INDICATORS OF NUTRITION AND REPRODUCTION IN FEMALE WOODLAND
CARIBOU RECEIVING SUPPLEMENTAL FEED IN BRITISH COLUMBIA
by
Lauren Rose Elviss
B.Sc., University of Northern British Columbia, 2021

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
IN
NATURAL RESOURCES AND ENVIRONMENTAL STUDIES
(BIOLOGY)

UNIVERSITY OF NORTHERN BRITISH COLUMBIA
September 2025
© Lauren Rose Elviss, 2025

Abstract
Adequate nutrition is closely associated with animal survival, reproduction, and
population growth. Supplementary feeding, the provisioning of food to offset reduced
availability of natural food and / or improve individual and population attributes, is often used as
a conservation tool to reverse the decline of threatened populations. Most woodland caribou
(Rangifer tarandus caribou) populations in British Columbia are declining throughout their
range due to habitat alterations caused by human activities. In response to these declines, various
management strategies have been employed to halt population declines and increase the number
of caribou in BC. Supplemental feeding for caribou conservation began in 2014, with highquality food pellets provided to free-ranging caribou in the Kennedy Siding herd each fall. The
goal of this program is to improve body condition, increase winter survival, and improve
reproductive outcomes. Feeding combined with predator reduction appears to have had a positive
effect on caribou, with the herd growing by approximately 16% per year.
I used multiple indicators of nutritional condition and reproduction to investigate the
effects of supplemental feeding on body condition and population growth among female caribou
in the Kennedy Siding herd. I used body mass, visual body condition scores, fecal and hair
cortisol, fecal triiodothyronine, and hair δ15N to evaluate the changes in physiological and
nutritional condition over the feeding period and to assess interannual variation in condition. I
compared indicators measured in female caribou with a weaned calf in the fall to those without a
calf to assess the effects of lactation and calf rearing on condition. I used a combination of fecal
progesterone measurements and GPS collar movement data to assess the potential effects of fall
supplemental feeding on ovulation, parturition, and calf survival.

ii

My results suggest that females with calves may be more vulnerable to nutritional
limitation compared to females without calves due to the nutritional costs of lactation and calf
rearing. Furthermore, I found differences in indicators of condition between years, highlighting
the need to account for environmental conditions when considering nutrition. Indicators of
nutrition suggested that all caribou had the capacity to become pregnant, regardless of age, body
condition, or whether they had a calf at heel or not. Total parturition rate was 87.5% and calf
survival was 74.3% in the Kennedy Siding herd, which are at the high end of comparable rates
reported in unfed caribou herds in BC. Combined, my findings suggest the mechanism that in
which supplemental feeding contributes to population growth is by supporting the deposition of
body fat and protein reserves, thus increasing the proportion of females that have successful
pregnancies and potentially also the proportion of calves that survive until the fall. My research
represents some of the first longitudinal monitoring of endocrine activity in woodland caribou
and contributes to understanding the role of nutrition in caribou conservation.

iii

Table of Contents

ABSTRACT.................................................................................................................................... II
TABLE OF CONTENTS .............................................................................................................. IV
LIST OF TABLES ........................................................................................................................ VI
LIST OF FIGURES ................................................................................................................... VIII
ACKNOWLEDGMENTS ............................................................................................................ XI
CHAPTER ONE: INTRODUCTION ............................................................................................. 1
CHAPTER 2: PHYSIOLOGICAL BIOINDICATORS REFLECT NUTRITIONAL
DIFFERENCES AMONG FEMALE CARIBOU IN AN ENDANGERED HERD WITH
SUPPLEMENTAL FEEDING ..................................................................................................... 14
INTRODUCTION .......................................................................................................................... 14
Indicators of nutritional condition ........................................................................................ 17
METHODS .................................................................................................................................. 22
Study site ............................................................................................................................... 22
Feeding, identification of individuals, and body condition measurements .......................... 24
Animal care ........................................................................................................................... 25
Sample collection .................................................................................................................. 26
Identification of individuals and aging using DNA .............................................................. 26
Hair preparation for stable isotope analysis ........................................................................ 27
Fecal hormone enzyme linked immunosorbent assay (ELISA) ............................................. 27
Statistical analysis ................................................................................................................ 32
RESULTS .................................................................................................................................... 34
Body mass and condition ...................................................................................................... 34
Hair nitrogen stable isotope (δ15N) ..................................................................................... 36
Fecal glucocorticoid metabolites (FGM) ............................................................................. 40
Fecal T3 metabolites (FTM) ................................................................................................. 44
DISCUSSION ............................................................................................................................... 48
Nutritional limitation in females with calves ........................................................................ 48
Nutritional limitation in yearlings ........................................................................................ 53
Inter-annual variation in nutritional condition .................................................................... 55
Changes in nutritional condition over the feeding period .................................................... 56
Management implications ..................................................................................................... 59
CHAPTER 3: REPRODUCTIVE RATES AMONG FEMALE CARIBOU WITH FALL
SUPPLEMENTAL FEEDING ..................................................................................................... 62
INTRODUCTION .......................................................................................................................... 62
METHODS .................................................................................................................................. 69
Study site .............................................................................................................................. 69
Feeding, identification of individuals, and body condition measurements .......................... 70

iv

Animal care ........................................................................................................................... 72
Sample collection .................................................................................................................. 73
DNA analysis ........................................................................................................................ 73
Fecal progesterone enzyme linked immunosorbent assay (ELISA) ...................................... 74
Statistical analysis ................................................................................................................ 77
RESULTS .................................................................................................................................... 78
Body mass and nutritional condition .................................................................................... 78
Parturition and calf survival rates ........................................................................................ 83
DISCUSSION ............................................................................................................................... 84
Management implications ..................................................................................................... 94
CHAPTER 4: CONCLUSION ..................................................................................................... 96
RESEARCH SUMMARY ................................................................................................................ 96
MANAGEMENT IMPLICATIONS .................................................................................................. 100
REFERENCES ........................................................................................................................... 102
APPENDICES ............................................................................................................................ 120
SUPPLEMENTAL MATERIAL FOR CHAPTER TWO ....................................................................... 120
Linear mixed effects models ................................................................................................ 120
Hair cortisol ........................................................................................................................ 122
SUPPLEMENTAL MATERIAL FOR CHAPTER THREE .................................................................... 127
Linear mixed effects models ................................................................................................ 127
Individual fecal progesterone profiles ................................................................................ 127

