Noninvasive Collection of Saliva in Panthera leo : Creation and Validation of a Novel Technique for Health Assessment in Captive African Lions (Panthera leo)

Full text

Antioch University
AU - Antioch University Repository and Archive
*33&24"4*/.3;&3&3 45%&.4,5-.*$)/,"23)*0*.$,5%*.(
*33&24"4*/.3;&3&3

Noninvasive Collection of Saliva in Panthera leo :
Creation and Validation of a Novel Technique for
Health Assessment in Captive African Lions
(Panthera leo)
Elizabeth Sgambelluri
Antioch University, New England
/,,/74)*3".%"%%*4*/.",7/2+3"4 )<03"52"".4*/$)&%5&4%3
"24/'4)& */,/(8/--/.3".%4)& .6*2/.-&.4",45%*&3/--/.3
;*3;&3*3*3#2/5()44/8/5'/2'2&&".%/0&."$$&33#84)&45%&.4,5-.*$)/,"23)*0*.$,5%*.(*33&24"4*/.3;&3&3"4 :.4*/$)
.*6&23*48&0/3*4/28".%2$)*6&4)"3#&&."$$&04&%'/2*.$,53*/.*.*33&24"4*/.3;&3&3#8"."54)/2*9&%"%-*.*342"4/2/' :.4*/$)
.*6&23*48&0/3*4/28".%2$)*6&/2-/2&*.'/2-"4*/.0,&"3&$/.4"$4 %0&.2/3&".4*/$)&%57-$(2"4)".4*/$)&%5
&$/--&.%&%*4"4*/.
("-#&,,52*,*9"#&4)/.*.6"3*6&/,,&$4*/./'",*6"*.".4)&2",&/2&"4*/.".%!",*%"4*/./'"/6&,&$).*15&'/2&",4)
33&33-&.4*."04*6&'2*$".*/.3".4)&2",&/ Dissertations & eses
)<03"52"".4*/$)&%5&4%3

Department of Environmental Studies
THESIS COMMITTEE PAGE
The undersigned have examined the thesis entitled:
Noninvasive Collection of Saliva in Panthera leo
Creation and Validation of a Novel Technique for Health Assessment in Captive African Lions
(Panthera leo)
presented by: Elizabeth Sgambelluri
candidate for the degree of Master of Science, and hereby certify that it is accepted.*
Committee chair name: Beth Kaplin, Ph.D.
Title/Affiliation: Antioch University New England, Environmental Studies Department
Committee member name: Lisabeth Willey, Ph.D.
Title/Affiliation: Antioch University New England, Environmental Studies Department
Committee member name: Michelle Kneeland, DVM
Title/Affiliation: Biodiversity Research Institute, Wildlife Health Program Director
Date Approved by all committee members:
*Signatures are on file with the Registrar’s Office at Antioch University New England.

Noninvasive Collection of Saliva in Panthera leo
Creation and Validation of a Novel Technique for Health Assessment in Captive African
Lions (Panthera leo)
Elizabeth Sgambelluri, M.S., PGCert., B.S
Master’s Candidate, Conservation Biology
Department of Environmental Studies
May 6, 2018
Under the Supervision of:
Beth Kaplin, PhD., Thesis Advisor, Antioch University New England
Lisabeth Willey, PhD., Antioch University New England
Michelle Kneeland, DVM, Biodiversity Research Institute

2
Table of Contents
LIST OF FIGURES iii
LIST OF TABLES iv
ACKNOWLEDGEMENTS v
ABSTRACT 7
CHAPTER 1. INTRODUCTION 8
CHAPTER 2. Literature Review 10
LITERATURE REVIEW …………………………….……………………………….... 10
Major Stressors for Wild / Free-Range African Predators……....................……. 10
Major Stressors for Captive African Predators .…………...........................…….. 14
Evidence of Stress Induced Immunosuppression …………….….....………........ 16
CHAPTER 3. Creation of a Novel Technique for Saliva Collection 23
INTRODUCTION ……………….…………………..…………………………………. 23
Goals and Objectives ………………………………..……………..…………… 24
METHODS ………………………………………………………….………………….. 25
Study Location and Individuals ……………………..……………..…………… 25
Experimental Design and Data Collection ...……………………………..…...… 26
Data Analysis………………………………………..……………………..…… 30
Statistical Analysis………………………………………...…………….……… 31
RESULTS ……………………..…………………………………….………………….. 32
DISCUSSION ………………………………...…………………….………………….. 46
CHAPTER 4. Conclusions and Recommendations 49
REFERENCES……………………………..…..……………………………………………… 52
APPENICES …...………………..…………..………………………………………………… 56
APPENDIX 1: Collection Protocols (pg. 57)
APPENDIX 2: Device Design and Outer Shell (pg. 59)
APPENDIX 3: Metrics of Success (pg. 60)
APPENDIX 4: Library Permission Slip (pg. 61)
ii
APPENDIX 5: Patent Filing Form and Description (pg. 62)

3
List of Figures
Figure 1. Bucket in a Bucket Structure of Device …………….................…………………....... 26
Figure 2. Time Budget of Total Behaviors for Each Individual During Trial #1 …………....... 34
Figure 3. Time Budget Plots: Trial #1 ………….…………….................…………………....... 39
Figure 4. Time Budget Plots: Trial #3 ………….…………….................…………………....... 40
iii

4
List of Tables
Table 1. Degrees of Success and Determinants ...............................………………………....... 31
Table 2. Working Ethogram Generated for African Lions ..............………………………....... 36
Table 3. Time Budget Summaries: Trial #1 ……...............................…………………..…....... 37
Table 4. Time Budget Summaries: Trial #3 ……...............................…………………..…....... 38
Table 5. Full Ethogram for Male Lion During Trial #1 ...............................………………....... 41
Table 6. Trial Outcomes and Variable Analysis for Success Evaluation ..…………………....... 45
iv

5
ACKNOWLEDGEMENTS
I would like to thank all the individuals that contributed to this study and provided
mountains of support. First and foremost, I would like to thank my thesis advisor, Beth Kaplin,
and committee for their enduring support, guidance, and stress relief. This group of wonderful
human beings helped keep my head above water during the most trying of times in this study and
were extremely motivational throughout the process. In particular, I would like to thank Lisabeth
Willey for always being there when I was having a meltdown, or simply needed reassurance in my
own potential. You all helped fuel this study, but more importantly you all helped me to stay
confident in my potential as a researcher and devotion to my aspirations.
A huge thanks is rightly due to my family, whom have always supported my academic
endeavors and dreams for the future. Thank you to my amazing father, Richard Sgambelluri, who
constructed the beautiful outer shell of the device in his free time. I cannot express in words how
grateful I am, and how appreciative I am for all the help and advice you provided regarding device
design and construction. I also am forever grateful for your confidence in me, and the comedic
relief you provided. To my sister, Jessica, and my mother, Carol: thank you for listening to me
whenever I hit a roadblock or was at my maximal frustration threshold. Thank you for keeping me
grounded, motivated, and bright minded. Thank you also to Figaro, my cat, for the snuggle breaks
and gentle reminders when it was time to stop for dinner.
To Tristan Anderson, a recent senior from the Engineering department at Keene State
College. This amazing, brilliant minded individual enabled what had started as a vision and
drawing on a piece of paper to become reality. I cannot thank him enough for his dedication to this
project, like the time he looked up the average teeth measurements and jaw dimensions for lions
v

6
to help guide the dimensions of the grating. He took the time out of his extremely busy schedule
to help design and build the internal collection structure of this device. Tristan, if it was not for
you, I would probably still be sitting in my basement with plastic funnels and duct tape! I wish
you the best for whatever you pursue in the future, and know that any engineering firm is blessed
to have you on their team. Thank you for your dedication, your sense of wonder, your contagious
enthusiasm, and work ethic.
Thank you to the staff – specifically Adrienne Whiteley – and community at Rosamond-
Gifford Zoo, for permission to work with your two lions and your support in this project. This
community served a crucial role in this research, providing the means to test the device in a captive
setting. Your willingness to support this project, and support novel research when many others did
not, truly meant the world and such should not go unnoticed.
Lastly, I would like to thank all of the individuals that contributed financially to the
GoFundMe page that supported the entirety of expenses for this project. Your faith in this project
was inspirational and served as a huge proponent in allowing this research to occur. I would also
like to give thanks to all of the members of the Antioch community, faculty and colleagues alike,
as well as my boyfriend Travis Smith, who offered reassurance, self-confidence, support,
guidance, and – perhaps most importantly – many laughs.

7
ABSTRACT
Very little research exists concerning direct correlations between alterations in stress
stimuli and direct effects on carnivore physiology. Many studies have researched and documented
what factors correlate to increased stress levels in predator species. Lacking, however, are studies
observing the specific physiologic influences of increased glucocorticoid concentrations in those
individuals. Furthermore, methods for monitoring and assessing health or immune function in
many carnivore species often involve invasive techniques, such as physical restraint, blood
sampling, and chemical immobilization. This study aimed to create, evaluate, and describe a new
device for noninvasive collection of saliva for health assessment. The device itself was constructed
on a “bucket in a bucket” concept, wherein saliva was deposited into a smaller, inner container
and attractants housed in the larger, outer container. Two attractants, catnip extract and bear bait,
were placed inside the device for elicitation of salivation and passive drooling response. Efficacy
of the device, as well as evidence regarding its accuracy and safety, were evaluated through testing
trials conducted at the Rosamond-Gifford Zoo in Syracuse, NY (USA) using two captive lions.
Success was determined and scored using several metrics, including the ability to elicit behaviors,
whether or not lion(s) approached the device, and whether or not saliva was collected. Three trials
were conducted, with two of the three yielding a success value greater than three (maximal score
being four). Saliva was successfully deposited and collected in the device during the first trial.
Urine and hair samples were also collected using the device. The results obtained substantiate that
such can be utilized in collection of saliva. Moreover, they highlight the potential of this device in
reforming or providing an alternative approach to current techniques, while providing a
springboard for the use of saliva in wildlife health assessment.
Disclosure: Provisional patent pending for device (Ser. No 62/720,961) by Wendy W. Koba,
Esq. (see Appendix 5)

8
CHAPTER 1
INTRODUCTION
Stress responses are controlled by the hypothalamo-pituitary-adrenal axis (HPA). This axis
comprises the hypothalamus, pituitary gland, and adrenal gland (Chrousos, 2013). When
vertebrates are exposed to stress factors, large amounts of glucocorticoids (GC) are released into
the body system (Busch, 2009). In mammals, the glucocorticoid hormone released is cortisol. The
degree or magnitude of the GC response differs according to stimuli and duration of exposure
(Sapolsky, 2002; Romero, 2004; Busch, 2009). A stress response is a natural process; it is the
mechanism through which animals evolved for survival. However, that applies only to short term.
Chronic exposure to stress stimuli causes overproductions of glucocorticoids that can actually
impair health (Randall, 2012).
The notion of health relates to the vitality/vigor of an individual and/or population, and is
tied directly to production of glycoprotein molecules called immunoglobulins (French, 2012).
Immunoglobulins function in production of antibodies within body systems. Mammalian systems
rely on these antibodies, as well as a specialized suite of immune proteins called lymphocytes, for
detection and defense against deleterious pathogens. Studies confirm, however, that cortisol and
other glucorticoids inhibit this protein, consequently suppressing the body’s immune system,
increasing susceptibility to infections (Hansen, 2014), and exacerbating severity of infections and
inflammation.
Various external and internal factors – such as competition, predation, and habitat change
– influence physiological systems in large carnivores. There is a dearth, however, in research
regarding how alterations in stress stimuli specifically could be detrimental to individual
physiology and population health. In the following section I will review the existing research on

9
sources of stress in wildlife, focusing on African predator species, and propose several sources of
evidence to explore the immunological impacts of stress.

