult 1 to 2.99 years; adult 3 years or older) and based on data-sheet classification
otherwise. All the Québec herd collections were cow-calf pairs, and a few such pairs
were among the Greenland collections also. Calves younger than 6 months were excluded
from the serology analyses because of potential confounding by transfer of maternal
antibodies to calves during lactation. Seasons of collection were defined as fall
(September through November), winter (December through February), spring (March
through May), and summer (June through August). Given the caribou reproductive cycle,
field crews were only able to ascertain pregnancy status grossly during spring or winter
collections. For females (all but calves) sampled in these seasons, pregnancy status was
classified as positive or negative.
Blood Samples
Whole blood and/or filter-paper (FP) blood samples (Nobuto filter strips: Toyo Roshi
Kaisha, Ltd., Tokyo, Japan) were obtained from each animal. Whole blood was collected
into a 10-mL Vacutainer® tube without anticoagulant and was processed in standard
fashion by the collectors for each herd using centrifugation in most cases. The exceptions
were the Québec herds, where whole blood (i.e., the 2009 blood samples for both herds)
was obtained at a field camp remote from laboratory facilities. Each of these samples was
allowed to clot and separate, and then serum was drawn off and maintained in cool
conditions (in a dugout cavity underground) for several days in the field before they
could be transferred to a –20 °C freezer. All sera were stored at –20 °C until analysis.
127
The FP samples were dried as quickly as possible at room temperature or ambient
temperature in the field, and were stored dry at room temperature (approximately 15-22
°C) with desiccant pouches until analysis. In most cases, FP samples were eluted (i.e.,
recovery of serum from the paper matrix) just prior to testing and the eluates were
shipped to laboratories frozen. The only exceptions were with WNV and T. gondii
testing, where the laboratory personnel eluted the FP samples in-house using specific
reagents (WNV; see Chapter 3) or a slightly different elution protocol (T. gondii; 4 hours
immersed in buffer as per Nobuto manufacturer specifications). All methods and
commercial products involved in the processing, storage, and eluting of FP samples are
detailed in Chapter 2.
Testing
Pathogen tests were prioritized as listed in Table 1. Serum was used if available and FP
samples were used alternatively. Given that FP eluates are estimated to be 1:10 serum,
protocol steps were adjusted as needed to ensure that serum and FP results were
comparable (see Chapters 2 and 3). All samples were tested in either July-August 2010
(Brucella, N. caninum, West Nile virus, BHV-1, PI-3, and BRSV assays) or FebruaryMay 2011 (T. gondii and Pestivirus assays).
Antibody tests were done at veterinary diagnostic laboratories in Canada, the United
States, and Norway. They were as follows:
•
A competitive enzyme immunosorbent assay (cELISA) for Brucella spp. that has
been validated for caribou (threshold value ≥30% inhibition; 99% sensitivity and 93%
128
specificity; Gall et al., 2001 and K. Nielsen, unpublished data [see threshold details in
Chapter 2]);
•
A commercial cELISA kit for N. caninum in cattle (VMRD Inc., WA, USA;
threshold ≥30% inhibition) that has been used for N. caninum serodiagnostics in
semi-domesticated reindeer (see Chapter 3);
•
A cELISA for WNV in domestic mammals (threshold ≥30% inhibition; Blitvich et
al., 2003) that has been used for WNV serodiagnostics in semi-domesticated reindeer
(see Chapter 3);
•
Indirect ELISAs for BHV-1, PI-3, and BRSV that have been adapted for cervids
(serum threshold ≥14 ELISA Units [EU]; FP threshold ≥14 EU for BHV-1 and ≥10
EU for PI-3 and BRSV; original test described by Durham and Hassard, 1990).
Curry et al. validated FP samples from Rangifer and tested their robustness (after 6
months and ≥12 months’ storage dry at room temperature, and under different
processing/storage conditions) in these six assays (see Chapters 2, 3, and 4). The same
authors listed all details for the laboratories, protocol steps, and threshold values used in
the current study (see Chapters 2, 3, and 4).
Testing for T. gondii was conducted at the United States Department of Agriculture,
Parasite Biology and Epidemiology Laboratory in Beltsville, Maryland using a modified
agglutination test (MAT) described by Dubey and Desmonts (1987). Though this test has
129
not been specifically validated for Rangifer samples (serum or FP), it has been used
previously for caribou serology based on results from domestic pigs naturally infected
with T. gondii (Zarnke et al., 2000; Kutz et al., 2001). As in earlier studies of caribou, a
MAT titer of ≥1:25 was considered to indicate T. gondii exposure.
Antibodies against pestiviruses were assayed by the Tryland research laboratory at the
Norwegian School of Veterinary Science in Tromsø, Norway using a commercial
cELISA kit developed for ruminants (SERELISA® BVD p80 Ab Mono Blocking,
Synbiotics Corp., France). This test detects antibodies to a protein that is common to all
Pestivirus strains. While not validated for Rangifer samples, competition percentages and
thresholds were determined according to the manufacturer’s instructions for ovine
samples, because the Pestivirus strain in sheep is genetically closest to a strain that has
been isolated from reindeer (Becher et al., 1999, 2003). Also, the kit thresholds were
evaluated using 14 reindeer sera that were Pestivirus-positive on virus neutralization. The
kit thresholds for sheep (as opposed to those for cattle) categorized these Pestiviruspositive reindeer samples most accurately (C. Das Neves, personal communication). As
in all the ELISAs noted above (see usage of test controls in Chapter 3), kit controls were
run with each test plate. Results were classified as positive if the percentage competition
was greater than 40% and “doubtful” if between 20% and 40%. Doubtful samples were
retested and classified negative if they were in the doubtful range a second time.
130
Analysis
All antibody test results were categorized as positive or negative, and overall estimates of
seroprevalence were calculated for each pathogen by herd. A stratified analysis was
conducted involving five variables: sex, age class, season of sample collection, pregnancy
status, and herd geographic location. Univariate analysis was performed using
crosstabulation. Proportions of seropositivity generated from the contingency tables were
considered to represent risk of exposure or infection, and were thus reported as
percentage “risk.”
For each pathogen, overall one-way contingency tables were first generated for the test
results (positive or negative) relative to each of the five variables. In each of these sets of
results, relative risk (RR) of being seropositive was calculated based on a referent (i.e.,
the group with lowest prevalence in all cases except geographic location, where Rivièreaux-Feuilles was selected as referent [lowest of the mid-range prevalence values]). In
cases where the risk result was 0%, RR was calculated using adjusted risk. This was done
by adding 0.5 to the numerator and denominator for the risk calculation. Chi-square
analysis (Pearson’s chi-square or Fisher’s exact test, as appropriate) was used to assess
for significant differences in risk within each stratum.
Based on the findings from these initial analysis steps, variables were then used to create
two- and three-way contingency tables. This allowed for comparisons across herds and
for examination of possible effects of individual variables on seropositivity (risk). In
131
accord with patterns of prevalence results, herd findings were grouped regionally and chisquare comparisons were done for each stratum.
Correlations were also calculated to assess whether environmental or demographic
parameters of potential relevance to disease transmission (mean June and July
temperatures on herd range, mean animal density on herd summer range; Table 2) were
associated with pathogen exposure/infection. The estimates of animal density on summer
range were obtained from government records. The two Greenland caribou herds in this
study have been consistently surveyed in small groups of approximately three animals
(Cuyler, 2007), whereas the other five herds aggregate in the thousands on summer range.
The Greenland herds were excluded from the animal density correlation because of this
difference.
The software PASW Statistics 18 (version 18.0, IBM Corporation, NY, USA) was used
for all statistical analyses. P values <0.05 were considered to indicate statistical
significance. Clopper-Pearson Exact 95% confidence intervals were calculated (EpiTools
calculators; Sergeant, 2009) for herd seroprevalence estimates and for risk results.
Results
Fifty-six FP samples were excluded from the analyses because they did not elute properly
according to the Nobuto filter paper manufacturer’s specification of “dark red wine
132
colour.” Of these 56 excluded FP samples, 28 were from each of the Québec herds (40
from 2007 collections and 16 from 2008).
Table 3 summarizes the numbers of samples tested in each assay and the seroprevalence
estimates for each of the eight pathogens/pathogen groups by herd and overall. The
highest overall prevalence estimates were for Pestivirus (28%), BHV-1 (25%), and PI-3
(7%), and seropositivity to these pathogens was generally higher in the western herds and
lower in the Québec and Greenland herds.
Of the total 4,420 samples tested, 2,782 (63%) were serum and 1,638 (37%) were FP
samples. Proportions of FP samples and serum analyzed were generally consistent across
all the pathogen tests (ranges: serum 62% to 66%, FP samples 34% to 38%); however,
these proportions differed considerably by herd (Table 4). Approximately 220 animals in
the study had FP samples tested for one or more pathogens (i.e., there was insufficient
serum available for one or more of the assays). Of these 220 animals, approximately 180
were collected in 2007 and 2008 from Bluenose-West and both Québec herds (roughly 90
per year), and the rest were collected in 2009-10 (data not shown).
Initial calculations revealed that calves and subadults generally had similar risk of
positivity for the pathogens relative to adults; thus, these two classes were combined and
analyzed as “Young.”
One-way contingency table analyses for the overall sample set revealed significant
differences (Table 5):
133
•
Males were more likely to be seropositive for BHV-1 and PI-3 than females
(though this did not hold true when further analysis was done);
•
Adults were more likely to be positive for BHV-1, PI-3, and Pestivirus than
young;
•
Caribou sampled in fall and summer were more likely to be positive for Pestivirus
than those sampled in spring;
•
Pregnant females were more likely to be positive for Pestivirus than non-pregnant
females;
•
Compared to the Rivière-aux-Feuilles herd (the chosen referent), herds located
west of Québec were more likely to have animals seropositive for BHV-1 and
Pestivirus, and both Greenland herds were less likely to have animals seropositive
for BHV-1 and Pestivirus.
Pregnancy data were only available for 93 females, and 41 of these were from the AkiaManiitsoq herd. None of the animals sampled from this herd were seropositive for
Pestivirus. The remaining 52 females (all from Bathurst herd) included 4 non-pregnant
and 48 pregnant females, and these groups included 25% (n=1) and 52% (n=25)
Pestivirus-positive animals, respectively (p>0.05; Fisher’s exact test).
Cross-Herd Findings: BHV-1
Initial analysis indicated that, compared to females, males were approximately twice as
likely to be seropositive for BHV-1 (Table 5). Stratification of results for all assays by
sex revealed overrepresentation of males in samples from the west and overrepresentation
134
of females in samples from the east. Males were sampled almost exclusively from the
Porcupine and Bluenose-West herds, females almost exclusively from the Greenland
herds, and both sexes but relatively fewer males from the Québec herds (males
approximately 17% of total samples per herd). More than 30 caribou of each sex were
sampled from the Bathurst herd, and the proportion of BHV-1 positives in those males
only slightly exceeded that in females (41% vs. 35%, respectively). When herd results
were combined according to region (West = Porcupine, Bluenose-West, Bathurst; Québec
= Rivière-aux-Feuilles, Rivière-George; Greenland = Akia-Maniitsoq, KangerlussuaqSisimiut), there were significant differences in BHV-1 prevalence across regions (p<0.01
for males and for females) and the highest prevalence for each sex was in the West (Table
6).
Adults (bulls and cows) were overrepresented in the samples from all herds. Young
caribou were scant in the Porcupine and Bluenose-West sample sets and comprised a
maximum of approximately 30% in all other herds except Québec’s. In the Rivière-auxFeuilles and Rivière-George herds, the young age class represented 45% and 39% of the
BHV-1 samples, respectively, and there were no animals seropositive for BHV-1 in either
of these groups. To effectively control for the effect of age, risk of seropositivity to BHV1 for adults alone was compared across the regions. There was a significant difference in
BHV-1 seroprevalence for adults across the regions (p<0.01) and the highest prevalence
was in the West (Table 6).
Seasons of sample collection were not uniform across herds, and the resultant uneven
distribution of results and relatively small n values created challenges for evaluations and
135
statistical comparisons by season. The Greenland herds were collected only in spring
(Akia-Maniitsoq in 2008 only and Kangerlussuaq-Sisimiut in 2009 only), whereas all
other herds were collected in two or more seasons and multiple years, but the seasons
differed (Table 2). The most frequent collection time was fall (five herds; all except
Greenland) and the least frequent was winter (Bluenose-West herd, n=8 in 2010 only).
There was no significant difference among the overall BHV-1 risk results by season.
Only the fall collections in the West had reasonable sample sizes to compare BHV-1
prevalence in bulls versus cows, and here the confidence intervals indicated no statistical
difference (in contrast to the significant higher risk observed for males in the overall
analysis; Table 5). The overrepresentation of fall-collected samples in the West and in
Québec suggests that season (fall collections) may have confounded the regional
differences in BHV-1 prevalence to some extent (Table 7). This held for all the pathogens
tested.
Three-Way Stratification
For BHV-1, PI-3, and Pestivirus, three-way stratification for sex, age, and season by
region reduced sample sizes such that, for virtually all comparisons by season, confidence
intervals overlapped (Table 7; only adults’ results shown). However, recognizing this
lack of statistical robustness, it is worth noting some generally similar seasonal
prevalence trends for these three pathogens/pathogen groups (Table 7). For PI-3 and
Pestivirus, the within-region comparisons for sex strata with the largest n values (i.e., fallcollected females vs. spring-collected females in the West) suggest that fall-collected
caribou may have higher risk of exposure than spring-collected caribou. For BHV-1, PI-
136
3, and Pestivirus, cross-region comparisons for sex strata with the largest n values (i.e.,
fall-collected females across regions, spring-collected females across regions) suggest
that seroprevalence in these seasons may be somewhat higher in the West than in other
regions.
Cross-Herd Findings: PI-3
One-way stratification of the overall PI-3 results by sex and by age revealed higher risk
of PI-3 seropositivity in males (compared to females) and in adults (compared to young),
respectively. These same patterns were observed for BHV-1; however, adults had a lower
RR of exposure for PI-3 than for BHV-1 (RR: 4.5 vs. 17.5, respectively) (Table 5). The
estimated risk of PI-3 exposure in young caribou overall was identical to that for BHV-1
(2%). Herd results stratified by sex, age, and season (individually and combined) showed
the same general patterns that were observed for BHV-1, but with lower risk estimates
and smaller differences (higher p values) in the cross-region comparisons (Table 6). Chisquare analysis within the sex and age strata across the West, Québec, and Greenland
regions revealed significant differences for risk of PI-3 seropositivity in males, females,
and adults (Table 6); however, the risk in western females was comparable to that in
Québec females (5% vs. 7%, respectively; chi-sq.=0.02, p>0.05).
Cross-Herd Findings: Pestivirus
One-way stratification of the overall Pestivirus results by sex revealed no significant
difference in Pestivirus prevalence for males and females. Stratification by sex across
herds suggested somewhat higher risk of Pestivirus seropositivity in Bathurst females
137
(47%, n=92) and Québec females (Rivière-aux-Feuilles 23%, n=102; Rivière-George
40%, n=98) than was observed for BHV-1 (35%, 16%, and 22%, respectively; data not
shown). Stratification of the overall results by age showed significantly higher Pestivirus
prevalence in adults, but relatively lower risk of Pestivirus seropositivity than of BHV-1
seropositivity in adults (RR: 5.4 vs. 17.5, respectively; Table 5). Sample size precluded
cross-herd or cross-region comparisons based on pregnancy status; however, there were
significant cross-regional differences in risk of Pestivirus seropositivity for adults, fallcollected caribou, and spring-collected caribou, with the highest prevalence in the West
(Table 6). Since there were generally fewer samples available for Pestivirus testing
compared to the other assays, it was only possible to compare across two regions for
season (Table 6) and two collection seasons were not represented in the three-way
stratified analysis for pestiviruses (Table 7).
Correlation
None of the relationships between risk of pathogen exposure and the environmental or
demographic parameters (mean June and July temperatures on herd range, mean animal
density on herd summer range, respectively) were statistically significant.
138
Discussion
Overall Prevalence Findings
As the first multi-pathogen serosurvey across seven migratory caribou herds of the North,
this study is an unique opportunity to assess Rangifer pathogen exposure at a broad scale.
Sampling a free-ranging species presents numerous challenges, one of which is obtaining
optimal sample size. The herd seroprevalence estimates are based on relatively limited
sample sizes, which restricts the study’s power for determining precise estimates and
discriminating differences. For example, roughly 140 caribou from the Bathurst herd
(most current estimate of population size: 32,000) were tested for each of BHV-1 and
Pestivirus, and the seroprevalence estimates for these pathogens were 37% and 40%,
respectively (Table 3). However, determination of 40% seroprevalence (with 95%
confidence and 5% precision) in a population of this size requires a considerably larger
sample size of 365 (EpiTools calculators: Sergeant, 2009). Further, while almost all the
CARMA sampling was scientist-driven (not biased towards visibly healthy animals), the
Bluenose-West herd sampling was mainly driven by local subsistence hunters
(community hunt setting). These hunters select what appear to be the healthiest and most
robust animals for food, thus, this subset (approximately 53 [10%] of the approximately
550 total caribou tested per assay; Table 3) was a convenience sample that might have
been somewhat biased towards less pathogen exposure. These sampling issues suggest
that the calculated prevalence estimates are likely to be underestimates of the true
population levels of exposure/seropositivity for these pathogens (or for closely related
Rangifer-specific agents that are circulating in herds and may be cross-reacting on some
139
assays). This should be kept in mind for the overall findings and the herd findings.
Recognizing the limitations of sample size and non-random sampling, the CARMA
survey effort was substantial and unprecedented, and the results provide possible
baselines and shed light on potential differences among herds and regions.
In general, the highest overall seroprevalence estimates observed were for Pestivirus,
BHV-1, and PI-3, respectively. Minimal reactivity was detected for Brucella, N. caninum,
and T. gondii, and none of the animals tested positive for WNV or BRSV. The latter
findings establish baselines for WNV and BRSV across these herds, whose annual ranges
span extensive parts of the Arctic. To date, there have been no published serological
findings for WNV in free-ranging Rangifer, but seropositivity without lethality has been
documented for captive reindeer in North America (Palmer et al. 2004a,b; see Chapter 3).
The absence of seropositivity for BRSV should be considered in context. Antibody
testing for this paramyxovirus in free-ranging caribou and reindeer populations has been
relatively limited. To the author’s knowledge, serological testing for BRSV in caribou
has only been reported in the grey literature in Canada (see Kutz, 2007 for list of testing)
and no seropositives have been detected in free-ranging caribou populations. Aguirre et
al. (1995) detected antibodies to RSV in cervids (mule deer [Odocoileus hemionus] and
wapiti [Cervus elaphus]) in national parks of the Western United States. The current
study’s result of no positive caribou on an indirect ELISA for BRSV with a large overall
sample size may be a true finding, but several considerations deserve mention. First and
possibly most important, the indirect ELISA for BRSV that was used was originally
developed for cattle (Durham and Hassard, 1990) and was adapted for cervids (see details
140
in Chapter 3) but not validated for cervids or Rangifer specifically. Consequently, the
assay’s efficacy for detecting antibodies to BRSV (or a closely related virus) in naturally
exposed caribou is unknown. The test was an indirect ELISA for BRSV, and it was based
on a bovine virus antigen and employed a protein G conjugate. It is likely that Rangifer
seropositivity on a bovine assay signals exposure to a closely related virus that is
reindeer- or caribou-specific rather than the bovine virus itself; however, cross-reacting
antibodies may bind differently from species-specific ones. Weak binding of Rangifer
antibodies cross-reacting to BRSV might have decreased the sensitivity of the assay.
Second, in domestic cattle, calves develop the most severe clinical BRSV infections and
peak incidence occurs at 2 to 6 months of age (Valarcher and Taylor, 2007). In the
current serosurvey, caribou calves younger than 6 months of age were excluded owing to
potential confounding by maternal antibodies during lactation. If RSVs infect Rangifer
and their epidemiological pattern is similar to that for cattle and humans with RSV
(Valarcher and Taylor, 2007), then i) the survey might have missed calves exposed or
infected with BRSV (or a closely related RSV) and ii) diseased young calves might be
more likely to die from infection, thus reducing the number of infectious and seropositive
animals in a herd. Also, according to some authors, immunity to natural RSV infections
in mammals is considered to be short-lived but re-infections are thought to be common
(Williams and Barker, 2001). A third factor to consider regarding the absence of antibody
reactivity to BRSV is that, as noted, all prevalence estimates in the study are likely
conservative owing to factors mainly linked to sample size.
Even accounting for possible underestimations, the very low reactivity for Brucella, N.
141
caninum, and T. gondii observed across all seven herds was unexpected. Cases of
rangiferine brucellosis, cases of human brucellosis related to caribou meat consumption,
and caribou seropositive for Brucella have been reported across arctic North America for
decades (Zarnke, 1983; Tessaro and Forbes, 1986; Forbes, 1991; Gunn et al., 1991;
Ferguson, 1997; Zarnke et al., 2006; Kutz, 2007). The 11 Brucella-positive caribou
detected in the current survey were in the Bluenose-West and Bathurst herds only.
Previous and recent studies have also shown higher seropositivity for N. caninum and T.
gondii in North American caribou than was detected in the herds survey (Kutz et al.,
2001; Zarnke et al., 2000, 2006; Stieve et al., 2010). A recent report documented 11.5%
seropositivity for N. caninum (45/390 caribou) in Alaska using an indirect fluorescent
antibody test, and the authors suggested that this might be explained by the many
potential definitive hosts in the state, including coyotes and possibly wolves (Stieve et al.,
2010). Kutz et al. (2001) surveyed five barrenground caribou (R. t. groendlandicus) herds
of northern Canada for antibodies to T. gondii using the same MAT protocol and
laboratory as were used in the current study. Of 117 caribou tested from three mainland
populations (including Bathurst and the population now designated as Bluenose-West),
43 (37%) were positive and 9 of these had very high titers (1:500). In the current
serosurvey, only 8 (1%) of 558 caribou tested were positive for T. gondii. None of these
was from the Bluenose-West herd (where Kutz et al. [2001] had detected 6/15 caribou
positive; 40%) and only 1 of the 8 had a relatively high titer (1:200). Although the USDA
laboratory had tested FP samples in the MAT for T. gondii previously (O. Kwok,
personal communication), during the current study, laboratory personnel reported some
difficulty reading these tests due to colouration of FP eluates. Thus, it is possible that
142
some FP tests for T. gondii were misclassified as negative.
Today, numerous caribou and reindeer herds are in serious decline worldwide and some
are thought to be stabilizing after declines (Vors and Boyce, 2009; Fiesta-Bianchet et al.,
2011; CARMA, 2012). There is concern about the resilience of these populations given
the significant, rapid environmental changes and other perturbations that are impacting
Rangifer habitat at local and global scales (Anisimov et al., 2007; Vors and Boyce, 2009;
Fiesta-Bianchet et al., 2011; CARMA, 2012). As it is not feasible to do aerial population
surveys of migratory caribou herds every year, there are sizeable gaps in population
estimates for virtually all populations. Regarding trends for “migratory tundra” caribou
(which include the barrenground and migratory woodland herds in this study)
specifically, many of these herds peaked in the 1980s and 1990s, and some have since
exhibited spectacular declines but may be stabilizing currently (Festa-Bianchet et al.,
2011; CARMA, 2012; Table 2). While it is known that caribou populations tend to cycle
over a span of decades, each period of decline is a complex dynamic process that is
multifactorial and specific to each population (Bergerud, 1996; Whitten, 1996; Gunn,
2003). Similarly, while pathogens are thought to play various roles in the population
dynamics of Rangifer and other cervids, it is not possible to make sweeping statements
about disease impacts that apply universally (Conner et al., 2008). However, considering
that some of the herds in this study were declining during the time of sampling or may
have been stabilizing after a decline, mortality of seropositive/diseased animals might
partially explain low overall seroprevalence findings.
The finding of <1% overall N. caninum seropositivity in the current survey was
143
considerably lower than the noted recent report of 11.5% prevalence in Alaskan caribou
(Stieve et al., 2010). That same study revealed 0.4% (2/452) prevalence of T. gondii in
Alaskan caribou, which is similar to the current serosurvey finding of 1% (8/556). For
both these parasites, transmission occurs when caribou ingest oocysts that are shed into
their environment by definitive hosts (i.e., coyotes and possibly wolves for N. caninum,
and most likely lynx for T. gondii) (Dubey, 2003). Stieve et al. (2010) detected 9% N.
caninum seroprevalence in 324 Alaskan wolves and high seroprevalence for N. caninum
observed in red fox (Vulpes vulpes) in Europe are evidence that foxes may also be
definitive hosts for this parasite (Buxton et al. 1997; Dubey, 2003). Stieve et al. (2010)
detected no N. caninum seropositives in a small sample of Alaskan red foxes (n=9).
Relevant to vertical transmission, in caribou Stieve et al. (2010) observed no significant
difference between the N. caninum seroprevalence results for young and older caribou
(12% vs. 10% for groups <1 year and ≥1 year of age, respectively). They pointed to this
as a signal of the potential importance of transplacental transmission of this parasite in
caribou populations. Regarding T. gondii, as noted, it is possible that other vectors such
as migratory fowl are involved in the cycling of this pathogen in the North.
Comparing findings for N. caninum and T. gondii seroprevalence in caribou, Stieve et al.
(2010) identified the higher density of definitive hosts for N. caninum in Alaska as a
likely explanation for the differential. The same might be expected across the ranges of
the North American caribou herds in the current study, whereas Greenland has no
predators and absence of seropositivity for both these parasites is not surprising.
However, virtually none of the North American caribou surveyed were positive for N.
144
caninum and the findings for T. gondii were lower than documented previously. The
former might be partially explained by a difference in test platforms. The Alaskan study
used an immunofluorescent antibody test (IFAT) for N. caninum (with no sensitivity or
specificity for caribou samples noted), whereas the current study used a cELISA kit
developed for cattle. While the cELISA kit had not been validated for Rangifer, it
performed reliably in experiments with reindeer that had been vaccinated with a cattle
vaccine for N. caninum (P. Curry and S. Kutz, unpublished data). It also performed
reliably over multiple sample-storage periods in unvaccinated reindeer that had tested
positive for N. caninum (see Chapter 4). However, this cELISA may be less sensitive for
caribou anti-N. caninum antibodies than the IFAT.
The apparent drop in T. gondii prevalence in the western herds might reflect a change in
the population dynamics or movements of lynx, the suspected definitive host in arctic
North America. Other possible contributors are altered range overlap between lynx and
caribou due to caribou migration route shifts, weather/snow depth issues, availability of
forage for caribou, or a decline in snowshoe hares (Lepus americanus). As snowshoe hare
are the primary food-web influence on Canadian lynx density, these species’ population
trends are closely tied (Poole, 1994; Stenseth et al., 1997). To the author’s knowledge,
there is no information available on recent lynx population changes in the areas relevant
to the herds surveyed.
Novel Biodiversity: Potential Research Directions
The finding of relatively high seroprevalence for BHV-1, PI-3, and Pestivirus overall
145
compared to the other pathogens investigated is in accord with most previous serological
surveys for these virus groups in Rangifer worldwide (Elazhary 1979, 1981; Zarnke,
1983; Rehbinder et al., 1992; Stuen et al., 1993; Farnell et al., 1999; Lillehaug et al.,
2003; Johnson et al., 2010; Evans et al., 2012; Kautto et al., 2012). In 1996, a distinct
Pestivirus (Reindeer-1) was isolated from a reindeer in Germany and this has been
sequenced (Becher et al., 1999, 2003; Avalos-Ramirez et al., 2001). Canadian woodland
caribou (R. t. caribou) have been found to have antibodies to a distinct alphaherpesvirus
isolated from elk (Cervus elaphus nelsoni) (Tessaro et al., 2005). Cervid alphaherpesvirus
is a separate virus from the elk virus. This agent has been isolated from Norwegian
reindeer (R. t. tarandus) and experimental work has demonstrated viremia, latency, and
horizontal and vertical transmission (Das Neves et al., 2009a, 2009b, 2010). Very
recently, Evans et al. (2012) reported high seroprevalence for alphaherpesvirus in
Alaskan caribou and reindeer (47% and 60%, respectively), as well as evidence that the
specific circulating virus is CvHV2. Confirmation of this isolate in North America would
be the first proof that it is cycling in Rangifer outside Europe. The current survey’s
results for the Porcupine, Bluenose-West, and Bathurst herds point to the Western Arctic
as a potential target location for future research on Rangifer-specific isolates and strains
of alphaherpesviruses and pestiviruses.