v

List of Tables
Table 1: Hypothesis category and structure of eight candidate models compared to predict fecal
glucocorticoid, fecal triiodothyronine (T3), and hair δ15N in female caribou (Rangifer tarandus
caribou) sampled in 2021and 2022 at the Kennedy Siding feeding site in British Columbia,
Canada........................................................................................................................................... 33
Table 2: Number of female caribou (Rangifer tarandus caribou) in each reproductive age class
with visible ribs at the Kennedy Siding feeding site used in our analysis. Most of the individuals
with visible ribs were adults with a calf........................................................................................ 35
Table 3: Top 2 of 8 linear mixed-effects models used to predict δ15N concentrations in hair
samples (n=52) from female caribou (Rangifer tarandus caribou) sampled in fall and winter of
2021 and 2022 at the Kennedy Siding feeding site in British Columbia, Canada. Competing
models within two ∆AICc are shown, as well as the null for reference. ...................................... 37
Table 4: Top 3 of 8 linear mixed-effects models used to predict fecal glucocorticoid metabolites
in female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021 and 2022 at the
Kennedy Siding feeding site in British Columbia, Canada. Competing models within two ∆AICc
are shown, as well as the null for reference. ................................................................................. 41
Table 5: Top 2 of 8 linear mixed-effects models used to predict female caribou (Rangifer
tarandus caribou) FTM concentrations in caribou sampled in fall and winter of 2021 and 2022 at
the Kennedy Siding feeding site in British Columbia, Canada. Competing models within two
∆AICc are shown, as well as the null for reference. ..................................................................... 45
Table 6: Body condition (assessed as visibility of ribs) of female caribou (Rangifer tarandus
caribou) in each reproductive age class at the Kennedy Siding feeding site in British Columbia,
Canada, in 2021 and 2022. ........................................................................................................... 80
Table 7: Ranking of four candidate general linear mixed-effects models used to predict fecal
progesterone metabolite concentrations in fecal samples (n = 397) from female caribou (Rangifer
tarandus caribou) sampled between fall and winter of 2021 and 2022 at the Kennedy Siding
feeding site in British Columbia, Canada. .................................................................................... 83
Table 8: Parturition and calf survival status of 40 female caribou (Rangifer tarandus caribou)
with radio collars in the Kennedy Siding herd, central British Columbia, Canada, between 2020
and 2024. Parturition and calf mortality was estimated using a model developed by DeMars et al.
(2013). ........................................................................................................................................... 83
Table 9: Parturition rates of caribou (Rangifer tarandus caribou) in five populations of woodland
caribou in British Columbia and Alberta, Canada. ....................................................................... 89
Table 10: Caribou (Rangifer tarandus caribou) calf survival in four populations of woodland
caribou in British Columbia and Alberta, Canada. ....................................................................... 91

vi

Table 11: The eight linear mixed-effects models used to predict fecal glucocorticoid
concentrations in female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021
and 2022 at the Kennedy Siding feeding project site in British Columbia, Canada. .................. 120
Table 12: The eight linear mixed-effects models used to predict fecal triiodothyronine
concentrations in female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021
and 2022 at the Kennedy Siding feeding project site in British Columbia, Canada. .................. 120
Table 13: The eight linear mixed-effects models used to predict hair δ15N concentrations in
female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021 and 2022 at the
Kennedy Siding feeding project site in British Columbia, Canada. ........................................... 121
Table 14: The eight linear mixed-effects models used to predict fecal progesterone metabolite
concentrations in female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021
and 2022 at the Kennedy Siding feeding project site in British Columbia, Canada. .................. 127
Table 15: Mean fecal progesterone metabolite concentrations (± standard error) of the follicular
(baseline) and luteal phases from 14 individual female caribou (Rangifer tarandus caribou)
sampled between fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in British
Columbia, Canada. The reproductive age classes are adult without a calf at heel (A), adult with a
calf at heel (AC), and yearling females (Y). ............................................................................... 130

vii

List of figures
Figure 1: Fecal ample collection at the feeding site at Kennedy Siding, British Columbia,
Canada, in 2021 and 2022. A) collecting fecal pellets from female caribou (Rangifer tarandus
caribou). B) A female caribou accessing a feeder by standing on the platform. C) Hair sample
collection from snares set on the feeder ........................................................................................ 13
Figure 2: Examples of female caribou (Rangifer tarandus caribou) at the Kennedy Siding
feeding site considered in poor body condition and female caribou considered in good body
condition. The caribou in images A and B are considered in poor condition based on their visible
ribs, spine, and hip bones in the trail camera images. The caribou in images C and D are
considered in good condition based on the lack of visible ribs or other bones. ........................... 25
Figure 3: Parallelism of cortisol (A) and T3 (B) standards (solid line) and a serially diluted
extract from caribou (Rangifer tarandus caribou) fecal samples (dashed line). .......................... 32
Figure 4: Body mass (kg) of adult female caribou (Rangifer tarandus caribou) without a calf at
heel (A) and with a calf at heel (AC) compared with yearling females (Y) using the feeding site
at Kennedy Siding, British Columbia, Canada, in 2021 and 2022. The centre line of each box
represents the median and the outer limits represent the 25th and 75th percentiles. The lines
extending below and above each box represent 1.5 times the interquartile range for each
percentile. ...................................................................................................................................... 35
Figure 5: Coefficients and 95% and 85% confidence intervals (CI) for variables in the top three
models (< 2 ∆AICc) predicting hair δ15N concentrations in female caribou (Rangifer tarandus
caribou) sampled in fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in
British Columbia, Canada (n = 52). The baseline category for reproductive age class is adult
without a calf at heel. The baseline category for body condition is poor (ribs visible). ............... 38
Figure 6: (A) The highest hair δ15N concentrations measured in caribou (Rangifer tarandus
caribou) sampled at Kennedy Siding, BC, Canada, were found in adults without a calf (A) and 1year olds (Y). Adults with a calf (AC) had the lowest values. The center point represents the
adjusted prediction, and the error bars represent the adjusted predictions ± the standard error. (B)
Hair δ15N concentrations were higher in 2021 compared to 2022. The center point represents the
adjusted prediction, and the error bars represent the adjusted predictions ± the standard error. .. 40
Figure 7: Coefficients and 95% and 85% confidence intervals (CI) for variables in the top three
models (< 2 ∆AICc) predicting FGM concentrations in female caribou (Rangifer tarandus
caribou) sampled in fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in
British Columbia, Canada (n = 432). The baseline category for reproductive age class is adult
without a calf at heel. The baseline category for body condition is poor (ribs visible). ............... 42
Figure 8: (A) Concentrations of FGMs decreased over time during the 2021 and 2022 feeding
seasons in female caribou (Rangifer tarandus caribou) using the Kennedy Siding site, however
the decline was more significant in 2022. The relationship between log transformed FGM
concentrations and sampling date was predicted from the top model with other terms held