10
CHAPTER 2
LITERATURE REVIEW
This review addresses the link between stress and immunosuppression, as well as the deficit in
noninvasive methods for assessing health. Recent evidence from human medicine and laboratory
studies suggests that stress influences neuroendorcrine mechanisms that catalyze
immunosuppression (Dohms et al., 1991; Glaser et al., 1987; Råberg et al., 1998; Roth, 1985). In
the sections that follow I provide background on two major stressors cited for African predators,
how stress influences body systems, and both human and wildlife based evidence for stress induced
immunosuppression. This section is organized into two principal parts: (1) major stressors in
African predators, and (2) evidence of stress induced immunosuppression.
Major Stressors for Wild / Free-Range African Predators
A stressor is any adverse factor (biotic or abiotic) that disrupts normal body system processes
(Sapolsky, 2002). When exposed to stress stimuli, individuals will exhibit stress responses, which
are various behavioral and/or hormonal changes evolutionary adapted for survival and re-
establishment of internal homeostasis (Aich et al., 2009; Young et al., 2004). The adrenal gland,
part of the endocrine system, governs this response through release of two hormones: the core
releases epinephrine, or adrenaline, whereas glucocorticoids are expelled by the adrenal cortex
(Sapolsky, 2002; Busch, 2009).
Anthropogenic Stressors
Over the past 20 years, populations of African lions have been reduced by at least 30 percent
(Rabinowitz, 2010), and current population size is estimated somewhere within the range of 16,500

11
to 30,000 individuals. This estimate is half of that quantified during the 1950s (Hayward et al.,
2005). The demise of this predator has largely been driven by anthropogenic activities such as
poaching. Increased human development has aggravated fatal competition with hyena populations,
as well as disease due to population overlap with other species (i.e. olive baboons and African wild
dogs). Between 1993 and 1997, over one thousand lions were estimated to have died from
outbreaks of canine distemper, a disease transmitted from wild dog species through spotted hyena
populations (Hayward et al., 2007).
Anthropogenic activities have often served as sources of wildlife stress. In one study, stress
responses in lions were negatively correlated with proximity to human settlements (Creel et al.,
2013). In this study, glucocorticoid levels were measured through immunoassay surveys on lion
fecal samples to evaluate stress levels. Results indicated that lions within human settlement study
zones had a 25% increase in fecal glucocorticoid concentration, as compared to those measured
from lions inhabiting conservation areas (Creel et al., 2013). As distance from settlement zones
increased, glucocorticoid metabolite concentrations within fecal samples declined by 22% per
kilometer (Creel et al., 2013). Glucocorticoid concentrations also increased if lions were unable to
move away from settlement zones (Creel et al., 2013). Observations on movement patterns,
through direct observation and radio telemetry, revealed that lions shift occupancy and utilize other
habitats to evade human interaction (Creel et al., 2013). Analogous findings were established in a
similar study with spotted hyenas inhabiting the Masai Mara National Reserve in Kenya. Fecal
glucocorticoids from both males and females in four social groups were used to delineate major
stressors (Van Meter et al., 2009). Interestingly, ecological stressors such as climate factors and
prey availability did not influence stress levels. Instead significant increases in glucocorticoid
concentrations correlated with increases in density of human populations bordering the

12
conservation area (Van Meter et al., 2009). Samples further revealed that agricultural practices
augmented stress levels in hyenas (Van Meter et al., 2009). All of these findings provide strong
evidence towards anthropogenic activities as major stressors.
Interspecific Competition
Interspecific killing
Accounting for roughly 70% of mortalities in some carnivore populations (Palomares et al., 1999),
interspecific killing typically involves large carnivorous predators purposefully hunting and
eliminating smaller carnivores (Palomares et al., 1999). Interspecific killing can be observed
within Africa’s large carnivore guild, where repeated incidents of deleterious interactions between
lions, hyenas, cheetahs, and African wild dogs have been documented (Caro et al., 2003; Creel et
al., 1996; Laurenson 1995; Palomares et al., 1999).
Lions and hyenas compete with each other for food due to high dietary overlap (Mills,
2005; Trinkle, 2005). While both are adept hunters, they also act as opportunists or scavengers.
Spotted hyenas are notorious for stealing carcasses acquired by lions. In Uganda, studies from
Rwenzori National Park have indicated that 33% of lion kills were stolen by hyenas (Palomares et
al., 1999; Trinkle, 2005) through the use of mobbing behavior. This behavior strategy causes
hyenas to aggressively attack and taunt lions in an attempt to drive them away (Hayward et al.,
2007; Palomares et al., 1999). In Saviti, Botswana, lionesses were mobbed at a kill site and chased
up trees by hyenas (Joubert, 1992; Nat Geo Wild, 2015). Instances such as this are fueled by food
competition. However, accounts of hostile interactions and randomized attacks between both
species have been documented that are not food related. Lionesses whose young were killed by
hyenas act most aggressively towards hyenas (Joubert, 1992). In the Serengeti, male lions have

13
sought and killed female matriarch hyenas when food was not concerned (Joubert, 1992; Nat Geo
Wild, 2015). Approximately 71% of hyena mortality was linked to lion predation within the Etosha
National Park (Trinkle, 2005).
There may be another purpose for hyena mob behavior: offspring protection. Though the
individual fitness is at risk, by engaging in aggressive behavior that may drive away predator
species, a social group’s inclusive fitness is preserved by protecting the offspring (kin selection
view) (Trinkle et al., 2005). This notion arises from observation that all mobbing events
documented in Ngorongoro Crater where in close proximity to hyena dens housing pups (Trinkle
et al., 2005). In these particular instances, mobbing not only drove away lions, but lions avoided
those regions for several weeks (Trinkle et al., 2005).
Similar dispersion patterns of these two species occurred in Addo Elephant National Park,
South Africa. When lion populations traveled to and inhabited northern regions of the park area,
hyenas would move divergently to inhabit southern regions (Hayward, 2007). Such behavior is
reflective of avoidance techniques described in the following section.
Avoidance behavior
Species react in different ways to counteract interspecific competition or predation, and often
display adaptive behaviors. In the savannahs of Africa, many of the carnivore guild species – lions,
hyenas, cheetahs, leopards, and African wild dogs – have modified behavioral patterns or altered
movement in order to avoid interaction with other predators (Valeix et al., 2012). Some species
migrate to new habitats to avoid confrontations (Palomares et al., 1999).
Cheetahs illustrate excellent examples of such predator induced adaptations. Interspecific
competition by lions serves as a large limiting agent in cheetah densities (Kelly et al., 1998;

14
Laurenson, 1995; Palomares et al., 1999). Studies have documented negative correlations between
cheetah reproductive outputs and proximity to lions – the greater the proximity to lions, the lower
the reproductive success of cheetah individuals (Kelly et al., 1998; Palomares et al., 1999). In the
Serengeti plains, this interaction contributed huge impacts on the population dynamics of cheetahs,
specifically, cub mortality dramatically increased with lion predation (Kelly et al., 1998;
Woodroffe et al., pg 165). In the Serengeti, lions accounted for 70% of overall cub mortality in
cheetah populations (Kelly et al., 1998; Laurenson, 1995). By adapting avoidance strategies and
modifying activity patterns, cheetahs are able to maintain or increase relative fitness (Palomares
et al., 1999; Ritchie, 2009). Research in Kruger National Park showed that cheetahs modified
hunting periods under pressures of interspecific competition. They changed to hunting during mid-
day hours to avoid kleptoparasitism by lions and hyenas (Palomares et al., 1998).
Adjustments in activity patterns to avoid larger predators can also be observed in African
wild dogs (Creel et al., 1996). In Hluhluwe-imfolozi National Park, South Africa, African wild
dogs deliberately altered movement patterns and modified hunt times to avoid lions (Darnell,
2014). As referenced previously, altered habitat use patterns by lions and hyenas in Addo Park,
South Africa, could be the results of avoidance behaviors. Several lionesses were observed to alter
activity patterns to become completely diurnal, with peak hunts occurring around sunrise
(Hayward, 2007). Such behavioral changes are believed to represent an avoidance method in
response to mobbing by spotted hyenas (Hayward, 2007).
Major Stressors for Captive African Predators
Discussions concerning the ethics and efficacy of maintaining captive populations are often
controversial, largely in regards to animal welfare. The purpose of captive environments, such as
those within zoological settings, is three-fold: (1) to provide a means of preserving threatened or

15
endangered species, (2) rehabilitation of orphaned or ill individuals, and (3) to increase public
education of species conservation. Some additionally strive to maintain healthy populations that
will ultimately be reintroduced into the wild. Unfortunately, the reality of the situation is that many
captive environments fail to provide the optimal conditions required for species flourishment
(Sajjad et al., 2011), thus serving more as a detriment rather than benefit in scope. Such failures
include lack of natural habitat, confined enclosures, noise levels, and overloaded human exposure.
Maintenance of healthy and sustainable captive populations is largely thwarted by stress,
and thus serves as the principle concern for such organizations (Conforti et al., 2012). Many
stressors stem from those failures described previously. However, the type and number of stressors
vary across species and captive setting. For captive felids, the primary stressors are chemical
restraint, translocation, and inadequate enclosures. Clouded leopards that were housed in confined,
size restricted enclosures exhibited higher fecal corticoid metabolites (FCMs) than those in large
enclosures (Wielebnowski et al., 2002). For captive cheetahs, translocation and public exposure
serve as the primary stressors. Individuals moved “on-exhibit” (housed in an enclosure space for
public viewing) were 20 times more likely to have elevated corticoid levels that were at least two
standard deviations from their baseline, compared to those that were “off-exhibit” (Franklin, 2014;
Munson et al., 2005; Terio et al., 2004; Wells et al., 2004). These elevated levels were not transient
in nature; in fact, these levels remained elevated on average for 30 days (Franklin, 2014; Wells et
al., 2004). A study by Nogueira et al (1997) found a strong correlation between stress (increased
cortisol level) and restraint in Panthera onca and Felis pardalis. Felines that were physically
restrained exhibited plasma cortisol levels that were, on average, four to ten times greater than
those that were not restrained (333 ± 47 nmol/l compared to 35 nmol/l and 87 ± 16 nmol/l;
Nogueira et al., 1997).

16
Numerous studies document behaviors believed to be representative of stress – known as
“stereotypies” – in many captive taxa (Vaz et al., 2017). Such behaviors include pacing, head
bobbing, anorexia, variations in skull morphology (Duckler, 1998), and excessive grooming.
However, such do not serve as a definitive indicator of increased stress (such could rather be a
factor of boredom), and fail to provide explicit quantitative measure of variations in stress level.
The behavioral adaptations previously described all stem from avoidance of or exposure to
a deleterious factor. While these factors trigger behavioral modifications or adaptations to
minimize costs and risks (Adamo et al., 2013), they also trigger a stress response. Exposure and
adaptation to stress factors impacts body system functions in several ways. In small doses slight
alterations in body systems may occur, though innocuous short term (Adamo et al., 2013). What
happens, however, when African predators are exposed to multiple stress sources on a chronic
basis?
Evidence of Stress Induced Immunosuppression
Stress Physiology in Humans
In humans, stress responses are known to suppress immune function and pathogen defense
(Sapolsky, 2002; Chrousos, 2013). This occurs largely through impairment of immunoglobulin
production. Immunoglobulins function in the production of antibodies within body systems. There
are five types found in all organisms: IgA, IgG, IgM, IgE, and IgD. Important to this study are
IgA, as such is indicative of both immune health and susceptibility. IgA antibodies protect body
surfaces that are exposed to foreign substances, and assists with immune function of mucosal
membranes. They are found in the intestinal tract, breathing passageways (e.g. nose and mouth),
and blood (eBioscience, 2015).