Geographic Differences
This study revealed generally higher seroprevalence for BHV-1, PI-3 and BRSV in the
West, with lower reactivity in Québec and almost none in Greenland (Table 3). One
plausible explanation for this is the different evolutionary origins of the caribou in these
146
regions. During the Pleistocene Epoch, the continental ice sheet created a massive
geographical barrier to animal movements and mixing. Phylogenetic analysis points to
two main original groupings of Rangifer, one that was isolated north of the ice (the
Beringian-Eurasian lineage [BEL]) and one that was isolated in the southern refugium
(the North American lineage [NAL]) (COSEWIC, 2011). With the retreat of glacial ice,
the BEL lineage essentially spread north into the Arctic Islands and eastward as the
ancestors of the caribou subspecies R. t. groenlandicus, granti, and pearyi (and sharing
lineage with Scandinavian reindeer [R. t. tarandus]) (COSEWIC, 2011). The NAL
lineage spread north and west through the boreal region of North America, including
Québec and Labrador, and north into the Rocky and Mackenzie Mountains as the
ancestors of R. t. caribou (COSEWIC, 2011).
Genetic analysis suggests there was some introgression at contact sites between NAL and
BEL; however, the two groupings of clades are distinct (COSEWIC, 2011). Thus,
evidence suggests that the caribou from the three “regions” sampled in this serosurvey are
descendents of primarily the BEL (western herds), primarily the NAL (the Québec
herds), and likely descendents of BEL animals that crossed from arctic North America
(Ellesmere Island) to what is western Greenland today (Roed, 2005). There is known to
be some introgression between populations in North America, yet the caribou in these
regions remain largely distinct from each other based on geography, life
history/behaviour characteristics, and different movement patterns (COSEWIC, 2011). It
is possible that groups of pathogens have remained circulating within and among
populations that originated from the ancestors of these distinct lineages, and that the
147
seroprevalence results in this study reflect this.
Prevalence of disease and pathogen exposure within caribou populations likely reflects a
dynamic mix of evolutionary and ecological factors. In some cases, Rangifer
translocations may also be involved. In this study, the western herds had the highest
seroprevalence for BHV-1, PI-3, and Pestivirus, whereas the Québec herds had somewhat
lower seropositivity on these assays, and the Greenland samples included only two
animals seropositive for BHV-1 (see discussion of Greenland herds below) (Table 3).
Introductions of Rangifer from Eurasia to North America may be relevant to this pattern.
Reindeer were first imported to North America (from Siberia to Alaska) in 1891, and
subsequent imports and husbandry generated a burgeoning population that exceeded
600,000 by the 1930s (Olson, 1969). The current population of Alaskan reindeer is
estimated at 15,000, and explanations for the overall decline include co-mingling with
and emigration to native migratory caribou herds, and increased predation by wolves (refs
in Evans et al., 2012). The Canadian government purchased a subpopulation of Alaskan
reindeer in 1929 and re-located these animals overland to the Mackenzie River Delta
region (Scotter, 1969). Only 10% of the original 3,000 animals were estimated to have
survived the 63-month journey, whereas the rest of the >2000 that arrived were born en
route (Scotter, 1969). Similar to events in Alaska, these reindeer introgressed with and
emigrated to native barrenground caribou herds of the Mackenzie Delta area. Although
there may have been pathogen species loss from the Eurasian host population during their
translocation and arduous overland trek (Torchin et al., 2003), pathogens could have
transmitted from these reindeer to ancestors of what is today the Bluenose-West herd, and
148
to other herds of the Western Arctic as well. Broughton et al. (1970) reported serological
evidence of brucellosis in reindeer from the Mackenzie Delta region, and clinical
brucellosis has been documented in Rangifer from Siberia and Alaska (refs in Tessaro
and Forbes, 1986). As well, alphaherpesviruses, paramyxoviruses, and pestiviruses are
known to be endemic to Scandinavian reindeer, which are managed/herded Rangifer
populations (Lillehaug et al. 2003). In this study, the three western herds had, in general,
the highest seroprevalence estimates for BHV-1, PI-3, and Pestivirus (Tables 3 and 5).
The minimal pathogen reactivity observed in the southwest Greenland herds might reflect
three key phenomena: i) Loss of pathogen species when, presumably, few Rangifer hosts
survived the passage from the arctic islands of North America to colonize Greenland,
effectively as “invaders”; ii) the genetic bottleneck that would have occurred for the hosts
and their accompanying parasites at this evolutionary juncture; iii) isolation on an arctic
island. Comparisons of parasites in species in native and introduced ranges have revealed
an “escape” or lag phenomenon of reduced parasitization in individuals who are at the
frontier of a host range advance or range expansion (Torchin et al., 2003; Phillips et al.,
2010). Similar to the current study’s findings in these two Greenland herds, a 1993
survey for three pathogens (reindeer herpesvirus [an alphaherpesvirus], PI-3, and BVDV)
in 40 reindeer (R. t. platyrhynchus) isolated on the arctic Norwegian island of Svaalbard
revealed no seropositivity for these agents (Stuen et al., 1993).
Genetic studies have demonstrated reduced genetic variability in the Greenland caribou
herds (Jepsen et al., 2002; Roed, 2005), and Cuyler (2007) has documented significant
fluctuations in population size for both the Kangerlussuaq-Sisimiut and Akia-Maniitsoq
149
herds. These findings are evidence of one or multiple genetic bottlenecks for Greenland
caribou over time.
With respect to isolation on Greenland and influences of geographic features, the
Kangerlussuaq-Sisimiut and Akia-Maniitsoq populations are the two largest herds in
southwest Greenland and are effectively separated from each other by a glacier (Cuyler et
al., 2011). The two populations are also separated from other Greenlandic herds by deep
fiords (Cuyler et al., 2011). Two hundred sixty-three Norwegian reindeer were introduced
to the area in 1952 to supplement the small number of caribou on Greenland at the time
(Cuyler and Haugerud, 1999; Jepsen et al., 2002). Genetic studies have demonstrated that
that these animals introgressed with the Akia-Maniitsoq herd almost exclusively (as
opposed to Kangerlussuaq-Sisimiut), and the Akia-Maniitsoq herd has also had more
contact with domestic animals historically (Rose et al., 1984). Alphaherpesviruses are
endemic in Norwegian reindeer (Das Neves et al., 2010), and mixing with reindeer and
other domestic species might explain the current serosurvey finding of BHV-1 reactors (2
of 47 caribou sampled) in the Akia-Maniitsoq herd. As well, a government report from
the 1990s noted serological evidence of alphaherpesvirus presence in Greenland caribou
(Anon., 1999 cited in Das Neves et al., 2009).
Sex, Age, Season, Pregnancy
Initial analyses in this study suggested that male caribou had twice the risk of
seropositivity for BHV-1 (i.e., alphaherpesvirus/es) and PI-3 compared to females (Table
5); however, males were overrepresented in the West and in fall collections, and further
150
stratification revealed that season (fall collection) was a likely confounder (Table 7).
Comparison of males versus females in the fall (the only season with reasonable sample
sizes for both sexes) suggested no sex difference. Lillehaug et al. (2003) tested close to
4,000 Norwegian reindeer sera and observed 28.5% positivity for BHV-1 on virus
neutralization, with higher prevalence in males. Scandinavian reindeer are generally
managed populations that are not necessarily isolated from domestic hosts and their
pathogens (Tryland, 2012). They also differ from free-ranging caribou with respect to
other factors that may impact disease transmission; for example, i) contact rates related to
migration (Altizer et al., 2011); ii) potential negative energy balances linked with
migration, calving time, and foraging stress, all of which might affect immunity (Russell
et al., 1993; Altizer et al., 2011); iii) other behavioural and life-history characteristics.
Evans et al. (2012) found no significant association between sex and seroprevalence for
BHV-1 in their results for 292 Alaskan reindeer. The current study’s results for herds
surveyed in arctic North America and Greenland do not provide evidence for a
conclusion about sex differences in exposure to alphaherpesviruses and PI-3 in migratory
caribou herds.
Compared to young caribou (calves and subadults combined), adults had greater risk of
exposure for BHV-1, PI-3, and Pestivirus. The age strata comparisons for the Québec
herds, with sample sets of cow-calf pairs exclusively, did not differ from others. The
finding of higher seropositivity for adults is consistent with extended exposure to
pathogens or, specifically, longer experience with repeated annual cycles of encountering
risk factors for transmission, which some authors categorize loosely as animal
151
movements, animal behaviour, life-history/demographic structure, and domestic animal
contact (Conner et al., 2008). Although age-class definitions vary and alphaherpesviruses
and pestiviruses are endemic in Norwegian reindeer, serosurveys for various members of
these virus groupings in Scandinavian reindeer have shown minimum four-fold higher
prevalence in adults (Stuen et al., 1993; Lillehaug et al., 2003; Kautto et al., 2012).
Again, the age structure, demography, and disease transmission factors in semidomesticated reindeer herds likely differ from those of caribou and should be considered.
Evans et al. (2012) tested 292 samples from Alaskan reindeer in an ELISA for BHV-1
and observed 47% prevalence overall, with adults four times more likely to be positive
than calves.
The current serosurvey results indicated approximately four to five times greater
likelihood of seropositivity for PI-3 and Pestivirus in adult caribou than in young (Table
5). Although no other published age-class comparisons could be found for Rangifer
positivity to PI-3, the documented evidence for other pathogens and the large sample size
of this survey suggests that this finding in caribou is a reasonable indicator of prevalence
(though a possible underestimate, as noted). Based on the findings for age with all three
pathogens/groups, the author controlled for age in the analysis. The risk for BHV-1
positivity in adults relative to young (RR=17.5; Table 5) was much higher than observed
for PI-3 and pestiviruses. One possible contributor to this might be latent infection, which
is a feature of herpesviruses, has been proven for CvHV2 (Das Neves et al., 2010), and
could increase the number of long-term seropositive adults in a population through
recurrent active infections causing repeated waves of contact with susceptible animals.
152
Fall collections were overrepresented in the sampling, and season (fall collection) was
thus considered a potential confounder. Though confidence intervals overlapped and no
strong conclusion can be made, three-way stratification suggested that, for all three
pathogens (BHV-1, PI-3, and Pestivirus), bulls and cows collected in fall might have
somewhat higher risk of exposure (Table 7). This was observed across all five individual
herds that had fall collections, and was almost fully consistent across herds’ fall-cows and
fall-bulls, respectively, as well (Table 7; comparisons by herd not shown). The
collections for the study were done in early fall prior to the rut. Contact rates in spring
and summer are potentially high, with increased animal movement and aggregations,
increased potential for contacts during migration stopovers and on calving grounds,
where pathogen-infected placental and genital fluids may be present. However, there is a
lag time for antibody production after exposure and pathogen prevalence may be
relatively lower in spring than in fall because clinically infected animals have reduced
overwinter survival or older animals, many of which are seropositive, tend to die during
winter. The very low sample numbers for summer and winter precluded any specific
conclusions for these sampling times. Overall (and cautiously because sample sizes after
stratification were small), the study findings for season suggest that, for long-term
monitoring programs, fall sample collection for serology might provide the most
comprehensive picture of the pathogens that are circulating in migratory caribou herds of
the North.
The overall one-way analysis indicated that pregnant females were approximately
fivefold more likely to be Pestivirus-positive than non-pregnant females (Table 5). This is
153
of interest considering that pestiviruses and alphaherpesviruses may be vertically
transmitted. Such a result might signal an important route of transmission for pestiviruses
in caribou; however, pregnancy status data were only available for 93 animals and sample
sizes precluded robust herd and multi-strata analyses. This potential difference linked
with pregnancy warrants future investigation with larger collections and underlines the
importance of recording pregnancy data during collections where caribou pregnancy can
be determined (spring and winter).
Herd / Ecological Factors
The correlations of herd demographic and environmental/ecological factors (mean
temperatures on herd range in June and July, respectively; mean animal density on
summer range) with herd prevalence for BHV-1, PI-3, and Pestivirus yielded no
significant associations; however, the number of data pairs was small. Based on this
study, no conclusions can be made about pathogen prevalence relative to these
parameters.
Surveillance Methodologies
The seroprevalence estimates in this study (Table 3) can be considered somewhat
conservative for methodological reasons as well as the other factors discussed above.
Filter-paper samples were tested after more than 2 years’ storage time in some cases.
Curry et al. found that, after 12-17 months of dry storage at room temperature and
compared to serum stored for the same period, FP samples in the antibody assays for
Brucella spp., WNV, N. caninum, BHV-1, PI-3, and BRSV retained specificity of 90% or
154
higher but lost some sensitivity (which was ≥88% in all but the BHV-1 test
[approximately 75%]) (see Chapter 4). As documented in studies on FP-based serology
for humans, long-storage FP samples from animals with antibody levels that were close
to the assay threshold at time of collection might have tested negative in this serology
survey due to antibody degradation (Behets et al., 1992). This would be most relevant for
herds in which higher proportions of FP samples from 2007 and 2008 collections were
tested (Rivière-aux-Feuilles, Rivière-George, Bluenose-West; Tables 3 and 4). While
there were no patterns of concern with the FP results compared to the serum results (data
not shown) and FP samples from Rangifer had been validated for all but two tests (T.
gondii and Pestivirus), the efficacy of FP samples in these two tests is not known. This is
an additional reason why true herd seroprevalence to some pathogens in this study might
be higher than observed.
The risk analyses for BHV-1, PI-3, and Pestivirus highlighted some challenges of the
CARMA sample set and some apparent differences within strata, within and across herds,
and across regions. Some of these might inform future long-term or large-scale
monitoring efforts. Collection of a more even distribution of males and females would
facilitate more robust analysis. The overall one-way stratified analyses suggested higher
risk of BHV-1 in males, yet this was largely an effect of overrepresentation of males in
the samples from western herds and it was not possible to examine sex differences
robustly. Initial analyses also indicated that pregnant females were more likely to be
Pestivirus-positive than non-pregnant females. When this comparison was tested
specifically within the Bathurst herd (the only Pestivirus-positive herd with definitive
155
pregnancy data), there was no statistical difference between these two groups and this
may have been related to sample size. In general, in-depth analyses of the Pestivirus
findings by herd and other strata were limited by sample size. More even representation
of females in the database and more consistent recording of females’ pregnancy status
during spring and winter collections would have helped clarify whether the initial
Pestivirus finding (possible association of seropositivity with pregnancy) was
meaningful.
Synopsis and Conclusion
The establishment of baselines and new prevalence landmarks for seroprevalence to
pathogens during International Polar Year 2007-2009, the patterns and differences within
and among herds and regions that were observed in this serosurvey, and the discovery of
potential new directions for research are all steps towards fuller understanding of disease
ecology in migratory caribou herds of North America and Greenland. As well, the
epidemiological findings from this study provide evidence for making decisions about
when and how best to sample and monitor migratory caribou herds serologically to
achieve desired goals.
Acknowledgements
The sacrifice of the caribou in this study is acknowledged and the author conducted this
research bearing responsibility for these lives. The extensive sampling of each of these
animals through CARMA has contributed to advancements in the knowledge of caribou
156
health through this and other studies. Meat from the caribou sampled from the Porcupine,
Bluenose-West, and Bathurst herds went to local northern communities for consumption.
Extensive collaborative efforts by the CARMA network, local logistics teams, graduate
students, hunters, and community-hunt participants of Colville Lake and Fort Good
Hope, Northwest Territories, Canada enabled the collection of samples and data for this
study. Each of these contributions is recognized and appreciated, and special thanks to the
leads and participants in the herd collections, including but not limited to the following:
Dorothy Cooley and Martin Kienzler (Porcupine herd); Marsha Branigan, Tracy Davison,
Susan Kutz, and Nathan deBruyn (Bluenose-West herd); Bruno Croft and Brett Elkin
(Bathurst herd); Steeve Côté and Joelle Taillon (Rivière-aux-Feuilles, Rivière-George);
Christine Cuyler (Akia-Maniitsoq, Kangerlussuaq-Sisimiut). Technical expertise was
provided by laboratory personnel Linda Kelly, Lori Hassard, Laura Bond, Roberta
Yemen, Antonia Dibernardo, and Dr. Oliver Kwok. Carlos Das Neves and the Tryland
lab team also contributed laboratory expertise, and much gratitude goes to Dr. Klaus
Nielsen’s laboratory at the Brucella Centre of Expertise, Canadian Food Inspection
Agency, Nepean, Ontario, Canada for in-kind support and graduate student training. Don
Russell, Ryan Brook, and Julie Ducrocq worked extensively on the CARMA database
and provided herd data and graphics. Thanks also to Carl Ribble, Susan Kutz, Karin
Orsel, and Morten Tryland for their valuable support and comments on this chapter.
Funding for this research was provided by the Circum-Arctic Rangifer Monitoring and
Assessment Network (CARMA, www.carmanetwork.org), International Polar Year
Funding from the NSERC Special Research Opportunity Program, Environment
Canada/Natural Resources Canada, Nasivvik Centre for Inuit Health and Changing
157
Environments (Canadian Institutes of Health Research), Alberta Innovates Technology
Futures, and the University of Calgary Faculty of Veterinary Medicine.
158
TABLES
Chpt 5. Table 1. Pathogens and their effects
Information about the agents in the study (adapted from Curry, 2010; sources: Tessaro
and Forbes, 1986; Rockborn et al., 1990; Forbes, 1991; Gunn et al., 1991; Oksanen et al.,
1997; Dubey et al., 2002; Palmer et al., 2004a; Dubey and Jones, 2008).
a
AGENT TYPE EFFECTS IN Rangifer ZOONOTIC IMPACT Brucella spp. Bacteria Abortion, weak calves, orchitis, joint disease, abscesses Multisystemic chronic disease Neospora caninum Protozoan Unknown West Nile virus Virus Neurological, death
(reindeer) Toxoplasma gondii Protozoan Abortion, lethal enteritis (reindeer) Abortion, birth defects Bovine herpesvirus‐1 Alphaherpesvirus Oral and ocular lesions, enteritis (reindeer) None known Bovine respiratory syncytial virus Paramyxovirus Unknown c
None known Parainfluenza virus‐3 Paramyxovirus Unknown c
None known Pestivirus Pestivirus Unknown c
None known a
None known b
b
Neurological, death Causes spontaneous abortion in cattle. Requires mosquito vector. c
Risk of transmission from domestic ruminants; virus groups endemic in Scandinavian reindeer (Lillehaug et al., 2003) b
159
Chpt 5. Table 2. Herd parameters and collection dates
Herd parameters and collection dates, with herds listed west to east and identified within
a region for epidemiological analysis (www.carmanetwork.org; MRNF, 2012; Taillon et
al., 2012; C. Cuyler, personal communication).
HERD POP ESTIMATE (year/s) CURRENT TREND CARMA COLLECTIONS (season, yr) REGION MEAN TEMP SUMMER RANGE* June/July (˚C) ANIMAL DENSITY SUMMER RANGE* 2
(per km ) Porcupine 169,000 (2010) Increasing Fall 08, 09 Summer 09 West 8.9/11.5 2.2 Bluenose‐
West 17,897 (2009) Stable Fall 07, 08, 10 Spring 10 Winter 10 West 8.9/12.1 0.6 Bathurst 32,000 (2009) Declining Fall 07, 08 Spring 08, 09 West 9.7/14.0 0.5 Rivière‐
aux‐
Feuilles 430,000 (2011) Declining Québec 6.3/11.2 1.9 Rivière‐
George 27,600 (2012) Declining Québec 8.2/12.2 0.8 Akia‐
Maniitsoq 36,000 (2005‐10) Slowly declining Spring 08 Greenland 7.1/9.9 12.5 Kang‐
Sisimiut 100,000 (2005‐10) Stable Spring 09 Greenland 9.6/12.3 103.0 Fall 07, 08, 09 Summer 07, 08, 09 Fall 07, 08, 09 Summer 07, 08, 09 * Average 2007‐09 Abbreviations: Kang: Kangerlussuaq; Pop: population; Temp: temperature 160
Chpt 5. Table 3. Prevalence estimates by herd and overall
Herd seroprevalence (Prev) and n values for each pathogen test in each herd. Herds are
listed west to east geographically left to right. Clopper-Pearson Exact 95% confidence
intervals (CI) are shown for the prevalence estimates.
ASSAY Brucella N. caninum WNV T. gondii BHV‐1 PI‐3 BRSV Pesti POS n Prev CI (%) POS n Prev CI (%) POS n Prev CI (%) POS n Prev CI (%) POS n Prev CI POS n Prev CI (%) POS n Prev CI (%) POS n Prev CI (%) PCH BW BA 0 33 0% (0‐11) 0 33 0% (0‐11) 0 33 0% (0‐11) 0 33 0% (0‐11) 15 32 47% (29‐65) 15 32 47% (29‐65) 0 32 0% (0‐11) 15 27 56% (35‐75) 5 53 9% (2‐17) 0 53 0% (0‐7) 0 52 0% (0‐7) 0 50 0% (0‐7) 32 52 62% (47‐75) 5 52 10% (3‐21) 0 52 0% (0‐7) 11 33 33% (18‐52) 6 143 4% (2‐9) 2 142 1% (0‐5) 0 141 0% (0‐3) 6 141 4% (2‐9) 52 140 37% (29‐46) 4 140 3% (1‐7) 0 140 0% (0‐3) 59 141 42% (43‐50) HERD R‐F 0 119 0% (0‐3) 0 119 0% (0‐3) 0 119 0% (0‐3) 1 120 <1% (0‐5) 16 119 13% (8‐21) 9 119 8% (4‐14) 0 119 0% (0‐3) 23 119 19% (13‐28) R‐G AK KA ALL HERDS 0 116 0% (0‐3) 0 115 0% (0‐3) 0 116 0% (0‐3) 1 115 <1% (0‐5) 21 111 19% (12‐27) 4 111 4% (1‐9) 0 111 0% (0‐3) 39 116 34% (25‐43) 0 47 0% (0‐8) 0 47 0% (0‐8) 0 47 0% (0‐8) 0 47 0% (0‐8) 2 47 4% (1‐15) 0 47 0% (0‐8) 0 47 0% (0‐8) 0 47 0% (0‐8) 0 50 0% (0‐7) 0 50 0% (0‐7) 0 50 0% (0‐7) 0 50 0% (0‐7) 0 50 0% (0‐7) 0 50 0% (0‐7) 0 50 0% (0‐7) 0 50 0% (0‐7) 11 561 2% (1‐3) 2 559 <1% (0‐1) 0 558 0% (0‐1) 8 556 1% (1‐3) 138 551 25% (21‐29) 37 551 7% (5‐9) 0 551 0% (0‐1) 147 533 28% (24‐32) TOTAL TESTS: 4,420 Abbreviations: AK: Akia‐Maniitsoq; BA: Bathurst; BHV‐1: Bovine herpesvirus‐1; BW: Bluenose‐West; KA: Kangerlussuaq‐Sisimiut; PCH: Porcupine; Pesti: Pestivirus; PI‐3: parainfluenza virus‐3; R‐F: Rivière‐aux‐
Feuilles; R‐G: Rivière‐George
161
Chpt 5. Table 4. Proportions of filter papers and serum tested
Proportions of filter-paper (FP) and serum samples from each herd that were tested for
Brucella spp. (the highest priority test) and for the three pathogens/pathogen groups with
the highest overall seroprevalence.
162
Table 4.
163
Chpt 5. Table 5. Results of overall one-way analyses
Significant findings for risk of exposure/infection (seropositive status) in the overall oneway analyses. Herds are numbered in ascending order from west to east. Relative risk
(RR) of being seropositive was calculated using adjusted risk values as required.
Referents are denoted by a double dash.
PATHOGEN VARIABLE GROUP No. POS n Risk RR p BHV‐1 SEX Female 72 381 19% ‐‐ <0.01 Male Adult Young 1 (PCH) 2 (BW) 3 (BA) 4 (R‐F) 5 (R‐G) 6 (AK) 7 (KA) 64 132 3 15 32 52 16 21 2 0 160 378 166 32 52 140 119 111 47 50 40% 35% 2% 47% 62% 37% 13% 19% 4% 0% 2.1 17.5 – 3.6 4.6 2.8 ‐‐ 1.4 0.3 *0.1 <0.01 <0.01 Female Male Adult Young 18 19 33 3 381 160 378 166 5% 12% 9% 2% ‐‐ 2.4 4.5 – <0.01 <0.01 AGE Adult Young 136 11 363 169 38% 7% 5.4 ‐‐ <0.01 SEASON Fall 87 300 29% 1.7 <0.01 Spring 27 160 17% – Summer Winter 33 0 73 0 45% / 2.7 / Non‐pregnant 1 17 6% ‐‐ 0.03 Pregnant 25 76 33% 5.5 1 (PCH) 15 27 56% 2.9 <0.01 2 (BW) 11 33 33% 1.7 3 (BA) 59 141 42% 2.2 4 (R‐F) 23 119 19% ‐‐ 5 (R‐G) 39 116 34% 1.7 6 (AK) 0 47 0% *0.1 7 (KA) 0 50 0% *0.1 AGE LOCATION SEX PI‐3 AGE PESTIVIRUS PREGNANCY LOCATION * Calculated using adjusted risk / Not calculated or analyzed Abbreviations: AK: Akia‐Maniitsoq; BA: Bathurst; BW: Bluenose‐West; KA: Kangerlussuaq‐Sisimiut; PCH: Porcupine; R‐F Rivière‐aux‐Feuilles; R‐G Rivière‐George 164
Chpt 5. Table 6. Comparisons of risk across regions
Comparisons of risk of exposure/infection (seropositive status) across the three regions
for groups identified in the one-way analyses. Sample size precluded comparisons for
pregnancy. P values and Clopper-Pearson Exact 95% confidence intervals are shown.
PATHOGEN GROUP BHV‐1 Males Females Adults Males PI‐3 Females Adults PESTIVIRUS Adults Fall Spring WEST QUEBEC GREENLAND p 64 124 0 32 0 3 Risk CI (%) POS n Risk CI (%) POS n Risk CI (%) 52% (42‐61) 33 95 35% (25‐45) 93 176 53% (45‐60) 0% (0‐11) 37 198 19% (14‐25) 37 132 28% (21‐37) 0% (0‐71) 2 89 2% (0‐8) 2 70 3% (0‐10) <0.01 POS n Risk CI (%) POS n Risk CI (%) POS n Risk CI (%) 19 124 15% (9‐23) 5 94 5% (2‐12) 21 176 12% (8‐18) 0 32 0% (0‐11) 13 198 7% (4‐11) 12 132 9% (5‐15) 0 4 0% (0‐60) 0 89 0% (0‐4) 0 70 0% (0‐5) POS n Risk CI (%) POS n Risk CI (%) POS n Risk CI (%) 81 161 50% (43‐58) 58 138 42% (34‐51) 27 63 43% (30‐56) 55 132 42% (36‐49) 29 133 22% (13‐30) * * * 0 70 0% (0‐5) * * * 0 97 0% (0‐4) POS n <0.01 <0.01 0.04 0.05 0.01 <0.01 <0.01 <0.01 * no data available for calculation (i.e., not sampled in the season or no pregnancy data from that region) 165
Chpt 5. Table 7. Results of three-way stratification by region
Three-way stratification of risk of exposure/infection (seropositive status) for BHV-1, PI3, and Pestivirus by region (West, Québec, [Que] and Greenland [Grld]) and with
Clopper-Pearson Exact 95% confidence intervals (CI) shown. None of the samples from
the West region that were tested for Pestivirus were collected in summer or winter.
166
Table 7.