viii

constant at their mean value. Sampling date was standardized to a mean of zero and a standard
deviation of one prior to analysis. (B) The highest FGM concentrations were found in yearlings
(Y), followed by adults with calves (AC). Adults without calves (A) had the lowest values. The
center point represents the adjusted prediction value, and the error bars represent the adjusted
predictions ± the standard error. ................................................................................................... 44
Figure 9: Coefficients and 95% and 85% confidence intervals (CI) for variables in the top three
models (< 2 ∆AICc) predicting FTM concentrations in female caribou (Rangifer tarandus
caribou) sampled in fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in
British Columbia, Canada (n = 385). The baseline category for reproductive age class is adult
without a calf at heel. The baseline category for body condition is ribs visible. .......................... 46
Figure 10: The relationships between sampling day, body mass, and reproductive age class on
the concentration of fecal triiodothyronine metabolites (FTM) in feces collected from female
caribou (Rangifer tarandus caribou) at Kennedy Siding, British Columbia. The distribution of
FTMs was normalized with a square root transformation prior to analysis. The distribution of
body mass was normalized using a square root transformation and both sampling day and
normalized body mass were standardized to a mean of zero and a standard deviation of one prior
to analysis. The concentration of FTMs was predicted from the top model with all other terms
held constant at their mean value. (A) The concentration of FTMs declined over time during both
the 2021 and 2022 feeding periods. (B). The concentration of FTMs declined with increasing
body mass of caribou. (C) Adults without a calf at heel had lower FTM concentrations than
yearlings and similar concentrations to adults with a calf. The center point represents the
adjusted prediction, and the error bars represent the adjusted predictions ± the standard error. .. 48
Figure 11: Examples of female caribou (Rangifer tarandus caribou) at the Kennedy Siding
feeding site considered in poor body condition and female caribou considered in good body
condition. The caribou in images A and B are considered in poor condition based on their visible
ribs, spine, and hip bones in the trail camera images. The caribou in images C and D are
considered in good condition based on the lack of visible ribs or other bones. ........................... 72
Figure 12: Parallelism of progesterone standards (solid line) and a serially diluted pooled sample
extract from female caribou (Rangifer tarandus caribou) fecal samples (dashed line) collected in
2021 and 2022 at the Kennedy Siding feeding site in British Columbia, Canada. ....................... 76
Figure 13: Body mass (kg) of adult female caribou (Rangifer tarandus caribou) without a calf at
heel (A) and with a calf at heel (AC) compared with yearling females (Y) using the feeding site
at Kennedy Siding, British Columbia, Canada, in 2021 and 2022. The centre line of each box
represents the median and the outer limits represent the 25th and 75th percentiles. The lines
extending below and above each box represent 1.5 times the interquartile range for each
percentile. The dashed line represents the median adult female body mass (132.5 kg). The dotted
line represents 70% of the median adult female body mass (92.75 kg). Sample sizes are shown at
the bottom of the panel. ................................................................................................................ 80
Figure 14: Fecal progesterone metabolite concentrations adult female caribou (Rangifer tarandus
caribou) without a calf at heel (A) and with a calf at heel (AC) compared with yearling females

ix

(Y) using the feeding site at Kennedy Siding, British Columbia, Canada, in 2021 and 2022. The
centre line of each box represents the median and the outer limits represent the 25th and 75th
percentiles. The lines extending below and above each box represent 1.5 times the interquartile
range for each percentile. The dotted line represents the baseline progesterone metabolite
concentration derived from a pooled male sample mass (162.8 ng/g). Sample sizes are shown at
the top of the panel. ....................................................................................................................... 82
Figure 15: Model validations from each top model set for caribou (Rangifer tarandus caribou)
fecal glucocorticoids, fecal T3, and hair δ15N. The histograms denote the frequency of residuals,
and the Normal Q-Q plots denote standardized residuals against the values that would be
expected under normality from each top model ......................................................................... 122
Figure 16: Parallelism of cortisol standards with a serially diluted extract from caribou (Rangifer
tarandus caribou) guard hair (dashed line). ............................................................................... 124
Figure 17: Cortisol concentrations in hair samples collected from female caribou (Rangifer
tarandus caribou) at Kennedy Siding in 2021 and 2022 (n = 23). The boxes represent the median
value (center line) and the 25th and 75th percentiles. The highest hair cortisol concentrations
were found in yearlings (Y) and adults with calves (AC). Adults without calves (A) had the
lowest values. The lines represent 1.5 times the interquartile range for each percentile. Sample
sizes are shown at the bottom of the panel. ................................................................................ 126
Figure 18: Longitudinal fecal progesterone metabolite profiles for 14 female caribou (Rangifer
tarandus caribou) sampled between fall and winter of 2021 and 2022 at the Kennedy Siding
feeding site in British Columbia, Canada. Individual identities are displayed at the top of each
plot. The dashed line indicates an individual’s baseline value used to differentiate between the
follicular and luteal phases.......................................................................................................... 132

x

Acknowledgments
This thesis was a collaborative effort that would not have been possible without the
considerable help and support from multiple organizations and people along the way. Funding
for my research was provided by the Fish and Wildlife Compensation Program (FWCP). The
Kennedy Siding project is located on the traditional territory of the McLeod Lake Indian Band
and operates with their vital support and participation.
I am grateful for the guidance and support of my supervisory committee. Thank you to
my supervisor Heather for your unwavering support, guidance, and kindness throughout the ups
and downs of this thesis. I am forever grateful to have had you as a mentor. Thank you to Doug
for sharing your enthusiasm and expertise. I appreciate thoughtful comments along with our
many discussions on the road to Kennedy Siding that encouraged me to always think deeply and
critically. Thank you to Laura for all your support and mentorship in the lab and for sharing all
your knowledge on hormones. Without you I’d probably still be running assays. Thank you to
Scott for sharing your expertise on caribou ecology and for all your thoughtful comments
throughout the writing process.
Thank you to my field assistants Merisa and Des for all your hard work and for enduring
some pretty cold and snowy conditions to help me collect hundreds of poop samples. Thank you
to Carlie, Alaysia, Andrew, and Caroline for all your help in the lab washing, weighing, and
grinding all that poop and hair. Thank you to the communities of McLeod Lake and Mackenzie
for being so welcoming and kind while I was there for field work. Thank you to all the many
people who helped interpret the thousands of trail camera photos.
To my wonderful friends that made my personal life at UNBC and around Prince George
fun, entertaining and rewarding, I will always be thankful. Special thanks go to Shaye and Madi
for keeping me sane and making sure I remembered to eat and to Carlie, Ben, Andrew, and
Caroline for all the essential lab treat times.
I could not have completed this study without the support of my family. To my parents,
Gill and Dave, I made it to this point because of you and your constant encouragement and love.
Finally—the caribou. I feel incredibly fortunate to have had the opportunity to learn so
much from these animals. This project has only increased my respect for these beautiful
creatures and the landscape they call home.