17
Chronic stress not only compromises health through immunosuppression, but through
impairment of multiple organ systems. For example, overproduction of glucocorticoids catalyzes
muscle mass attrition and bone marrow loss through calcium reductions (Young et al., 2004).
Cortisols can act as atherogenics, meaning they increase plaque in both coronary and cerebral
arteries (Dr. Sgambelluri, MD). Thus, stress increases risk for hypertension, osteoporosis, and
Cushing’s syndrome (Chrousos, 2013; Young et al., 2004) in humans. If these health detriments
are present in humans, there is a possibility that similar reactions manifest in other mammalian
species.
Evidence in Free-Ranging African Carnivore Populations
There is some evidence that stress in African carnivores correlates with
immunosuppression and disease susceptibility. Infections are typically asymptomatic and common
to African carnivore species (Williams et al., 2014; Goller, 2011). However, they can be
pathogenic under certain circumstances (e.g. unnatural hosts or habitat degradation) (Williams et
al., 2014; Goller, 2011). For example, Babesia leo, a form of babesiosis in lions that is usually
tolerated, will become pathogenic if an individual is immunocompromised (Williams et al., 2014).
Another similar example is with the protozoan Hepatozoon (Williams et al., 2014; Goller, 2011).
Adults of many African carnivorous species, such as spotted hyenas, display high prevalence
(>90%) of this infection with no clinical signs (Williams et al., 2014; Goller, 2011). However,
Hepatozoon was the cause of mortality in 18% of infected juvenile spotted hyenas. (Aguirre et al.,
2012; Williams et al., 2014). I believe it may be a function of stress induced by sibling rivalry.
There is also evidence that stress may increase susceptibility to parasitic infections.
Hookworms are common extracellular gastrointestinal parasites of mammals. Spotted hyenas are
known to be infected with two hookworm species called Cystoisospora and Ancylostoma (East et

18
al., 2013). Lactating females were found to have higher infections of these hookworms than non-
lactating females (East et al., 2013). Furthermore, infections also increased in females nursing
twins rather than singletons (East et al., 2013).
Cheetahs provide an interesting example of how stress may correlate with
immunosuppression. A study observing stress impacts on captive breed cheetahs indicted that
feline panleukopesvirus (FPV) and canine parvovirus were most prevalent in these individuals
during times of stress (Munson et al., 2004). It also showed that even with vaccinations, captive
populations were highly susceptible to infections (Munson et al., 2004) - indicated by the
consistently elevated corticoid levels under unmodified captive care (Munson et al., 2005; Munson
et al., 2004).
The negative impacts of stress induced immunosuppression extend beyond an individual
scope, in that it can harm the health of the population. Stress induced disease from intensive
handling and vaccination against canine distemper caused massive declines in wild dog
populations between 1965 and 1991 (Goller, 2011). Mycobacterium bovis, a strain of tuberculosis
or bovine virus, can go into dormancy and later be reactivated in a host (Drewe et al., 2009). In
the Kalahari, dormant bovine viruses were reactivated in male meerkats due to stress from
aggressive relations and finding mates (Drewe, 2009; Drewe et al., 2009). This event was
responsible for the extinction of four social groups between 1995 and 2005.
Evidence in Captive Populations
Evidence of immunosuppression becomes increasingly apparent in captive populations.
Captive cheetah populations often fail to thrive, exhibiting increased disease prevalence and
diminished reproductive success (Terio et al., 2004). Interestingly, many of the diseases that

19
manifest in captive populations are either absent or avirulent in free ranging individuals (Terio et
al., 2004). Furthermore, captive individuals exhibit atypical immune responses to common
pathogens, divulged by recurrent viral infections and high mortality associated with degenerative
and/or inflammatory diseases (i.e., glomerulosclerosis and amyloidosis). In 2004, Terio et al. used
fecal corticoid metabolites to evaluate degrees of stress between captive and free ranging cheetah
populations. Baseline corticoid levels were significantly higher in the fecal samples of captive
individuals then free-range (Terio et al., 2004). Samples from midsagittal region of adrenal glands
indicated that corticomedullary ratios were significantly larger in captive individuals then wild,
with prolonged elevations in corticosteroid concentrations (Munson et al., 2005). A study by
Munson et al. (2005) showed the presence of severe, multiple organ lesions in captive cheetahs
with elevated plasma cortisol levels – dissimilar to the wild counterparts. A large majority of the
individuals with these lesions also presented cardiac fibrosis, splenic lymphoid depletion, and
glomerulosclerosis (Munson et al., 2005). All findings were correlated with significant elevations
in adrenal cortical levels, indicating a high prevalence of adrenal cortical hyperplasia in the captive
populations (Munson et al., 2005). These findings provide concrete evidence for the physiologic
influences of stress on the health of these individuals.
Discrepancies in disease virulence between captive and free-range populations are also
apparent in African lions (Panthera leo). In the wild, many lions are carriers for different subtypes
of feline immunodeficiency virus (FIV), though the lentiviral strains are relatively nonpathogenic
or seemingly latent (Hoover et al., 1991; Roelke et al., 2009). Synonymous with trends reflected
in cheetah populations, pathogenesis of FIV specific to African lions (FIV-Ple or LLV) will alter
under certain conditions (Roelke et al., 2009). For example, the lentivirus will become

20
pathogenic/virulent during periods of environmental stress, in captive populations, or when the
individual has an underlying immunologic or neurologic disorder.
Current Methods for Health Assessment in Wildlife
Modern methods of health assessment in both captive and wild populations vary,
depending on the specific variables targeted by the research for analysis. For evaluation of stress,
most studies now employ the use of fecal samples and metabolites; a preferable, noninvasive
alternative. When it comes to monitoring and assessing health or immune function in carnivore
species, however, the techniques employed are far more invasive. These techniques include
physical restraint, blood sampling, and chemical immobilization. While these practices enable
acquisition of body fluids for health examination, they carry a high degree of risk for both the
researcher and the animal.
The majority of the time blood is collected and used for health analysis. However, this is
not the only medium that yields information relative to the health of an animal. Both urine and
saliva can provide a wealth of information, ranging from health indicators to reproductive function
to genetics (Chiappin et al., 2007; Laudenslager et al., 2006). Two variables in particular that can be
evaluated through saliva as health measures are secretory immunoglobin A (IgA) and C-Reactive
protein (CRP) levels. Secretory immunoglobulin A (SIgA) is a type of IgA antibody found in
mucous secretions that prevents pathogen invasion in oral, lung, and gut systems. C-Reactive
Protein (CRP) is a protein produced in the liver that indicates inflammation.
Again, a significant amount of information can be gleaned from saliva and/or urine.
However, at the current time, obtaining these secretions is difficult. Collection of urine in the wild
is more of a serendipitous event, usually only occurring when the animal is actively followed by

21
the researcher or research team. When it is collected, only small amounts or swabs are obtained;
never a fresh catch. Furthermore, noninvasive methods for retrieving saliva and/or urine from
wildlife populations are absent. With saliva, the only method that exists are oral swabs. Thus to
obtain saliva from an animal, it must either be physically restrained or immobilized.
Summary
As stated previously, very little research exists concerning direct correlations between
alterations in stress stimuli and direct effects on carnivore physiology. Many studies have
researched and documented what factors correlate to increased stress levels in predator species
(Adamo et al., 2013; Creel et al., 1996; Hayward, 2007; Kelly et al., 1998; Palomares et al., 1999;
Van Meter et al., 2009). Lacking, however, are studies observing the specific immunological
impacts of increased glucocorticoid concentrations on those individuals (Martin et al., 2011), as
well as noninvasive methods for assessing health in wild felid and canid species. The research
described thus far demonstrates that stress influences certain physiological systems, particularly
movement or behavior, of African carnivores (Adamo et al., 2013; Creel et al., 1996; Hayward,
2007; Kelly et al., 1998; Palomares et al., 1999; Van Meter et al., 2009). More importantly,
evidence from human subjects reveals high degrees of, or chronic exposure to, stress augments
immunosuppression (Chrousos, 2013; Sapolsky, 2002; Young et al., 2004), with some evidence
indicative of similar trends in wildlife (Drewe et al., 2009; East et al., 2013; Goller, 2011; Munson
et al., 2004). The findings that some infections become pathogenic under stress, and several disease
outbreaks coincide with stress events, demonstrates that high levels of stress could negatively
impact health and individual fitness.

22
The current methods employed for assessing health in wildlife populations – both captive
and wild – are highly invasive. This is concerning as such techniques burden a high level of risk
to both the animal and the researcher. Even more concerning, however, is the impact these
techniques may have on the health of an animal, particularly if exposure to stress events augment
immunosuppression. If stress does in fact jeopardize an animal’s immune function, it is imperative
that noninvasive techniques for health assessment be devised and employed in current practice.

23
CHAPTER 3
CREATION OF A NOVEL TECHNIQUE FOR SALIVA COLLECTION
Faunal species are exposed to various sources of stress, whether from environmental change,
competition, or habitat degradation. These stressors influence physiological systems largely in
terms of behavioral modifications. However, based on current and historic evidence in human
medicine, it seems plausible to consider immunological impacts of stress as well, particularly if
stressors are burdened on a chronic basis.
There is a current deficit in research and knowledge regarding stress and
immunosuppression, and how alterations in stress stimuli could detriment individual and
population health. While there is evidence of stress induced immunosuppression in wildlife, there
have been little to no studies to date specifically identifying this relationship, nor studies assessing
the immunologic impacts directly caused by a predetermined stressor (Buchanan, 2000; Martin et
al., 2011). Such research is extremely important, as it can provide information about how incessant
exposure to stress stimuli may impact population health in the future.
There is also a deficit in regard to noninvasive techniques for wildlife health assessment.
Currently, methods for collecting body fluids to be used in health assessment are nonexistent. The
absence of noninvasive techniques for saliva, as well as urine, collection is likely attributed to the
fact that it is difficult, if not impossible, to predict where these body fluids will be secreted by
individual of interest. Furthermore, there are currently no methods existent in the literature that
enable a controlled collection of body fluids for evaluation, and none that do not require the
presence of the researcher nor physical interaction with the study individual.

24
GOALS AND OBJECTIVES
The goal of this research was to devise a novel technique for saliva collection in African
lions (Panthera leo), while providing a new way to assess wildlife health without requiring
immobilization or invasive procedures (and the resulting stress). Of particular interest to this study,
is its potential for use in evaluating stress-induced immunosuppression. This goal was achieved by
designing a device, and evaluating the efficacy of a device for efficient and effective saliva
collection from lions.