167
FIGURES
Chpt 5. Figure 1. Herds surveyed
a) Ranges of migratory caribou and reindeer herds, with inset showing the general area
encompassed by the study herds; b) Close-up of the ranges of the seven herds in the
serology study. (source: D. Russell, www.carmanetwork.org)
168
CHAPTER SIX
HUNTER-BASED WILDLIFE SAMPLING AND
FILTER-PAPER BLOOD COLLECTION:
PERCEPTIONS OF CARIBOU HARVESTERS FROM ACROSS
NORTHERN CANADA
169
Abstract
Two community-based caribou health-monitoring programs were launched in Canada’s
North between 2003 and 2008, but implementation and acceptance of these programs has
not been formally evaluated at community level. The aim was to assess attitudes and
beliefs of northern subsistence harvesters regarding wildlife disease and hunter-based
wildlife health monitoring in general, and hunter use of a field sampling tool (blood
collection on filter paper [FP]) in particular. In 2009 and 2010, face-to-face semistructured interviews were conducted with 25 caribou harvesters in three communities of
Nunavut, Canada (n=16) and four communities of the Sahtu Settlement Region of the
Northwest Territories, Canada (n=9). All participants volunteered. Results were analyzed
quantitatively and qualitatively. The main findings were that most harvesters viewed
wildlife disease as important, most were somewhat interested in hunter-based sampling,
and very few had issues with FP sampling. The most frequent critique was perceived lack
of results reporting. Two hunters described cultural beliefs that could be barriers to
sample collection from caribou. Community-based wildlife monitoring in the Arctic is
complex and challenging, and the two programs evaluated are still new. Clues from other
Aboriginal or northern community-engagement models and focus on the identified areas
for improvement could enhance the impact and sustainability of these and other similar
initiatives.
170
Introduction
Many communities in Canada’s North are intimately tied to wildlife for species’ integral
and cultural values, and for their traditional, preferred, and rich nutritional benefits
(Receveur et al., 1997; Receveur and Kuhnlein, 1998; Batal et al., 2005; Gordon, 2003,
2005; Kendrick and Lyver, 2005; Lambden et al., 2007; Wesche et al., 2011; Giroux et
al., 2012). Today, food security is considered an urgent concern for Canadian Aboriginal
people because of low income levels, reliance on the nutrient quality of traditional foods,
traditional harvest-sharing systems, effects of environmental change on food-source
sustainability, and other factors (Chan et al., 2006; Furgal and Seguin, 2006; Lambden et
al., 2007; Power, 2008; Wesche et al., 2011). For those who reside in the Arctic, the cost
of market food is particularly high and can be prohibitive, and this means that many
Northerners depend heavily on wildlife food sources (Batal et al., 2005; Giroux et al.,
2012). Specific to the environment, there is unease about climate-change impacts on
wildlife availability as sustainable, secure, and safe food sources in the North (Furgal and
Seguin, 2006; Guyot et al., 2006; GRID-Arendal, 2009; Wesche and Chan, 2010; Wesche
et al., 2011). In addition to spiritual and other values that wildlife embody, these
nutritional and food safety/security concerns underline the importance of monitoring
wildlife population health for Northerners.
Wildlife harvesting occurs year round across the Arctic, and typically shifts in accordance
with the seasonal cycles of the different species (e.g., Fig. 1). This means that subsistence
hunters from remote northern towns and hamlets may harvest in all seasons, and can
access the ranges of caribou (Rangifer tarandus ssp.) and moose (Alces alces) in ways
171
and frequencies that few scientists are logistically able to undertake. The concept of local
hunters collecting samples for wildlife health assessment in the North is attractive from a
science perspective and is not new. In Canada, wildlife “co-management” initiatives are
mechanisms of power sharing between government and local resource users (Berkes,
2009) that are linked with Aboriginal land claims agreements. Various such initiatives, as
well as federal and provincial government agencies, and have employed communitybased sampling to monitor marine and terrestrial species (DFO, 2009; Peacock et al.,
2009; Armitage et al., 2011).
Researchers have cited a number of benefits that hunter-based monitoring can offer
communities, including active participation in assessing stocks that local people value
and depend upon, raised local awareness and ownership of wildlife management issues,
trust-building between wildlife managers and users, and means to include indigenous
understandings of wildlife species’ ecology in future management considerations
(Harwood et al., 2002; Kofinas et al., 2003; Lyver and Gunn, 2004; DFO, 2009; Peacock
et al., 2009; DFO, 2010). However, formal assessments of community/hunter
perspectives on these programs are rare. Further, it is not uncommon for local people to
be engaged in so-called “community-based” programs only peripherally (Brook and
McLachlan, 2008). Examples are involvement strictly as paid sample collectors or
observers, roles in which community members remain mostly at arm’s length from the
research/analysis and, in some cases, may only be contacted by organizers remotely or by
a third party.
172
In contrast to this, indigenous people in Canada’s North have expressed interest in being
integrally involved with community-based research (Kendrick and Lyver, 2005; ITK and
NRI, 2007; Maar et al., 2011). Further, new regulations under Aboriginal land claims
agreements mandate community engagement, consultation, and accessible results
reporting (Byers, 1999; Gearheard and Shirley, 2007). In addition, researchers have
begun to focus more on the convergence of western science with traditional knowledge
(Furgal et al., 2006; Berkes et al., 2007; Brook and McLachlan, 2008). Harvester-based
approaches for caribou monitoring have been described by Kofinas et al. (2003;
assessment of body condition), Lyver and Gunn (2004; prediction of pregnancy),
Kendrick and Lyver (2005; caribou movements), and others (Veitch et al., 2005; Veitch
and Kutz, 2006; and Brook et al., 2009; multiple health indicators). However, even
though more wildlife managers and ecological and conservation scientists are now
seeking to incorporate insights/understandings from local or traditional ecological
knowledge in their studies (Brook and McLachlan, 2008), this is a complex challenge
with no one optimum approach (Raymond et al., 2010). Many efforts are failing to
actively and fully engage community members in the research process (Brook and
McLachlan, 2008). Despite documented support for community participation in northern
research, communities have also raised significant concerns about the status quo, citing
lack of fundamental community involvement (i.e., consultation/input on research
questions and study design), lack of maximal engagement in the research process, lack of
local relevance, token inclusion of local expertise, and inadequate reporting of results,
among other issues (ITK and NRI, 2007).
173
Specific to hunter-based wildlife health monitoring, Brook et al. (2009) noted the benefits
of unified approaches that encompass a mix of science and community elements.
Considering the above-noted concerns, northern communities may benefit most from
these programs if hunter participants are invested for reasons beyond remuneration, if
their expertise is valued, if test results are reported back in timely and accessible ways,
and if the findings shed light on food safety and on the health of wildlife populations that
communities rely upon and co-manage. Fundamentally, such monitoring initiatives hinge
on certain key determinants (Kofinas et al., 2003), including hunters’ desire to take part,
methods that fit with hunting practices (i.e., selection of sampling tools and techniques
that are acceptable to collectors), and methods that yield robust scientific data. These
were the central issues in the current study.
Study Context: Hunter-Based Caribou Sampling Programs
This research took a case-study approach to assess the subjective experiences, attitudes,
and beliefs of subsistence harvesters of northern wildlife relative to hunter-based wildlife
sampling (Mayan, 2001). Prior to the study, health-monitoring programs for caribou (and
caribou/moose) had been implemented to different degrees in the North Baffin Region
(Qiqiktaaluk) of Nunavut Territory (NU), Canada and in the Sahtu Settlement Region of
the Northwest Territories (NT), Canada (Figs. 2 and 3).
Program in the North Baffin Region:
Nunavut’s Caribou Health Monitoring Program (CHMP) was initiated in 2007. This
initiative was funded by the Government of Nunavut’s Department of Environment
174
(GNU DoE) and by the Nunavut Wildlife Management Board, a public government
agency that is the main instrument of wildlife management in the territory. The program’s
overall aim was to implement and assess a subsistence-hunter-based system of
information and sample collection for caribou health monitoring and genetic analysis
(Jenkins, 2009). The primary disease concern for the CHMP was brucellosis, a zoonotic
bacterial illness caused by Brucella spp. that is known to exist in caribou on Baffin Island
and that can cause serious illness in humans who contact infected tissues or consume
undercooked meat from an infected animal (Forbes, 1991; Ferguson, 1997; Jenkins,
2009). Community consultation for the CHMP was undertaken in 2007-08, and the
program was launched in autumn 2008 in three isolated hamlets of North Baffin Region
(Pond Inlet, Arctic Bay, and Clyde River; Fig. 2). The CHMP implementation process
involved a presentation to scientific personnel affiliated with the Nunavut Wildlife
Management Board, as well as key in-community meetings with hunters that were
arranged collaboratively by local Hunters and Trappers Organizations (HTOs) and the
GNU DoE (Jenkins, 2009). The hunter meetings were held in Arctic Bay and Pond Inlet,
and involved approximately 25 caribou hunters. Organizers introduced the process of
hunter-based caribou data collection and sampling, and the concept of taking hunter
sampling kits on caribou hunts. Part of each meeting was devoted to hands-on training;
however, methods were demonstrated without an actual caribou carcass (Fig. 4) due to
the general scarcity of caribou for hunting in North Baffin Region in 2008. Wildlife
Conservation Officers within the DoE represented the CHMP in each of the three
respective communities. Posters were placed in DoE offices and in the offices of local
HTOs, and program organizers shipped prepared hunter sampling kits to the conservation
175
officers. Each officer was responsible for making kits available for harvesters to take
caribou hunting, for receiving and temporarily storing any data/samples and signing off
on hunter submissions, and for forwarding submissions to the Regional Wildlife Office
for processing and storage. The arrangement was that hunters who submitted materials
would be compensated immediately by the HTO (CAD 25 or CAD 30 depending on the
sample set submitted), and the HTO would then invoice the DoE.
At the time the NU interviews for this study were conducted, hunter kits had been made
available in Arctic Bay, Pond Inlet, and Clyde River, and also in Sanikiluaq, NU, which
harvests a small population of reindeer on the Belcher Islands, NU. Some hunters in
Arctic Bay had carried kits while hunting, but caribou were scarce at the time and no
samples had been submitted (Jenkins, 2009). No hunters in Pond Inlet or Clyde River had
picked up or used CHMP sampling kits.
All of Nunavut’s 26 communities are remote from each other and only accessible by air
throughout the year. While the CHMP has not physically reached all locations, news
tends to disseminate rapidly throughout NU via various communication modes.
Community radio, the Internet, various news media, and updates from representatives at
HTO meetings are major ways that people throughout the territory keep abreast of
wildlife events, programs, and issues that they deem important.
Program in the Sahtu Settlement Region:
The Sahtu Monitoring Project for wildlife health assessment, now known as the Sahtu
Wildlife Health Monitoring Program (WHMP) was launched in the Sahtu Settlement
176
Region of NT in 2003. In addition to the aim of monitoring general health of wildlife
populations, the program focused on disease concerns in NT caribou and also in moose,
which are abundant and widely hunted for subsistence in the region. Two of these
concerns, brucellosis and toxoplasmosis (caused by Toxoplasma gondii), are potentially
serious zoonotic illnesses. Antibodies against the agents that cause these diseases have
been detected in caribou from herds that are hunted by Sahtu residents (Tessaro and
Forbes, 1986; Kutz et al., 2001).
Between 2003 and 2005, six subsistence hunters of caribou and/or moose from three of
the five total communities in the Sahtu Region expressed interest in being part of the
WHMP and were trained as “Wildlife Health Monitors” (WHMs) (Neimanis, 2005; Fig.
3). Training to collect wildlife data and samples was done in classroom and hands-on
sessions. Initially, program organizers gave an overview of the project and showed
trainees images of visibly recognizable caribou diseases. In some cases, trainees were
then shown how to collect data and samples from a freshly killed caribou (Neimanis,
2005; Fig. 5). Hunter sampling kits were provided to WHMs via the Norman Wells office
of the Department of Environment and Natural Resources, Government of Northwest
Territories (ENR GNWT). The WHMs submitted caribou and moose samples to the
department for processing, storage, and testing, and initially received the equivalent of
CAD 100 in fuel per animal. In later years, this was paid in cash or in fuel.
The WHMP remains in operation, with funding from ENR GNWT and the Sahtu
Renewable Resources Board, the main body for wildlife and fisheries co-management in
the region. Given the program’s duration, the relatively small number of hunters in each
177
community, the annual presentations of program results to the Sahtu Renewable
Resources Board, and communication among Sahtu communities, numerous Sahtu
harvesters are familiar with the program, although they are not necessarily trained
participants. Samples have been submitted since the program’s inception and all but those
archived were processed annually. The archived specimens were i) blood samples
collected on filter paper strips (see below), which were held until this method was
validated for a series of serological tests (see Chapters 2, 3, and 4); ii) tissue samples that
were archived for potential contaminant testing in future. The yearly presentations to the
Board reported findings for parasites and the body condition, sex, and age of the animals
that WHMs sampled. As of 2012, two of the trained WHMs in the Sahtu Region were
regularly submitting caribou and moose samples as part of the original format of the
WHMP (the format to which the interviews in this study relate).
Moose are abundant in the Sahtu Region and very highly valued by hunters and
communities. Recently (since the interviews for this study were conducted), the WHMP
has focused intensively on moose health monitoring and has increased hunter awareness
about an emerging disease issue (infestation by winter tick [Dermacentor albipictus]) that
affects moose and is therefore of concern for Sahtu communities. The program has also
implemented considerably higher compensation for each complete moose sample set and,
as opposed to the former system, all sample collection/submission logistics are now being
managed by the Sahtu Renewable Resources Board and local Renewable Resource
Councils. Sample submissions for the WHMP have increased considerably, with 14
178
hunters currently submitting moose samples and others submitting caribou samples (S.
Kutz, personal communication).
Filter-Paper Blood Collection: A Tool for Hunter-Based Sampling
Collection of blood on filter paper (FP) has been used in human diagnostics for decades,
initially for neonatal screening and today for a wide range of applications and research
(Guthrie and Susi, 1963; Mei et al., 2001, McDade et al., 2007). The simplicity of this
method lends it well to sampling of caribou and other wildlife by subsistence hunters.
Use of the FP method in wildlife that are killed for food does not require live-animal
handling or manipulation, which many Aboriginal cultures consider disrespectful and/or
cruel (Byers, 1999). This technique offers several advantages for field collections in the
North and elsewhere, as the papers are lightweight, non-breakable, compact to carry on
multi-day hunts, and users require no special blood-collection skills. The sampling steps
involve simply dipping and saturating the FP media (Nobuto filter-paper strips, Toyo
Roshi Kaisha, Ltd., Tokyo, Japan) in a pool of clean whole blood from a large vein, such
as the jugular vein. Important considerations for conventional tube blood sampling in the
field are the requirements for a cold chain (cool conditions but avoiding freezing) and
processing equipment to spin the samples and collect serum. These are not issues with FP
samples. To preserve blood analytes in the FP matrix, samples can be dried as per the
Nobuto manufacturer’s recommendation and stored dry. Alternatively, if climatic
conditions demand, research suggests that FP samples may be frozen at collection and
kept at –20 ºC until analysis (see Chapter 4).
179
Hunters were trained in FP blood sampling as part of both the CHMP and the WHMP. In
the former, hunters at the program implementation meetings tried the method by dipping
Nobuto FP strips into simulated caribou blood (Fig. 4b). In the WHMP training sessions,
the FP technique was demonstrated using a caribou carcass whenever possible (Neimanis,
2005; Fig. 5). Experiments have shown that FP samples from caribou and reindeer (both
Rangifer tarandus ssp.) are effective in antibody assays for eight pathogens of known or
potential relevance to this species (see Chapters 2 and 3), including Brucella spp. and
others. However, proof of efficacy in wildlife serology does not guarantee uptake and
validity of the method in a hunter-based wildlife-monitoring program. To be useful in this
type of system, the FP method must be implemented successfully at the community level
and accepted by harvesters.
Research Aim
This study is the first formal qualitative evaluation of certain aspects of the CHMP and
WHMP. The aim was to assess the attitudes and beliefs of subsistence harvesters in
Canada’s North regarding wildlife disease and hunter-based wildlife health monitoring in
general, and hunter use of the FP technique for caribou (and/or moose) sampling
specifically. The ultimate goal was to identify challenges or concerns that might inform
these two monitoring programs and potentially bolster future hunter-based wildlife
disease-surveillance programs in northern Canada and elsewhere.
180
Methodology
Descriptive qualitative methodology was used, as the purpose was to fundamentally
describe and analyze phenomena (people’s opinions, beliefs, and attitudes) and was not
theory-based. Twenty-five face-to-face semi-structured interviews were conducted with
key informants, experienced subsistence harvesters of caribou (and also moose), in
remote communities of Baffin Region, NU and the Sahtu Region, NT, Canada between
July 2009 and September 2010. Ethics approval for research with humans was obtained
from the University of Calgary Conjoint Research Ethics Board. Interviews in NU were
conducted in cooperation with the CHMP and interviews in NT were conducted as part of
the Sahtu WHMP. Neither territory required special scientific research licensing for the
interviews because strictly attitudes, personal beliefs, and opinions—not traditional
knowledge—were collected. The seven communities where interviews took place are
located in Canada’s Eastern, Western, and High Arctic, with populations ranging from
approximately 130 to 6,700 residents (Table 1; Figs. 2 and 3).
Sampling: Participants and Recruitment
Local subsistence harvesters were the appropriate sample group to provide the knowledge
sought because these people hunt, prepare, and share wildlife as food in their
communities. Further, local hunters would be potential participants in hunter-based
wildlife sampling programs and would be the most informed individuals to comment on
the practicality and feasibility of FP collection during hunts.
181
“Harvester” was defined as a man or woman who had extensive life experience (even at
relatively young age) hunting caribou and/or moose for food, or butchering and preparing
meat for consumption, or using these hunted wildlife in other ways (sewing clothing,
fabricating instruments, etc.). It was recognized that the involved communities were
generally small (most populations 500 or less) and that the respective pools of harvesters
were even smaller. Efforts were made to recruit individuals who were known as
experienced or “specialist” caribou or caribou/moose harvesters (moose are more widely
hunted than caribou in some Sahtu communities). All participants volunteered and none
was directly invited to participate by the researcher. In Pond Inlet, NU, snowball
sampling was employed. A recruitment letter in English and Inuktitut was sent to the
chairperson of the Mittimatalik (Pond Inlet) HTO. As well, a third-party employee in the
GNU DoE at Pond Inlet, who was a local caribou hunter, invited known caribou hunters
from the community. In Iqaluit and Grise Fiord, harvesters volunteered through thirdparty contacts in the Nunavut Wildlife Management Board (after a recruitment letter was
sent to the Board’s director), the GNU DoE, and the Grise Fiord HTO. In the Sahtu
communities, third parties linked with the WHMP invited known caribou and moose
harvesters to volunteer. In both territories, the intent was to recruit some harvesters who
had been trained in and/or collected FP samples as part of the CHMP or WHMP;
however, one well-known, active WHM in the Sahtu was unable to be interviewed.
Researcher Information, Data Collection
The researcher (PC) is a veterinarian and doctoral student whose project focus is FP
blood collection as a method for wildlife health monitoring. This study was her first
182
experience with semi-structured interviews, and her technique improved during the
course of the study. She had no direct relationship with any of the harvester volunteers.
The researcher conducted all interviews and made efforts to put participants at ease. Each
session began with the interviewer providing some personal background and explaining
the research as outlined in a written consent document. Confidentiality was emphasized,
as was the fact that there were no “right answers” to any question and that only the
person’s opinions/beliefs were sought. The interviewee gave signed consent and received
a copy of the form. A native-language interpreter was provided according to the
participant’s wish, and a background information sheet (Appendix 6A) was completed.
Interviews were carried out following a general script. Owing to some differences
between NU and the Sahtu Region with respect to implementation of hunter-based FP
blood sampling and species hunted, the scripts for the two groups were slightly different
(Appendices 6B, 6C). To improve the study iteratively, one question posed during the
NU interviews (Section II, Qn. 8 in NU interview script [Qn 2 in Appendix 6E]) was
omitted from the Sahtu (SA) interviews because it was layered and called for speculation,
and was thus considered to have generated questionable data.
Each interview comprised approximately 20 quantifiable primary questions (i.e., Yes/No
answers or scale items, which are considered best for collecting data on attitudes or
beliefs [Robson, 2002]). The scale items categorized responses into one of five
unweighted categories. There were also secondary open-ended probing questions related
to the primary ones. Aboriginal communities have identified open-ended questions as
important with regard to respecting people’s desire to express answers from their own
183
perspective (Maar et al., 2011). The researcher formulated the interview questions (and
the specific plain language used) in close consultation with two university researchers
who were experienced with focus groups and conducting wildlife-related interviews with
Aboriginal Northerners and others. The interviews were divided into two sections. The
questions in Section I were identical in the NU and SA interviews and collected
harvesters’ views on wildlife and food in general. Section II included a demonstration of
the FP method with a mock hunter kit and FP blood sampling supplies. The items in this
section focused on participants’ opinions about hunter-based wildlife sampling and the
FP technique, with some questions directed only at “hunters” (those who actually stalked
and shot animals in addition to butchering or other activities) and others only at “nonhunters” (those who strictly engaged in butchering, meat preparation, and activities such
as sewing etc.). Due to the generally small populations of the communities, the very small
cluster of hunters in each community, and logistical barriers associated with testing these
interviews over the phone, it was not possible to test-run the interview scripts on a sample
of the study population prior to interviewing on site.
The majority of interviews lasted 50-60 minutes (range, 40 to 100 minutes), which is
considered the maximum effective time (Robson, 2002; Maar et al., 2011). Some were
extended due to participants’ lengthy answers and comments around probing questions.
Upon completing the session, each harvester received an honorarium (CAD 75 cash or
equivalent in gasoline vouchers) and had a photo taken if they had consented to this. The
researcher took field notes throughout each interview and audio-recorded each session as
per signed consent. A transcriptionist bound by a signed confidentiality agreement
184
transcribed the data recordings using word-processor computer software. Respondents’
identities, voice files, transcribed data, and hard copies of questionnaires and notes were
stored securely as per ethics board requirements and were accessed only by those named
in the ethics approval.
Data saturation was challenging to assess. In most locations, the number of experienced
caribou or caribou/moose harvesters per community was very small and this strongly
influenced the sample size. In addition, the logistics and cost of on-site interviewing in
the North are considerable and it is typically not possible to schedule or plan ahead to any
great extent. Flexibility of timing and a researcher’s availability for interviews are key. In
numerous communities, the researcher only had 1 or 2 days in which to recruit and
conduct all interviews. Despite these constraints, several elements helped ensure that the
interviews formed a reasonable basis for addressing the research questions: i) the targeted
focus of the study; ii) the total sample size of 25; iii) representation of harvester views
from the Eastern, Western, and High Arctic; iv) the respondents’ range of experience with
sampling programs.
Analysis
Quantitative and systematic qualitative analyses were carried out. After quantitative
results for all questions were calculated, root data (all those relevant to the research aim)
were selected for qualitative analysis. The voice files for these portions of each interview
were reviewed and compared to the matching transcripts to clean the transcribed data and
ensure understanding of answer tone and details. Latent content analysis was then applied
185
(Mayan, 2001), with coding and categorization of statements, including assignment of
statements to multiple categories as appropriate. Once coded statements were assigned to
each category, links between categories were considered and related categories were
grouped to form several distinct main categories with subcategories. Results for these
were reported, using scale-item rankings to group comments as appropriate. Themes that
wove throughout and emerged from the results were identified for discussion.
Results
Table 2 lists each interviewee’s key background information and characteristics, and
Table 3 summarizes the participants’ demographic characteristics. Sixteen of the 25
harvesters were Inuit from NU and 9 were Dené First Nation people from NT. The
majority of respondents had been harvesting caribou and/or moose for more than 30
years. Eight participants self-identified as Elders, and all of these were older than 50
years.
Nineteen of the 25 interviewees were naïve to the FP method prior to the interview. Of
the six who were familiar with the technique, two had been recently trained in a CHMP
meeting but had not used the method on a hunt, and four had been trained as monitors
(WHMs) in the WHMP. Of the four trained monitors, only one was still collecting and
submitting all WHMP samples, including FP samples. Another was no longer submitting
samples due to age/compromised eyesight limiting his hunting, and two were still
actively hunting but were not submitting samples.
186
Table 4 lists the interview questions that generated the selected root data and the
quantitative findings for these in the NU interviews, the SA interviews, and both sets
combined. Quantitative results for the entire interviews are presented in Appendices 6D,
6E, and 6F. An example schematic of the qualitative analysis steps is in Figure 6.
1. Importance of Wildlife Disease
Twenty-two (88%) of the 25 total harvesters ranked wildlife disease as important or very
important, and the proportions of respective rankings in the NU and SA interviews were
similar (Table 4). Three harvesters (two from NU, one from SA) had no opinion or
viewed wildlife disease as only somewhat important, and all three indicated that this was
because they observed minimal disease in wildlife. One of the three also commented that
predators “remove all the sick ones” from caribou/moose populations.
The 88% who deemed wildlife disease important voiced three main categories of reasons
for this.
•
Food: Safety, Taste, and Human Health
By far the most common concern linked with wildlife disease was food safety. Many of
the 22 respondents who considered wildlife disease important/very important indicated
that they did not actually observe a lot of sick animals and never had throughout their
lives as harvesters, and yet 13 (59%) still referred to food safety and health as their prime
concerns. One NU harvester said, “I’ve had one brother in the past who contacted
brucellosis and it changed him forever, like he was never the same. I don’t see much
187
[wildlife] disease … I haven’t seen anything [lesions/abnormalities in wildlife] in awhile.
Maybe because there isn’t much caribou [in this area].” Another from NU said, “ [It is]
very important dealing with diseased animals. We have heard about people getting sick
from walrus. It hasn’t happened here, but we are concerned about that.” A harvester from
SA underlined Dené people’s reliance on wildlife: “[Wildlife disease is] important for the
Dené because we depend on wildlife. We depend on the moose, we depend on the
caribou …”
The tissue abnormalities most often noted by the hunters were “pus” (abscesses) and
“white rice” (tapeworm cysts) in caribou and moose meat but, in general, interviewees
indicated that they rarely encountered overt disease. Two harvesters (one from each
region) noted that their disease concerns stemmed mainly from abnormal meat taste. An
older hunter described issues with moose meat: “Mostly it’s the taste, sometimes just like
eating a piece of rag or something; that’s how it tastes. [The animals] have different kinds
of disease now … I mean [the meat is] changed, different, it tastes funny. Sometimes
moose tastes good and sometimes it doesn’t. The dark meat doesn’t look as dark, looks
grey color, and it doesn’t taste that good. You know, it depends what [the animals are]
eating … Maybe it’s the climate change too, the food they eat. Too much chemicals
going onto the land I think.”
Two participants also commented on the need to cook food “harder” today, suggesting
either greater awareness of disease or higher prevalence compared to the past, or both.
Relating to this category and also the Science, Training, Disease category below, one
respondent noted, “People should be aware of … what kind of [wildlife] sicknesses are
188
out there. This way they can … make sure they handle their meat properly so they won’t
get sick.”
•
Science, Training, Disease: Ability to Detect and Monitor
Several of the 22 interviewees who viewed wildlife disease as important/very important
referred to some degree of increased wildlife disease over time and to scientific
approaches and skills in disease detection. Four (18%) had noted some change in the
frequency of wildlife disease during their lifetime and 4 (18%) mentioned the importance
of being able to recognize abnormalities and illness in caribou/moose they hunt. One NU
harvester said “It’s very important because I never got to [know about] wildlife disease
because we’ve never been trained to see [hunted] animals with disease … We’ve never
been taught that because they were always good. [Now] you don’t know which one might
be carrying disease; you have to gamble.” Two (9%) of the 22 also spoke of the
importance of monitoring caribou population health. One said of wildlife disease, “… it
would be very important because if you have a diseased caribou then [the population is]
in trouble. I think they should be monitored quite well. I think it should be a full-time job
to monitor your caribou … your stocks should be monitored on a daily or monthly basis.”
•
Humaneness and Disease Concerns for Wildlife at Large
Four (18%) of the harvesters indicated that wildlife disease was important/very important
to them in relation to the integral value of wildlife and the need to end a sick animal’s
suffering. Three harvesters (14%) also mentioned worries about disease in caribou or
moose being transmitted to other species that are human food sources or that prey on
189
ungulates. Two (9%; one person from each region) referred to disease concerns
connected with anthropogenic effects of climate change and industrial contaminants, and
one (5%) highlighted injury by humans as his prime concern: “That’s what we watch out
for, whether it [wildlife disease] is naturally caused or has this animal been wounded by
[a] human being … We are very conscious of those types of situations so this is an
important topic.”