xi

Chapter One: Introduction
“So that made it quite important to us … if they looked after us way back then now I guess it’s
our turn to take care of them and we always try to help out and in our own way to increase
herds.”
- Alec Chingee, Elder, McLeod Lake Indian Band
Habitat alteration, including conversion, degradation, and fragmentation, is often the
primary cause of species endangerment and a leading factor hindering species recovery
worldwide (Myers et al. 2000, Newbold et al. 2015). Habitat alteration has been implicated in the
extinction or endangerment of almost 800 terrestrial mammal species globally (IUCN 2025) and
in Canada, has played a major role in declines in woodland caribou (Rangifer tarandus caribou;
(McLoughlin et al. 2003a, Wittmer et al. 2007, Wasser et al. 2011).
Once abundant across Canada, woodland caribou have experienced population declines
over the past 50 years and in 2019 were among the most recently extirpated large mammals in
the contiguous United States (Festa-Bianchet et al. 2011, Moskovitz 2019). The decline of
caribou has important ecological and cultural consequences. Caribou are part of the normal
functioning of boreal and mountain ecosystems and have immense cultural and spiritual value to
Indigenous people (Lamb et al. 2022). The decline of caribou presents a significant threat to a
way of life, culture, and identity (Lamb et al. 2022). For example, many Indigenous nations in
BC have ceased hunting caribou completely (Tŝilhqot’in National Government 2019, Nak’azdli
Whut’en 2021, Lamb et al. 2022, McLeod Lake Indian Band n.d.), “making the sacrifice of
sustenance for conservation” (Tŝilhqot’in National Government 2019).
Among woodland caribou, the Mountain ecotype in British Columbia (BC) has
experienced declines of 50–60% (Environment Canada 2014, Environment and Climate Change

1

Canada 2020, McNay et al. 2022). The Mountain ecotype is divided into two populations—–the
Southern Mountain and Northern Mountain populations (Zimmerman et al. 2025). The Southern
Mountain population is further divided into three designatable units—the Northern Group,
Central Group, and Southern Group—and currently consists of 32 subpopulations or herds
(Zimmerman et al. 2025). The Southern Mountain population is particularly imperiled, with nine
herds fully extirpated and two functionally extirpated between 2000 and 2017 (Zimmerman et al.
2025). Many of the remaining populations are small (<100 individuals) and currently continue to
decline (Johnson et al. 2015, Environment and Climate Change Canada 2018, Moskovitz 2019,
Serrouya et al. 2019).
The primary cause of the decline in caribou populations has been attributed to alteration
of boreal and mountain ecosystems caused by human resource extraction activities such as
clearcut logging, oil and gas development, and mining (Festa-Bianchet et al. 2011, Wasser et al.
2011, Johnson et al. 2015). The proximate cause of the decline in numbers is thought to be
unsustainable predation caused by habitat changes that have increased the abundance, access,
and mobility of predators such as wolves (Canis lupus); Wasser et al. 2011, Hervieux et al. 2014,
DeMars and Boutin 2018). While early seral stage forests that result from habitat disturbance
tend to have higher primary productivity and higher vascular plant abundance, they also support
higher predator abundances and caribou will trade off access to forage for predator avoidance
(McGreer et al. 2015, Viejou et al. 2018). Caribou declines are further exacerbated by climate
change and the associated changes in snowpack, fire regimes, and forage availability and quality
(Festa-Bianchet et al. 2011). Caribou tend to select mature forests with high lichen abundance in
winter (Fortin et al. 2008, Joly et al. 2010, McGreer et al. 2015). Variation in winter temperature
and precipitation can alter the depth and hardness of snowpack and, therefore, the accessibility of

2

both arboreal and terrestrial lichens (Kinley et al. 2007, Tyler 2010, DeMars et al. 2021).
Warmer and drier conditions have led to larger wildfires and extended fire seasons over the past
30 years (Gustine et al. 2014b), decreasing the area of mature forests and the availability of
lichens for caribou (Cichowski et al. 2022, Stanton et al. 2023, Russell et al. 2025). Climate
change may shift the timing of plant emergence, reduce nutrient content, and alter species
composition, cover, and growth, creating nutritional phenological mismatches between caribou
and their forage resources (Zamin et al. 2017, Mallory and Boyce 2018, Twining et al. 2022,
Webber et al. 2022, Stanton et al. 2023).
Habitat changes and associated changes in predator-prey dynamics are contributing to
declines among the Central Group of Southern Mountain Caribou The Central Group occurs on
the east side of the Rocky Mountains, north of Kakwa Provincial Park in the south, to the
southern shore of the east arm of Williston Lake in the north (Environment Canada 2014,
Environment and Climate Change Canada 2020). There are currently five herds defined in the
Central Group, the Quintette, Narraway, Burnt Pine, Klinse-Za (also known as Moberly Scott),
and Kennedy Siding herds (COSEWIC 2014, Environment and Climate Change Canada 2018,
2020). The Kennedy Siding herd declined from a high of 120 in 2007 to 41 in 2012, the
Quintette herd from 265 in 2002 to 106 in 2014 and the Klinse-Za and Scott East herds
combined from 191 in 1997 to 16 in 2013 (Seip and Jones 2018, Serrouya et al. 2019). The Burnt
Pine subpopulation was confirmed extirpated in 2014 (COSEWIC 2014).
Various management strategies have been used to halt declines and increase the
population of herds in the Central Group. Federal recovery strategies for caribou identify habitat
restoration and protection as the key management actions for recovering self-sustaining
populations, however, habitat restoration is expected to take decades, and with continued