25
METHODS
Study Location and Individuals
Three 20-30-minute trials were conducted between the months of August and October 2017 using
captive lions at the Rosamond-Gifford Zoo in Syracuse, New York. During these trials, samples
were collected and device efficacy tested. Testing dates were contingent on both the availability
of zoo personnel and personal availability. Two of the three trials (trials #1 and #3) were conducted
in the morning hours immediately following the morning feeding. The other trial (trial #2) was
conducted in the late afternoon after their evening feeding. All three were conducted in the “off
exhibit” enclosure – that which was inaccessible for public viewing. Such enabled a more
controlled test setting by avoiding potential disturbances from zoo visitors, as well as enabled
better observation and/or monitoring of behavior.
Two lions at the Rosamond Gifford Zoo were used for this study; one male and one female.
Both subjects were adults between 15 to 16 years of age. They were obtained by the zoo separately;
they are not related. Physical, external examination of each individual yielded healthy body and
coat condition, normal activity, and no apparent abnormalities (in both behavior and physical
condition).
As stated previously, all three trials were conducted in the enclosure that was off-exhibit.
This was large, caged area underground where the lions would sleep, eat, and be housed during
colder weather. The entire enclosure was composed of four conjoined enclosures, separated from
the other by a grated wall with a sliding door. The doors could be opened or closed by zoo
personnel outside of the enclosure. Several staggered horizontal platforms were suspended inside
the enclosure to provide enrichment for the lions. A large hammock was installed at the far end,

26
and various other enrichment toys were scattered throughout the enclosure. Ambient temperature
in the enclosure ranged from was on the cooler side, ranging between an estimated 64 to 73 degrees
Fahrenheit.
Experimental Design and Data Collection
Use of non-invasive methods, though potentially more costly and time consumptive, prevents
artificial increases in stress levels. Though I did not test whether or not my presence alone elicited
a stress response, by enforcing non-invasive techniques the burden of my presence should be
minimal.
Saliva Collection
Salivary samples were collected as whole saliva from passive drool samples using a “bucket in a
bucket” device (Figure 1). The device was comprised of two stainless steel buckets: a larger outer
Figure 1. The two buckets comprising the “bucket in a bucket” collection device. Left:
smaller, inner bucket serving as collection basin for drooled saliva. A grate, that can
be removed when deposited saliva is being collected, covers opening. Right: larger,
outer bucket housing ice and scent wicks. Bolts and acorn nuts hold buckets together,
while also securing device to wooden housing. (Sgambelluri, 2017)

27
bucket and a smaller inner bucket. The outer bucket served as a holding chamber for ice and the
attractant. Use of ice in the outer bucket facilitated cooling of the sample to prevent bacterial
growth. Sponges soaked in attractants (see Attractants section) were housed in the bottom of the
outer bucket. Several holes in the wall of the smaller bucket (inner bucket; held within larger
bucket) enabled the scent to waft through from outer to inner bucket so that it may be detected.
The inner bucket served as the collection chamber for drooled saliva, which was then collected for
testing. Several metal rods that grate the mouth of the smaller bucket facilitated licking and thus
salivation.
The entire contraption was housed in a reinforced wooden box (Appendix 2, Figure 3).
This provided another layer of protection for the device, as well as enabled secure attachment to
the enclosure wall. Two durable ratchet straps with industrial grade polyester webbing (breaking
weight 1,200 lbs.) were run through four slits made in the back wall of the box. The straps were
then drawn through the grated enclosure wall, and tightened on the enclosure exterior. This enabled
the box to be secured to the wall of the enclosure, while keeping the straps completely out of reach
by the lions. Furthermore, zoo personnel could easily and safely access the ratchet for tightening
without having to enter the enclosure.
Samples were retrieved within 30 minutes of collection to avoid bacterial growth. Addition
of ice in the outer bucket would provide a slightly more flexible timeline in case the retrieval could
not occur within the 30-minute timeframe. Deposited saliva was collected using 1 to 5 cc (1 – 5
mL) syringes (Bstean syringes; model #: SY-10) with blunt tipped needles (14, 16, and 18 gauge),
which were then capped and labeled according to collection protocol (Appendix). Those allocated
for biological validations were stored in a cooler (short term storage for four hours) until they
could be placed for long term storage in a freezer.

28
Attractants
Two attractants were used and tested in this study: (1) catnip extract and (2) bear bait. Catnip
extract (SynergyLabs Xtreme Catnip concentrated catnip spray; model: FG00004) was employed
to elicit a licking response and salivation from the lions – as found in catnip study by Hill et al.
(1975). The bear bait (Moultrie Bear Magnet® Liquid Bear Attractant; bacon flavored) served as
a secondary attractant, playing more of a role in motivating a lion to approach the device. Ability
of such to provide such functions were derived from whether or not a lion approached the device,
whether drooling was observed, and observation of the behaviors displayed by the lions in the
presence of the device.
Four kitchen sponges were placed in two airlocked Pyrex containers (two sponges per
container) filled with a specific attractant; one containing catnip extract, and the other filled with
bear bait. Concentration of attractant during time trial could be mediated by number of sponges
soaked and used; the more soaked sponges, the higher the concentration of attractant. The sponges
were soaked in each attractant, separately, for at least 15 hours prior to time trial.
During the trials, sponges were moved to aluminum baking trays, and then placed in the
outer bucket. The aluminum trays prevented the attractants from reacting with the metal bucket
itself, as well as enable addition of extra liquid attractant if desired. Sponges were replaced for
each trial.
Sample Retrieval and Storage
Deposited saliva was collected following the procedure detailed in Appendix 1 using 1 to 5 mL
capped syringes, depending on the amount of saliva visible. Each syringe contained a cryolabel
indicating sample number, date of collection (mnth/day/yr), and gender (M / F). Sample number

29
was constructed based on a self designed system. Each sample began with the abbreviated zoo
name (i.e. RGZ for Rosamond Gifford Zoo). This provided information on the location of the
sample. This acronym was then followed with either M or F, depending on which lion deposited
the sample. Since there were only two lions at this zoo, simply M or F was sufficient. Had there
been several individuals of the same gender, a number assigned to each individual (i.e. 1 or 2)
would have followed. A line of code contiaining 00 followed by the sample number for that
individual followed RGZM or RGZF text.
Information was also transcribed to an Excel sheet for each collection/sample. This
information included: sample number (matching that on syringe), date of collection, collection
time (military time), gender, approximate amount of saliva collected (mL), internal basin
temperature at time of collection (according to reading on digital thermometer), ambient
temperature of enclosure, general zoo activity (busy, moderate, quiet), and outdoor weather. Base
and walls of basins, as well as rods of grating, were wiped dry with paper towels between each
collection to prevent mixing of saliva or contamination. Scent sponges were replaced for each trial.
Collected samples were then stored in a cryofreezer until time of laboratory analysis. At
that time, the frozen saliva samples were thawed and then centrifuged at 3000 rpm for 15 minutes
to remove mucins and particulate matter (Salimetrics, 2015). Samples were then left out after
centrifugation until they reached room temperature. Once at room temperature, the samples were
pipetted into dilution tubes for 1) Corticosterone, 2) SIgA, and 3) C-Reactive Protein (CRP). Any
remaining saliva was re-frozen. The salivary samples in the SIgA and CRP tubes were used for
biological validation of these two test variables (Salimetrics, 2015).

30
Data Analysis
Method / Device Success
Efficacy and success were evaluated using several metrics listed in Appendix 3. The immediate
success of the device was ranked on a numeric scale based on the behavior or reaction exhibited
by the lion(s) towards the device. This method was employed to account for the fact that the actual
degree of success was dependent on the degree of behaviors elicited. The behaviors were
categorized as either being level 1 or level 2 behaviors. Level one behaviors comprised those
associated with general interaction with the device. Such behaviors included investigation,
sniffing, and playing with box, as well as pawing ground next to the device. Level two behaviors,
on the other hand, pertain to responsive behaviors. In other words, these were the behaviors
exhibited in response to the device, often following level one behavior(s). Such behaviors deemed
as level two included scent marking (rubbing of head, flank, or hind limbs) on device, scent
marking on enclosure, urine spraying, salivation/drooling, and licking.
In general, there were four predominant determinants of success value: (1) whether or not
the lion(s) approached the device, (2) whether or not the lion(s) interacted with the device (level
1 behavior), (3) whether or not the device elicited a level 2 behavior, and (4) whether or not any
saliva was collected in the basin. Degree of success was then ranked accordingly (Table 1).
Approach refers simply to the lion walking up to the device, not necessarily interacting with it; it
does not involve level one or level two behaviors. Interaction or investigation differs from
approach in that the lion made a direct and explicit effort to interact with the box, rather than just
looking at it. In general, it involved the lion standing at the device for more than five seconds and
exhibition of an investigative behavior, such as sniffing.

31
Degree of Success
Description of Value Determinants
1
lion(s) approached device, but failed to interact with the device any
further than the initial approach
2
lion(s) approached and interacted with the device (elicitation of level 1
behaviors). No elicitation of saliva or scent marking (failed to elicit
level 2 behaviors).
3
lion(s) approached and interacted with the device (elicitation of level 1
behaviors). Successfully elicited salivation or scent marking (level 2
behaviors).
4
All prior success values were displayed, and actual collection of saliva
in the basin occurred.
Statistical Analysis
Variable data from each trial were compiled and organized using Microsoft Excel Software (V.
2017). This included documentation of trial demographics (i.e. time, date, and duration), as well
as success determinants (i.e. level 1 behaviors, level 2 behaviors, time to approach device,
deposition of samples, amount collected). Such records were generated for each trial and individual
separately. Ethograms and time budget plots were generated using BORIS software program, as
well as calculation of trial statistics.
Table 1. Degrees of success yielded according to trial outcome, and the determinants founding
each evaluation. Failure to approach device is reported as a 0 – no success.

32
RESULTS
A total of three time trials were conducted to evaluate device efficacy. Trial #1 and #2 were both
conducted on September 17th, 2017. Trial #3 occurrred on October 1st, 2017. Both the first and
third trials were conducted early in the morning immediately following morning feeding. The
second trial was conducted during the evening of the first trial day.
Two of the three trials yielded a success rating of greater than or equal to three; trials one
and three (Table 6). The first trial yielded a success rating of four for the male and three for the
female, with a mean success value of 3.5 (Table 6). No success was achieved in the second trial
for either individual, yielding an overall success value of zero for that trial (Table 6). The third
trial yielded a success value of three for the male and one for the female, for a mean success value
of two (Table 6). In general, a higher success in elicitation of behavioral responses and actual
collection of saliva was observed with the male.
Of note, ice was not used in any of the trials, since such were tailored strictly to 20-30
minute periods. Thus, the potential for bacterial growth prior to collection was not of concern.
Trial #1 Results and Observations
During the first trial, the male approached the device nine minutes after entering the enclosure
(Figure 3). Four level one behaviors were exhibited within the 20 minute trial. These were (1)
investigation, (2) sniffing of box, (3) pawing at box or ground next to box, and (4) playing with
the box (Table 6). The device held strong when played with by the male lion, and no damages
occurred. Level two behaviors observed included salivation/drooling along corners of mouth, and
rubbing sides of face (cheeks) against box (Table 6). This type of head rubbing serves as scent
marking. Although actual drooling of saliva into the device was not observed, pooled saliva was

33
present at the bottom of the basin following termination of trial. Approximately 0.1 cc (100 µL)
of saliva was retrieved. Observation of level one behaviors, elicitation of level two behaviors, and
the collection of saliva yielded this trial – for the male individually – as extremely successful,
earning the highest success value of four. Furthermore, the amount of saliva collected exceeded
the minimum amount required for conduction of ELISAs – about 10-20 µL – indicating the device
was successful in collecting adequate sized samples for salivary analyses.
Approach by the female occurred approximately 10 minutes after entry into enclosure
(Figure 3). Both level one and two behaviors were observed. Level one behaviors included (1)
investigation of box, (2) sniffing, and (3) pawing ground next to box (Table 6). Frequent head ru
vn bbing – scent marking and level two behavior - occurred, though such was performed on the
enclosure rather than the box (Figure 3, Table 3). Several behaviors were of particular interest, and
appeared to be a direct response to the device. These such behaviors – which are discussed further
in the following chapter – included frequent head rubbing on enclosure, frequent sneezing fits,
and pawing at ground next to box. Since no saliva was collected, but both level one and two
behaviors were observed, this individual trial for the female was ranked as a success value of three.
Overall success for this trial, or the mean value between scores for male and female, was a 3.5.
Time Budget
Behaviors in both individuals were allocated between stationary, vocalization, and grooming
(Figure 2, Table 3). For the male, 70% of total trial time was spent in stationary behavior – either
standing or resting/laying down – comprising a total duration of approximately 1345 seconds for
stationary (frequency = 10; Table 3). This was followed by vocalization, which comprised 13.8%
of the trial period for a total duration of 265 seconds (frequency of 4), and grooming activities,
comprising 9.9% of the total trial with a total duration of 191 seconds (frequency of 4) (Table 3).