2. Importance of and Interest in Hunter-Based Wildlife Sampling
Of the total 22 hunters who participated, 15 (68%) said that they viewed hunter-based
sampling as important or very important, and the proportion was slightly higher in NU
than in SA (Table 4). One of these individuals (7%) expressed interest in getting involved
with sampling for the payment alone if it was “good money.” The one youth participant,
who had done some hunter-based sampling but selected “no opinion” on this in the
interview, said that at least the sampling had paid something and that it gave him
“something to do.”
The main categories of comments made by the 15 hunters who were positive about
hunter-based sampling are listed below, and the concerns and problems expressed by the
remaining eight hunters are summarized in the Other Issues and Pitfalls section.
•
Food Safety, Human Health, and Science/Training
Concerns about food safety and health were the most frequent reasons cited for viewing
hunter-based sampling as important, and for interest in doing this type of sampling (8/15;
190
53%). A few hunter responses also spoke to having observed or heard about changes in
wildlife disease and to the science/training points noted above (population health
monitoring, training in disease detection). The same number also expressed interest in
this kind of sampling if it could assess for contaminants and effects of climate change.
One individual expressed his belief in scientific testing, stating, “I want to make sure the
health of the animals that I eat are healthy enough to eat in general. [I am concerned
about] the overall health of all animals like caribou. I want to be sure they are safe, tested
scientifically.” Five (5/15; 33%) of the interviewees’ comments about hunter-based
wildlife sampling alluded to greater interest in sampling because of broader awareness
(not necessarily higher prevalence) of disease nowadays through media and scientific
study.
•
Hunter-Based Sampling, Researchers, and Communities
Two (13%) of the 15 respondents who ranked hunter-based wildlife sampling as
important or very important offered critique and tips aimed at wildlife researchers and
hunter-based sampling programs in the North. One referred to the pervasive problem of
research fatigue throughout NU and the need for authorities (government departments in
particular) to genuinely participate alongside Inuit in wildlife research. He expanded:
“You [the biologist/researcher] want to be out in the field where you’re going to get good
samples rather than relying on [hunters]. You know, the hunter might be getting sick of
providing samples to a department that doesn’t get out [alongside him/her] hunting. I
think it would mean more to Inuit if, let’s say, a department would go out there, not just
be an authoritative figure. You know, [show that] they care for what the hunter is going
191
to be eating.” The same participant voiced the opinion that hunter-sampling programs and
other research “with Inuit” are colonizing: “You’re promoting, telling an Inuit to get into
research. You’re creating a thought process to that particular individual.” He also
underlined the need for more thought around appropriate, useful, meaningful reporting of
results, and, as an example, explained that mailing a follow-up card to a hunter with an
image of the animal he shot and reporting age, sex, and measurements might be regarded
as offensive (“the hunter already knows it was a big white male” [beluga whale,
Delphinapterus leucas]) or even as a “tactic” on the part of the scientist. The hunter
emphasized that the people value practical information about health/food safety.
This same individual also stressed that researchers should collaborate closely with a local
community member (a champion of sorts) who is connected to the hunters/people and can
communicate, advocate, and advise. Another respondent indicated that the best way to
maximize hunter interest and participation in hunter-based sampling programs was for
researchers to go on community hunts with the people.
•
Other Issues and Pitfalls
Eight (36%) of the 22 hunters mentioned lack of results/feedback to the community as a
problem with hunter-based sampling programs. Seven of these eight were among the 15
hunters who ranked such programs as important or very important, and six of them
expressed anger, frustration, and other strong negative feelings about the issue. One
hunter said, “It seems like a waste of time because we don’t get any feedback. We won’t
like you any more at all if we don’t get any feedback from you. Sometimes we stop doing
192
it completely.” Another pointed out that HTOs are now writing requirements for results
reporting into agreements with biologists and others who engage hunters in sampling and
said, “And when they’re not producing any results we can go after them now.”
A related issue raised by 2 (9%) of the hunters was that the time lag of analysis negates
the value of testing from a food-safety perspective. Several others pointed to poverty in
the North and the high cost of living (including the cost of fuel for hunts, etc.), and some
interviewees indicated that the payment for WHMP sampling was not sufficient or worth
their trouble (often in the cold) to complete the process. One respondent noted that
hunters are collecting samples for biologists, yet biologists’ pay far exceeds the incomes
of subsistence harvesters. Two of the hunters (9%) also mentioned distrust of a “science
agenda” and specifically of authorities who might “use information [from the sampling]
against us” relative to caribou harvesting regulations.
From a practical viewpoint, 2 hunters (9%) noted that caribou are relatively scarce in both
regions currently and that people were reluctant to submit samples that are traditional
delicacies. Four hunters (18%) perceived no need/value for hunter-based sampling
programs. One of them indicated that the concept of health monitoring was flawed for
caribou and moose (i.e., predators take the diseased animals, therefore it is the health of
predators that should be monitored) and others noted that hunters are expert frequent
observers who do their own version of monitoring, and that sampling results are not
reported back anyways.
193
3. Concerns about Filter-Paper Blood Collection
Only 4 of the 25 total harvesters interviewed (16%; 2 NU and 2 SA) had any concerns
with FP blood collection from caribou and/or moose. One emphasized that he condoned
sampling strictly from subsistence kills only, and this hunter also underlined a demand
for results. Two mentioned issues of safety/survival and inconvenience (cold hands),
explaining that weather might preclude sampling if there were blizzard conditions
threatening or if it were too cold.
Two hunters mentioned that traditional Dené beliefs might prevent FP sampling and other
types of wildlife sampling. One indicated that he had no problems with the method but
that other hunters would refuse based on cultural beliefs that sampling should not be
done. The other hunter spoke of the value of caribou and people’s dependence on them:
“It’s not like you can go to the store and get caribou; you can’t do that. If you don’t watch
it [take care] you could lose it.” He discussed respect for animals, traditions around
blood/tissue handling when hunting and preparing meat. The same individual talked
about the potential impacts that doing wrong or “being smart” to caribou could have on
his hunting success and general wellbeing in future: “Well, us [Dené] we really watch
blood, especially around women. Women can’t [step] over blood like that … I could lose
my luck if I don’t watch how I take care of an animal. Later on I’m going to have harder
times [less luck hunting etc.], things like that. I’m going to give you [a woman] blood and
then never know what you’re going to do with it. [And if it were a male researcher] I
mean, still, I give him blood or [samples]; I don’t know how he’s going to take care of it
194
... That’s what I mean. [The FP method itself] is not too bad to do but you’ve got to make
sure you’re watching eh, being careful.”
Discussion
Emergent Themes
This unique study involved subsistence harvesters from different Aboriginal groups and
regions of northern Canada, and explored their views on wildlife disease, hunter
sampling, and an innovative blood-sampling tool in the context of wildlife healthmonitoring programs.
Four main themes emerged:
•
STRONG INTEREST IN WILDLIFE DISEASE: The majority of harvesters viewed
wildlife disease as important even though they encountered sick animals rarely and
most had observed relatively little change in wildlife disease during their lifetime.
•
GENERAL INTEREST IN HUNTER-BASED SAMPLING: Most hunters deemed hunterbased wildlife sampling important and were interested in participating in such
programs to some degree. This despite the fact that those with prior sampling
experience said they had never received results.
195
•
KEY ISSUES WITH HUNTER-BASED SAMPLING: Hunters who did not value hunterbased wildlife sampling highly shared several common reasons for this.
•
BROAD ACCEPTANCE OF FP SAMPLING, BUT WITH SOME EXCEPTIONS: Most
harvesters had minimal or no concerns about FP blood sampling, though one
condoned it only because it was done on subsistence hunts and two cited Dené
traditions and beliefs that could prevent collection of blood and other samples from
caribou.
Basic Interest and Motivation
Several interviewees alluded to greater awareness of wildlife disease and issues today
because of modern media communications and the Internet. Other authors have noted that
the Internet and other global media have become significant sources of health information
in many First Nation and Inuit communities (Richmond and Ross, 2009). This
“connectedness” coupled with general concern about food safety, may underpin some of
the interest in wildlife disease that was evident from the interviews. Whatever its origins,
this interest is a basic requirement for hunter-based wildlife sampling programs and the
study results indicate that both NU and SA harvesters possess it. General interest in
hunter-based wildlife sampling is also essential to such initiatives and it was noteworthy
that some hunters who were visibly frustrated or even angry about the perceived lack of
results reporting still indicated willingness to participate. Though these levels of interest
in no way guarantee a successful hunter-based monitoring program, they do suggest that
196
both regions have some of the fundamentals for potentially effective community-based
wildlife monitoring programs (Kofinas et al., 2003).
Results Reporting, Communication in General
Lack of feedback or results from hunter-based wildlife health monitoring was the
negative issue that harvesters voiced most frequently and most strongly regarding their
experiences. This complaint has also been reported by others working in northern, mainly
Aboriginal communities (Pufall et al., 2011). According to local scientists in NU and
scientists involved with the WHMP, results of wildlife research and monitoring have
been formally disseminated frequently in Pond Inlet and have been presented directly to
harvesters and/or harvester organizations in the Sahtu Region annually (D. Jenkins, S.
Kutz, and A. Veitch, personal communication). Clearly, there is a disconnect that needs
to be examined in this regard because the interviewees were unaware of any such
reporting, yet news that is of interest generally circulates quickly among locals and
communities in the North. It appears that wildlife health results are not being recognized
or understood as such in communities, or that findings may not be delivered in accessible
or meaningful ways at the community level. Three main points are salient here, all related
to communication and effective knowledge translation.
First, authors examining local and specifically Aboriginal community engagement in
research have noted the need to work closely with communities to understand locals’
perceptions and attitudes, and to define and reiterate the purpose of a research project,
thus ensuring that outcomes and expectations are aligned (Lavery et al., 2010; Maar et al.,
197
2011). Misperceptions of research or program goals and scope can be problematic (Pufall
et al., 2011) and may be relevant to the current study. The majority of NU and SA
interviewees indicated that food safety and human health were their main concerns with
wildlife disease, whereas wildlife health monitoring programs have considerably broader
goals, such as assessing animal population demographics and health. It is possible that
community members anticipate feedback in a more targeted form (for example, “Report
on Food Safety”), whereas scientists tend to deliver more diverse results that may not be
what communities or participants expect or value.
Second, it is possible that cultural and worldview disparities within community-based
wildlife monitoring programs (for example, worldview differences between southern
scientists and indigenous hunters in the Arctic) may muddy the communication of results,
purpose, and other key program elements. Authors have described grappling with this in
various wildlife and environmental resource management initiatives (Berkes et al., 2007;
Raymond et al., 2010; Armitage et al., 2011). In addition to the results-reporting issues,
some interviewees expressed confidence in and preference for their own culturally/locally
based wildlife knowledge and expertise as opposed to that which might come from
hunter-based monitoring. This may reflect worldview differences and misunderstandings
of purpose and program elements as well. Reports on co-management and
monitoring/environment-related initiatives in the North and elsewhere have noted a need
to understand the generally more holistic approach that indigenous people take; that is, to
understand the indigenous focus on the process (ways of knowing; for example, Inuit
meanings that come from multiple simultaneous observations on the land) as opposed to
198
the thing (fragments of knowledge provided by individual indicators) (Catley and
Leyland, 2001; Berkes et al., 2007; Raymond et al., 2010; Maar et al., 2011; Wesche et
al., 2011). Moreover, some have expressed that the interconnectedness of ecosystem and
social wellness points to a need to approach these issues from a holistic, systems-type
perspective, which is in line with Aboriginal ways of knowing (Wesche et al., 2011).
Gaining such understanding requires collaborative, cyclic, reflexive learning as well as
long time horizons (Raymond et al., 2010; Armitage et al., 2011), and the CHMP and
WHMP are both relatively new programs. For any collaborative initiative in the North,
bringing all participants together face-to-face and/or conducting focus groups or
community meetings involves significant financial and logistic hurdles. However, others’
experience and findings (Maar et al., 2011; Wesche et al., 2011) suggest that time and
funding invested in jointly reframing and revisioning goals and communications might
pay dividends in terms of results accessibility, increased local trust and interest, and
greater sustainability, impact, and overall success of a hunter-based wildlife healthmonitoring program. One place to begin thinking around this process might be the
concept of “integrative science” described by Bartlett et al. (2012), who promote
collaborative efforts to educate by weaving indigenous ways of knowing with mainstream
knowledge. A final point related to cultural differences is consideration of the mode and
format of results delivery. Maar et al. (2011) stressed that tangible knowledge translation
of health research is key for Aboriginal communities, and that too-technical terms and
inappropriate literacy level are two common issues. In line with others’ experiences with
results reporting in Canada’s North, Pufall and colleagues (2011) found that members of
an Inuit community identified open houses with interactive displays and active
199
involvement of children as some approaches to results dissemination that resonated for
them.
Third, harvesters’ perceptions about results reporting might, in part, reflect lack of
understanding of the time lags inherent in sample testing and data analysis. One way to
help address this would be to involve participants in the testing/analysis steps of the
research as much as possible, which has been promoted by other authors (FernandezGimenez et al., 2008; see discussion below). As well, community-based wildlife health
monitoring programs might benefit from creative presentations to local participants on
“The Life of a Sample,” one that would follow an individual specimen in a detailed
chronology through all the steps, storage phases, technicians’ hands, and detailed
protocols that are required in order to obtain results.
Gauging uptake or barriers to use of the FP method was a primary goal of this research
and it was evident from comments (or lack thereof) that participants generally found the
technique to be straightforward and practical. One hunter pointedly stated that he would
only condone FP collection from hunter-killed animals. This is in accord with the
widespread belief among Aboriginal cultures that handling/manipulation of or attaching
devices to wildlife is wrong or cruel (Byers, 1999). Only a few hunters noted issues of
inconvenience (cold hands) and safety (foregoing collection in risky weather conditions).
Two interviewees indicated that Dené traditions and beliefs, specifically issues around
respect for animals and impacts of potentially “disrespectful” practices, could affect
caribou/moose sample collection. Other researchers have also documented the value
placed on respect for animals in Aboriginal communities of northern Canada (Kendrick
200
and Lyver, 2005). It is a pervasive belief among the majority of North American
Aboriginal cultures that animals can perceive when they are demeaned or ill treated
through human action or speech, and that there are consequences (e.g., general bad luck,
animal attacks, unsuccessful hunts) when humans fail to show due respect to wildlife that
are subsistence species (Byers, 1999).
The interviews may not have detected all potential issues with hunter sampling that might
be linked with Aboriginal traditions or worldview. In Inuit and First Nations
communities, there may be additional culture-related barriers to FP collection or to
certain aspects of hunter-based wildlife monitoring, or even to the “informing wildlife
management” aspect of these monitoring programs (Byers, 1999). It would be valuable to
understand any such impediments as fully as possible in order to be respectful and
sensitive to these issues, demonstrate respect for culture and openness to solutions, and
hopefully earn deeper trust with participants and communities. In the realm of human
medical practice, authors have examined ways to achieve “cultural competence,” defined
generally as exploration, empathy, and responsiveness to people’s needs, values, and
preferences, and development of skills that are particularly useful in cross-cultural
settings (Betancourt, 2004). There is consensus that cultural competence improves
communication (Betancourt, 2004), and it would be beneficial to ensure this proficiency
for all non-local participants in wildlife health monitoring programs in the North.
201
Engagement, Critiques, and Context
Regarding the views of experienced monitors in the WHMP, it should be mentioned that
not all trained WHMs were included in the study sample. Two trained WHMs were
unable to be interviewed, and one of these hunters is an active and interested sample
collector who was an original trainee in the WHMP. This individual has submitted
maximum caribou and moose samples consistently since the program’s inception, and has
made extra efforts to collect additional samples when requested for special studies. His
attitudes and engagement are not accounted for in the study data. Regarding the WHMs
who were interviewed for the study, of the three who were still actively hunting, only one
was still submitting samples to the program.
Reasons cited for lack of enthusiasm about hunter-based wildlife sampling were i) poor
pay relative to the cost of fuel/hunting/living and the labour/inconvenience of sample
collection, ii) trust issues with government/scientists/authorities, and iii) lack of belief in,
or misunderstanding of, the program’s methodology (i.e., the view that the program is
wrongly monitoring the health of prey instead of the predators who keep populations
“clean”). Other politely but firmly delivered critiques and tips for researchers and hunter
sampling programs pointed to the need for more collaborative efforts that are
meaningful—and not disrespectful—to program participants and to local people and
communities.
Some of these issues relate to the above discussion of communication/knowledge
translation, and all of them reflect the complexities of community-based monitoring
202
programs and co-management initiatives in Canada’s North (Berkes et al., 2007;
Raymond et al., 2010; Armitage et al., 2011). In general, the literature on community
engagement in research is relatively young, and differences among disciplines, study
settings, and research aims, along with other disparities, present real challenges for
making comparisons. Generalizing about community engagement in research is difficult,
and in some cases impossible, because communities’ needs, attitudes, and motivational
factors may differ greatly (Khadka and Nepal, 2010).
To date, only a few studies have considered basic frameworks and processes for
community-based research on human health, natural resources, or animal health (Catley
and Leyland, 2001; Danielsen et al., 2005, 2009; Berkes et al., 2007; Lavery et al., 2010).
Catley and Leyland (2001) analyzed seven different levels of community participation
within veterinary initiatives for livestock health in Africa, and concluded that programs
which foster self-mobilization (collective action) are most likely to have maximal impact
and sustainability. In contrast, the CHMP and WHMP models feature lower levels of
participation; they are closest to Catley and Leyland’s (2001) material-based
(reumunerative) or function-based (cooperative) categories of community engagement,
which the authors suggest characterize programs less likely to thrive. Though these levels
of participation are preferable to others based on manipulation, compliance/passivity, or
one-way consultation with no consideration of local benefits or concerns (Catley and
Leyland, 2001), community-based wildlife monitoring programs could be improved with
more interactive local participation (co-learning, with joint planning and data analysis,
greater community control over decisions, etc.) and independence. These goals are in line
203
with the documented desires of Canadian Aboriginal communities regarding engagement
in research (ITK and NRI, 2007; Maar et al., 2011), and some success with them has been
borne out in examples of wildlife co-management in the North (Berkes et al., 2007;
Armitage et al., 2011).
Comparisons to Other Community-Based Initiatives
Fernandez-Gimenez et al. (2008) analyzed multiple community-based forestry
monitoring initiatives in the United States and found that the most benefits and
sustainability were achieved in programs with i) large numbers of diverse participants, ii)
local engagement in many aspects of the assessment process, and iii) large budgets. The
authors especially highlighted the gains in ownership and understanding that were made
when community members were involved with analysis. Efforts to incorporate this into
community-based wildlife monitoring programs may be beneficial and some of the multifaceted program activities led by Brook et al. (2009) are a step in this direction. In line
with Berkes et al. (2007), Fernandez-Gimenez et al. (2008) also noted the trust-building
and valuable social learning that occurred when people and groups with divergent
opinions and beliefs worked collaboratively over the long term. Similar processes and
outcomes have been noted with some community-based wildlife monitoring and comanagement programs in Canada as well (Berkes et al., 2007; Armitage et al., 2011);
however, the complexities of each situation are different. For example, the situation with
caribou health and management is perhaps more of a hot-button issue than the situation
for some other Arctic species. In some communities and regions of Canada (including the
Sahtu and parts of NU), concerns about caribou declines and sustainability have led to
204
significant political and power tensions between Aboriginal communities and government
authorities regarding harvesting and quotas (Festa-Bianchet et al., 2011). In the current
study, a small number of interviewees voiced negative opinions with political and
historical overtones and mistrust regarding hunter-based caribou monitoring. Although
these tensions were not directly related to CHMP or WHMP processes or personnel,
documenting these views is important because they reveal that hunters or communities
may contextualize “new” initiatives as mere continuations of past experiences with
scientists or authorities. Depending on species and management issues involved, such
tensions might create issues for a community-based wildlife health-monitoring program.
The Western United States study settings of Fernandez-Gimenez et al. (2008) differed
greatly from the Arctic context, yet these authors still cited challenges of recruiting and
sustaining participants, and communicating program results effectively. This indicates
that such issues are not unique to community-based monitoring programs in Canada’s
North; however, demographics, socioeconomics, and other factors particular to the North
may significantly impact programs such as the CHMP and WHMP. These include small
populations of potential participants, poverty (i.e., the need to focus on higher wage
activities as opposed to hunting/sampling), research fatigue (exacerbated by small
community populations), cultural beliefs, and the limited number of community leaders
in the North, most of whom are in great demand (which means that few may have time to
devote as wildlife monitoring program champions or advocates). Low income is a serious
issue in the North and some individuals who sign on for health-monitor training may do
so strictly out of financial need. These trainees may or may not ultimately collect samples
205
during hunts, as has been observed in the WHMP (personal communication, S. Kutz). As
well, some harvesters who view wildlife disease as important will simply not be
interested in participating as samplers for a variety of personal reasons.
Research Fatigue and Capacity – Others’ Experience:
The reality of research fatigue is a major challenge for Aboriginal community
engagement in Canada (Bull, 2010; Maar et al., 2011). In circumstances where there is
sufficient community capacity and human resources, this might be mitigated or shifted
significantly on a project-by-project basis if researchers and program leaders listen and
attend to the voiced concerns of communities (ITK and NRI, 2007; Bull, 2010; Maar et
al., 2011). Wesche et al. (2011) recently described a productive community-based health
research initiative on food security and climate change in a remote Gwich’in community
of Yukon Territory, Canada. The authors cited a positive community history of working
with researchers, reliance on and capacity for leadership by community members,
alignment of community and researcher interests, workshops to develop accessible
research “products” (results), a multidisciplined research team, and substantial funding as
some of the factors that led to active local engagement and overall success. Maar et al.
(2011) undertook focus-group discussions in Aboriginal communities of Ontario and
received suggestions from residents on how to conduct health studies in Aboriginal
communities. The main themes that emerged were i) the need to invest considerable time
in building trust relationships, ii) the importance of culturally congruent research
approaches, and iii) the need to foster researchers’ respect for local knowledge, realities,
and needs. These findings underline the need for cultural competency, echo some of the
206
issues documented by Inuit in Canada’s North (ITK and NRI, 2007), and are in line with
findings of the current study and others on community monitoring of animal health
(Kofinas et al., 2003). Authors have also highlighted the value of and ethical
requirements for extensive collaboration on all aspects of a project, as well as multiple
research methods, flexible research design, and genuine engagement/training of
community members (Bull, 2010; Maar et al., 2011).
Some published reports on successful community-based research experiences in the
North provide wisdom, hope, and important process targets/ingredients to strive for.
Ethical research in Aboriginal communities demands “authentic research relationships”
and these require time, sensitivity, and knowledge to build (Bull, 2010). Yet, it is also a
reality that socioeconomic circumstances, capacity, and human resources differ
significantly from community to community across the North and this can impact
research initiatives. Compared to others in Canada, Aboriginal Canadians bear a
disproportionate share of the burden of physical disease and mental illness, and have
higher rates of numerous chronic diseases, alcoholism, substance abuse, suicide, and
shorter average life spans (MacMillan et al., 1996). These health inequities have been
documented extensively and linked broadly to “unfavourable economic and social
conditions” (MacMillan et al., 1996). Searching for more specific causes, authors have
tied these issues to historical and ongoing environmental dispossession (Richmond and
Ross, 2009), a concept which is directly linked to wildlife conservation, preservation of
traditional foods, and cultural practices related to hunting and animistic spiritual beliefs.
There is a need to avoid vast assumptions and generalizations about Aboriginal people
207
(i.e., to instead recognize the individuality of people and communities) and to consider
research questions and practices in both historical and contemporary contexts (Bull,
2010). However, it remains a fact that, in many northern communities, the capacity for
involvement and leadership in programs and research initiatives is often severely
constrained by small population size and the significant health and socioeconomic
concerns of residents and extended families.
Synopsis of Findings
In summary, hunters who participate in wildlife health monitoring programs can offer
unique expertise, observations, and feedback in addition to sample collection and efforts
are needed to engage community participants fully in research (ITK and NRI, 2007;
Brook et al., 2009; Brook and McLachlan, 2008). However, while the concept of western
science and traditional/indigenous knowledge coming together is potentially positive, and
while mutual benefits can be gained from initiatives such as collaborative hunter-based
wildlife monitoring (Furgal et al., 2006; ITK and NRI, 2007), there are some real
challenges. In order for hunter-based wildlife health programs in Canada’s North, such as
the CHMP and WHMP, to be sustainable and have impact, this study suggests the need to
focus on three main areas: i) Effective, accessible communication of program elements
and results, ii) mitigation or bypass of research fatigue with new approaches, and iii)
long-term, iterative collaboration to build trust and authentic research relationships. Each
of these is particularly challenging to address or transform in the North.
208
Conducting the Study: Flexibility, Biases, Lessons Learned
Most recruitment for this study was done in Pond Inlet, NU and some of the potential
volunteers there declined interviews. Two reasons given for this were occupation with
income-related pursuits (such as commercial guiding/tourism) and lack of interest in
participating. The latter fit with the researcher’s observations of significant research
fatigue during her 8 weeks of fieldwork in this community (July-August 2009) and one of
the interviewees raised this point as a pan-Nunavut problem with research currently. One
individual who was invited for an interview initially scheduled one but then indicated
some hesitation. This person ultimately declined to participate, citing having had “bad
experiences with White people/Southerners” as a younger person. A fourth reason why
harvesters did not participate was the seasonal timing of data collection. Spring and
summer is typically a time for spiritual renewal after the long dark season, and many
Inuit families move out on the land to outpost hunting camps or to attend summer retreats
or events (NPC, 2000; Boult, 2006). As well, Pond Inlet is a coastal hamlet and the
interview period there coincided with the first safe open water for hunting seal (ringed
seals, Pusa hispida, and others) and narwhal (Monodon monoceros), prime food sources
in the community (Fig. 1). Given changeable weather, some harvesters were reluctant to
commit to interviews in order to be free to capitalize on calm seas. To adapt to this, the
researcher built extensive flexibility into the process/timing of interviews so as to
accommodate people with short notice if they found time to participate. One experienced
hunter who had initially declined an interview heard more details about the topic from
friends, became interested, and decided to participate given the flexible timing.
209
While effects of potential researcher biases cannot be ruled out, these and power
inequities were considered to the greatest extent possible. As well, care was taken to
deliver questions with neutrality and as professionally and succinctly as possible, and to
listen more than speak during interviews. The event of the one individual declining to
participate because of prior “bad experiences with White people” spoke to the
transgenerational power inequity, colonization, and assimilation-related issues that affect
many Inuit and First Nations people in Canada, often linked with the Indian Residential
Schools System (Richmond and Ross, 2009; Igloliorte, 2010). This event and a personal
encounter of similar flavour with another community member in Pond Inlet heightened
the researcher’s awareness of her location and power position within the study, and
fundamentally informed her data-collection approach throughout. Though exploration of
power dynamics was beyond the scope of this study, White (2006) identified power
issues linked with the mechanisms of wildlife co-management boards in the North. These
could be relevant for wildlife monitoring programs based in Aboriginal communities as
well. There is a general need for all researchers working with Aboriginal communities to
be aware of the history of colonialism and its impact on research today (Bull, 2010; Maar
et al., 2011).
A few issues arose—some potentially related to power inequity—that might have
affected data collection to some degree. In both regions, interviewees tended to hesitate
when asked to rank answers, and some expressed unease about the validity/worth of their
answers. In NU, the reluctance to rank items was fairly general and might have been
related to Inuit cultural values of equality and non-interference in others’ lives (Boult,
210
2006). Although the questions did not require ranking/judging of other people, the
concept of ranking in general might have been somewhat foreign and uncomfortable for
Inuit. The researcher made efforts to ease tension, ensure each participant felt unrushed,
and reframe questions as needed. In all but a few questions total throughout the NU
interviews, the researcher was confident that the interviewee felt comfortable with the
rank ultimately assigned. In SA, reluctance to rank answers arose mainly when
interviewees were asked to state a hierarchy of species (for example, which animals they
viewed as most important for food). In line with a holistic worldview, most SA harvesters
indicated that all species are equally important and that it is contrary to Dené beliefs to
rank them or “place one in front of another.” The animal-ranking issue did not affect the
data collection for the research aim of this specific study.