3

declines, more immediate actions are needed to keep caribou on the landscape (Environment
Canada 2014, Environment and Climate Change Canada 2020, Lamb et al. 2024b). The Province
of BC began a wolf reduction program in 2015 and, over five years, the density of wolves in the
area was reduced from 12.6 wolves/1000 km2 to less than 2 wolves/1000 km2 (Bridger 2019). In
2014, a supplemental feeding program was initiated for the Kennedy Siding herd with the goal of
improving overwinter survival and reproduction by improving nutritional condition in autumn
(Heard and Zimmerman 2021). These programs have increased Central Group caribou numbers
since their implementation. The Quintette herd, which only receives wolf reduction, had a
population growth rate (λ) of 1.08/yr and a change in lambda (Δλ) of 0.13 (Heard and
Zimmerman, 2021). Combining management treatments appears to have an added beneficial
effect on populations. Supplemental feeding and predator control resulted in the Kennedy Siding
herd having a λ of 1.16/yr and a Δλ of 0.25 (Heard and Zimmerman, 2021).
Supplemental feeding is a wildlife management tool that uses nutrition to maintain or
increase a population (Dunkley and Cattet 2003, Murray et al. 2016). For example, supplemental
feeding has been used extensively to maintain and increase populations of endangered avian
scavenger species worldwide (Wilbur et al. 1974, Gilbert et al. 2007, Cortés-Avizanda et al.
2016). Nutrition underpins essentially all life processes and plays a direct role in caribou
population dynamics by influencing reproduction and survival (Chan-McLeod et al. 1994, Parker
et al. 2009). As environmental conditions change, the survival of endangered species such as
caribou will depend on whether habitats can meet nutritional demands while also accounting for
other environmental pressures.
Most studies of nutritional ecology in northern ungulates, including caribou, have
focused on winter foraging conditions, as winter is the season when food is assumed to be most

4

limiting. There is, however, increasing evidence to suggest that late spring to early autumn may
be when nutritional limitation occurs for caribou (Crête and Huot 1993, Dale et al. 2008,
Couturier et al. 2009, Cook et al. 2013, 2021, Kelly 2020, Heard and Zimmerman 2021, Denryter
et al. 2022). In northeastern BC, mortality of adult female Boreal caribou due to starvation is
highest between June and October (Cook et al. 2021). During the summer, the energy costs of
lactation plus that of recovering from previous winter weight loss can more than double daily
energy requirements compared to those of non-lactating females in the summer/autumn (Barboza
and Parker 2008, Parker et al. 2009). Female caribou lose body mass for approximately 3 weeks
following calving because of the high costs of lactation (Parker et al. 2009). Adequate nutrition
is needed for a caribou to both lactate and replenish body reserves before winter (Cook et al.
2004, 2013, Dale et al. 2008, Parker et al. 2009). The body reserves accumulated over the
summer influence the probability of conceiving and carrying a fetus to term, the growth of
offspring in utero and post partum, and over winter survival (Crête and Huot 1993, Cameron and
Hoef 1994, Chan-McLeod et al. 1994, Adams and Dale 1998, Barboza and Parker 2006, Parker
et al. 2009). Nutrition also plays a major role in determining the age of puberty in caribou and
other large herbivores (Festa-Bianchet et al. 1994, Gaillard et al. 2000). Maternal condition
affects neonatal birth weights, which in turn impacts calf viability and susceptibility to predation
(Cameron et al. 1993, Chan-McLeod et al. 1994, Adams 2005, Taillon et al. 2012). Caribou are
generally predator sensitive foragers and will trade off high quality foraging opportunities as
vegetation greens up at lower elevations to reduce predation risk, especially when they have a
calf (Gustine et al. 2006, Jones et al. 2007, Parker et al. 2009, Viejou et al. 2018). If caribou
cannot recover body stores lost to lactation and calf rearing, they may forgo reproduction that
year in favour of investing energy into self maintenance (Cameron 1994, Chan-McLeod et al.

5

1994, Parker et al. 2009). Widespread nutritional inadequacies will result in a gradual reduction
in population productivity that increases with increasing nutritional inadequacy (Crête and Huot
1993). For an already imperilled population, any additional effect on population growth rate
could compromise recovery and conservation actions.
Providing food to declining populations and/or populations that are food limited can
have a positive effect on population growth via improvements in individual reproduction,
survival, and body condition (Milner et al. 2014, Murray et al. 2016, Horstkotte et al. 2022).
Supplemental feeding may enhance female reproductive success via improvements in body
condition resulting in increased pregnancy and parturition rates, decreased age of reproductive
maturity, and/or improved milk production which in turn improves calf survival (Hata et al.
2021, Horstkotte et al. 2022, Heffelfinger et al. 2023, Cao et al. 2025). The effect of
supplemental feeding on body condition and reproduction can be assessed using multiple
physical and chemical indices. Measurements of body mass and visual body condition scoring
can be combined with chemical techniques, such as stable isotope and hormone analysis, to
provide a more comprehensive assessment of nutritional condition (Delgiudice et al. 1990,
Barboza et al. 2004, Parker et al. 2009, Taillon et al. 2011, Pérez-Flores et al. 2016, Schiffmann
et al. 2017, Found et al. 2022).
The naturally occurring stable isotope ratios of N (15N/14N relative to atmospheric N;
δ15N) in tissues can provide useful insights into animal nutrition (Gustine et al. 2011, 2014a,
Barboza et al. 2020). Nitrogen (N) is essential for the synthesis of body protein and when dietary
N is inadequate, animals use N from their own body proteins and microbial protein in the rumen
for the maintenance of critical tissues(Lavery and Ferris 2021). Animals using endogenous
protein reserves for maintenance undergo nitrogen recycling and preferentially return 15N to their

6

N pool. This results in tissues such as hair becoming more enriched in 15N because amino acids
are now being synthesized from a more 15N rich source than the diet (Minagawa and Wada 1984,
Kelly 2000, Fuller et al. 2005, Habran et al. 2010). Nutritional stress can thus increase δ15N
values in tissues when an animal is in a net catabolic state associated with lean muscle
degradation (N intake < N excretion or negative N balance; Fuller et al. 2005, Parker et al. 2005,
Habran et al. 2010, Gustine et al. 2011). Changes in physiological status can also influence δ15N
values of various tissues in herbivores. Many herbivores rely on body stores of fat and protein to
meet the high N demands of fetal development and lactation (Parker et al. 2009, Taillon et al.
2013, Gustine et al. 2014a). The net anabolic state associated with protein synthesis (N intake >
N excretion or positive N balance) during gestation and lactation causes the 15N isotope ratio of
the maternal body protein pool to decline (Fuller et al. 2004). The body N pool of lactating
animals may be further depleted in 15N as it is transferred to offspring though milk, though the
level of N fractionation between mother and offspring is species specific and may be influenced
by the protein quality or concentration of the mother’s diet (Jenkins et al. 2001, Habran et al.
2010, Endo and Kobayashi 2022). Predation and disturbance stress is correlated with lower δ15N
in herbivores because stress hormones such as glucocorticoids can supress protein assimilation
while promoting protein degradation and amino acid export. (De Feo 1996, Deschner et al. 2012,
Found et al. 2022, Jimeno and Verhulst 2023).
Nutritional and physiological condition, as well as reproduction, can be further examined
using hormones. Glucocorticoids (GCs) are a class of steroid hormones that are released as part
of the stress response — physiological and behavioral coping mechanisms that help to neutralize
the effects of the stressor and to reestablish homeostasis (Romero 2004, Reeder and Kramer
2005). A stressor is a noxious or unpredictable stimulus caused by any event or situation that