34
In total, device specific behaviors constituted 8.5% of the total trial period (0.3% play, 3.6%
investigation of box, 0.7% roll, 3.9% scent marking) (Table 3), for total dhuration of approximately
163.5 seconds.
Synonymous with the male, the majority of trial period for the female was spent in
stationary behavior (Figure 2). Such comprised 73% of total trial for a total duration of
approximately 1400 seconds (frequency = 18; Table 3). Also synonymous with the male,
vocalization was the second largest time contributive behavior for the female, constituting over
9% of the trial period (total duration of approximately 184 seconds) (Table 3). Such is expected,
as most vocalization events occur as a call and response between individuals of a small group or
Figure 2. Percentage of time spent per activity throughout the entire 20 minutes of trial one. Activity time
percentages are displayed for both individuals; male (top) and female (bottom). Ranking of behaviors do
not correspond to first occurrence in trial. Each color correlates to a different behavior or grouping of
behaviors.

35
pride, as a means to strengthen family bonds and/or secure territory. Of all the vocalization events
documented, there was only one instance where the female did not join the male in chorus. Though
not the greatest contributer for time, the most frequent behavior displayed by the female was
locomotion – observed 21 times (Table 3). A significant amount of scent marking behavior was
observed, often enacted as facial scent marking or head rubbing on the enclosure grating (Figure
3). A total of nine different scent marking events were observed for the female, often followed or
paired with grooming activities and short sneezing fits (Table 3, Figure 3). A significantly smaller
amount of time was allotted to device specific behaviors; roughly 6.3% of the total trial period,
with a total duration of 121 seconds. Device specific behaviors for the female included
investigation of box (0.8%), pawing (1.6%), and scent marking/head rubbing (3.9%) (Table 3).

36
Table 2. Ethogram of all recorded and/or major behaviors for captive African lions, with
operational definitions. Behavioral measures grouped according to their assumed “type”; solitary,
ambulatory, food, social, aggression/territorial, and miscellaneous. Of note: urination and urine
spraying are two different behaviors. Urination involves the lion squatting to dispel urine, and
typically no external stimulus is involved. Urine spraying, on the other hand, represents scent
marking and/or territorial behavior. Unlike urination, the animal is standing with tail erect in air.
A jet of urine, measuring up to or over 1 meter in length, is expelled onto a vertical object. Tail
will often quiver during such, and the animal may be seen hindlimb scraping.
Type of
Behavior
Behavior
Code Des cription of Behavior
Solitary Sleep S
Animal assumes species-specific position for sleep; stays in one place and is not
alert to environmental changes
Stationary R
Animal stays in one place, but may be roused easily by environmental changes;
may be either laying down (R-L) or standing (R-S); eyes may be open or closed
Yawn Y
Common usage
Groom Self GS
Animal engages in washing its fur, scratching, and/or grooming activities using its
tongue and forelimbs
Play P
Animal engages in solitary play- based interactions (exhibits no intention to harm)
with enrichment toys or with box
Investigation I-B; I-O
Animal engages in solitary investigation of box (I-B) or other object (I-O) in
enclosure
Head Shake HS
Animal shakes head from side to side
Maintenance M
Animal urinates (in squatting position) or defecates
Ambulatory Locomotion L
Animal moves from one location to another in a nonrepetitive manner; with
purpose (i.e. walk, trotting)
Pacing PC
Repetitive movement between two locations, with no apparent purpose
Climbing CL
Animal ascends and/or descends an object or structure
Hindlimb Scraping HH
Animal scrapes hind feet backwards along ground surface, shuffling between feet
Roll RL
Animal moves onto its back and exposes belly. Playful intention
Pawing PW
Animal pats at ground surface using forelimb, with no clear purpose; only one
paw used and claws are retracted
Food Related Eat E Animal consumes food
Drink D
Animal consumes water
Social
Communal Groom / Ear
Suckling
CG
Animal engages in washing fur, grooming, or ear suckling of another individual in
environment
Play GP
Animal engages in interactions with other individuals that may involve locomotion,
climb, manipulating objects, or other activities that show a relationship between
two or more interacting animals (i.e. wrestling)
Contact CO
Animal comes in physical contact with another individual
Vocalization V
Animal engages in non-aggressive vocal activities (i.e. roaring, mewing, chatting)
Aggressive
or Territorial
Fight F
Animal engages in aggressive physical contact with another animal in its
environment (i.e. biting, hit, slap, chase, snarl)
Scent Marking / Rubbing SM Animal rubs glands near tail, legs, or on face against surface or object.
Urine Spray US
Jet of urine released onto vertical surface or object while standing. Tail is raised
vertically, and may quiver.
Misc Other O Animal engages in behavior not explicitly detailed or identified above

37
Behavior Frequency Total duration (s) Duration Mean (s) IEI Mean (s)
% of total
length
Investigation - O 1 9.10 9.1 NA 0.5
Stationary 10 1345.50 134.55 ± 139.31 49.72 ± 42.37 70.1
Yawn 3 19.00 6.33 ± 2.31 307 ± 74.95 1
Groom Self 4 191.00 47.75 ± 47.82 323 ± 254.21 9.9
Play 1 5.00 5NA 0.3
Investigation - B 4 69.20 17.3 ± 20.86 240.93 ± 274.67 3.6
Locomotion 13 102.59 7.89 ± 7.5 132.62 ± 179.77 5.3
Climbing 1 2.00 2NA 0.1
Roll 1 14.10 14.1 NA 0.7
Eat 1 14.00 14 NA 0.7
Vocalization 4 265.10 66.28 ± 11.76 192.63 ± 148.66 13.8
Scent Marking / Rubbing 4 75.20 18.8 ± 16.0 185.27 ± 308.86 3.9
Behavior Frequency Total duration (s) Duration Mean (s) IEI Mean (s)
% of total
length
Stationary 18 1400.78 77.82 ± 83.87 22.66 ± 26.02 73
Groom Self 7 92.00 13.14 ± 10.92 236.33 ± 284.05 4.8
Investigation - B 1 16.00 16 NA 0.8
Maintenance 1 10.00 10 NA 0.5
Locomotion 21 160.39 7.64 ± 9.96 76.68 ± 89.66 8.4
Climbing 2 5.00 2.5 ± 2.12 1537 0.3
Paw 2 30.00 15 ± 7.07 8 1.6
Vocalization 3 184.00 61.33 ± 16.29 234.5 ± 195.87 9.6
Scent Marking / Rubbing 9 75.00 8.33 ± 3.67 188.75 ± 279.95 3.9
Other 1 10.00 10 NA 0.5
Table 3. Time budget summaries for male (top) and female (bottom) African lions during trial #1, at Rosamond-Gifford Zoo. Frequency
represents the number of times a behavioral event was observed. Total duration represents cumulative time (in seconds) allocated to a
specific behavior. Such valuation was utilized to evaluate the percentage of total trial period spent per behavior. Average duration and
inter-event intervals (IEI) reported as mean ± SD. There were no inter-event intervals for those behaviors occurring a single time
(frequency of 1), and thus IEI means for such were reported as NA. Time trial was conducted on September 17, 2017 for a total time of
20 min (1200 sec). Data constructed using BORIS software (Friard et al., 2016). (Sgambelluri, 2018)

38
Behavior Frequency Total duration (s) Duration Mean (s) IEI Mean (s)
% of total
length
Investigation - O 1 1.10 1.1 NA 0.1
Stationary 28 1410.17 50.36 ± 89.6 13.33 ± 20.86 79.7
Investigation - B 1 8.00 8NA 0.5
Locomotion 28 164.28 5.87 ± 6.4 36.59 ± 44.9 9.3
Climbing 1 4.00 4NA 0.2
Hindlimb Scraping 1 6.00 6NA 0.3
Contact 1 2.00 2NA 0.1
Vocalization 6 306.00 51 ± 29.79 186.2 ± 118.9 17.3
Scent Marking / Rubbing 3 24.00 8 ± 4.36 2.5 ± 0.71 1.4
Urine Spray 1 12.00 12 NA 0.7
Behavior Frequency Total duration (s) Duration Mean (s) IEI Mean (s)
% of total
length
Investigation - O 1 1.00 1NA 0.1
Stationary 11 1695.19 154.12 ± 307.72 7.48 ± 6.70 95.8
Yawn 1 2.00 2NA 0.1
Groom Self 1 32.00 32 NA 1.8
Locomotion 10 46.09 4.61 ± 3.26 83.66 ± 185.93 2.6
Climbing 1 2.00 2NA 0.1
Contact 1 2.00 2NA 0.1
Vocalization 4 180.00 45 ± 13.78 318 ± 159.62 10.2
Urine Spray 1 15.00 15 NA 0.8
Table 4. Time budget summaries for male (top) and female (bottom) African lions during trial #3, at Rosamond-Gifford Zoo. Frequency
represents the number of times a behavioral event was observed. Total duration represents cumulative time (in seconds) allocated to a
specific behavior. Such valuation was utilized to evaluate the percentage of total trial period spent per behavior. Average duration and
inter-event intervals (IEI) reported as mean ± SD. There were no inter-event intervals for those behaviors occurring a single time
(frequency of 1), and thus IEI means for such were reported as NA. Time trial was conducted on October 1,2017 for a total time of 30
min (1800 sec). Data constructed using BORIS software (Friard et al., 2016). (Sgambelluri, 2018)

39
Figure 3. Time budget diagrams for both male (top) and female (bottom) African lions during trial #1 at Rosamond Gifford Zoo. Behavior
parameters include only state events (frequency and duration). Point events (frequency only) were not included. Trial length was 20
minutes, starting at 9:42 AM and ending at 10:12 AM.

40
Figure 4. Time budget diagrams for both male (top) and female (bottom) African lions during trial #3 at Rosamond Gifford Zoo.
Behavior parameters include only state events (frequency and duration). Point events (frequency only) were not included. Trial length
was 30 minutes, starting at 9:36 AM and ending at 10:06 AM.