A second issue that emerged (during the SA interviews only) was the concept of it being
“bad luck to talk about” caribou/moose; that this is a sign of disrespect and could affect
hunting success and general providence. As noted, these beliefs have been documented
previously and are common among Aboriginal cultures in Canada (Byers, 1999;
Kendrick and Lyver, 2005), though community members do follow cultural traditions to
different degrees. It was difficult to gauge whether or how much these and other cultural
beliefs affected data collection in this study. These issues need to be considered in future
assessments of any hunter-based wildlife programs with Aboriginal communities. In
general, for both regions, more in-depth cultural knowledge might have contributed to the
formulation of interview questions and optimized the researcher’s approach to some
questions.
211
Big-Picture Applications, Importance
This study provides insights into northern Canadian subsistence harvesters’ views on
wildlife disease, hunter-based wildlife sampling, and hunter acceptance of FP blood
collection from caribou and/or moose. It also contributes to the relatively limited existing
literature on community-based research, and on hunter-based wildlife health monitoring
programs specifically. Some of its elements might be generalized to other northern or
remote populations and hunter-based wildlife programs. Documenting and analyzing
caribou harvesters’ views on these topics in their own voices is important and, to the
author’s knowledge, has not been done previously. Several key challenges were
highlighted, and it is hoped that these might inform the two wildlife monitoring programs
examined here, as well as potential future endeavours by wildlife researchers and others
who seek to team with hunters and communities in northern Canada.
Acknowledgements
Qujannamik, Mahsi, Thank you! to all the participants who gave their time to speak with
me for this research, and to their communities that were, in turn, part of the study as well:
Pond Inlet, Iqaluit, Grise Fiord, Colville Lake, Fort Good Hope, Tulita, and Norman
Wells. Debbie Jenkins and Grigor Hope (Department of Environment, Government of
Nunavut), and Susan Kutz and Ryan Brook (Department of Ecosystem and Public Health,
Faculty of Veterinary Medicine, University of Calgary) contributed multi-faceted support
and expertise to this research. Recruiting efforts by George Koonoo (Department of
212
Environment, Government of Nunavut) and Alasdair Veitch (Department of Environment
and Natural Resources, Government of the Northwest Territories) were extraordinary and
instrumental; the study would not have been possible without them. Thank you also to the
primary Inuktitut interpreter and translator (both anonymous), to Jeffrey Qaunaq who
interpreted Inuktitut in Grise Fiord, and to transcriptionist Kiran Pandher. Carl Ribble and
Theresa Burns (Department of Ecosystem and Public Health, Faculty of Veterinary
Medicine, University of Calgary) generously guided the qualitative analysis. Many
organizations contributed to this research: Government of Nunavut; Nunavut Wildlife
Management Board; the Hunters and Trappers Organizations of Pond Inlet, Arctic Bay,
Clyde River, and Grise Fiord; the Sahtu Renewable Resources Board; Band Councils in
the Sahtu communities; Renewable Resource Councils of the Sahtu Settlement Region. I
also wish to recognize and express appreciation to the funders: International Polar Year
Funding from the NSERC Special Research Opportunity Program, Environment
Canada/Natural Resources Canada, the Nasivvik Centre for Inuit Health and Changing
Environments (Canadian Institutes of Health Research), Alberta Innovates Technology
Futures, the Northern Scientific Training Program (Indian and Northern Affairs,
Government of Canada), the Arctic Institute of North America, and the University of
Calgary Faculty of Veterinary Medicine.
213
TABLES
Chpt 6. Table 1. Interview locations
Geographic locations east to west and approximate population sizes during the study
period for the Nunavut (NU) and Northwest Territories (NT) communities where
interviews were conducted.
COMMUNITY Iqaluit Pond Inlet Grise Fiord Colville Lake Fort Good Hope Tulita Norman Wells TERRITORY NU NU NU NT NT NT NT LATITUDE/LONGITUDE 63°44' N, 68°30' W 72°35' N, 77°46' W 76°25' N, 82°54' W 67°02' N, 126°05' W 66°15' N, 128°37' W 64°54' N, 125°34' W 65°16' N, 126°50' W (sources: Nunavut Bureau of Statistics, Canada Census 2011) POPULATION 6,700 1,500 130 150 515 480 730 214
Chpt 6. Table 2. Background information – Interviewees
Background information for the harvester interviews.
No. DATE (d/m/y) 1 31/7/09 2 3 4/8/09 4/8/09
4 41 YEARS IN COMMUNITY 38 YEARS CARIBOU HARVESTING 32 YEARS MOOSE HARVESTING n/a M 59 59 40 n/a PI ‐ Inuit M 65 40 56 n/a 5/8/09 PI ‐ Inuit M 66 65 55 n/a 5 6/8/09 PI ‐ Inuit M 43 43 27 n/a 6 7/8/09 PI ‐ Inuit M 20 15 8 n/a 7 7/8/09 PI ‐ Inuit M 37 37 12 n/a 49 a
6 n/a a
8 7/8/09 COMMUNITY
and CULTURE
PI ‐ Inuit SEX AGE (yrs) M PI ‐ Inuit PI ‐ Inuit F 67 9 10/8/09 PI ‐ Inuit F 72 38 35 n/a 10 11 11/8/09 PI ‐ Inuit M 62 62 39 n/a 13/8/09 PI ‐ Inuit M 54 37 47 n/a 12 14/8/09 PI ‐ Inuit M 57 29 40 n/a 13 14 21/8/09 M 43 35 37 n/a 1/2/10 IQ ‐ Inuit TU ‐ qDené
M 42 42 30 30 15 2/2/10 TU ‐ Dené M 58 50 1 48 16 3/2/10 FGH‐ Dené F 56 50 21 31 17 6/2/10 NW ‐ Dené M 39 6 25 25 18 6/2/10 NW ‐ Dené M 70 16 58 58 19 8/2/10 CL ‐ Dené M 43 43 30 30 20 8/2/10 CL ‐ Dené M 44 44 32 32 a
10/2/10 FGH ‐ Dené F 50 50 45 45 22 11/2/10 FGH ‐ Dené M 69 69 53 54 23 24/9/10 GF ‐ Inuit M 52 44 30 n/a 24 24/9/10 GF ‐ Inuit M 54 44 42 n/a 25 24/9/10 GF ‐ Inuit M 64 40 45 n/a a
a
21 Butchering and meat preparation only Abbreviations: CL: Colville Lake, Northwest Territories (NT); FGH: Fort Good Hope, NT; GF: Grise Fiord, Nunavut (NU); IQ: Iqaluit, NU; NW: Norman Wells, NWT; PI: Pond Inlet, NU;
TU: Tulita, NT
215
Chpt 6. Table 3. Demographic characteristics – Interviewees
Demographic characteristics of the 25 harvester interviewees.
NUNAVUT n=16 NORTHWEST TERRITORIES n=9 COMBINED n=25 Sex n (%) Males Females 14 (88%) 2 (12%) 8 (89%) 1 (11%) 22 (88%) 3 (12%) Age Mean (range) Community n (%) 54 yrs (20‐72) PI: 12 (75%) GF: 3 (19%) IQ: 1 (6%) 52 yrs (39‐70) CL: 2 (22%) FGH: 3 (34%) TU: 2 (22%) NW: 2 (22%) 53 (20‐72) Totals: Residence in Community Mean (range) 16 (64%) 9 (36%) 42 yrs (15‐65) 41 yrs (6‐69) 42 yrs (6‐69) a
Elder status 8/25 (32%) a Self‐identified as elders; all were 50 years or older. Abbreviations: CL: Colville Lake; FGH: Fort Good Hope; GF: Grise Fiord; IQ: Iqaluit; NW: Norman Wells; PI: Pond Inlet; TU: Tulita 216
Chpt 6. Table 4. Interview questions relevant to the study
The primary and probing interview questions (bullets) that generated the qualitative data
that were analyzed. Quantitative results for Yes/No and scale responses are also shown
with overall totals, n Nunavut [N] / n Sahtu [S], and respective percentages.
217
Table 4.
218
FIGURES
Chpt 6. Figure 1. Example of wildlife harvesting cycles – Nunavut
Wildlife harvesting cycles for North Baffin Region, Nunavut (source: NPC, 2000).
219
Chpt 6. Figure 2. North Baffin Region, Nunavut
The North Baffin Region of Nunavut with the communities of Arctic Bay, Pond Inlet,
Clyde River, Resolute Bay, and Grise Fiord. Harvester interviews took place in Pond
Inlet, Grise Fiord, and also in Nunavut’s capital Iqaluit (not shown). (source: NPC, 2000)
220
Chpt 6. Figure 3. Sahtu Settlement Region, Northwest Territories
The Sahtu Settlement Region of the Northwest Territories with its five communities:
Colville Lake, Fort Good Hope, Norman Wells, Tulita, and Deliné. Harvester interviews
took place in all except Deliné. (source: Brook et al., 2009)
221
a) b) Chpt 6. Figure 4. Hunter meeting and training – CHMP
The Nunavut Caribou Health Monitoring Program implementation meeting with hunters
in the hamlet of Arctic Bay, NU, October 2008: a) Government of Nunavut Wildlife
Biologist Debbie Jenkins introduces the data and sample collection system to hunters; b)
As part of the training sessions, hunters dipped filter-paper strips into fake caribou blood
to simulate blood collection in the field (photos: S. Kutz).
222
a) b) Chpt 6. Figure 5. Hunter training and sample set – WHMP
a) Hunter training sessions for the Sahtu Wildlife Health Monitoring Program (WHMP)
used a fresh caribou carcass whenever possible; b) The sample set that Wildlife Health
Monitors collect for the WHMP: Clockwise from top left, the lower leg bone (metatarsal)
with hide, the jaw, a liver sample, kidney with associated fat, fecal pellets, and bloodsaturated filter paper strips (centre) (photos: S. Kutz).
223
Chpt 6. Figure 6. Schematic – Latent content analysis
Schematic outlining an example of the qualitative analysis steps for one of the topics:
Importance of wildlife disease. Numbers of harvester comments (n) for each of the
initially derived categories are shown.
224
Figure 6.
225
CHAPTER SEVEN
WHERE DO WE GO FROM HERE?
KNOWLEDGE GAINED AND FUTURE DIRECTIONS
226
Knitting It All Together
Filter-paper (FP) blood sampling is a method well suited to the harsh conditions of the
Arctic, and one that can facilitate long-term surveillance of caribou pathogen exposure.
The spiritual and physical wellbeing of many Canadian Northerners depend on healthy
caribou populations, and collection of FP blood samples by subsistence harvesters can
expand the scope of caribou disease surveillance. Use of this technique in wildlife that are
killed for food involves no manipulation of live animals, and this could help make FP
sampling acceptable for Aboriginal hunters of caribou. However, engaging northern
hunters and communities in caribou health-monitoring programs is not a given. Part of
the essence of these programs is to make communities central to the monitoring of a
species that many residents value deeply. As well, these initiatives allow scientists and
hunters to exchange expertise that can benefit research and communities. Still, there are
challenges to implementing and sustaining these endeavours.
Summarized below are the main findings and future research directions for the three
phases of this project.
PHASE I: Efficacy of Rangifer Filter-Paper Samples for Serology
Chapters 2 and 3 described experiments that compared results for Nobuto FP samples
from Rangifer to results for matched serum (as the relative standard) in serological tests.
227
Analyses within 2 months after collection revealed that FP samples have comparable
sensitivity and specificity to serum in nine antibody assays: competitive enzyme-linked
immunosorbent assays (cELISAs) for Brucella spp., Neospora caninum, and West Nile
virus (WNV); indirect ELISAs (iELISAs) for Brucella spp., bovine herpesvirus-I (BHV1), parainfluenza virus type 3 (PI-3), and bovine respiratory syncytial virus (BRSV); and
virus neutralization (VN) assays for bovine viral diarrhea virus types I and II (BVDV-I
and –II, respectively). In most cases, FP specificity was almost equal to that of serum,
whereas sensitivity was somewhat lower but still comparable (above 85% for all tests
when FP test thresholds were adjusted slightly in two of the assays). In the iELISAs for
PI-3 and BRSV, decreasing the FP threshold slightly resulted in similar sensitivity and
specificity to those observed for all other tests. In the VNs for BVDV-I and –II, toxicity
of FP samples to the cell layer reduced the overall utility of this platform for this sample
type (see further discussion of this below). Correlation analysis demonstrated minimal
variability of test results among multiple FP samples from an individual animal.
Chapter 4 examined how FP samples from Rangifer performed in the above-listed tests
after the samples were subjected to conditions that mimicked potential collection and
storage scenarios in the Arctic. Again, serum results were used as the relative standard for
test performance analysis. The effects of different periods of storage at room temperature
(6 months and 1 year or longer) and different processing/storage conditions (i.e., freezing
vs. dry at room temperature) were assessed. In general, when comparing to serum stored
for the same period, FP specificity remained above 90% at 1 year, whereas sensitivity
was somewhat lower (above 88% in most tests) and the lowest sensitivity after this period
228
of storage was in the BHV-1 test (approximately 70%). For most but not all assays, FP
sensitivity was higher at 6 months than at 1 year of dry/room temperature storage.
Though sample sizes were small, preliminary testing suggested that freezing FP samples
from Rangifer at time of collection and storing them frozen (without performing a drying
step) does not negatively affect FP performance at 1 year. This scenario is relevant for
winter sample collections in the Arctic.
In addition to these evaluations, several considerations are noteworthy regarding the use
of Nobuto FP blood samples. Principal among these is that it is necessary to test the
efficacy of FP samples in each assay in which they will be used. Also, though FP
collection demands no special skills, some processing of the strips is required prior to
analysis. The amount of serum that can be absorbed and recovered from each strip is
limited (i.e., the final FP-eluate fluid volume per strip is approximately 200-250 uL). As
well, Nobuto FP eluates are estimated to be a 1:10 dilution of serum (Toyo Roshi Kaisha,
Ltd., Tokyo, Japan), and this may limit test sensitivity in some platforms. Nobuto FP
eluates are dark-red, which may hamper the readability of some bioassays (e.g., modified
agglutination test results for Toxoplasma gondii, as noted in Chapter 5). The eluates also
contain fragments of blood-cell membranes as well as free hemoglobin and particulates
of the FP matrix. Such components might impair the efficacy of FP samples in
serological tests. Specifically, one or more of them might have contributed to the failure
of FP samples in our trial of the fluorescence polarization assay for detecting antibodies
to Brucella spp. (Chapter 2); however, detailed investigation of this was beyond the scope
of this research. We also observed issues with FP samples in the VN assays. Compared to
229
the ELISAs, FP samples were less effective in VN because of toxicity (i.e., unreadable
results) and this meant that numerous samples were excluded from analysis. This
effectively reduced sample size (statistical power), which is concerning because sample
size is a significant challenge in most wildlife studies. Toxicity occurred more often with
FP samples than with serum, and was observed at initial testing (Chapter 3) and after
longer periods of storage (Chapter 4). In addition to the toxicity issue with VN, this
platform was an example where sensitivity of FP samples could have been reduced (but
was not affected in this study; see Chapter 3) due to the initial sample dilutions that are
inherent to VN protocols. Overall, the findings from this research suggest that the VN
platform may be less preferable for FP samples than ELISAs.
Future Directions:
Wildlife serology tends to be challenging to interpret because few tests are speciesspecific and many have only been validated for domestic species (Gardner et al., 1996).
In this project, only one of the eight antibody tests for which FP samples were assessed
had been validated for Rangifer. Although this limits the serological utility of FP samples
from Rangifer to some extent, the same limitations apply for Rangifer serum. There is a
general need for validated serological tests for caribou and other wildlife, and for speciesspecific reagents or test components in some tests.
Recognizing the noted limitations, the demonstration of FP efficacy in nine serological
tests is still valuable for caribou health surveillance and monitoring. To the author’s
knowledge, this breadth of FP validation is beyond what has been done for any wildlife
230
species to date. These findings for FP samples from Rangifer are not only useful for
caribou and reindeer, but also provide clues to how FP samples from other cervids might
perform in similar tests. Such guideposts can be particularly valuable for wildlife
applications because sample sizes for wildlife collections tend to be limited and
researchers often need bases on which to prioritize testing options. Possible next steps
would be validation of FP samples in serological tests for other pathogens that are of
relevance for caribou and/or food safety, such as T. gondii.
The experimental results suggested that freezing FP samples at time of collection
(without a drying step) and storing them frozen does not hamper FP performance;
however, investigation of this with larger sample sizes is needed.
The project included four different antibody assay platforms. As noted, there were
apparent functional problems with the VN assays for BVDV as compared to the iELISAs
and cELISAs for the other pathogens. Further investigation of the toxicity issues that
occurred with FP samples (and serum) from Rangifer in these and potentially other VN
tests is warranted. In the meantime, given that these toxicity problems decreased n values
substantially and given that wildlife sample sizes are notoriously small and results need to
be maximized, this research points to ELISA as the preferable assay platform. The fourth
platform explored was fluorescence polarization assay for Brucella spp., and FP samples
failed in this field-friendly test. Exploring the specific reasons for this was beyond the
scope of this project; nonetheless, this finding could help guide future users and
researchers working with FP samples.
231
PHASE II: Serological Snapshot in Time: Eight Pathogens, Seven Migratory Herds
Chapter 5 described a large serosurvey that was conducted across seven migratory tundra
caribou herds simultaneously as part of an initiative by the CircumArctic Rangifer
Monitoring and Assessment Network (CARMA). Collections were carried out during
International Polar Year 2007-2009, and the populations studied were five North
American herds (Porcupine, Bluenose-West, and Bathurst in the west; Rivière-auxFeuilles and Rivière-George in Québec) and two Greenland herds (Akia-Maniitsoq and
Kangerlussuaq-Sisimiut). Serum or FP was tested from each animal and the pathogens (or
pathogen groups) assayed were Brucella spp., N. caninum, WNV, T. gondii, BHV-1, PI3, BRSV, and Pestivirus. Samples from approximately 550 animals were tested in each of
these antibody assays. Of the total 4,420 samples analyzed, 2,782 (63%) were serum and
1,638 (37%) were FP samples; thus, the study demonstrated the capacity of FP collection
for increasing sample size and its potential for bolstering statistical power.
The limitations of serological tests in wildlife noted above (Gardner et al., 1996), some
degree of bias towards “healthier-looking” animals in subsistence-hunter convenience
samples for one herd (approximately 10% of the totals tested), and extended storage of
some FP samples likely resulted in underestimation of seroprevalence. As well, sample
sizes limited the accuracy and precision of estimates given the large population sizes of
the herds.
The main findings of the herd serosurvey centered on the overall estimates. In general,
the highest seroprevalence was observed for BHV-1, PI-3, and Pestivirus. There was
232
minimal reactivity to Brucella, N. caninum, and T. gondii, and the complete absence of
seropositivity for WNV and BRSV establishes baselines for these pathogens. However,
this finding for BRSV is difficult to interpret due to possible issues with the sensitivity of
the iELISA and other unknowns regarding BRSV infection in Rangifer. The overall
prevalence estimates for Brucella, N. caninum, and T. gondii were lower than reported
previously (Zarnke et al., 1983, 2000, 2006; Tessaro and Forbes, 1986; Kutz et al., 2001)
or lower than detected in other relevant contemporary studies (Stieve et al., 2010); thus,
these results were surprising. These differences may reflect die-offs of seropositive
animals, decreased population size resulting in reduced density-dependent transmission
(less frequent contact), and/or various broader ecological changes linked with caribou
population density, climate, shifts in populations of definitive hosts, or other possible
factors. In the case of N. caninum, the cELISA kit used in this study (which performed
reliably in reindeer vaccinated for N. caninum [P. Curry and S. Kutz, unpublished data])
was different from the indirect fluorescent antibody test used to test caribou sera in a
recent study from Alaska (Stieve et al., 2010).
The findings of higher seroprevalence for BHV-1, PI-3, and Pestivirus are similar to
those in previous reports for Rangifer worldwide (Elazhary 1979, 1981; Zarnke, 1983;
Rehbinder et al., 1992; Stuen et al., 1993; Farnell et al., 1999; Lillehaug et al., 2003;
Johnson et al., 2010; Evans et al., 2012; Kautto et al., 2012). Among the seven herds
surveyed, there was a general pattern of higher prevalence for these pathogens in the
western and Québec herds, and virtually no animals seropositive for these agents in the
Greenland herds. The latter is consistent with a hypothesis of pathogen species loss
233
during the caribou colonization of Greenland (Roed, 2005). The geographic differences
observed may also reflect pathogen circulation within the lineages of caribou
(COSEWIC, 2011) that were the ancestors of those sampled from the regions in this
study. As well, pathogens introduced with Eurasian reindeer translocated to Alaska and
the Mackenzie Delta region in the early 1900s (Scotter, 1969) might have transmitted to
native migratory caribou herds, and this might have contributed to the higher
seroprevalence for these pathogens.
Regarding additional patterns of BHV-1, PI-3, and Pestivirus exposure, there was no
evidence for a difference between males and females; however, there was evidence for
higher risk of exposure in adults versus young caribou, and in fall collections compared
to other seasons. Higher prevalence in adults would be expected with greater probability
of exposure to agents over time. The greater risk of exposure in caribou collected in fall
might be partially explained by increased animal movements and mass aggregations in
spring and summer, which set the stage for relatively high pathogen contact rates.
Although robust testing for pregnancy-related patterns was not possible due to sample
size and data-recording issues, overall analysis (n=93, including 41 females from a herd
in which no sampled animals tested positive for Pestivirus) indicated that pregnant
females might be at higher risk for Pestivirus exposure than non-pregnant females. The
data did not support further analysis and the reasons for this potential disparity in
Pestivirus prevalence are unclear.
234
Future Directions:
The high seroprevalence estimates for BHV-1 and Pestivirus are of particular interest for
future investigation. Recently, Evans et al. (2012) reported high seroprevalence for cervid
alphaherpesvirus in Alaskan caribou, and testing suggested that the circulating agent is
likely cervid herpesvirus 2. This isolate has not yet been confirmed in North America.
The current survey’s results point to the Western Arctic as a potential hot spot for future
research in Rangifer-specific isolates and strains of alphaherpesviruses and pestiviruses.
Regarding survey methodologies, the results also suggest that fall might be the best
season for screening the range of pathogens in caribou herds. This could be useful when
considering aims and design of surveillance schemes or other scientific collections.
Further, the possibility that season of collection might influence caribou serology results
should be taken into account when comparing across herds, years, and different studies in
the literature. Other issues were also highlighted relevant to the CARMA sampling
efforts. It is recognized that a subset of the collections was likely and unavoidably
influenced by seasonal hunting patterns (e.g., in the community hunts of the BluenoseWest herd, for example, as hunters prefer to take bulls in fall). However, that sampling
strategy could be improved to permit more robust analysis by i) obtaining more even
distributions of males and females, and ii) consistently and accurately recording pregnant
vs. non-pregnant status for females. In general, the results demonstrate that sampling with
the ultimate goal of being able to stratify the analysis maximally (i.e., separate and
compare groups by age class, time of collection, etc.) is the best approach for advancing
the understanding of disease ecology and trends across space and time.
235
PHASE III: Filter Papers and Community-Based Wildlife Monitoring
Chapter 6 assessed key aspects of two community-based caribou health-monitoring
programs that were launched in Canada’s North between 2003 and 2008: the Caribou
Health Monitoring Program (CHMP) in Nunavut and the Wildlife Health Monitoring
Program (WHMP) in the Northwest Territories. Both these programs feature hunterbased FP collection from caribou hunted for subsistence, and I contributed to community
implementation of the CHMP in Nunavut in 2008. The chapter reports the quantitative
and qualitative results from face-to-face interviews that were conducted with 25 caribou
harvesters (key informants) from the Western, Eastern, and High Arctic between 2009
and 2010. Some of the interviewees had been trained how to collect FP samples and other
samples and data in the CHMP and WHMP. The goal of the study was to assess attitudes
and beliefs of northern caribou harvesters regarding wildlife disease and hunter-based
wildlife health monitoring in general, and regarding FP blood collection in particular.
Most harvesters viewed wildlife disease as important, most were interested in hunterbased sampling to some degree, and very few had issues with FP sampling. There was
widespread perceived lack of results reporting. Two harvesters described cultural beliefs
that could be barriers to sample collection from caribou.
Reflection and Learning:
Implementing systems for community-based wildlife monitoring in the Arctic is
complicated. Many of the challenges are community-specific and reflect varied capacities
for engagement and leadership among communities. As a new, green researcher in the
236
North, I received a relatively small but important dose of ground-level understanding of
certain commonalities and traits among the Inuit and Dené people I met. Most notable of
these (and admirable to me) were an above-all practical approach to living (which
significantly informs attitudes related to wildlife, wildlife disease, and hunting), an
incredible sense of humour regardless of circumstance, and powerful, heartening
resilience and resignation to “fix it, handle it” and move forward in even the most
difficult situations. Perhaps two challenges had the most profound impact on me during
this part of my doctoral research. One centered around naïve assumptions I realized that I
had made about i) the novelty of my project in the North, ii) the level of interest it
c/would spark in local people, and (iii) the diagnostic approach of FP sampling as a
community-based tool that is “non-invasive” in the subsistence-hunting context. The
latter was of particular interest to me relative to animistic belief systems and honouring
Aboriginal traditions and culture. The other main impact was a relatively shocking
realization of the intensity and pervasiveness of research fatigue in Aboriginal
communities of the North. To me, the potential for this to deaden/muffle a research
initiative—regardless of a project’s genuine intent or potential community importance or
scientific worth—was extremely disquieting. As I had found when seeking to understand
the lives of arctic caribou, reading about research fatigue and communication issues was
different from “witnessing it on the tundra.” Although the reasons for this issue are
understandable, the fact that many of the barriers related to it were established through
past practices and experiences puts contemporary researchers in a difficult position. I
found myself feeling somewhat at a loss for what to do about this given the brief window
of a typical doctoral project. The logistical difficulties of spending the time required and
237
desired in the communities my project was touching was a major obstacle I could not
overcome. Illuminating of this predicament was the situation where, during a meeting in a
northern community, I (as the only PhD student in the room) was indirectly but very
publicly tagged as one of those PhD students “attempting to elevate themselves” through
research on the backs of the people. I did not take this personally as I was comfortable in
the genuineness of my project’s focuses on animal-human connections and communitybased engagement. However, the comment so starkly and bluntly expressed the damage
to local people that past experiences have caused. They left me (and still leave me)
wondering how to make a difference, make things better, as a researcher.
Future Directions:
I hope that forthcoming work with well-intended and potentially valuable wildlife healthmonitoring programs in the North, such as the CHMP and WHMP, can receive the
necessary funding and support that will allow researchers to begin to address, step by
step, some of the deep-rooted challenges that exist. In order for hunter-based wildlife
health programs in Canada’s North to be sustainable and have impact, this study and
relevant literature suggest the need to focus on three main areas: i) effective, accessible
communication of program elements and results; ii) mitigation or dissolution of research
fatigue with new approaches; iii) long-term, iterative collaboration and trust-building
with careful attention to cultural competency, project context, and—above all—ethics.
Each of these is challenging to address or transform in the North. In addition to this
research, several recent studies have gathered opinions directly from Aboriginal
238
community residents regarding health- or wildlife-health research and community
engagement (Byers 1999; Bull, 2010; Wesche et al., 2010; Maar et al., 2011; Pufall et al.,
2011). Adding to what was heard during the interviews in the current study, these articles
offer ideas on ways to proceed.
Final Words …
As a body of work, this research introduces a practical, effective, diagnostic tool for
Rangifer and provides new knowledge and insights about pathogen exposure in caribou.