7

disrupts homeostasis and poses an immediate or perceived threat to an animal (Romero 2004,
Reeder and Kramer 2005). Stressors can be internal, external, or psychological (Reeder and
Kramer 2005). GCs mobilize energy reserves by converting glycogen to glucose, which provides
an animal with the energy to respond to immediate stressors (Reeder and Kramer 2005, Baker et
al. 2013). A chronic elevation in GCs due to a persistent stressor can negatively impact an animal
by supressing the immune system, depleting body reserves, and reducing reproductive success
(Sapolsky et al. 2000, Romero 2004, Reeder and Kramer 2005, Karaer et al. 2023). GCs are also
closely associated with metabolism. When metabolic rate is low, GCs are maintained at low
levels (Sapolsky et al. 2000, Jimeno and Verhulst 2023). Increases in metabolic rate may be
anticipated (i.e., normal daily/seasonal fluctuations) or unanticipated (sudden temperature
changes, psychological stress, etc.; Sapolsky et al. 2000, Jimeno and Verhulst 2023). GCs will
rise with both gradual and acute increases in metabolic rate, resulting in an increase in circulating
glucose at a rate matching metabolic requirements (Sapolsky et al. 2000, Jimeno and Verhulst
2023).
GCs can be examined alongside thyroid hormones, such as triiodothyronine (T3), to
distinguish nutritional stress from other physiological stressors (Wasser et al. 2010, Ayres et al.
2012, Behringer et al. 2018). T3 is a modified amino acid that regulates metabolism, allowing an
animal to adapt to changes in environmental conditions and nutritional availability and
requirements. T3 levels decrease when food is scarce to decrease metabolic rate and conserve
energy (Blake et al. 1991, Wasser et al. 2010, Pasciu et al. 2024). T3 is also involved in the
progression and timing of developmental processes (Todini 2007, Behringer et al. 2018) and
plays an important role in thermogenesis (Todini 2007, Behringer et al. 2018, Hunninck et al.
2020). Unlike GCs, T3 is relatively unaffected by psychological stressors such as predation,

8

disease, disturbance, and social interactions, but is highly responsive to nutritional stress.
(Kitaysky et al. 2005, Wasser et al. 2010). Both thyroid and GC systems work closely together to
regulate energy availability (Wasser et al. 2011, Behringer et al. 2018, Jimeno and Verhulst
2023). Examining the relative differences between these two hormones can help understand the
relative importance of nutritional and non-nutritional stressors on wildlife (Wasser et al. 2011,
Behringer et al. 2018, Gort‐Esteve et al. 2024).
Understanding basic reproductive processes and components of individual fitness can aid
species management and conservation because they are important factors influencing population
dynamics in wild animals (Schwarzenberger 2007, Sontakke 2018). The general health of a
population is reflected in its capacity to reproduce, because reproduction is usually the first
physiological loss resulting from severe stress (Bleke et al. 2021). Progesterone is a steroid
hormone produced by the corpus luteum in the ovary after ovulation and is responsible for the
maintenance of pregnancy in mammals (Smith 2009, Crowe 2016, Shrestha 2021). Caribou are
seasonally polyestrous, having multiple estrous cycles during a single breeding season (McEwan
and Whitehead 1972, Bergerud 1975, Ropstad et al. 1995a). The mammalian estrous cycle
consists of a follicular and luteal phase (Crowe 2016). During the follicular phase an ovarian
follicle and its enclosed oocyte mature in preparation for ovulation and fertilization (Crowe
2016). Progesterone levels are low during this phase (O’Connor et al. 2013). During the luteal
phase is the corpus luteum, a transient endocrine organ formed from the remnants of a mature
follicle after ovulation, is formed and secretes progesterone (Asher 2011, O’Connor et al. 2013,
Shrestha 2021). If pregnancy does not occur, the corpus luteum degenerates and progesterone
levels decline and a new estrous cycle begins with another opportunity to establish a pregnancy
(Asher 2011, Shrestha 2021). If pregnancy is established, the luteal lifespan is prolonged and the

9

corpus luteum continues to produce progesterone along with the placenta through the gestation
period (Asher 2011). The effect of nutrition on progesterone secretion and ovulation has been
extensively studied in domestic ruminants. Improved pre-ovulatory nutrition increases the size of
the ovulatory follicle and the progesterone-secreting ability of the corpus luteum (Robinson et al.
2006). For cattle, post-partum nutrition plays a major role in the timing of the onset of oestrous
cyclicity after calving, the normality of its expression, and conception rate (Robinson et al.
2006). Negative energy balance is associated with a delay to first ovulation, cycles which are
longer than normal (prolonged corpus luteum), and failure to ovulate again at an appropriate time
in dairy cattle (Wathes et al. 2007). Negative energy balance and body condition loss are related
to reduced serum progesterone concentrations during the breeding period and to lower pregnancy
rates in dairy cattle (Butler 2003). In beef cows, those in good body condition at calving and
those receiving high food intakes after calving have shorter anoestrous periods than thinner and
less well-nourished animals (Robinson et al. 2006). In sheep, undernutrition reduces the number
of follicles that develop and therefore the number that are available to ovulate (Robinson et al.
2006). Measures of progesterone can be combined with analysis of movement patterns using
GPS radio-telemetry data to infer parturition and survival of neonatal offspring (DeMars et al.
2013).
Traditionally, blood samples were used for the analysis of hormones and isotopes,
however the stress and hazards of capture and restraint can be detrimental to an animals' health
and may bias the results of hormone analysis (Schwarzenberger 2007, Morden et al. 2011). For
example, GCs are especially sensitive to short term stressors such as capture (Schwarzenberger
et al. 2000, Hunt et al. 2019). Fecal and hair analysis has been developed as alternative, noninvasive techniques for studies of stress and reproductive physiology in wildlife