41
BEHAVIOR
TIME
(h:min:se c)
DURATION
(se c)
NOTES
L
9:42:04 3 male entered enclosure
R-S
9:42:07 12 looking around enclosure and sniffing air
L
9:42:19 17 moved to far side of enclosure by hammock
L
9:42:36 11 ambled around enclosure containing hammock
CL
9:42:47 3 jumped up onto/into hammock
R-L
9:42:50 56 laying in hammock
G-S
9:43:46 20
R-L
9:44:06 60
R-L
9:45:06 60
R-L
9:46:06 53
L
9:46:59 26
stood up and left hammock. Moved across enclosure,
and into adjacent enclosure, which contained saliva
device/box
R-S
9:47:15 50 standing in enclosure and sniffing air
L
9:48:05 5 walked to front of enclosure
R-S
9:48:10 43 standing at front of enclosure. Intermittent sniffing
L
9:48:53 5
walked towards back of enclosure, near platform where
female was resting
R-L
9:48:58 60 resting on ground
R-L
9:49:58 60
R-L
9:50:58 60
R-L
9:51:58 60
G-S; R-L
9:52:58 57
Y
9:53:55 5
R-L
9:54:00 35
G-S
9:54:35 23
R-L
9:54:58 7
V
9:55:05 82
male initiated vocalization. Female joined roughly five
seconds in
L
9:56:27 10 Vocalization ended. Walked over to log
I-O
9:56:31 6 investigating log
L
9:56:41 4 walked to food tray
E
9:56:45 15 eating food
L
9:57:00 3 Stopped eating. Walked over to box
I-B
9:57:03 7 sniffing (interacting with) box
I-B; SM
9:57:10 41
Rubbing face and body on box; sniffing box intermittent.
Apparent salivation observed, though uncertain if any
deposited.
I-B; P
9:57:51 5 trying to play with box. Box held strong
I-B; SM
9:57:56 4
TRIAL STARTED AT 9:42:00
Table 5. Observation and documentation of behaviors exhibited by male lion during first trial on
September 17, 2017. Total duration of trial was 20 minutes, beginning at 9:42 am and concluding
at 10:12 am. Time at which behavior was initiated (time) and the total duration of that behavior (in
seconds; duration) are reported in table. Behaviors recorded using coding provided in table 1.
Scent marking predominantly occurred using scent glands in cheeks; this is recorded as SM and
described as head rubbing or rubbing of face. As with female, rubbing face against an object was
documented and presumed as scent marking (SM), rather than grooming, based on current
knowledge of wild felid behavior as described in scientific literature (Gorman et al., 1989;
Marnewick et al., 2006; Pageat et al., 2003; Reiger, 1979; Soini et al., 2012). Refer to ethogram
(table 2) for behavioral coding.

42
L
9:58:00 4 Left box and moved to center of enclosure
V
9:58:04 54
In center of pen. Initiated vocalization and joined by
female.
R-L
9:59:00 26 Stopped vocalizing. Laid down in middle of enclosure.
G-S
9:59:26 32 Lots of grooming
R-L
9:59:58 2
Y
10:00:00 9
R-L
10:00:09 51 laying in center of enclosure
R-L
10:01:00 60 appears to be napping
R-L
10:02:00 3
joined by female. Lifted head briefly when female
approached, then lowered back to floor
R-L
10:02:03 60
G-S
10:03:03 6
Lifted head and repositioned body (shifted from laying
on side to sternal recumbant)
R-L
10:03:06 60
R-L
10:04:06 17
Y
10:04:23 5
R-L
10:04:28 35
V
10:05:03 60 male started vocal at 10:05:03. Still laying down
R-L
10:06:03 50
Male stopped vocalizing. Remained laying down, and
soon joined by female
L
10:06:53 5 Stood up and moved back over to box
I-B
10:06:58 3
SM; RL
10:07:01 15
Rubbed head against box. Then started rolling on his
back against box, continuing to rub cheeks (where scent
glands are located) against box.
I-B
10:07:16 8
Stopped rolling and stood up. Sniffed box. Seems to
have saliva pooling at corners of mouth. Licked left side
mouth. Did not lick grate on top of box.
SM
10:07:24 17 Rubbing head against top of box
R-S
10:07:41 7 Stopped rubbing. Standing at box. Looking to right side
L
10:07:48 10 moved to far side of enclosure by hammock
R-S
10:07:58 2 standing by hammock, facing front of enclosure
V
10:08:00 60 male vocalized. Female did not join vocalization
V
10:09:00 8 Still vocalizing
R-S
10:09:08 14
stopped vocalizing. Standing in same spot. Intermittently
moving head to shift gaze around enclosure.
R-S
10:09:22 6 Still standing, but now watching female investigate box
R-S
10:09:28 20
Stopped focus on female. Returned to standing with
intermittent shift in gaze around enclosure.
L
10:09:48 15
moved back across enclosure and approached saliva
device
I-B
10:10:03 12 sniffing box
L
10:10:15 3 moved to center of enclosure
R-S
10:10:18 60 laying down in center of enclosure
R-S
10:11:08 52
TRIAL ENDED AT 10:12:00

43
Trial #2 Results and Observations
Both lions failed to approach and/or interact with the device. In general, they seemed undisturbed
and uninterested in the box. Bulk of 20 minute trial period was allocated between grooming,
resting, and moving between areas of the enclosure for both individuals. No vocalizations or scent
marking activities were observed. Additionally, no level one or two behaviors were observed or
recorded. In accordance, the trial was deemed as unsuccessful for both individuals, yielding an
overall success value of zero.
Trial #3 Results and Observations
While the device was seen by the female, she failed to pursue any further interaction with such.
After approximately 13 minutes, this individual approached within one foot of the box. A total of
nine seconds were spent at that location, with three instances where the female looked directly at
the device; device was approached and observed from a distance. No level one behaviors occurred
at that time, nor during time following (Figure 4). However, a level two behavior did occur. Within
10 seconds of leaving the spot near the device, the female sprayed along the back wall of the
enclosure (Figure 4). Specific motivation behind such behavior was ambigous; such may have
been in direct response to the device or in response to human presence previously in the enclosure.
Remainder of trial period was allocated between resting and vocalization (Figure 4). Frequent head
rubbing and sneezing episodes that had been observed during trial one were absent in this trial.
Three mililiters (3000 uL) of urine – from urine spray – was collected upon trial termination (Table
6). Since the specific motivation behind the urine spray was not clearly defined, the behavior was
not recorded as a device-specific response in an attempt to avoid potential bias. Consequently, a
success value of one was assigned for this individual in this trial.

44
The male, conversely, exhibited a strong, favorable response to the device. Approach
occurred approximately nine minutes after entering the enclosure (Figure 4), with time spent at the
device exceeding one minute. Scent marking via facial rubbing against the device occured, as well
as scent marking via urine spray (Table 4). Duration of such behaviors totaled to 50 seconds,
representing 3% of trial (Table 4). The urine spray occurred immediately following cessation of
head rubbing, and was projected towards the front of the box, alluding territorial motivation.
Unlike that of the female, the male’s urine spray was directly associated with the device, in that it
was in response to and directed at the device specifically. Approximately 2 mL (2000 uL) of urine
was collected amongst three collection syringes, along with retrieval of mane hair fibers from the
wooden crate (Table 6). While this trial had a success value of three due to lack of saliva deposition
and collection, it should be noted that the device did succeed in eliciting deposition of body fluid
– the urine spray – from at least one individual.

45
Trial Date
Time (24 hr;
hr:min:sec)
Duration
(min.)
Gender
Approach
(Y/N)
Time to Approach
(min.)
Level 1 Behaviors Level 2 Behaviors
Elicitation of
Licking (Y/N)
Obvious
Salivation
(Y/N)
Body Fluid Deposited
and Collected
Amount deposited
(cc ; uL)
# of Samples
Collected
Damage
Occurrence
(Y/N)
Success
Value
1 9/17/2017 9:42:00 20 Male Y 9
Investigation Salivation/Drooling N Y Saliva 0.1 cc ; 100 uL 1 N 4
Sniffing Head Rubbing
Pawing
Play
1 9/17/2017 9:42:00 20 Female Y 10
Investigation Head Rubbing* N N None - 0 N 3
Sniffing
Pawing
2 9/17/2017 17:10:00 30 Male N X
X X N N None - 0 N 0
2 9/17/2017 17:10:00 30 Female N X
X X N N None - 0 N 0
3 10/1/2017 9:28:00 30 Male Y 9
Investigation Urine Spray N Y Urine 2 cc ; 2000 uL 3 (+ hair) N 3
Sniffing Head Rubbing
3 10/1/2017 9:28:00 30 Female Y 13
X Urine Spray N N Urine 3 cc ; 3000 uL 1 N 1
Table 6. Complete table of time trial data, as well as documentation of information relative to success ranking. Level 1 and level 2
behaviors observed for each individual per trial are described. Whether body fluid was deposited and collected during each trial is
illustrated, as well as the amount that was collected if deposition occurred (1 cc = 1 mL). Damage occurrence refers to whether or not
the device sustained any damage requiring repair during/after a trial period. This factor served as a metric in success evaluation as a
primary determinant of durability.
Approach refers simply to the lion walking up to the device and remaining within a 1 ft radius of device. Such behavior does not deal
with a direct interaction with the device; it does not involve investigation of device or any other behaviors elicited therein. Investigation
differs from approach in that lion made direct and explicit effort to interact with the box, typically through sniffing, and involved the
lion standing at the device for more than five seconds. Scent marking predominantly occurred using scent glands in cheeks; this is
described as head rubbing (level 2 behavior); described as scent marking based on current knowledge of wild felid behavior as described
in scientific literature (Gorman et al.,1989; Marnewick et al., 2006; Pageat et al., 2003; Reiger, 1979; Soini et al., 2012). A urine spray
is different from normal urination. Urination usually involves squatting and typically no stimulus involved. Urine spray done while
standing, with a jet of urine being released straight behind the animal. Tail is usually erect in air, and hindlimb scraping may be involved.
*Female head rubbing during first trial was against enclosure walls, not the device itself. Animal curator indicated that the frequency at
which the female exhibited this behavior was far greater than normally observed.
Table 6. Complete table of time trial data, as well as documentation of information relative to success ranking. Level 1 and level 2
behaviors observed for each individual per trial are described. Whether body fluid was deposited and collected during each trial is
illustrated, as well as the amount that was collected if deposition occurred (1 cc = 1 mL). Damage occurrence refers to whether or not
the device sustained any damage requiring repair during/after a trial period. This factor served as a metric in success evaluation as a
primary determinant of durability.
Approach refers simply to the lion walking up to the device and remaining within a 1 ft radius of device. Such behavior does not deal
with a direct interaction with the device; it does not involve investigation of device or any other behaviors elicited therein. Investigation
differs from approach in that lion made direct and explicit effort to interact with the box, typically through sniffing, and involved the
lion standing at the device for more than five seconds. Scent marking predominantly occurred using scent glands in cheeks; this is
described as head rubbing (level 2 behavior); described as scent marking based on current knowledge of wild felid behavior as described
in scientific literature (Clark et al., 2008; Marnewick et al., 2006; Rallis, 1971; Reiger, 1979; Soini et al., 2012). A urine spray is different
from normal urination. Urination usually involves squatting and typically no stimulus involved. Urine spray done while standing, with
a jet of urine being released straight behind the animal. Tail is usually erect in air, and hindlimb scraping may be involved.
*Female head rubbing during first trial was against enclosure walls, not the device itself. Animal curator indicated that the frequency at
which the female exhibited this behavior was far greater than normally observed.