It also sheds light on caribou harvesters’ perceptions of community-based wildlife-health
monitoring, and the range of complexities that such initiatives entail in the Canadian
North. Findings from this project set new paths for further research and program
transformation, and contribute to the literature as well. Fundamentally, the experimental
work to assess the efficacy of FP samples in nine serological tests opens new avenues for
caribou researchers, biologists, wildlife managers, and others to be able to gather reliable
caribou health data in challenging field conditions. Testing the performance of FP
samples for serology after prolonged storage and under various storage/processing
conditions may inform field workers and others about the robustness of FP samples and
help with decision-making about their handling and storage under different logistical
constraints. Results from the experiments also provide a guide with respect to which
types of test platforms might work optimally with FP samples in other wildlife species,
and therefore which tests might be priorities for validating FP samples. This doctoral
239
project also provides information for wildlife researchers and others worldwide about the
practicality, benefits, and possible pitfalls of FP sampling, and about the technique’s
applicability in different settings. Specifically, it identifies some possible barriers to FP
collection related to Aboriginal cultures and traditions. At a broad scale, this research is
relevant to other wildlife and communities globally, particularly where environmental
conditions are harsh, distances are great, and finances, expertise and practical
requirements for conventional tube blood collection/storage are lacking.
My doctoral research threaded veterinary medicine and human medicine together in an
“arctic ecosystem tapestry” that placed caribou health in context with pressing
environmental concerns (climate change and other anthropogenic disturbances linked
with resource extraction), social issues, Aboriginal cultures, and other elements. It was
also part of a team commitment to enhancing community engagement and capacity in the
North, honouring subsistence harvesters’ wildlife expertise, and weaving scientific and
local knowledge. For me, these many facets and challenges were very meaningful and
created a holistic undertaking that captured my interest from the moment I read the
project description in Dr. Kutz’s graduate student posting. Throughout the past 5 years,
the work has continued to hold its shine, and perhaps the only thing that tarnished (and
that was a good thing!) was some initial naïveté about the North and Northerners. This
tuning was a positive change that required up-close experience, that focused my approach
and thoughts more clearly, and that–I hope–informed the research well.
Being a PhD project, several iterations of the research unfolded that brought revelations
and new twists, that required me to be more flexible and take large (!) leaps of faith, and
240
that exacted typical graduate student costs (stress, angst, nightmares, potential life-years
lost, etc.) . All these happened en route to mostly positive outcomes and many “ahaaa”
moments. Somewhat miraculously—and thanks in great part to the contributions of
others in the North and South—the project managed to complete the arc that was so
important to me from the beginning: i) testing the efficacy of a practical non-invasive (in
subsistence hunting context) field tool for caribou diagnostics, ii) assessing its robustness
by mimicking collection and storage in real-world settings, iii) applying the tool in
scientific research on pathogens of caribou, and iv) putting it in the hands of caribou
harvesters, whose communities are connected to and depend on the animals most
intimately.
It has been an honour and immense life learning to do this work, to hopefully contribute
in some way to the lives and conservation of wild caribou, and to attempt to personally
act on the words of Inuit leader and visionary Sheila Watt-Cloutier:
“Learn from us; we know a little bit about sustainability …
Work with us to keep and maintain our way of life.”
241
BIBLIOGRAPHY
AFSHAR, A., F. C. THOMAS, P. F. WRIGHT, J. L. SHAPIRO, P. T. SHETTIGARA,
AND J. ANDERSON. 1987. Comparison of competitive and indirect enzyme-linked
immunosorbent assays for detection of bluetongue virus antibodies in serum and whole
blood. Journal of Clinical Microbiology 25: 1705-1710.
AGUIRRE, A. A., D. E. HANSEN, E. E. STARKEY, AND R. G. MCLEAN. 1995.
Serologic survey of wild cervids for potential disease agents in selected national parks in
the United States. Preventive Veterinary Medicine 21: 313-322.
AGUIRRE, A. A., R. S. OSTFELD, G. M. TABOR, C. HOUSE, AND M. C. PEARL,
(eds.) 2002. Conservation medicine: Ecological health in practice. Oxford University
Press, New York. 407 pp.
AGUIRRE, A. A. AND G. M. TABOR. 2008. Global factors driving emerging infectious
diseases. Annals of the New York Academy of Sciences 1149: 1-3.
ALBON, S. D., A. STIEN, R. J. IRVINE, R. LANGVATN, E. ROPSTAD, AND O.
HALVORSEN. 2002. The role of parasites in the dynamics of a reindeer population.
Proceedings of the Royal Society of London Series B-Biological Sciences 269: 16251632.
ALTIZER, S., D. HARVELL, AND E. FRIEDLE. 2003. Rapid evolutionary dynamics
and disease threats to biodiversity. Trends in Ecology & Evolution 18: 589-596.
242
ALTIZER, S., R. BARTEL, AND B. A. HAN. 2011. Animal migration and infectious
disease risk. Science 331: 296-302.
ANISIMOV, O., D. G. VAUGHAN, T. V. CALLAGHAN, C. FURGAL, H.
MARCHANT, T. D. PROWSE, H. VILHALMSSON, AND J. E. WALSH. 2007. Polar
regions (Arctic and Antarctic). Climate Change 2007: Impacts, Adaptation and
Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change, M. L. Parry, O. F. Canziani, J. P Palutikof,
P. J. van der Linden, and C. E. Hanson (eds.) 653-685 pp.
ARMITAGE, D., F. BERKES, A. DALE, E. KOCHO-SCHELLENBERG, AND E.
PATTON. 2011. Co-management and the co-production of knowledge: Learning to adapt
in Canada's Arctic. Global Environmental Change 21: 995-1004.
ÅSBAKK K., S. STUEN, H. HANSEN, AND L. FORBES. 1999. A serological survey
for brucellosis in reindeer in Finnmark county, northern Norway. Rangifer 19: 19-24.
ASTLES, R. 2010. Presentation: CLIA proficiency testing criteria for acceptable
performance. CLIAC (Clinical Laboratory Improvement Advisory Committee) meeting
September 1, 2010. Laboratory Practice Standards Branch, Division of Laboratory
Science and Standards at the Centers for Disease Control, USA.
http://wwwn.cdc.gov/cliac/pdf/Addenda/cliac0910/Addendum%20K_Astles.pdf.
Accessed August 2012.
243
AVALOS-RAMIREZ, R., M. ORLICH, H. J. THIEL, AND P. BECHER. 2001.
Evidence for the presence of two novel pestivirus species. Virology 286: 456-465.
BALMASEDA, A., S. SABORIO, Y. TELLEZ, J. C.MERCADO, L. PEREZ, S. N.
HAMMOND, C. ROCHA, G. KUAN, AND E. HARRIS. 2008. Evaluation of
immunological markers in serum, filter-paper blood spots, and saliva for dengue
diagnosis and epidemiological studies. Journal of Clinical Virolology 43: 287–291.
BANKS, M. 1985. Detection of antibodies to Aujeszky's disease virus in whole blood by
Elisadisc. Journal of Virological Methods 12: 41-45.
BARTLETT, C., M. MARSHALL, AND M. MARSHALL. 2012. Two-Eyed Seeing and
other lessons learned within a co-learning journey of bringing together indigenous and
mainstream knowledges and ways of knowing. Journal of Environmental Studies and
Sciences (accepted July 2012).
BATAL, M., K. GRAY-DONALD, H. V. KUHNLEIN, AND O. RECEVEUR. 2005.
Estimation of traditional food intake in indigenous communities in Denendeh and the
Yukon. International Journal of Circumpolar Health 64: 46-54.
BEARD, C. W. AND M. BRUGH, JR. 1977. Use of the Nobuto blood-sampling paper
strip for Newcastle disease serology. Avian Diseases 21: 630–636.
BECHER, P., M. ORLICH, A. KOSMIDOU, M. KONIG, M. BAROTH, AND H. J.
THIEL. 1999. Genetic diversity of pestiviruses: Identification of novel groups and
implications for classification. Virology 262: 64-71.
244
BECHER, P., R. A. RAMIREZ, M. ORLICH, S. C. ROSALES, M. KONIG, M.
SCHWEIZER, H. STALDER, H. SCHIRRMEIER, AND H. J. THIEL. 2003. Genetic
and antigenic characterization of novel pestivirus genotypes: implications for
classification. Virology 311: 96-104.
BEHETS, F., M. KASHAMUKA, M. PAPPAIOANOU, T. A. GREEN, R. W. RYDER,
V. BATTER, J. R. GEORGE, W. H. HANNON, AND T. C. QUINN. 1992. Stability of
human immunodeficiency virus type 1 antibodies in whole blood dried on filter paper and
stored under various tropical conditions in Kinshasa, Zaire. Journal of Clinical
Microbiology 30: 1179-1182.
BERGERUD, A. T. 1996. Evolving perspectives on caribou population dynamics, have
we got it right yet? Rangifer, Special Issue No. 9: 95-115.
BERKES, F. 2009. Evolution of co-management: Role of knowledge generation,
bridging organizations and social learning. Journal of Environmental Management 90:
1692-1702.
BERKES, F., M. K. BERKES, AND H. FAST. 2007. Collaborative integrated
management in Canada's North: The role of local and traditional knowledge and
community-based monitoring. Coastal Management 35: 143-162.
BETANCOURT, J. R. 2004. Cultural competence — Marginal or mainstream
movement? New England Journal of Medicine 351: 953-955.
245
BLITVICH, B. J., R. A. BOWEN, N. L. MARLENEE, R. A. HALL, M. L. BUNNING,
AND B. J. BEATY. 2003. Epitope-blocking enzyme-linked immunosorbent assays for
detection of West Nile virus antibodies in domestic mammals. Journal of Clinical
Microbiology 41: 2676-2679.
BOULT, D. 2006. The Inuit way: A guide to Inuit culture. Pauktuutit, Inuit Women's
Association, Ottawa, Ontario. 21 pp.
BRADLEY, M., S. J. KUTZ, E. JENKINS, AND T. M. O’HARA. 2005. The potential
impact of climate change on infectious diseases of Arctic fauna. International Journal of
Circumpolar Health 64: 468–477.
BROOK, R. K. AND S. P. M. MCLACHLAN. 2008. Trends and prospects for local
knowledge in ecological and conservation research and monitoring. Biodiversity &
Conservation 17: 3501-3512.
BROOK, R. K., S. J. KUTZ, A. M. VEITCH, R. A. POPKO, B. T. ELKIN, AND G.
GUTHRIE. 2009. Fostering community-based wildlife health monitoring and research in
the Canadian North. Ecohealth 6: 266–278.
BROUGHTON, E., L. P. E. CHOQUETTE, J. G. COUSINEAU, AND F. L. MILLER.
1970. Brucellosis in reindeer, Rangifer tarandus L., and the migratory barren-ground
caribou, Rangifer tarandus groenlandicus (L.), in Canada. Canadian Journal of Zoology
48: 1023-1027.
246
BRUGH, M. AND C. W. BEARD. 1980. Collection and processing of blood samples
dried on paper for microassay of Newcastle disease virus and avian influenza virus
antibodies. American Journal of Veterinary Research 41: 1495-1498.
BULL, J. R. 2010. Research with Aboriginal peoples: Authentic relationships as a
precursor to ethical research. Journal of Empirical Research in Human Research Ethics 5:
13-22.
BUTLER, D. 2006. Disease surveillance needs a revolution. Nature 440: 6-7.
BUXTON, D., S. W. MALEY, P. P. PASTORET, B. BROCHIER, AND E. A. INNES.
1997. Examination of red foxes (Vulpes vulpes) from Belgium for antibody to Neospora
caninum and Toxoplasma gondii. Veterinary Record 141: 308-309.
BYERS, T. 1999. Perspectives of Aboriginal peoples on wildlife research. Wildlife
Society Bulletin 27: 671-675.
CAFF (CONSERVATION OF ARCTIC FLORA AND FAUNA GROUP). 2010. Arctic
Biodiversity Trends 2010: Selected Indicators of Change..
http://www.arcticbiodiversity.is/index.php/en/species/wild-reindeer-and-caribou.
Accessed August 2012.
CARMA (CIRCUMARCTIC RANGIFER MONITORING AND ASSESSMENT
NETWORK). 2012. Herds status and trends.
http://www.carmanetwork.com/display/public/Herds. Accessed August 2012.
247
CASTRO, A. C., L. G. BORGES, S. SOUZA RDA, M. GRUDZINSKI, AND P. A.
D’AZEVEDO. 2008. Evaluation of the human immunodeficiency virus type 1 and 2
antibodies detection in dried whole blood spots (DBS) samples. Revista do Instituto de
Medicina Tropical de Sao Paulo 50: 151–156.
CATLEY, A. AND T. LEYLAND. 2001. Community participation and the delivery of
veterinary services in Africa. Preventive Veterinary Medicine 49: 95-113.
CDC (CENTERS FOR DISEASE CONTROL). 2003. Epidemic/epizootic West Nile
virus in the United States: Guidelines for surveillance, prevention, and control. Fort
Collins. 77 pp.
CHAN, H. M., K. FEDIUK, S. HAMILTON, L. ROSTAS, A. CAUGHEY, H.
KUHNLEIN, G. EGELAND, AND E. LORING. 2006. Food security in Nunavut,
Canada: Barriers and recommendations. International Journal of Circumpolar Health 65:
416-431.
CHASE, B. A., S. A. JOHNSTON, AND J. B. LEGUTKI. 2012. Evaluation of biological
sample preparation for immunosignature-based diagnostics. Clinical and Vaccine
Immunology 19: 352-358.
CHITAMBAR, S. D. AND M. S. CHADHA. 2000. Use of filter paper disks for hepatitis
A surveillance. Indian Journal of Gastroenterology 19: 165-167.
CHOMEL, B. B., Y. KIKUCHI, J. S. MARTENSON, M. E. ROELKE-PARKER, C. C.
CHANG, R. W. KASTEN, J. E. FOLEY, J. LAUDRE, K.MURPHY, P. K. SWIFT, V. L.
248
KRAMER, AND S. J. O’BRIEN. 2004. Seroprevalence of Bartonella infection in
American free-ranging and captive pumas (Felis concolor) and bobcats (Lynx rufus).
Veterinary Research 35: 233–241.
CLAGUE, A. AND A. THOMAS. 2002. Neonatal biochemical screening for disease.
Clinica Chimica Acta 315: 99-110.
CONNER, M. M., M. R. EBINGER, J. A. BLANCHONG, AND P. C. CROSS. 2008.
Infectious disease in cervids of North America: data, models, and management
challenges. Annals of the New York Academy of Sciences 1134: 146-172.
CORRAN, P. H., J. COOK, C. LYNCH, H. LEENDERTSE, A. MANJURANO, J.
GRIFFIN, J. COX, T. ABEKU, T. BOUSEMA, A. C. GHANI, C. DRAKELEY, AND E.
RILEY. 2008. Dried blood spots as a source of anti-malarial antibodies for
epidemiological studies. Malaria Journal 7: 195.
COSEWIC (COMMITTEE ON THE STATUS OF ENDANGERED WILDLIFE IN
CANADA). 2011. Special Report: Designatable Units for caribou in Canada. Committee
on the Status of Endangered Wildlife in Canada. Ottawa. 78 pp.
CRAY, C., M. RODRIGUEZ, J. ZAIAS, AND N. H. ALTMAN. 2009. Effects of storage
temperature and time on clinical biochemical parameters from rat serum. Journal of the
American Association for Laboratory Animal Science 48: 202-204.
CURRY, P. S. 2010. Caribou herds and arctic communities: Exploring a new tool for
caribou health monitoring. Arctic 62 (4): 495-499.
249
CUYLER, C. 2007. West Greenland caribou explosion: What happened? What about the
future? Rangifer 17: 219-226.
CUYLER, C. AND R. E. HAUGERUD. 1999. Reindrift i Gronland. [Success and failure
of reindeer herding in Greenland.]. Rangifer 3: 81-92.
CUYLER, C., M. ROSING, H. MØLGAARD, R. HEINRICH, AND K. RAUNDRUP.
2011. Status of two West Greenland caribou populations 2010; 1) KangerlussuaqSisimiut, 2) Akia-Maniitsoq. 162 pp.
CNWDS (CANADA’S NATIONAL WILDLIFE DISEASE STRATEGY). 2004.
Government of Canada, Environment Canada
http://www.ec.gc.ca/Publications/default.asp?lang=En&xml=A75CF606-AF6C-4C158181-596A0CE473B4. Accessed August 2012.
DANIELSEN, F., N. BURGESS, AND A. BALMFORD. 2005. Monitoring matters:
Examining the potential of locally-based approaches. Biodiversity & Conservation 14:
2507-2542.
DANIELSEN, F., N. D. BURGESS, A. BALMFORD, P. F. DONALD, M. FUNDER, J.
P. JONES, P. ALVIOLA, D. S. BALETE, T. BLOMLEY, J. BRASHARES, B. CHILD,
M. ENGHOFF, J. FJELDSA, S. HOLT, H. HUBERTZ, A. E. JENSEN, P. M. JENSEN,
J. MASSAO, M. M. MENDOZA, Y. NGAGA, M. K. POULSEN, R. RUEDA, M. SAM,
T. SKIELBOE, G. STUART-HILL, E. TOPP-JORGENSEN, AND D. YONTEN. 2009.
Local participation in natural resource monitoring: a characterization of approaches.
250
Conservation Biology 23: 31-42.
DAS NEVES, C. G., E. RIMSTAD, AND M. TRYLAND. 2009a. Cervid herpesvirus 2
causes respiratory and fetal infections in semidomesticated reindeer. Journal of Clinical
Microbiology 47: 1309-1313.
DAS NEVES, C. G., T. MORK, J. GODFROID, K. K. SORENSEN, E. BREINES, E.
HAREIDE, J. THIRY, E. RIMSTAD, E. THIRY, AND M. TRYLAND. 2009b.
Experimental infection of reindeer with cervid herpesvirus 2. Clinical and Vaccine
Immunology 16: 1758-1765.
DAS NEVES, C. G., S. ROTH, E. RIMSTAD, E. THIRY, AND M. TRYLAND. 2010.
Cervid herpesvirus 2 infection in reindeer: A review. Veterinary Microbiology 143: 7080.
DASZAK, P., A. A. CUNNINGHAM, AND A. D. HYATT. 2000. Emerging infectious
diseases of wildlife—threats to biodiversity and human health. Science 287: 443-449.
DASZAK, P., A. A. CUNNINGHAM, AND A. D. HYATT. 2001. Anthropogenic
environmental change and the emergence of infectious diseases in wildlife. Acta Tropica
78: 103-116.
DASZAK, P., G. M. TABOR, A. M. KILPATRICK, J. EPSTEIN, AND R.
PLOWRIGHT. 2004. Conservation medicine and a new agenda for emerging diseases.
Annals of the New York Academy of Sciences 1026: 1-11.
251
DASZAK, P., J. H. EPSTEIN, A. M. KILPATRICK, A. A. AGUIRRE, W. B. KARESH,
AND A. A. CUNNINGHAM. 2007. Collaborative research approaches to the role of
wildlife in zoonotic disease emergence. Current Topics in Microbiology and Immunology
315: 463-475.
DE BLAS, I., C. ORTEGA, K. FRANKENA, J. NOORDHUIZEN, AND M.
THRUSFIELD. 2005. Win Episcope 2.0. EPIDECON, Borland® and Delphi™,
www.clive.ed.ac.uk/winepiscope/. Accessed August 2012.
DFO (DEPARTMENT OF FISHERIES AND OCEANS), Government of Canada. 2009.
A Health check-up for the beluga whale (an Inuvialuit community works with scientist to
monitor the population). http://www.dfo-mpo.gc.ca/science/publications/article/2009/0414-09-eng.htm. Accessed August 2012.
DFO (DEPARTMENT OF FISHERIES AND OCEANS), Government of Canada. 2010.
Unravelling the food web (community support to trace the feeding patterns of seals)
http://www.dfo-mpo.gc.ca/science/publications/article/2010/07-21-10-eng.html.
Accessed August 2012.
DIETERICH, R. A., R. L. ZARNKE, J. MORTON, R. BARRETT, J. DAU, AND A. W.
FRANZMANN. 1981. Mammals: Viral diseases: Respiratory viruses. In: Alaskan
Wildlife Diseases. Institute of Arctic Biology, University of Alaska, Fairbanks. pp. 28-29
DUBAY, S. A., S. S. ROSENSTOCK, D. E. STALLKNECHT, AND J. C. DEVOS, JR.
2006. Determining prevalence of bluetongue and epizootic hemorrhagic disease viruses
252
in mule deer in Arizona (USA) using whole blood dried on paper strips compared to
serum analyses. Journal of Wildlife Diseases 42: 159–163.
DUBEY, J. P. 2003. Review of Neospora caninum and neosporosis in animals. The
Korean Journal of Parasitology 41: 1-16.
DUBEY, J. P., AND G. DESMONTS. 1987. Serological responses of equids
fed Toxoplasma gondii oocysts. Equine Veterinary Journal 19: 337-339.
DUBEY, J. P., J. RIGOULET, P. LAGOURETTE, C. GEORGE, L. LONGEART, AND
J. L. LENET. 1996. Fatal transplacental neosporosis in a deer (Cervus eldi siamensis).
Journal of Parasitology 82: 338-339.
DUBEY, J. P., B. LEWIS, K. BEAM, AND B. ABBITT. 2002. Transplacental
toxoplasmosis in a reindeer (Rangifer tarandus) fetus. Veterinary Parasitology 110: 131135.
DUBEY, J. P. AND P. THULLIEZ. 2005. Prevalence of antibodies to Neospora caninum
in wild animals. Journal of Parasitology 91: 1217-1218.
DUBEY, J. P. AND J. L. JONES. 2008. Toxoplasma gondii infection in humans and
animals in the United States. International Journal for Parasitology 38: 1257-1278.
DUBEY, J. P., M. C. JENKINS, O. C. KWOK, R. L. ZINK, M. L. MICHALSKI, V.
ULRICH, J. GILL, M. CARSTENSEN, AND P. THULLIEZ. 2009. Seroprevalence of
Neospora caninum and Toxoplasma gondii antibodies in white-tailed deer (Odocoileus
253
virginianus) from Iowa and Minnesota using four serologic tests. Veterinary Parasitology
161: 330-334.
DURHAM, P. J. AND H. M. SILLARS. 1986. Evaluation of an enzyme-linked
immunosorbent assay (ELISA) for serodiagnosis of infectious bovine rhinotracheitis
infection, with results of a preliminary survey. New Zealand Veterinary Journal 34: 2730.
DURHAM, P. J. AND L. E. HASSARD. 1990. Prevalence of antibodies to infectious
bovine rhinotracheitis, parainfluenza 3, bovine respiratory syncytial, and bovine viral
diarrhea viruses in cattle in Saskatchewan and Alberta. Canadian Veterinary Journal 31:
815-820.
DUSCHER, G., R. PESCHKE, W. WILLE-PIAZZAI, AND A. JOACHIM. 2009.
Parasites on paper—The use of FTA Elute® for the detection of Dirofilaria repens
microfilariae in canine blood. Veterinary Parasitology 161: 349–351.
DUSEK, R. J., J. S. HALL, S. W. NASHOLD, J. L. TESLAA, AND H. S. IP. 2011.
Evaluation of Nobuto filter paper strips for the detection of avian influenza virus antibody
in waterfowl. Avian Diseases 55: 674-676.
EK-KOMMONEN, C., P. VEIJALAINEN, M. RANTALA, AND E. NEUVONEN.
1982. Neutralizing antibodies to bovine herpesvirus 1 in reindeer. Acta Veterinaria
Scandinavica 23: 565-569.
254
EK-KOMMONEN, C., S. PELKONEN, AND P. F. NETTLETON. 1986. Isolation of a
herpesvirus serologically related to bovine herpesvirus 1 from a reindeer (Rangifer
tarandus). Acta Veterinaria Scandinavica 27: 299-301.
ELAZHARY, M. A., R. S. ROY, AND J. L. FRECHETTE. 1979. Serological evidence
of IBR and BVD infection in caribou (Rangifer tarandus). Veterinary Record 105: 336.
ELAZHARY, M. A., J. L. FRECHETTE, A. SILIM, AND R. S. ROY. 1981. Serological
evidence of some bovine viruses in the caribou (Rangifer tarandus caribou) in Quebec.
Journal of Wildlife Diseases 17: 609-612.
ELBIN, C. S., P. OLIVOVA, C. A. MARASHIO, S. K. COOPER, E. CULLEN, J. M.
KEUTZER, AND X. K. ZHANG. 2011. The effect of preparation, storage and shipping
of dried blood spots on the activity of five lysosomal enzymes. Clinica Chimica Acta
412: 1207-1212.
ELMORE, S. A., E. J. JENKINS, K. P. HUYVAERT, L. POLLEY, J. J. ROOT, AND C.
G. MOORE. 2012. Toxoplasma gondii in circumpolar people and wildlife. Vector-Borne
and Zoonotic Diseases 12: 1-9.
EVANS, A. L., C. G. DAS NEVES, G. F. FINSTAD, K. B. BECKMEN, E. SKJERVE,
I. H. NYMO, AND M. TRYLAND. 2012. Evidence of alphaherpesvirus infections in
Alaskan caribou and reindeer. BMC Veterinary Research 8: 5.
FARNELL, R., R. ZARNKE, AND G. KUZYK. 1999. Serological survey of Yukon
caribou 1988-1997: A look at disease prevalence. Department of Renewable Resources,
255
Government of Yukon. 13 pp.
FERGUSON, M. A. 1997. Rangiferine brucellosis on Baffin Island. Journal of Wildlife
Diseases 33: 536-543.
FERNANDEZ-GIMENEZ, M. E., H. L. BALLARD, AND V. E. STURTEVANT. 2008.
Adaptive management and social learning in collaborative and community-based
monitoring: A study of five community-based forestry organizations in the western USA.
Ecology & Society 13: 1-21.
FERRAZ, A. S., E. F. BELO, L. M. COUTINHO, A. P. OLIVEIRA, A. M. CARMO, D.
L. FRANCO, T. FERREIRA, A. Y. YTO, M. S. MACHADO, M. C. SCOLA, AND E.
DE GASPARI. 2008. Storage and stability of IgG and IgM monoclonal antibodies dried
on filter paper and utility in Neisseria meningitidis serotyping by Dot-blot ELISA. BMC
Infectious Diseases 8: 30.
FESTA-BIANCHET, M., J. C. RAY, S. BOUTIN, S. D. CÔTÉ AND A. GUNN. 2011.
Conservation of caribou (Rangifer tarandus) in Canada: An uncertain future. Canadian
Journal of Zoology 89: 419-434.
FORBES, L. B. 1991. Isolates of Brucella suis biovar 4 from animals and humans in
Canada, 1982-1990. Canadian Veterinary Journal 32: 686-688.
FORBES, L. B. AND S. V. TESSARO. 1993. Transmission of brucellosis from reindeer
to cattle. Journal of the American Veterinary Medical Association 203: 289–294.
256
FORBES, L. B. AND S. V. TESSARO. 2003. Evaluation of cattle for experimental
infection with and transmission of Brucella suis biovar 4. Journal of the American
Veterinary Medical Association 222: 1252–1256.
FORBES, L. B., L. MEASURES, AND A. GAJADHAR. 2009. Infectivity of
Toxoplasma gondii in northern traditional (country) foods prepared with meat from
experimentally infected seals. Journal of Food Protection 72: 1756-1760.
FURGAL, C. AND J. SEGUIN. 2006. Climate change, health, and vulnerability in
Canadian northern Aboriginal communities. Environmental Health Perspectives 114:
1964-1970.
FURGAL, C., C. FLETCHER, AND C. DICKSON. 2006. Ways of knowing and
understanding: Towards the convergence of traditional and scientific knowledge of
climate change in the Canadian North. Catalogue No. En4-69/2006E. Environment
Canada. 95 pp.
GALL, D., K. NIELSEN, L. FORBES, W. COOK, D. LECLAIR, S. BALSEVICIUS, L.
KELLY, P. SMITH, AND M. MALLORY. 2001. Evaluation of the fluorescence
polarization assay and comparison to other serological assays for detection of brucellosis
in cervids. Journal of Wildlife Diseases 37: 110-118.
GALL, D. AND K. NIELSEN. 2004. Serological diagnosis of bovine brucellosis: A
review of test performance and cost comparison. Revue Scientifique et Technique 23:
989–1002.
257
GARDNER, I. A., S. HIETALA, AND W. M. BOYCE. 1996. Validity of using
serological tests for diagnosis of diseases in wild animals. Revue Scientifique et
Technique 15: 323-335.
GATES, C. C., G. WOBESER, AND L. B. FORBES. 1984. Rangiferine brucellosis in a
Muskox, Ovibos moschatus moschatus (Zimmermann). Journal of Wildlife Diseases 20:
233–234.