10

(Schwarzenberger 2007). Stable isotopes and hormones in hair reflect physiological state over
the period the hair was grown and can therefore be related to longer-term life history stages such
as pregnancy and lactation (Fuller et al. 2004, Bryan et al. 2013b). Hormones like GCs, T3, and
progesterone are metabolized by the liver into metabolites that are excreted in feces and/or urine
(Wasser et al. 2010, Brown 2018, Hunt et al. 2019). Fecal metabolites reflect the endocrine
activity of an animal over the gut transit time prior to sampling (Karaer et al. 2023). Fecal
sampling also enables frequent data collection without disturbing the animal, allowing for
longitudinal studies that can be used to assess changes in reproductive activity and stress
(Schwarzenberger 2007). Very few longitudinal studies that monitor endocrine activity exist for
caribou, likely due to difficulties in obtaining repeated samples from wild populations (Messier
et al., 1990a, Morden et al., 2011, Joly et al., 2015).
As environments continue to change the survival of many species will depend on whether
habitats are able to provide the necessary nutritional resources. Since nutrition is fundamental to
both individual performance and population productivity, it should be a key consideration for
managers aiming to develop effective management policies and practices (Parker et al., 2009).
The information gained from measures of physiological and nutritional stress, as well as their
reproductive consequences, can inform conservation and recovery strategies for endangered
populations. In this thesis, I used multiple indicators to investigate the mechanisms by which
reproduction and supplemental feeding could affect female caribou and population growth of the
Kennedy Siding herd. I used noninvasive methods to collect body condition measurements, fecal
pellets, and hair samples from female caribou using the Kennedy Siding supplemental feeding
site during autumn and winter in 2021 and 2022 (Figure 1). Following sample collection, I
validated assays for fecal GC, T3, and progesterone metabolites and applied them to my research.

11

Finally, I drew on data from the literature and BC Caribou Recovery Program in addition to my
field-collected data to put my results in context with other herds in BC.
I predicted that indices of nutritional condition would differ among reproductive classes and age
classes. I also predicted that concentrations of fecal GC and T3 metabolites would change over
time in response to supplemental feeding. Finally, I predicted that improving female body
condition would increase the proportion of females that bred each year and improve parturition
and calf survival rates. My specific research objectives were to:
A. Compare body mass and condition, GCs and T3 in fecal pellets, and δ15N in hair samples
among reproductive age classes (yearlings, adults with calves, and adults without calves)
to assess the effects of reproduction on physiological and nutritional condition;
B. Determine the effect of supplemental by comparing temporal changes in fecal
glucocorticoid (GC) and triiodothyronine (T3) concentrations over the feeding period.
C. Examine progesterone in fecal pellets to assess ovarian activity and test for a relationship
between progesterone measurements and body condition, age, and previous reproductive
status and;
D. Use GPS location data to assess parturition and neonatal survival rates and compare these
results to other populations to assess potential impacts of autumn nutrition on
reproductive success.

12

Figure 1: Fecal ample collection at the feeding site at Kennedy Siding, British Columbia, Canada, in 2021 and 2022. A)
collecting fecal pellets from female caribou (Rangifer tarandus caribou). B) A female caribou accessing a feeder by standing on
the platform. C) Hair sample collection from snares set on the feeder

In addition to this introduction (chapter 1), my thesis includes two data chapters and a
research synthesis (chapter 4). I present methods and discuss results in chapters 2 (for objectives
A and B) and 3 (for objectives C and D). Both chapters 2 and 3 were written as manuscripts for
publication, thus there is some information repeated in both. Finally, in chapter 4, I integrate the
findings of both chapters 2 and 3. To acknowledge the contributions of many to this work,
chapters 2 and 3 are written in the first-person plural whereas chapters 1 and 4 are written in the
first-person singular. The results of this study contribute to an increased understanding of
physiology and reproduction in caribou and provide insight into the effects of nutritional
resources in the summer and autumn on caribou physiology. This study demonstrates the
potential for supplemental feeding to enhance caribou conservation and can be used to inform
conservation strategies and land-use plans for Central Mountain caribou.

13

Chapter 2: Physiological bioindicators reflect nutritional differences among female caribou
in an endangered herd with supplemental feeding
Introduction
Nutrition underpins essentially all animal life processes and plays a direct role in
population health by influencing reproduction and survival (Chan-McLeod et al. 1994, Parker et
al. 2009). As environments continue to change rapidly at both global and local scales, the
survival of many species will depend on whether habitats are able to provide the necessary
nutritional resources. Nutritional condition, the state of an animal’s body components (i.e., fat
and protein) affects many critical aspects of its life history such as juvenile growth rates, adult
mass gain, body condition, probability of pregnancy, over-winter survival, timing of parturition,
and neonatal birth mass and survival (Parker et al. 2009). Since nutrition is fundamental to both
individual performance and population productivity, it should be a key consideration for
managers aiming to develop robust and effective management policies and practices (Parker et
al. 2009).
One strategy used in endangered species management to improve nutritional condition is
supplemental feeding (Milner et al. 2014, Ossi et al. 2017). Supplemental feeding involves
intentionally providing additional food to wild populations beyond what is naturally available
(Dunkley and Cattet 2003). The practice is implemented at various scales and for multiple
reasons, including enhancing individual and population health, providing emergency food when
natural sources are scarce or inaccessible, and mitigating damage to crops and livestock
(Dunkley and Cattet 2003). Depending on the goal, supplemental feeding can be a short (<
5years) or long-term (˃ 5 years) management measure (Milner et al. 2014). Feeding can lead to

14

rapid population increases due to direct effects on survival and breeding success, which makes it
an attractive management option (Robb et al. 2008, Sergio et al. 2008, Bishop et al. 2009). In
addition, supplemental feeding is often more politically supported and socially acceptable
compared with alternatives population recovery actions such as predator control (Robb et al.
2008). Supplemental feeding also provides an opportunity for researchers to address ecological
and biological aspects of a species, such as survival, growth rates, behaviour, reproduction, and
physiology by removing or mitigating the effects of food as a limiting factor and providing
opportunities for data collection (Dubois and Fraser 2013).
The information gained from measures of physiological and nutritional stress, as well as
their reproductive consequences, can inform strategies to conserve and recover endangered
wildlife populations such as woodland caribou (Rangifer tarandus caribou). Once the most
abundant cervid species in Canada, woodland caribou have suffered dramatic population declines
over the past century, and are now among the most recently extirpated large mammals in the
United States of America (Festa-Bianchet et al. 2011, Moskovitz 2019). In British Columbia
(BC), caribou populations have declined by as much as 50–60% and are listed as Threatened in
Canada under the Species at Risk Act (Environment Canada 2014, Environment and Climate
Change Canada 2020, McNay et al. 2022). The Southern Mountain population is particularly
imperiled, with ten subpopulations being extirpated between 2000 and 2017, and most remaining
populations are small (<100 individuals) and continue to decline (Johnson et al. 2015,
Environment and Climate Change Canada 2018, Moskovitz 2019, Serrouya et al. 2019).
The Central Group of the Southern Mountain population consists of five herds, the
Quintette, Narraway, Burnt Pine, Klinse-Za, and Kennedy Siding, all of which have experienced
considerable declines in recent years. The Burnt Pine population was declared extirpated in