46
DISCUSSION
Overall, the device was successful, with a success rating of greater than or equal to three in two of
the three trials conducted. In general, there was a higher success in elicitation of behavioral
responses and actual collection of saliva with the male.
The female exhibited some interesting behaviors that may have been correlated with the
presence of the device, and/or a reaction towards one of the attractants. During the first trial, the
female exhibited excessive head rubbing against grating of enclosure, as well as frequent sneezing
fits – both of which were deemed as unusual for this individual by the animal curator. Furthermore,
after approaching the device, the female went alongside the device and began pawing at the
ground. This may have been a type of behavior called scratching, a variation of scent marking
wherein the individual marks territorial areas by scratching or pawing at the ground (Ghoddousi
et al., 2008). This typically is followed by spraying, which was not observed in this event. It is
unclear whether one of the attrancts was percieved positively by the male, but negatively by the
female, and if such was illiciting such behavior in the female. It is possible that such behavior was
conducted more as an act of submission, since the male seemed to claim the device during the first
trial. It would be interesting to know whether the female was indeed reacting negatively to the
device, or if one of the attractants used was perceived differently by the genders.
Both individuals were not nearly as interested in the device when they were reintroduced
to it on the evening of the first trial. It is difficult to determine causation for failed success in this
second trial – whether such was a function of time of day or lack of interest, and there was not
enough data to make such determination. However, based on the data obtained, it appears that the
device yielded most success when removed for a period of time and reintroduced for each trial,
rather than remaining in the enclosure. It is possible that the individuals habituated to the stimulus

47
when it was left in the enclosure between trials. For the third trial, it is possible that a decreased
response from the lions occurred because the attractants were less concentrated (only one sponge
soaked with each attractant was used, rather than two), relative to the first trial. The less
concentrated the attractant(s), the less noxoious the stimuli, and – potentially – the lower the degree
of response. Furthermore, installation of the device occurred while the individuals were already
inside the enclosure; per request of the zookeepers. Though they could not access the device during
installation, the lions were able to observe it being installed and were also exposed to the scents at
that time. It is possible that direct observation of such, rather than finding of device upon entry
into enclosure, detered or influenced the behavior of the lions towards the device; for instance,
such may explain why the female lion would only approach the device at a distance, and why less
time was spent at the device for both lions in general. It could be argued that both urine sprays
observed during the trial were, rather, territorial responses to an unknown human being having
been in the enclosure previously. Howerver, if such were the case, there likely would have been a
greater amount of strong scent marking behavior made throughout the entire enclosure by both
individuals. Furthermore, the general relaxed attitude of the lions belies such arguement.
Throughout the entire installation and time trial, both individuals appeared undisturbed and
unagitated. There were no aggressive behaviors observed or displayed towards myself or the
animal curator during device installation. Rather, the female was resting on the horizontal platform
and the male was sleeping in the hammock. Pacing, another behavior that would be expected in
such scenario, was not exhibited by either individual.
Based on the obsersvations and results of these trials, the device was indeed successful,
with the potential to not only elicit and collect saliva, but also collect other body fluids as well. In
this case, urine spray. The combination of attractants employed succeeded in eliciting approach by

48
the individuals, as well as a number of interesting behaviors. In this particular case, the almost
allergic like response of the female versus the territorial response by the male. The device itself
carries great potential for how it may be used in the future, particularly in the realm of wildlife
health assessment. However, before such can be enacted, followup studies must be conducted. The
replicative power, or its ability to reproduce saliva collection in other individuals, needs to be
evaluated for this device. It is not to say that, while it indeed worked with the lions used at the
Rosamond-Gifford Zoo, it could yield greatly different responses in lions at different zoos.

49
CHAPTER 4
CONCLUSIONS AND RECOMMENDATIONS
The goal of this thesis was to devise a novel technique for saliva collection in African lions
(Panthera leo), while providing a new way to noninvasively assess health in wildlife populations.
The fact that this device was able to elicit responses from the study individuals, and successfully
collected an adequate sample of saliva, indicates that this method was indeed successful for this
study. In one of the trials, not only were level 2 behaviors elicited (e.g. scent marking and head
rubbing), but the device succeeded in eliciting the deposition of body fluid – the urine spray.
Furthermore, this fluid was collectable. This finding is significant, in that it indicates (1) the device
has the potential to trigger a variety of responses, and (2) it can collect several different types of
samples. In the three trials performed in this study to test this device, the device was able to collect
saliva, urine, and hair samples, all of which can be utilized to gain information relative to
individual health. This highlights the potential for this device to be used beyond only saliva
collection, and thus broadens scope of how it may be used in the future.
Future Research and Considerations
Obviously, further research needs to be conducted to supplement the findings obtained
here. The device needs to be tested on different individuals in different locations. This can be taken
a step further, and tested on different felid species to compare responses. Next steps for the device
would also include the addition of a self-closing lid system, to ensure collection of individual
samples and avoid cross contamination between individuals.
Saliva can be used to provide a wealth of knowledge ranging from immunologic to
reproductive health. Based on this understanding, it could be used as a valuable medium for

50
assessing health across various spectra. Efforts should be, and need to be, made to begin
incorporating the use of saliva and employing it to regularly monitor health of captive species;
particularly if health is impaired by stress events.
Regarding the utilization of saliva, steps also need to be made to biologically validate feline
salivary health measures. This requires the acquisition and/or production of ELISA assays that are
validated for feline saliva. An original objective in this study was the biological validation of
salivary assays for feline secretory immunoglobin A (IgA) and c-reactive protein (CRP), and using
such for evaluation of stress-induced immunosuppression. During the progression of this study, it
became clear that such objective was better suited as a separate, ongoing project in and of itself.
At this current time, IgA and CRP ELISAs operative on domestic feline saliva are rare; for lions
they are nonexistent. Most of the ELISAs for evaluations of such levels in felines only detect –
and thus function – with serum or plasma samples. Attempts to contact individuals who analyzed
immunoglobin levels in saliva of domestic felines for evaluation of feline leukemia were made,
although such attempts were unsuccessful. Furthermore, valid and thorough completion of the
biological validations requires careful collaboration of several entities – diagnostic laboratories
and companies with appropriate ELISAs in particular.
Broader Impacts
The importance of this research is (1) the creation of a novel technique that provides a
noninvasive alternative to current health sampling methods, which can then (2) be employed in
investigating how incessant exposure to stress stimuli may impact population health in the future.
Understanding occurrence of physiologic issues augmenting immunosuppression could be crucial
in constructing successful and effective conservation strategies. If we are to obtain accurate data

51
relative to stress-induced immunosuppression, it is imperative that noninvasive techniques be
created and employed.

52
REFERENCES
Adamo, Shelley A., Ilya Kovalko, and Brianna Mosher. (2013). The behavioural effects of predator-
induced stress responses in the cricket (Gryllus texensis): the upside of the stress response. The Journal of
experimental biology 216.24: 4608-4614.
Agarwal, Ashok, et al., "Effect of oxidative stress on male reproduction." The world journal of men's
health 32.1 (2014): 1-17.
Aguirre, A. A., Ostfeld, R., & Daszak, P. (2012). New directions in conservation medicine: applied cases
of ecological health. OUP USA.
Aich, P., Potter, A. A., & Griebel, P. J. (2009). Modern approaches to understanding stress and disease
susceptibility: A review with special emphasis on respiratory disease. International journal of general
medicine, 2, 19.
Buchanan, K. L. (2000). Stress and the evolution of condition-dependent signals. Trends in Ecology &
Evolution, 15(4), 156-160.
Busch, D. Shallin, and Lisa S. Hayward. (2009) Stress in a conservation context: a discussion of
glucocorticoid actions and how levels change with conservation-relevant variables. Biological
Conservation 142.12: 2844-2853.
Caro, T. M., and C. J. Stoner. (2003). The potential for interspecific competition among African
carnivores. Biological Conservation 110.1: 67-75.
Caro, T. M., et al., (1987). Health of adult free‐living cheetahs. Journal of Zoology 212.4: 573-584.
Chiappin, S., Antonelli, G., Gatti, R., & Elio, F. (2007). Saliva specimen: a new laboratory tool for
diagnostic and basic investigation. Clinica Chimica Acta, 383(1), 30-40.
Chrousos, George, M.D. (2013). The Stress Response and its Implications for Pituitary Tumors. Pituitary
Network Association. Pituitary Network Association. Web.
<https://pituitary.org/medical-resources/pavilions/women-s-health/women-s-health-archive/the-stress-
response-and-its-implications-for-pituitary-tumors>.
Creel, Scott, and Nancy Marusha Creel. (1996). Limitation of African wild dogs by competition with
larger carnivores. Conservation Biology 10.2 (1996): 526-538.
Creel, S., Christianson, D., & Schuette, P. (2013). Glucocorticoid stress responses of lions in relationship
to group composition, human land use, and proximity to people. Conservation Physiology, 1(1), cot021.
Darnell, A. M., Graf, J. A., Somers, M. J., Slotow, R., & Gunther, M. S. (2014). Space use of African
wild dogs in relation to other large carnivores. PloS one, 9(6), e98846.
Dohms, J. E., & Metz, A. (1991). Stress—mechanisms of immunosuppression. Veterinary Immunology
and Immunopathology, 30(1), 89-109.
Drewe, J. A., et al., (2009). Pathology of Mycobacterium bovis infection in wild meerkats (Suricata
suricatta). Journal of comparative pathology 140.1: 12-24.
Drewe, Julian A. (2009). Who infects whom? Social networks and tuberculosis transmission in wild
meerkats. Proceedings of the Royal Society of London B: Biological Sciences: rspb20091775.

53
East, Marion L., et al. (2013). Factors influencing Dipylidium sp. infection in a free-ranging social
carnivore, the spotted hyaena (Crocuta crocuta). International Journal for Parasitology: Parasites and
Wildlife 2 (2013): 257-265.
eBioscience. (2015). Introduction to Immunoglobulins. Retrieved March 17, 2016, from affymetrix
website:
http://www.ebioscience.com/knowledge-center/antigen/immunoglobulin.htm
French, M. (Ed.). (2012). Immunoglobulins in health and disease (Vol. 1). Springer Science & Business
Media.
Friard, O. and Gamba, M. (2016), BORIS: a free, versatile open-source event-logging software for
video/audio coding and live observations. Methods Ecol Evol, 7: 1325–1330.
Ghoddousi, A., Hamidi, A. K., Ghadirian, T., Ashayeri, D., Hamzehpour, M., Moshiri, H., ... & Julayi, L.
(2008). Territorial marking by the Persian Leopard (Panthera pardus saxicolor) in Bamu National Park,
Iran. Zoology in the Middle East, 44(1), 101-103.
Glaser, R., Rice, J., Sheridan, J., Fertel, R., Stout, J., Speicher, C., & Kiecolt-Glaser, J. (1987). Stress-
related immune suppression: Health implications. Brain, behavior, and immunity, 1(1), 7-20.
Goller, Katja Verena. Pathogens in free-ranging African carnivores. Diss. Humboldt-Universität zu
Berlin, Mathematisch-Naturwissenschaftliche Fakultät I, 2011.
Gorman, M. L., & Trowbridge, B. J. (1989). The role of odor in the social lives of carnivores. In
Carnivore behavior, ecology, and evolution (pp. 57-88). Springer, Boston, MA.
Hansen, Fawne. (2014) How Does Stress Affect Your Immune System? The Adrenal Fatigue Solution.
PERFECT HEALTH, 20 July 2014. Web. http://adrenalfatiguesolution.com/stress-immune-system/.
Hayward, Matt W., and Gina J. Hayward. (2007). Activity patterns of reintroduced lion (Panthera leo)
and spotted hyaena (Crocuta crocuta) in the Addo Elephant National Park, South Africa." African journal
of ecology 45.2: 135-141.
Hill, J. O., Pavlik, E. J., Smith III, G. L., Burghardt, G. M., & Coulson, P. B. (1976). Species-characteristic
responses to catnip by undomesticated felids. Journal of Chemical Ecology, 2(2), 239-253.
Hoover, E. A., and James I. Mullins. (1991). Feline leukemia virus infection and diseases. Journal of the
American Veterinary Medical Association 199.10: 1287-1297.
Joubert, Dereck (1992). Eternal Enemies: Lions and Hyenas. Screenplay by Dereck Joubert. National
Geographic Society. Film.
Laudenslager, M. L., Bettinger, T., & Sackett, G. P. (2006). Saliva as a medium for assessing cortisol and
other compounds in nonhuman primates: collection, assay, and examples. In Nursery rearing of nonhuman
primates in the 21st century (pp. 403-427). Springer US
Laurenson, M. Karen. (1995). Implications of high offspring mortality for cheetah population
dynamics. Serengeti II: dynamics, management and conservation of an ecosystem: 385-399.
Marnewick, K. A., Bothma, J. D. P., & Verdoorn, G. H. (2006). Using camera-trapping to investigate the
use of a tree as a scent-marking post by cheetahs in the Thabazimbi district. South African Journal of
Wildlife Research-24-month delayed open access, 36(2), 139-145.