GEARHEARD, S. AND J. SHIRLEY. 2007. Challenges in community-research
relationships: Learning from natural science in Nunavut. Arctic 60: 62-74.
GIROUX, M.-A., M. CAMPBELL, M. DUMOND, AND D. JENKINS. 2012.
Availability of caribou and muskoxen for local human consumption across Nunavut.
Report to the Nunavut Anti-Poverty Secretariat. Department of Environment,
Government of Nunavut, Canada. 26 pp.
GLASER, A. 2004. West Nile virus and North America: An unfolding story. Revue
Scientifique et Technique 23: 557-568.
GORDON, B. 2003. Rangifer and man: An ancient relationship. Rangifer, Special Issue
No. 14: 15-28.
GORDON, B. C. 2005. 8000 years of caribou and human seasonal migration in the
Canadian Barrenlands. Rangifer, Special Issue No. 16: 155-162.
258
GRID-ARENDAL. 2009. Assessing the Impacts of Climate Change on Food Security in
the Canadian Arctic. United Nations Environment Program, Ottawa. 48 pp.
GUNN, A. 2003. Voles, lemmings and caribou - population cycles revisited? Rangifer,
Special Issue No. 14: 105-111.
GUNN, A., T. LEIGHTON, AND G. WOBESER. 1991. Wildlife diseases and parasites
in the Kitikmeot Region, 1894-90. Department of Renewable Resources, Government of
the Northwest Territories, Coppermine, NWT, Canada. 60 pp.
GUTHRIE, R. AND A. SUSI. 1963. A simple phenylalanine method for detecting
phenylketonuria in large populations of newborn infants. Pediatrics 32: 338-343.
GUYOT, M., C. DICKSON, C. PACI, C. FURGAL, AND H. M. CHAN. 2006. Local
observations of climate change and impacts on traditional food security in two northern
aboriginal communities. International Journal of Circumpolar Health: 403-415.
HARWOOD, L. A., P. NORTON, B. DAY, AND P. A. HALL. 2002. The harvest of
beluga whales in Canada's Western Arctic: Hunter-based monitoring of the size and
composition of the catch. Arctic 55: 10.
HELFAND, R. F., H. L. KEYSERLING, I. WILLIAMS, A. MURRAY, J. MEI, C.
MOSCATIELLO, J. ICENOGLE, AND W. J. BELLINI. 2001. Comparative detection of
measles and rubella IgM and IgG derived from filter paper blood and serum samples.
Journal of Medical Virology 65: 751-757.
259
HONOUR, S. AND K. M. HICKLING. 1993. Naturally occurring Brucella suis biovar 4
infection in a moose (Alces alces). Journal of Wildlife Diseases 29: 596–598.
HOPKINS, J. S., H. CHITAMBO, N. MACHILA, A. G. LUCKINS, P. F. RAE, B. P.
VAN DEN, AND M. C. EISLER. 1998. Adaptation and validation of antibody-ELISA
using dried blood spots on filter paper for epidemiological surveys of tsetse-transmitted
trypanosomosis in cattle. Preventive Veterinary Medicine 37: 91–99.
HUDSON, P. J., A. P. DOBSON, AND D. NEWBORN. 1998. Prevention of population
cycles by parasite removal. Science 282: 2256-2258.
IGLOLIORTE, H., (ed.). 2010. We were so far away. The Inuit experience of residential
schools. Legacy of Hope Foundation, Ottawa. 190 pp.
ITK and NRI (INUIT TAPIRIIT KANATMI AND THE NUNAVUT RESEARCH
INSTITUTE). 2007. Negotiating Research Relationships with Inuit Communities: A
Guide for Researchers. Scot Nickels, Jamal Shirley, and Gita Laidler (eds). Inuit Tapiriit
Kanatami and Nunavut Research Institute: Ottawa and Iqaluit. 38 pp.
JENKINS, D. 2009. Caribou health monitoring: Caribou sample collection for disease
monitoring and genetic analysis. Department of Environment, Government of Nunavut,
Interim Progress Report December 2009 NWMB # 2-09-03, Pond Inlet, Nunavut. 9 pp.
JENKINS, D. A. AND J. GOORTS. 2011. Space use and movement patterns of North
Baffin caribou. Field summary and progress report. Nunavut Wildlife Management Board
Project No. 03-09-01. 48 pp.
260
JEPSEN, B. I., H. R. SIEGISMUND, AND M. FREDHOLM. 2002. Population genetics
of the native caribou (Rangifer tarandus groenlandicus) and the semi-domestic reindeer
(Rangifer tarandus tarandus) in southwestern Greenland: Evidence of introgression.
Conservation Genetics 3: 401-409.
JOHNSON, D., N. J. HARMS, N. C. LARTER, B. T. ELKIN, H. TABEL, AND G. WEI.
2010. Serum biochemistry, serology, and parasitology of boreal caribou (Rangifer
tarandus caribou) in the Northwest Territories, Canada. Journal of Wildlife Diseases 46:
1096-1107.
JONES, K. E., N. G. PATEL, M. A. LEVY, A. STOREYGARD, D. BALK, J. L.
GITTLEMAN, AND P. DASZAK. 2008. Global trends in emerging infectious diseases.
Nature 451: 990-993.
JORDAN, C. N., T. KAUR, K. KOENEN, S. DESTEFANO, A. M. ZAJAC, AND D. S.
LINDSAY. 2005. Prevalence of agglutinating antibodies to Toxoplasma gondii and
Sarcocystis neurona in beavers (Castor canadensis) from Massachusetts. Journal of
Parasitology 91: 1228–1229.
JORDAN, L. T., W. J. RETTIE, AND S. V. TESSARO. 2003. Evidence of herpesvirus
infection in Woodland Caribou in Saskatchewan. Journal of Wildlife Diseases 39: 216220.
KALAYOU, S., H. TADELLE, A. BSRAT, N. ABEBE, M. HAILESELASSIE, AND H.
D. SCHALLIG. 2011. Serological evidence of Leishmania donovani infection in
261
apparently healthy dogs using direct agglutination test (DAT) and rk39 dipstick tests in
Kafta Humera, north-west Ethiopia. Transboundary and Emerging Diseases 58: 255-262.
KAUTTO, A. H., S. ALENIUS, T. MOSSING, P. BECHER, S. N. BELÁK, AND M.
LARSKA. 2012. Pestivirus and alphaherpesvirus infections in Swedish reindeer
(Rangifer tarandus tarandus L.). Veterinary Microbiology 156: 64-71.
KEESING, F., L. K. BELDEN, P. DASZAK, A. DOBSON, C. D. HARVELL, R. D.
HOLT, P. HUDSON, A. JOLLES, K. E. JONES, C. E. MITCHELL, S. S. MYERS, T.
BOGICH, AND R. S. OSTFELD. 2010. Impacts of biodiversity on the emergence and
transmission of infectious diseases. Nature 468: 647-652.
KENDRICK, A. AND P. LYVER. 2005. Denésoliné (Chipewyan) knowledge of barrenground caribou (Rangifer tarandus groenlandicus) movements. Arctic 58: 175-191.
KHADKA, D. AND S. K. NEPAL. 2010. Local responses to participatory conservation
in Annapurna conservation area, Nepal. Environmental Management 45: 351-362.
KOFINAS, G., P. LYVER, D. RUSSELL, R. WHITE, A. NELSON, AND N.
FLANDERS. 2003. Towards a protocol for community monitoring of caribou body
condition. Rangifer, Special Issue No. 14: 43-52.
KREEGER, T. J., W. E. COOK, W. H. EDWARDS, AND T. CORNISH. 2004.
Brucellosis in captive Rocky Mountain bighorn sheep (Ovis canadensis) caused by
Brucella abortus biovar 4. Journal of Wildlife Diseases 40: 311–315.
262
KUIKEN, T., F. A. LEIGHTON, R. A. FOUCHIER, J. W. LEDUC, J. S. PEIRIS, A.
SCHUDEL, K. STOHR, AND A. D. OSTERHAUS. 2005. Public health. Pathogen
surveillance in animals. Science 309: 1680-1681.
KUTZ, S. 2007. Climate Change Action Fund, Project A760. December 2007. An
Evaluation of the Role of Climate Change in the Emergence of Pathogens and Diseases in
Arctic and Subarctic Caribou Populations. Prepared for the Climate Change Action Fund,
Government of Canada, by: Susan Kutz, DVM PhD, Research Group for Arctic
Parasitology (RGAP). 132 pp.
KUTZ, S. J., B. T. ELKIN, D. PANAYI, AND J. P. DUBEY. 2001. Prevalence of
Toxoplasma gondii antibodies in barren-ground caribou (Rangifer tarandus
groenlandicus) from the Canadian Arctic. Journal of Parasitology 87: 439-442.
KUTZ, S., E. P. HOBERG, L. POLLEY, AND E. JENKINS. 2005. Global warming is
changing the dynamics of arctic host-parasite systems. Proceedings of the Royal Society
of Biological Sciences 272 (1581): 2571-2576.
KUTZ, S. J., E. J. JENKINS, A. M. VEITCH, J. DUCROCQ, L. POLLEY, B. ELKIN,
AND S. LAIR. 2009. The Arctic as a model for anticipating, preventing, and mitigating
climate change impacts on host-parasite interactions. Veterinary Parasitology 163: 217228.
KUTZ, S. J., J. DUCROCQ, G. G. VEROCAI, B. M. HOAR, D. D. COLWELL, K. B.
BECKMEN, L. POLLEY, B. T. ELKIN, AND E. P. HOBERG. 2012a. Parasites in
263
ungulates of Arctic North America and Greenland: A view of contemporary diversity,
ecology, and impact in a world under change. Advances in Parasitology 79: 89-252.
KUTZ, S. J., J. DUCROCQ, C. CUYLER, B. ELKIN, A. GUNN, L. KOLPASHIKOV,
D. RUSSELL, AND R. G. WHITE. 2012b. Standardized monitoring of Rangifer health
during International Polar Year. Rangifer, in press.
LAMBDEN, J., O. RECEVEUR, AND H. V. KUHNLEIN. 2007. Traditional food
attributes must be included in studies of food security in the Canadian Arctic.
International Journal of Circumpolar Health 66: 308-319.
LAVERY, J. V., P. O. TINADANA, T. W. SCOTT, L. C. HARRINGTON, J. M.
RAMSEY, C. YTUARTE-NUÒEZ, AND A. A. JAMES. 2010. Towards a framework for
community engagement in global health research. Trends in Parasitology 26: 279-283.
LEDERMAN, E. R., M. G. REYNOLDS, K. KAREM, Z. BRADEN, L. A. LEARNEDOROZCO, D. WASSAWASSA, O. MOUNDELI, C. HUGHES, J. HARVEY, R.
REGNERY, J. V. MOMBOULI, AND I. K. DAMON. 2007. Prevalence of antibodies
against orthopoxviruses among residents of Likouala region, Republic of Congo:
Evidence for monkeypox virus exposure. American Journal of Tropical Medicine and
Hygiene 77: 1150–1156.
LEIGHTON, F. A. 2002. Health risk assessment of the translocation of wild animals.
Revue Scientifique et Technique 21: 187-195.
LILLEHAUG, A., T. VIKOREN, I. L. LARSEN, J. AKERSTEDT, J. THARALDSEN,
264
AND K. HANDELAND. 2003. Antibodies to ruminant alpha-herpesviruses and
pestiviruses in Norwegian cervids. Journal of Wildlife Diseases 39: 779-786.
LINDSAY, D. S., E. J. KELLY, R. D. MCKOWN, F. J. STEIN, J. PLOZER, J.
HERMAN, B. L. BLAGBURN, AND J. P. DUBEY. 1996. Prevalence of Neospora
caninum and Toxoplasma gondii antibodies in coyotes (Canis latrans) and experimental
infections of coyotes with Neospora caninum. Journal of Parasitology 82: 657-659.
LINDSAY, D. S., J. L. WESTON, AND S. E. LITTLE. 2001. Prevalence of antibodies to
Neospora caninum and Toxoplasma gondii in gray foxes (Urocyon cinereoargenteus)
from South Carolina. Veterinary Parasitology 97: 159-164.
LYVER, P. AND A. GUNN. 2004. Calibration of hunters’ impressions with female
caribou body condition indices to predict probability of pregnancy. Arctic 57: 233-241.
MAAR, M. A., N. E. LIGHTFOOT, M. E. SUTHERLAND, R. P. STRASSER, K. J.
WILSON, C. M. LIDSTONE-JONES, D. G. GRAHAM, R. BEAUDIN, G. A.
DAYBUTCH, B. R. DOKIS, M. T. LESAGE, M. RAYMOND, AND P. WILLIAMSON.
2011. Thinking outside the box: Aboriginal people’s suggestions for conducting health
studies with Aboriginal communities. Public Health 125: 747-753.
MACMILLAN, H. L., A. B. MACMILLAN, D. R. OFFORD, AND J. L. DINGLE.
1996. Aboriginal health. Canadian Medical Association Journal 155: 1569-1578.
265
MAYAN, M. 2001. An introduction to qualitative methods: A training module for
students and professionals. Qual Institute Press, International Institute for Qualitative
Methodology, Edmonton. 171 pp.
MCDADE, T. W., J. F. STALLINGS, A. ANGOLD, E. J. COSTELLO, M. BURLESON,
J. T. CACIOPPO, R. GLASER, AND C. M. WORTHMAN. 2000. Epstein-Barr virus
antibodies in whole blood spots: A minimally invasive method for assessing an aspect of
cell-mediated immunity. Psychosomatic Medicine 62: 560-567.
MCDADE, T. W., J. BURHOP, AND J. DOHNAL. 2004. High-sensitivity enzyme
immunoassay for C-reactive protein in dried blood spots. Clinical Chemistry 50: 652654.
MCDADE, T. W., S. WILLIAMS, AND J. J. SNODGRASS. 2007. What a drop can do:
Dried blood spots as a minimally invasive method for integrating biomarkers into
population-based research. Demography 44: 899-925.
MCDONALD, J. C., T. W. GYORKOS, B. ALBERTON, J. D. MACLEAN, G.
RICHER, AND D. JURANEK. 1990. An outbreak of toxoplasmosis in pregnant women
in northern Quebec. Journal of Infectious Diseases 161: 769-774.
MCLEAN, G. AND F. HILBINK. 1989. Use of dried blood on filter paper in the ELISA
for Brucella abortus. Journal of Immunological Methods 123: 39–43.
266
MEI, J. V., J. R. ALEXANDER, B. W. ADAM, AND W. H. HANNON. 2001. Use of
filter paper for the collection and analysis of human whole blood specimens. Journal of
Nutrition 131: 1631S-1636S.
MEI, J. V., L. LI, S. A. RASMUSSEN, S. COLLIER, J. L. FRIAS, M. A. HONEIN, G.
M. SHAW, F. LOREY, R. MEYER, S. CHAING, M. A. CANFIELD, J. JONES, AND
W. H. HANNON. 2011. Effect of specimen storage conditions on newborn dried blood
spots used to assess Toxoplasma gondii immunoglobulin M (IgM). Clinica Chimica Acta
412: 455-459.
MERIANOS, A. 2007. Surveillance and response to disease emergence. Current Topics
in Microbiology and Immunology 315: 477-509.
MESSIER, V., B. LEVESQUE, J. F. PROULX, L. ROCHETTE, M. D. LIBMAN, B. J.
WARD, B. SERHIR, M. COUILLARD, N. H. OGDEN, E. DEWAILLY, B. HUBERT,
S. DERY, C. BARTHE, D. MURPHY, AND B. DIXON. 2009. Seroprevalence of
Toxoplasma gondii among Nunavik Inuit (Canada). Zoonoses Public Health 56: 188-197.
MORTON, J. K., J. F. EVERMANN, AND R. A. DIETERICH. 1990. Experimental
infection of reindeer with bovine viral diarrhea virus. Rangifer 10: 75-77.
MORENS, D. M., G. K. FOLKERS, AND A. S. FAUCI. 2004. The challenge of
emerging and re-emerging infectious diseases. Nature 430: 242-249.
267
MÖRNER, T., D. L. OBENDORF, M. ARTOIS, AND M. H. WOODFORD. 2002.
Surveillance and monitoring of wildlife diseases. Revue Scientifique et Technique 21:
67-76.
MRNF (MINISTÈRE DES RESSOURCES NATURELLES ET DE LA FAUNE). 2012.
Québec Ministry of Natural Resources press release, August 16, 2012: George River
caribou herd. Ministry publishes results of [July 2012 population] survey.
http://www.mrnf.gouv.qc.ca/english/press/press-release-detail.jsp?id=9880. Accessed
September 2012.
MUBARAK, H. S., S. YUKSEL, O. M. MUSTAFA, S. A. IBRAHIM, A. D.
OSTERHAUS, AND R. L. DE SWART. 2004. Surveillance of measles in the Sudan
using filter paper blood samples. Journal of Medical Virology 73: 624-630.
NCCLS (NATIONAL COMMITTEE FOR CLINICAL LABORATORY
STANDARDS). 2003. Blood collection on filter paper for newborn screening programs.
4th Ed. NCCLS Wayne.
NEIMANIS, A. 2005. The Sahtu Wildlife Health Monitor Program: Summary and
assessment. Unpublished report. Dept. of Veterinary Pathology, Western College of
Veterinary Medicine, Saskatoon, SK, Canada. 33pp.
NIELSEN, K. 1990. The serological response of cattle immunized with Yersinia
enterocolitica O:9 or O:16 to Yersinia and Brucella abortus antigens in enzyme
immunoassays. Veterinary Immunology and Immunopathology 24: 373–382.
268
NIELSEN, K., L. KELLY, D. GALL, P. SMITH, J. BOSSE, P. NICOLETTI, AND W.
KELLY. 1994. The use of divalent cation chelating agents (EDTA/EGTA) to reduce nonspecific serum protein interaction in enzyme immunoassay. Veterinary Research
Communications 18: 433–437.
NIELSEN, K., D. GALL, W. KELLY, D. VIGLIOCCO, D. HENNING, AND M.
GARCIA. 1996. Immunoassay development: Application to enzyme immunoassay for
diagnosis of brucellosis. Monograph, Agriculture and Agri-Food Canada, Animal
Diseases Research Institute, Nepean, Ontario, Canada. 188 pp.
NIELSEN, K., D. GALL, P. SMITH, S. BALSEVICIUS, F. GARRIDO, M. D. FERRER,
F. BIANCIFIORI, A. DAJER, E. LUNA, L. SAMARTINO, R. BERMUDEZ, F.
MORENO, T. RENTERIA, AND A. CORRAL. 2004. Comparison of serological tests
for the detection of ovine and caprine antibody to Brucella melitensis. Revue Scientifique
et Technique 23: 979–987.
NPC (NUNAVUT PLANNING COMMISSION). 2000. North Baffin Regional Land Use
Plan. Approved by Governments of Canada and Nunavut in June 2000. 130 pp.
OKSANEN, A., K. GUSTAFSSON, A. LUNDEN, J. P. DUBEY, P. THULLIEZ, AND
A. UGGLA. 1996. Experimental Toxoplasma gondii infection leading to fatal enteritis in
reindeer (Rangifer tarandus). Journal of Parasitology 82: 843-845.
OKSANEN, A., K. ÅSBAKK, M. NIEMINEN, H. NORBERG, AND A. NÄREAHO.
1997. Antibodies against Toxoplasma gondii in Fennoscandian reindeer — association
269
with the degree of domestication. Parasitology International 46: 255-261.
OLSON, D. F. 1969 Alaska reindeer herdsmen: A study of native management in
transition. Institute of Social, Economic and Government Research, University of
Fairbanks, Alaska.
http://www.alaskool.org/projects/reindeer/history/iser1969/rdeer_1.html Accessed
September 2012.
PALMER, M. V., W. C. STOFFREGEN, D. G. ROGERS, A. N. HAMIR, J. A. RICHT,
D. D. PEDERSEN, AND W. R. WATERS. 2004a. West Nile virus infection in reindeer
(Rangifer tarandus). Journal of Veterinary Diagnostic Investigation 16: 219-222.
PALMER, M. V., W. R. WATERS, D. D. PEDERSEN, AND W. C. STOFFREGEN.
2004b. Induction of neutralizing antibodies in reindeer (Rangifer tarandus) after
administration of a killed West Nile virus vaccine. Journal of Wildlife Diseases 40: 759762.
PEACOCK, L., ORLANDO, AND A., COXON. 2009. Polar bear research in Nunavut:
Presentation from the 2009 Wildlife Symposium in Rankin Inlet, Nunavut, Canada.
http://env.gov.nu.ca/sites/default/files/Polar%20Bear%20Research%20in%20Nunavut_E
NGLISHsmall.pdf. Accessed July 2012.
PHAC (PUBLIC HEALTH AGENCY OF CANADA). 2007. Public Health Agency of
Canada surveillance reports for West Nile Virus: http://www.phac-aspc.gc.ca/wnvvwn/nsr-rns2007-eng.php). Accessed August 2012.
270
PHILLIPS, B. L., C. KELEHEAR, L. G. PIZZATTO, G. P. BROWN, D. BARTON,
AND R. SHINE. 2010. Parasites and pathogens lag behind their host during periods of
host range advance. Ecology 91: 872-881.
POOLE, K. G. 1994. Characteristics of an unharvested lynx population during a
snowshoe hare decline. Journal of Wildlife Management 58: 608-618.
POWER, E. 2008. Conceptualizing food security for Aboriginal people in Canada.
Canadian Journal of Public Health 99: 95-97.
PRESTRUD, K. W., K. ASBAKK, E. FUGLEI, T. MORK, A. STIEN, E. ROPSTAD, M.
TRYLAND, G. W. GABRIELSEN, C. LYDERSEN, K. M. KOVACS, M. J. J. E.
LOONEN, K. SAGERUP, AND A. OKSANEN. 2007. Serosurvey for Toxoplasma
gondii in arctic foxes and possible sources of infection in the high Arctic of Svalbard.
Veterinary Parasitology 150: 6-12.
PUFALL, E. L., A. Q. JONES, S. A. MCEWEN, C. LYALL, A. S. PEREGRINE, AND
V. L. EDGE. 2011. Community-derived research dissemination strategies in an Inuit
community. International Journal of Circumpolar Health 70: 532-541.
PUNNARUGSA, V. AND V. MUNGMEE. 1991. Detection of rubella virus
immunoglobulin G (IgG) and IgM antibodies in whole blood on Whatman paper:
Comparison with detection in sera. Journal of Clinical Microbiology 29: 2209-2212.
RADOSTITS, O., C. GAY, K. HINCHCLIFF, AND P. CONSTABLE. 2007. Veterinary
Medicine, 10th edition: A textbook of the diseases of cattle, horses, sheep, pigs and goats.
271
Saunders Ltd. Philadelphia. 2065 pp.
RAYMOND, C. M., I. FAZEY, M. S. REED, L. C. STRINGER, G. M. ROBINSON,
AND A. C. EVELY. 2010. Integrating local and scientific knowledge for environmental
management. Journal of Environmental Management 91: 1766-1777.
RECEVEUR, O., M. BOULAY, AND H. V. KUHNLEIN. 1997. Decreasing traditional
food use affects diet quality for adult Dené/Métis in 16 communities of the Canadian
Northwest Territories. Journal of Nutrition: 2179-2186.
RECEVEUR, O. AND H. V. KUHNLEIN. 1998. Benefits of traditional food in
Dené/Métis communities. International Journal of Circumpolar Health 57 Suppl 1: 219221.
REHBINDER, C., S. BELAK, AND M. NORDKVIST. 1992. A serological,
retrospective study in reindeer on five different viruses. Rangifer 12: 191-195.
RICHMOND, C. A. AND N. A. ROSS. 2009. The determinants of First Nation and Inuit
health: A critical population health approach. Health Place 15: 403-411.
ROBSON, C. 2002. Real world research. 2nd edition. Blackwell Publishers, Oxford. 599
pp.
ROCKBORN, G., C. REHBINDER, B. KLINGEBORN, M. LEFFLER, K.
KLINTEVALL, T. NIKKILA, A. LANDEN, AND M. NORDKVIST. 1990. The
demonstration of a herpesvirus, related to bovine herpesvirus 1, in reindeer with
272
ulcerative and necrotizing lesions of the upper alimentary tract and nose. Rangifer,
Special Issue No. 3: 373-384.
ROED, K. H. 2005. Refugial origin and postglacial colonization of holarctic reindeer and
caribou. Rangifer 25: 19-30.
ROMICH, J. A. 2008. Understanding zoonotic diseases. Thomas Delmar Learning,
Clifton Park, New York. 701 pp.
ROSE, C. H., P. NANSEN, R. J. JORGENSEN, AND D. E. JACOBS. 1984. Sheep
farming in Greenland: Its history, managemental practices and disease problems.
Scandinavian Journal of Veterinary Science 36: 65-76.
RUANGTURAKIT, S., S. ROJANASUPHOT, A. SRIJUGGRAVANVONG, S.
DUANGCHANDA, S. NUANGPLEE, AND A. IGARASHI. 1994. Storage stability of
dengue IgM and IgG antibodies in whole blood and serum dried on filter paper strips
detected by ELISA. Southeast Asian Journal of Tropical Medicine and Public Health 25:
560-564.
RUSSELL, D. E., A. M. MARTELL, AND W. A. C. NIXON. 1993. Range ecology of
the porcupine caribou herd in Canada. Rangifer, Special Issue No. 8: 1-168.
SACKS, B. N., B. B. CHOMEL, R. W. KASTEN, C. C. CHANG, R. K. SANDERS,
AND S. D. LETERME. 2002. Validation for use with coyotes (Canis latrans) of a
commercially available enzyme-linked immunosorbent assay for Dirofilaria immitis.
Veterinary Parasitology 109: 45–58.
273
SCOTTER, G. W. 1969. Reindeer husbandry as a land use in Northern Canada. In
conference proceedings: Productivity and Conservation in Northern Circumpolar Lands,
Edmonton, Alberta, Canada. pp. 159-169
SERGEANT, E. G. S. 2009. Epitools epidemiological calculators. AusVet Animal Health
Services and Australian Biosecurity Cooperative Research Centre for Emerging
Infectious Disease Available at: http://epitools.ausvet.com.au. Accessed August 2012.
SMITH, K. F., K. ACEVEDO-WHITEHOUSE, AND A. B. PEDERSEN. 2009. The role
of infectious diseases in biological conservation. Animal Conservation 12: 1-12.
STALLKNECHT, D. E. 2007. Impediments to wildlife disease surveillance, research,
and diagnostics. Current Topics in Microbiology and Immunology 315: 445-461.
STALLKNECHT, D. E. AND W. R. DAVIDSON. 1992. Antibodies to bluetongue and
epizootic hemorrhagic disease viruses from white-tailed deer blood samples dried on
paper strips. Journal of Wildlife Diseases 28: 306–310.
STENSETH, N. C., W. FALCK, O. N. BJORNSTAD, AND C. J. KREBS. 1997.
Population regulation in snowshoe hare and Canadian lynx: asymmetric food web
configurations between hare and lynx. Proceedings of the National Academy of Sciences
of the United States 94: 5147-5152.
STIEVE, E., K. BECKMEN, S. A. KANIA, A. WIDNER, AND S. PATTON. 2010.
Neospora caninum and Toxoplasma gondii antibody prevalence in Alaska wildlife.
Journal of Wildlife Diseases 46: 348-355.
274
STUEN, S., J. KROGSRUD, B. HYLLSETH, AND N. J. C. TYLER. 1993. Serosurvey
of three virus infections in reindeer in northern Norway and Svalbard. Rangifer 13: 215219.
TAILLON, J., M. FESTA-BIANCHET, AND S. D. CÔTÉ. 2012. Shifting targets in the
tundra: Protection of migratory caribou calving grounds must account for spatial changes
over time. Biological Conservation 147: 163-173.
TAKKOUCHE, B., J. IGLESIAS, J. R. ALONSO-FERNANDEZ, C. FERNANDEZGONZALEZ, AND J. J. GESTALOTERO. 1995. Detection of Brucella antibodies in
eluted dried blood: A validation study. Immunology Letters 45: 107–108.
TESSARO, S. V. AND L. B. FORBES. 1986. Brucella suis biotype 4: a case of
granulomatous nephritis in a barrenground caribou (Rangifer tarandus groenlandicus L.)
with a review of the distribution of rangiferine brucellosis in Canada. Journal of Wildlife
Diseases 22: 479-483.