54
Martin, L. B., Andreassi, E., Watson, W. & Coon, C. (2011) Stress and Animal Health: Physiological
Mechanisms and Ecological Consequences. Nature Education Knowledge 3(6):11
Mills MGL. (1990). Kalahari Hyaenas: The Behavioural Ecology of Two Species. Unwin
Hyman, London.
Mills, M. G. L. (2005). Large carnivores and biodiversity in African savanna ecosystems. Large
carnivores and the conservation of biodiversity, 208-229.
Munson, Linda, et al., (2004). Serosurvey of viral infections in free-ranging Namibian cheetahs (Acinonyx
jubatus). Journal of Wildlife Diseases 40.1: 23-31.
Munson, L., Terio, K. A., Worley, M., Jago, M., Bagot-Smith, A., & Marker, L. (2005). Extrinsic factors
significantly affect patterns of disease in free-ranging and captive cheetah (Acinonyx jubatus) populations.
Journal of wildlife diseases, 41(3), 542-548.
M. W. Hayward, and G. Kerley. (2005). Prey preferences of the lion (Panthera leo), Journal of Zoology
267 (issue 3): 309–322
Nat Geo Wild (2015) Hyenas and Lions Face-Off. Nat Geo Wild. National Geographic Channel, 1996-
2015. Web. 8 Dec. 2015.
<http://channel.nationalgeographic.com/wild/built-for-the-kill/videos/hyenas-and-lions-face-off/>.
Pageat, P., & Gaultier, E. (2003). Current research in canine and feline pheromones. Veterinary Clinics:
Small Animal Practice, 33(2), 187-211.
Palomares, Francisco, and Tim M. Caro. (1999). Interspecific killing among mammalian carnivores. The
American Naturalist 153.5: 492-508.
Råberg, L., Grahn, M., Hasselquist, D., & Svensson, E. (1998). On the adaptive significance of stress-
induced immunosuppression. Proceedings of the Royal Society of London B: Biological Sciences,
265(1406), 1637-1641.
Rabinowitz, Alan. (2010). A New Strategy for Saving the World's Wild Big Cats. Interview by E360.
Yale Environment 360 17 Feb. 2010: Print.
Randall, M. (2012, February 3). The Physiology of Stress: Cortisol and the Hypothalamic-Pituitary-
Adrenal Axis. Retrieved from Dartmouth Undergraduate Journal of Science website:
http://dujs.dartmouth.edu/2011/02/the-physiology-of-stress-cortisol-and-the-hypothalamic-pituitary-
adrenal-axis/#.WAWNbOgrKhc
Reiger, I. (1979). Scent rubbing in carnivores. Carnivore, 2(1/2), 17-25.
Ritchie, Euan G., and Christopher N. Johnson. (2009). Predator interactions, mesopredator release and
biodiversity conservation. Ecology letters 12.9: 982-998.
Roelke, M. E., Brown, M. A., Troyer, J. L., Winterbach, H., Winterbach, C., Hemson, G., & Alexander, K.
A. (2009). Pathological manifestations of feline immunodeficiency virus (FIV) infection in wild African
lions. Virology, 390 (1), 1-12.

55
Roth, J. A. (1985). Cortisol as mediator of stress-associated immunosuppression in cattle. In Animal stress
(pp. 225-243). Springer New York.
Salimetrics, L. L. C., & SalivaBio, L. L. C. (2011). Saliva collection and handling advice. Available at
www. salimetrics.com.
Salimetrics, L. L. C., & SalivaBio, L. L. C. (2015). SALIVARY SECETORY IgA INDIRECT ENZYME
IMMUNOASSAY KIT. Available at www. salimetrics.com. https://www.salimetrics.com/assets/-
documents/1-1602n.pdf
Sapolsky, Robert M. (2002). Endocrinology of the Stress Response. Behavioral Endocrinology. By Jill B.
Becker. MIT Press. 409-49. Print.
Sapolsky, Robert M., L. Michael Romero, and Allan U. Munck. (2000). How do glucocorticoids
influence stress responses? Integrating permissive, suppressive, stimulatory, and preparative
actions. Endocrine reviews 21.1: 55-89.
Soini, H. A., Linville, S. U., Wiesler, D., Posto, A. L., Williams, D. R., & Novotny, M. V. (2012).
Investigation of scents on cheeks and foreheads of large felines in connection to the facial marking
behavior. Journal of chemical ecology, 38(2), 145-156.
Terio, K. A., Marker, L., & Munson, L. (2004). Evidence for chronic stress in captive but not free-ranging
cheetahs (Acinonyx jubatus) based on adrenal morphology and function. Journal of Wildlife Diseases,
40(2), 259-266.
Trinkel, Martina, and Gerald Kastberger. (2005). Competitive interactions between spotted hyenas and
lions in the Etosha National Park, Namibia. African Journal of Ecology 43.3: 220-224.
Valeix, M., Hemson, G., Loveridge, A. J., Mills, G., & Macdonald, D. W. (2012). Behavioural
adjustments of a large carnivore to access secondary prey in a human‐dominated landscape. Journal of
Applied Ecology, 49(1), 73-81.
Van Meter, P. E., French, J. A., Dloniak, S. M., & Watts, H. E. (2009). Fecal glucocorticoids reflect socio-
ecological and anthropogenic stressors in the lives of wild spotted hyenas. Hormones and Behavior, 55,
329-337.
Vaz, J., Narayan, E. J., Kumar, R. D., Thenmozhi, K., Thiyagesan, K., & Baskaran, N. (2017). Prevalence
and determinants of stereotypic behaviours and physiological stress among tigers and leopards in Indian
zoos. PloS one, 12(4), e0174711.
Wells, A., Terio, K. A., Ziccardi, M. H., & Munson, L. (2004). The stress response to environmental
change in captive cheetahs (Acinonyx jubatus). Journal of Zoo and Wildlife Medicine, 35(1), 8-14.
Williams, Brianna M., et al., (2014). Prevalence and diversity of Babesia, Hepatozoon, Ehrlichia, and
Bartonella in wild and domestic carnivores from Zambia, Africa. Parasitology research 113.3: 911-918.
Woodroffe R, Ginsberg JR. (2005). King of the beasts? Evidence for guild redundancy among large
mammalian carnivores. See Ray et al., 2005, pp. 154-175
Young, K. M., et al., (2004). Noninvasive monitoring of adrenocortical activity in carnivores by fecal
glucocorticoid analyses. General and comparative endocrinology 137.2: 148-165.

56
APPENDICES

57
APPENDIX 1. COLLECTION PROTOCOLS
Collecting saliva
1. Remove grate from top of device. This can be done either by 1) unfastening one of the rounded
bolts and pivoting the grate to reveal the bucket opening, or 2) unfastening both bolts and
removing the grate.
2. Collect deposited saliva from basin using a 1 to 5 mL syringe, depending on the amount of
saliva visible. Cap with black cap to prevent loss of contents.
3. Label each syringe using a cryolabel, recoding the following information using a sharpie:
▪ Sample number (refer to number 5)
▪ Date of Collection (mnth/day/yr)
▪ Gender (M / F)
4. Record, on the Excel sheet provided, the following information for each sample:
▪ Sample number (matching that on syringe)
▪ Date of Collection
▪ Collection time
▪ Gender of lion depositing sample (M / F)
▪ Approximate amnt of saliva collected (mL)
▪ Internal basin temperature @ time of collection (according to reading on digital
thermometer)
▪ Ambient temperature of enclosure
▪ Zoo activity – in general (busy, moderate, quiet)
▪ Outdoor weather
5. Sample numbering system: label for each sample will begin with SPZ. This provides
information on the location of the sample (Seneca Park Zoo). This will then be followed with
either M or F, depending on which lion deposited the sample. If there are only two lions at this
zoo, just M or F is sufficient. If there are several individuals of the same gender, then a number
would follow assigned to each individual. Following this code (either SPZM or SPZF), put 00
followed by whatever sample number this is collected for that individual. If this is the sixth
sample deposited by the male lion, then the sample number/label would look as follows:
SPZM006 if only one male, or SPZM2_006 if it was male number two of several males.
6. Wipe base and walls of basin dry with paper towel between each collection to prevent mixing
of saliva or contamination. Do this for the grate as well, making sure to wipe each rod
individually and thoroughly.

58
Sample Storage
1. Immediately following collection, samples/syringes should be placed in zoo freezer until they
can be transported either by zoo veterinarian or by myself to Cornell University.
Device Care
1. Replace ice every six to eight hours during testing periods (times the device is out for saliva
collection). When saliva is not being collected, or lions are absent from enclosure, check ice
and replace if necessary. Old ice that has melted can be discarded (remove both buckets to
access the outer bucket housing the ice) and replaced with new ice. Scent wicks should be
replaced biweekly.
2. Wipe base and walls of basin dry with paper towel between each collection to prevent mixing
of saliva or contamination. Do this for the grate as well, making sure to wipe each rod
individually and thoroughly.
Camera Care
1. Batteries and SD cards should be changed on a biweekly basis. At each exchange, please verify
that cameras are set to the following settings: 30 second video; 30 seconds between each
recording.

59
Appendix 2. Device Design and Outer Shell
Upper left: building designs for wooden holding container with attachment detail for bolt and wing nut or acorn nut. Upper right:
rectangular wooden box with collection device attached internally. Both buckets can be removed and replaced from the outside. Lower
left: top of collection device bolted into wooden box. Locking washer holds bolts in place to prevent slipping/dislodging when either
bucket (inner or outer) are removed. Lower right: back wall of wooden box with four slits for threading industrial grade ratchet strap.
Strap will secure this back-panel flush to grated wall of enclosure, while keeping strap outside of enclosure out of reach of lions.
(Sgambelluri, 2017)

60
Appendix 3. Metrics of Success
Lists of (left) measures to determine success of device and (right) possible factors that may
influence success. (Sgambelluri, 2017)

61
Antioch University New England
Environmental Studies Department
Library Permission Slip
I give my permission for one copy of this thesis to be permanently housed in the Antioch New England
Graduate School Library and to circulate to patrons as a part of the library collection. I understand that the
thesis circulation process will be subject to all the rules and regulations of the Antioch New England
Graduate School Library.
Title of Thesis:
Noninvasive Collection of Saliva in Panthera leo
Creation and Validation of a Novel Technique for Health Assessment in Captive African Lions
(Panthera leo)
Student Signature: (on file)
Student Name (Print): Elizabeth Sgambelluri___________________________
Date: __5/8/18__________________

62

63

64

65

66

67