TESSARO, S. V., D. DEREGT, E. DZUS, C. ROHNER, K. SMITH, AND T.
GABOURY. 2005. Herpesvirus infection in woodland caribou in Alberta, Canada.
Journal of Wildlife Diseases 41: 803-805.
THANGAVELU, A., G. D. RAJ, S. ELANKUMARAN, AND A. KOTEESWARAN.
2000. Evaluation of a filter paper blood sampling technique for quantitative assessment of
antibodies to infectious bursal disease virus. Tropical Animal Health and Production 32:
179–182.
275
THURMOND, M. C. 2003. Conceptual foundations for infectious disease surveillance.
Journal of Veterinary Diagnostic Investigation 15: 501-514.
TORCHIN, M. E., K. D. LAFFERTY, A. P. DOBSON, V. J. MCKENZIE, AND A. M.
KURIS. 2003. Introduced species and their missing parasites. Nature 421: 628-630.
TRUDEAU, S., P. MINEAU, G. S. CARTIER, G. FITZGERALD, L. WILSON, C.
WHELER, AND L. D. KNOPPER. 2007. Using dried blood spots stored on filter paper
to measure cholinesterase activity in wild avian species. Biomarkers 12: 145–154.
TRYLAND, M. 2012. Are we facing new health challenges and diseases in reindeer in
Fennoscandia? Rangifer 32: 35-47.
TRYLAND, M., T. MORK, K. A. RYENG, AND K. K. SORENSEN. 2005. Evidence of
parapox-, alphaherpes- and pestivirus infections in carcasses of semi-domesticated
reindeer (Rangifer tarandus tarandus) from Finnmark, Norway. Rangifer 25 (2): 75-83.
TRYLAND, M., C. G. DAS NEVES, M. SUNDE, AND T. MORK. 2009. Cervid
herpesvirus 2, the primary agent in an outbreak of infectious keratoconjunctivitis in
semidomesticated reindeer. Journal of Clinical Microbiology 47: 3707-3713.
TRYLAND, M., S. J. KUTZ, AND P. S. CURRY. 2012. Wildlife health in a changing
North: A model for global environmental change. In: Aguirre, A. A., R. S. Ostfeld, and P.
Daszak (eds.). New directions in conservation medicine: Applied cases of ecological
health. Oxford University Press, New York. 639 pp.
276
ULVEVADET, B. AND K. KLOKOV. 2004. Family-based reindeer herding and hunting
economies, and the status and management of wild reindeer/caribou populations. Centre
for Saami Studies, University of Tromso, Tromso, Norway. 169 pp.
VALARCHER, J. F. AND G. TAYLOR. 2007. Bovine respiratory syncytial virus
infection. Veterinary Research 38: 153-180.
VEITCH, A., S. KUTZ, B. ELKIN, R. POPKO, A. NEIMANIS, E. JENKINS, AND G.
GUTHRIE. 2005. Northwest Territories Cumulative Impact Monitoring. Communitybased monitoring of wildlife populations and health in the Sahtu: Report to the NWT
Cumulative Impact Monitoring Program: 2004-2005.
VEITCH, A. AND S. KUTZ. 2006. Northwest Territories Cumulative Impact
Monitoring. Community-based monitoring of wildlife populations and health in the
Sahtu: Report to the NWT Cumulative Impact Monitoring Program: 2005-2006.
VIKOREN, T., J. THARALDSEN, B. FREDRIKSEN, AND K. HANDELAND. 2004.
Prevalence of Toxoplasma gondii antibodies in wild red deer, roe deer, moose, and
reindeer from Norway. Veterinary Parasitology 120: 159-169.
VORS, L. S. AND M. S. BOYCE. 2009. Global declines of caribou and reindeer. Global
Change Biology 15: 2626-2633.
WALDNER, C. L. AND J. R. CAMPBELL. 2005. Use of serologic evaluation for
antibodies against bovine viral diarrhea virus for detection of persistently infected calves
in beef herds. American Journal of Veterinary Research 66: 825-834.
277
WESCHE, S. AND H. CHAN. 2010. Adapting to the impacts of climate change on food
security among Inuit in the western Canadian Arctic. Ecohealth 7: 361-373.
WESCHE, S., R. C. SCHUSTER, P. TOBIN, C. DICKSON, D. MATTHIESSEN, S.
GRAUPE, M. WILLIAMS, AND H. M. CHAN. 2011. Community-based health research
led by the Vuntut Gwitchin First Nation. International Journal of Circumpolar Health 70:
396-406.
WHITE, G. 2006. Cultures in collision: Traditional knowledge and Euro-Canadian
governance processes in northern land-claim boards. Arctic 59: 401-414.
WHITTEN, K. R. 1996. Ecology of the Porcupine caribou herd. Rangifer, Special Issue
No. 9: 45-51.
WILLIAMS, E. S. AND I. K. BARKER, (eds.) 2001. Infectious diseases of wild animals.
3rd edition. Blackwell Publishing, Ames. 558 pp.
WOBESER, G. AND A. G. WOBESER. 1992. Carcass disappearance and estimation of
mortality in a simulated die-off of small birds. Journal of Wildlife Diseases 28: 548-554.
WOBESER, G. A. 2006. Essentials of disease in wild animals. 1st edition. Blackwell
Publishing. Ames. 244 pp.
WOODRUM, D. AND L. YORK. 1998. Two-year stability of free and total PSA in
frozen serum samples. Urology 52: 247-251.
WOODS, L. W., M. L. ANDERSON, P. K. SWIFT, AND K. W. SVERLOW. 1994.
278
Systemic neosporosis in a California black-tailed deer (Odocoileus hemionus
columbianus). Journal of Veterinary Diagnostic Investigation 6: 508-510.
YAMAMOTO, K., B. B. CHOMEL, L. J. LOWENSTINE, Y. KIKUCHI, L. G.
PHILLIPS, B. C. BARR, P. K. SWIFT, K. R. JONES, S. P. RILEY, R. W. KASTEN, J.
E. FOLEY, AND N. C. PEDERSEN. 1998. Bartonella henselae antibody prevalence in
free-ranging and captive wild felids from California. Journal of Wildlife Diseases 34: 56–
63.
YU, C., C. ZIMMERMAN, R. STONE, R. E. ENGLE, W. ELKINS, G. A. NARDONE,
S. U. EMERSON, AND R. H. PURCELL. 2007. Using improved technology for filter
paper-based blood collection to survey wild Sika deer for antibodies to hepatitis E virus.
Journal of Virological Methods 142: 143–150.
ZARNKE, R. L. 1983. Serologic survey for selected microbial pathogens in Alaskan
wildlife. Journal of Wildlife Diseases 19: 324-329.
ZARNKE, R. L., J. P. DUBEY, O. C. H. KWOK, AND J. M. VER HOEF. 2000.
Serologic survey for Toxoplasma gondii in selected wildlife species from Alaska. Journal
of Wildlife Diseases 36: 219-224.
ZARNKE, R. L., J. P. DUBEY, J. M. VER HOEF, M. E. MCNAY, AND O. C. KWOK.
2001. Serologic survey for Toxoplasma gondii in lynx from interior Alaska. Journal of
Wildlife Diseases 37: 36-38.
279
ZARNKE, R. L., J. M. VER HOEF, AND R. A. DELONG. 2006. Geographic pattern of
serum antibody prevalence for Brucella spp. in caribou, grizzly bears, and wolves from
Alaska, 1975-1998. Journal of Wildlife Diseases 42: 570-577.
280
APPENDIX 4A
Remaining estimates for performance of filter-paper (FP) blood samples relative to serum
for each respective storage period. Prevalence (prev.) based on serum and FP results,
positive and negative predictive values (PPV and NPV), and confidence intervals (CI) are
shown in duplicate (Run A and B). All FP samples were dried overnight after collection
and stored dry at room temperature. AGENT a
ESTIMATE and TEST Brucella cELISA Brucella. iELISA S prev % CI F prev % CI PPV % CI NPV % CI S prev % CI F prev % CI PPV % CI NPV % CI N. caninum cELISA S prev % CI F prev % CI PPV % CI NPV % CI 51.6 44.2‐59.0 46.2 38.8‐53.7 98.8 93.6‐100 88.9 81.0‐94.3 42.4 35.2‐49.9 42.9 35.7‐50.4 98.7 93.2‐100 99.3* 97.2‐100 1 mo. A (n=45) FP / SERUM STORAGE TIME and RUN (n sample pairs tested) 2 mo. B 12 mo. A 12 mo. B 24 mo. A 24 mo. B b
b
(n=184) (n=184) (n=184) (n=88) (n=88) 50.5 46.2 51.1 45.5 n/a 43.1‐58.0 38.8‐53.7 43.6‐58.5 34.8‐56.4 45.7 48.4 43.5 40.0 n/a 38.3‐53.1 41.0‐55.8 36.2‐51.0 29.5‐50.8 98.8 92.1 97.5 97.1 n/a 93.5‐100 84.5‐96.8 91.3‐99.7 85.1‐99.9 90.0 96.8 84.6 88.7 n/a 82.4‐95.1 91.1‐99.3 76.2‐90.9 77.0‐95.7 42.4 44.6 45.1 38.6 n/a 35.2‐49.9 37.3‐52.1 37.8‐52.6 28.4‐49.6 42.4 43.5 42.4 38.6 n/a 35.2‐49.9 36.2‐51.0 35.2‐49.9 28.4‐49.6 99.1* 93.8 98.7 98.0* n/a 96.2‐100 86.0‐98.0 93.1‐100 91.6‐100 99.4* 93.3 94.3 98.7* n/a 97.2‐100 86.6‐97.3 88.1‐98.0 94.6‐100 1 mo. B 6 mo. A 6 mo. B 17 mo. A 17 mo. B (n=45) (n=45) (n=45) (n=45) (n=45) 73.3 58.1‐85.4 75.6 60.5‐87.1 97.1 84.7‐99.9 93.9* 76.2‐100 73.3 58.1‐85.4 75.6 60.5‐87.1 97.1 84.7‐99.9 93.9* 76.2‐100 2 mo. A (n=184) 73.3 58.1‐85.4 64.4 48.8‐78.1 97.6* 90.2‐100 75.0 47.6‐92.7 71.1 55.7‐83.6 60.0 44.3‐74.3 97.5* 89.5‐100 72.2 46.5‐90.3 73.3 58.1‐85.4 71.1 55.7‐83.6 97.9* 91.1‐100 92.3 64.0‐99.8 71.1 55.7‐83.6 66.7 51.1‐80.0 97.7* 90.5‐100 86.7 59.5‐98.3 a Where the initial estimate derived was 100%, the median unbiased estimate (denoted with *) was reported instead. b After 24 months’ dry storage, a subset of 88 FP samples was compared to corresponding sera (see Methods). Abbreviations: CI: Clopper‐Pearson 95% exact confidence interval; F prev: filter paper prevalence; NPV: predictive value of a negative filter paper test; PPV: predictive value of a positive filter paper test; S prev: serum prevalence 281
APPENDIX 4A (cont’d)
AGENT a
ESTIMATE and TEST BHV‐1 iELISA PI‐3 iELISA BRSV iELISA S prev % CI F prev % CI PPV % CI NPV % CI S prev % CI F prev % CI PPV % CI NPV % CI S prev % CI F prev % CI PPV % CI NPV % CI FP / SERUM STORAGE TIME and RUN (n sample pairs tested) 1 mo. A (n=36) 1 mo. B (n=36) 6 mo. A (n=36) 6 mo. B (n=36) 12 mo. A (n=36) 12 mo. B (n=36) 58.3 40.8‐74.5 50.0 32.9‐67.1 *96.2 84.7‐100 83.3 58.6‐96.4 52.8 35.5‐69.6 55.5 38.1‐72.1 94.7 74.0‐99.9 88.2 63.6‐98.5 66.7 49.0‐81.4 80.6 64.0‐91.8 79.3 60.3‐92.0 85.7 42.1‐99.6 58.3 40.8‐74.5 55.6 38.1‐72.1 *96.6 86.1‐100 93.8 69.8‐99.8 55.6 38.1‐72.1 50.0 32.9‐67.1 94.4 72.7‐99.8 83.3 58.6‐96.4 66.7 49.0‐81.4 69.4 51.9‐83.7 96.0 79.7‐99.9 *93.3 76.2‐100 58.3 40.8‐74.5 50.0 32.9‐67.1 94.4 72.7‐99.9 77.8 52.4‐93.6 44.4 27.9‐61.9 47.2 30.4‐64.5 94.1 71.3‐99.9 *96.4 85.4‐100 61.1 43.5‐76.9 66.7 49.0‐81.4 91.7 73.0‐99.0 *94.4 77.9‐100 61.1 43.5‐76.9 44.4 27.9‐61.9 *95.8 82.9‐100 70.0 45.7‐88.1 36.1 20.8‐53.8 36.1 20.8‐53.8 84.6 54.6‐98.1 91.3 71.9‐98.9 66.7 49.0‐81.4 61.1 43.5‐76.9 *96.9 87.3‐100 85.7 57.2‐98.2 58.3 40.8‐74.5 41.7 25.5‐59.2 *95.5 81.9‐100 71.4 47.8‐88.7 50.0 32.9‐67.1 36.1 20.8‐53.8 *94.8 79.4‐100 78.3 56.3‐92.5 50.0 32.9‐67.1 52.8 35.5‐69.6 94.7 74.0‐99.9 *96.0 83.8‐100 61.1 43.5‐76.9 38.9 23.1‐56.6 *95.2 80.7‐100 63.6 40.7‐82.8 38.9 23.1‐56.5 36.1 20.8‐53.8 84.6 54.6‐98.1 87.0 66.4‐97.2 58.3 40.8‐74.5 58.3 40.8‐74.5 85.7 63.7‐97.0 80.0 51.9‐95.7 a Where the initial estimate derived was 100%, the median unbiased estimate (denoted with *) was reported instead. Abbreviations: CI: Clopper‐Pearson 95% exact confidence interval; F prev: filter paper prevalence; NPV: predictive value of a negative filter paper test; PPV: predictive value of a positive filter paper test; S prev: serum prevalence 282
APPENDIX 4A (cont’d)
AGENT a
ESTIMATE and TEST c
BVDV‐I VN S prev % CI F prev % CI PPV % CI NPV % CI c
BVDV‐II VN S prev % CI F prev % CI PPV % CI NPV % CI FP / SERUM STORAGE TIME and RUN (n sample pairs tested) 1 mo. A (n=32) 1 mo. B (n=32) 6 mo. A 6 mo. B 12 mo. A (n=26) 12 mo. B (n=26) 71.9 53.3‐86.3 65.6 46.8‐81.4 *96.8 86.7‐100 81.8 48.2‐97.7 71.9 53.3‐86.3 68.8 50.0‐83.9 *96.9 87.3‐100 90.0 55.5‐99.8 n/a n/a n/a n/a n/a n/a n/a n/a 61.5 40.6‐79.8 53.9 33.4‐73.4 *95.2 80.7‐100 83.3 51.6‐97.9 61.5 40.6‐79.8 53.9 33.4‐73.4 *95.2 80.7‐100 83.3 51.6‐97.9 1 mo. A (n=30) 1 mo. B (n=26) 6 mo. A 6 mo. B 12 mo. A (n=29) 12 mo. B (n=30) 66.7 47.2‐82.7 63.3 43.9‐80.1 *96.4 85.4‐100 90.9 58.7‐99.8 65.4 44.3‐82.8 57.7 36.9‐76.7 *95.5 81.9‐100 81.8 48.2‐97.7 n/a n/a n/a n/a n/a n/a n/a n/a 62.1 42.3‐79.3 58.6 38.9‐76.5 *96.0 83.8‐100 91.7 61.5‐99.8 63.3 43.9‐80.1 60.0 40.6‐77.3 *96.2 84.7‐100 91.7 61.5‐99.8 c
c
c
c
a Where the initial estimate derived was 100%, the median unbiased estimate (denoted with *) was reported instead. c Sample pairs with a toxic (unreadable) result on VN were excluded, altering the sample sizes for estimate calculations. Estimates for 6 months’ storage were not available due to severely reduced sample size after these exclusions. Abbreviations: CI: Clopper‐Pearson 95% exact confidence interval; F prev: filter paper prevalence; NPV: predictive value of a negative filter paper test; PPV: predictive value of a positive filter paper test; S prev: serum prevalence 283
APPENDIX 6A
Background information sheet for the harvester interviews.
284
APPENDIX 6B
Nunavut interview script (8 pages).
285
286
287
288
289
290
291
292
APPENDIX 6C
Sahtu interview script (8 pages).
293
294
295
296
297
298
299
300
APPENDIX 6D
Summary of results for the Section I items of the Nunavut and Sahtu interviews that had
quantifiable responses i.e., visual analog scale 1 through 5, or Yes/No answers). The 25
total harvesters interviewed included “Hunters” (those who actually hunted animals in
addition to other activities such as butchering etc.) and “Non-hunters” (those who
engaged strictly in butchering and meat preparation). The scale response options were 1 =
Not important; 2 = Somewhat important; 3 = No opinion; 4 = Important; 5 = Very
Important
Paraphrased Interview Item (n respondents) Scale Responses n (%) 1 2 3 4 5 What is the importance/value of wildlife to you? (n=25 ) 0 0 0 5 (20%) What is the importance/value of caribou, specifically, to you? (n=22) 0 0 0 Is wildlife important to your community overall? (n=25 ) Do you/your family eat wildlife? (n=25 ) Are you/your family eating more store‐bought food than you were 5 yrs ago? (n=25) What is the importance of wildlife disease to you? (n=25) Yes No 20 (80%) 4 (18%) 18 (82%) 25 (100%) 0 25 (100%) 0 19 (76%) 6 (24%) 0 1 (4%) 2 (8%) 7 (28%) 15 (60%) SECTION I Background: Wildlife and Food 301
APPENDIX 6E
Summary of results for the Section II items of the Nunavut interviews (n=16) that had
quantifiable responses (i.e., visual analog scale 1 through 5, or Yes/No answers). The
harvesters included “Hunters” (those who actually hunted animals in addition to
butchering etc.) and “Non-hunters” (those who engaged strictly in butchering and meat
preparation).
The response options for the scale items were coded as follows:
A: 1 = Don’t like it (not in favour); 2 = Somewhat negative (a few problems with it); 3 =
No opinion; 4 = Valuable; 5 = Very valuable
B: 1 = Not important; 2 = Somewhat important; 3 = No opinion; 4 = Important; 5 = Very
Important
C: 1 = Totally useless (none needed); 2 = Somewhat useless; 3 = No opinion; 4 =
Valuable (it added info); 5 = Very valuable (definitely needed/useful)
D: 1 = No concern (would collect FP samples for sure; would take a kit/s on hunts); 2 =
Minor concern (e.g., cold hands, but would still take FP kits on hunts); 3 = No opinion; 4
= Major concern (some problems with collecting FP samples; partly why wouldn’t take
FP kits on hunts); 5 = The main concern (my problems with collecting FP samples are the
main reason I didn’t take FP kits on hunts)
E: 1 = Very difficult; 2 = Somewhat difficult; 3 = No opinion; 4 = Easy; 5 = Very easy
302
Paraphrased Interview Item Response Code APPENDIX 6E (cont’d)
Scale Responses n (%) 1 2 3 4 5 SECTION II Filter‐paper (FP) Blood Sampling and Hunter‐based Wildlife Disease Sampling 1. Had you heard of the FP blood sampling method before this interview? (n=16) 2 (13%) 14 (87%) A 0 1 (6%) 0 For Hunters Only: 3. Have you done any wildlife sampling in the past? (n=14) 11 (79%) 3 (21%) 4. Are you currently interested in doing wildlife sampling? (n=14) 13 (93%) 1 (7%) B 1 (7%) 2 (14%) 1 (7%) 13 (93%) 1 (7%) 5. In your view, how important is hunter‐based wildlife sampling? (n=14) 6. Have you heard of or been involved in any organized program/s of hunter‐based wildlife sampling? (n=14) No 2. Based on the FP collection method demonstrated in the interview, and assuming FP blood samples can be tested the same as regular blood samples, what is your opinion of FP sampling for caribou disease monitoring? (n=16) Yes 10 5 (63%) (31%) 7 3 (50%) (22%) Code A: 1 = Don’t like it (not in favour); 2 = Somewhat negative (a few problems with it); 3 = No opinion; 4 = Valuable; 5 = Very valuable Code B: 1 = Not important; 2 = Somewhat important; 3 = No opinion; 4 = Important; 5 = Very Important 303
Paraphrased Interview Item Response Code APPENDIX 6E (cont’d)
Scale Responses n (%) 1 2 3 4 5 Yes No SECTION II (cont’d) Filter‐paper (FP) Blood Sampling and Hunter‐based Wildlife Disease Sampling 7. If Yes, was FP sampling part of this/these program/s? (n=13) 2 (15%) a
11 (85%) 8. Has anyone trained/shown you how to collect blood on FP before? (n=14) 2 (14%) a
12 (86%) 9. If Yes, what is your opinion of the FP training you received? (n=2) C 0 0 0 10. Did you go caribou hunting this past season? (n=14) 5 (36%) 9 (64%) 11. The Government of Nunavut Caribou Health Monitoring Program has hunter sampling kits available in communities. If you hunted caribou last season, did you take any of these kits with you? b
(n=3) 0 3 (100%) 1 1 (50%) (50%) Code C: 1 = Totally useless (none needed); 2 = Somewhat useless; 3 = No opinion; 4 = Valuable (it added info); 5 = Very valuable (definitely needed/useful) a Government of Nunavut Caribou Health Monitoring Program initiated in 2008 b
At time of interview, kits were only available to hunters in North Baffin Region communities (Pond Inlet, Arctic Bay, Clyde River). Only three hunter interviewees from North Baffin had hunted caribou during the past season 304
Paraphrased Interview Item Response Code APPENDIX 6E (cont’d)
Scale Responses n (%) 1 2 3 4 5 Yes No SECTION II (cont’d) Filter‐paper (FP) Blood Sampling and Hunter‐based Wildlife Disease Sampling For All Interviewees (Hunters and Non‐hunters): 12. If you did not take a hunter sampling kit/s with you, how did concerns about FP blood collection affect your decision b
not to take kits? (n=3) D 1 1 (33%) (33%) 1 (33%) 0 0 13. If you used a hunter sampling kit/s this past season, how was the FP collection process? (n=0) E 0 0 0 0 0 14. Do you have any concerns
about collecting FP blood
samples from caribou (i.e., the actual sampling steps, or taking blood from a harvested animal,
or any other aspects)? (n=16)
2 (13%) 14
(87%)
Code D: 1 = No concern (would collect FP samples for sure; would take a kit/s on hunts); 2 = Minor concern (e.g., cold hands, but would still take FP kits on hunts); 3 = No opinion; 4 = Major concern (some problems with collecting FP samples; partly why wouldn’t take FP kits on hunts); 5 = The main concern (my problems with collecting FP samples are the main reason I didn’t take FP kits on hunts) Code E: 1 = Very difficult; 2 = Somewhat difficult; 3 = No opinion; 4 = Easy; 5 = Very easy 305
APPENDIX 6F
Summary of results for the Section II items of the Sahtu interviews (n=9) that had
quantifiable responses (i.e., visual analog scale 1 through 5, or Yes/No answers). The
harvesters interviewed included “Hunters” (those who actually hunted animals in addition
to activities such as butchering etc.) and “Non-hunters” (those who engaged strictly in
butchering and meat preparation).
The response options for the scale items were coded as follows:
A: 1 = Not important; 2 = Somewhat important; 3 = No opinion; 4 = Important; 5 = Very
Important
B: 1 = Totally useless (none needed); 2 = Somewhat useless; 3 = No opinion; 4 =
Valuable (it added info); 5 = Very valuable (definitely needed/useful)
C: 1 = No concern (would collect FP samples for sure; would take a kit/s on hunts); 2 =
Minor concern (e.g., cold hands, but would still take FP kits on hunts); 3 = No opinion; 4
= Major concern (some problems with collecting FP samples; partly why wouldn’t take
FP kits on hunts); 5 = The main concern (my problems with collecting FP samples are the
main reason I didn’t take FP kits on hunts)
D: 1 = Very difficult; 2 = Somewhat difficult; 3 = No opinion; 4 = Easy; 5 = Very easy
306
Paraphrased Interview Item
Response Code APPENDIX 6F (cont’d)
Scale Responses n (%) 1 2 3 4 5
Yes No SECTION II Filter‐paper (FP) Blood Sampling and Hunter‐based Wildlife Disease Sampling 1. Had you heard of the FP blood sampling method before this interview? (n=9) For Hunters Only: 2. Have you done any wildlife sampling in the past? (n=8) 3. Are you currently interested in doing wildlife sampling? (n=8) 4. In your view, how important is hunter‐based wildlife sampling? (n=8) 5. Have you heard of or been involved in any organized program/s of hunter‐based wildlife sampling? (n=8) 4 (44%) 5 (56%) 5 (63%) 3 (37%) 5 (63%) a
3 (37%) 1 (13%) 0 2 (25%) 8 (100%) 0 b
4 (50%) A 2 3 (25%) (37%) 6. If Yes, was FP sampling part of this/these program/s? (n=8) 4 (50%) Code A: 1 = Not important; 2 = Somewhat important; 3 = No opinion; 4 = Important; 5 = Very Important a Three of the five hunters with past experience in wildlife sampling were still interested; two hunters with no prior experience in wildlife sampling were interested. b
The Sahtu Wildlife Health Monitoring Program provided hunter sampling kits in communities; three hunter interviewees were trained as “Wildlife Health Monitors” in the program and had hunter kits available to them; only two of these individuals hunted in the season prior to the interview. 307
Paraphrased Interview Item
Response Code APPENDIX 6F (cont’d)
Scale Responses n (%) 1 2 3 4 5
Yes No 3 (37%) b
5 (63%) SECTION II (cont’d) Filter‐paper (FP) Blood Sampling and Hunter‐based Wildlife Disease Sampling 7. Has anyone trained/shown you how to collect blood on FP before? (n=8) 8. If Yes, what is your opinion of the FP training you received? (n=3) B 0 0 9. Did you go caribou hunting this past season? (n=8) 1 (33%) 10. For hunters trained in the Wildlife Health Monitoring b
Program : If you went caribou (or moose) hunting this past year, did you take any hunter sampling kits with you? (n=2) 11. If you did not take a hunter sampling kit/s with you, how did concerns about FP blood collection affect your decision not to take kits? (n=1) C 1 (100%) 0 1 1 (33%) (33%) 5 (63%) 3 (37%) 1 (50%) 1 (50%) 0 0 0 Code B: 1 = Totally useless (none needed); 2 = Somewhat useless; 3 = No opinion; 4 = Valuable (it added info); 5 = Very valuable (definitely needed/useful) Code C: 1 = No concern (would collect FP samples for sure; would take a kit/s on hunts); 2 = Minor concern (e.g., cold hands, but would still take FP kits on hunts); 3 = No opinion; 4 = Major concern (some problems with collecting FP samples; partly why wouldn’t take FP kits on hunts); 5 = The main concern (my problems with collecting FP samples are the main reason I didn’t take FP kits on hunts)
b
The Sahtu Wildlife Health Monitoring Program provided hunter sampling kits in communities; three hunter interviewees were trained as “Wildlife Health Monitors” in the program and had hunter kits available to them; only two of these individuals hunted in the season prior to the interview. 308
Paraphrased Interview Item Response Code APPENDIX 6F (cont’d)
Scale Responses n (%) 1 2 3 4 5 Yes No SECTION II (cont’d) Filter‐paper (FP) Blood Sampling and Hunter‐based Wildlife Disease Sampling 12. If you used a hunter sampling kit/s this past season, how was the FP collection process? (n=1) D 0 0 0 0 1 (100%) For All Interviewees (Hunters and Non‐hunters): 13. Do you have any concerns about collecting FP blood samples from caribou or moose (i.e., the actual sampling steps, or taking blood from a harvested animal, or any other aspects)? (n=9) 2 (22%) 7 (78%) Code D: 1 = Very difficult; 2 = Somewhat difficult; 3 = No opinion; 4 = Easy; 5 = Very easy