15

2013. The Quintette herd declined from 265 individuals in 2002 to 106 in 2014, the Klinse-Za
herd fell from approximately 250 individuals in the 1990s to 38 by 2013, and the Kennedy
Siding herd declined from 120 in 2007 to 41 in 2012 (Seip and Jones 2013, 2016, 2018).
The decline of these populations is largely due to habitat disturbances caused by resource
extraction such as oil, gas, coal, hydroelectricity, and wood harvesting (Wittmer et al. 2005,
2007, McNay et al. 2022). Such disturbances create favorable conditions for other herbivores,
such as deer (Odocoileus spp.) and moose (Alces alces), which in turn support increased
populations of predators such as wolves which increase predation on caribou (Canis lupis; Holt
1977, Wittmer et al., 2013, Fortin et al., 2017). Habitat loss can also reduce caribou recruitment
(Johnson et al. 2020), and increase adult caribou mortality (McLoughlin et al. 2003b, Wittmer et
al. 2007).
The influences of predation and habitat also interact with nutrition (Cook et al., 2021).
Caribou are predator-sensitive foragers, and may choose to forgo high-quality foraging
opportunities at lower elevations to reduce predation risk (Bergerud and Elliott 1998, James et al.
2004, Leech et al. 2017). Caribou with calves especially may use nutritionally inferior habitats to
avoid predators while also sacrificing foraging time and efficiency due to increased vigilance
(Gustine et al. 2006, Viejou et al. 2018).
Several strategies have been implemented to mitigate caribou population declines in BC.
In 2014, a supplemental feeding program was initiated for the Kennedy Siding herd, providing
caribou with food pellets each fall (Heard and Zimmerman 2021). The goal of this program is to
increase survival and reproduction by improving the body condition of caribou (Heard and
Zimmerman 2021). Additionally, the Province of BC began an annual program to reduce wolf

16

density within the Kennedy Siding and adjacent herds ranges, reducing wolf densities by over
80% starting in winter 2015–16 (Bridger 2019).
Feeding appears to have an incremental effect on population growth above the effect of
wolf reduction alone, as indicated by comparisons of growth rates between the Kennedy Siding
and the adjacent Quintette herds, however the exact mechanism by which improved nutrition
leads to a population increase is unclear (Heard and Zimmerman 2021). Accordingly, our goal
was to assess the role of feeding on nutritional condition among female caribou in the Kennedy
Siding Herd through two inter-related questions. First, we asked whether there is evidence of
nutritional limitation within the herd, which is an important requisite for feeding to be beneficial.
Second, we assessed whether there was evidence that feeding leads to improved nutrition among
females that use the feeders in fall. We addressed these questions using indicators of
physiological and nutritional condition including body mass, body condition scores, δ15N values
in guard hair, and fecal glucocorticoid and T3 metabolites.
Indicators of nutritional condition
Multiple physical and chemical indices can be used to assess nutritional condition and the
effectiveness of supplemental feeding on individuals and populations. Measurements of body
mass and visual body condition scoring are commonly used to evaluate individual health. Body
mass is the most frequently used index, as it integrates several components of body condition,
including variations in protein and fat reserves (Parker et al. 2009, Taillon et al. 2011). Body
mass is limited in that it cannot differentiate among changes in body size, fat, and protein
reserves (Taillon et al. 2011). Visual scoring of body condition is a quick and non-invasive
method of assessing nutritional condition and body fat reserves (Pérez-Flores et al. 2016,
Schiffmann et al. 2017). It often involves assessing the visibility of skeletal structures such as the

17

ribs, spine, and hip bones (Schiffmann et al. 2017, Zielke et al. 2018). Visual scoring can be
influenced by species-specific characteristics (i.e., thick and dense hair), intestinal tract filling
and hydration status, or observer variability due to the inherent subjectivity of the technique
(Schiffmann et al. 2017, Smiley et al. 2020). Visual body condition also only provides
information on subcutaneous fat, which may not reflect high or low nutritional condition (Smiley
et al. 2020)
Body mass and condition scoring can be combined with chemical techniques, such as
stable isotope and hormone analysis, to provide a more comprehensive assessment of nutritional
condition. Measurements of body protein are useful for evaluating nutritional condition because
nitrogen (N) is essential for the synthesis of body protein . When dietary N is inadequate,
animals use N from their own body proteins for the maintenance of critical tissues and to support
pregnancy and lactation. Whereas body fat is the main labile source of energy, body protein is
also catabolized to provide some energy.
The naturally occurring stable isotope ratios of N (15N/14N relative to atmospheric N;
δ15N) in tissues can be used to describe nitrogen balance in animals (Gustine et al. 2011, 2014a,
Barboza et al. 2020). Nitrogen balance can be used to assess the status of protein nutrition and
metabolism in an individual and represents the net result of protein anabolism and catabolism in
the body. The 14N isotope is the most abundant form of N on Earth. Less energy is required to
break bonds with 14N than 15N (Lavery and Ferris 2021, Fraser et al. 2025). During digestion and
N metabolism in the liver and kidney, 14N-containing substrates are preferentially used by
enzymes over 15N (Kelly 2000, Fuller et al. 2005, Lavery and Ferris 2021, Fraser et al. 2025).
This results in urine being enriched in 14N and body tissues such as plasma, milk, and hair being
enriched in δ15N than relative to dietary δ15N (Kelly 2000, Fuller et al. 2005, Lavery and Ferris

18

2021, Fraser et al. 2025). Diet as well as physiological changes influence δ15N values in hair as it
is grown. Even during periods of nutritional restriction, protein synthesis must continue in vital
organs such as the liver and heart (Lee et al. 2012). If animals are not consuming enough protein
from their diet, they will break down their fat and muscle reserves to synthesize new protein
molecules (Barboza and Parker 2006, Lee et al. 2012). Protein synthesis will result in elevated
δ15N values in the hair when an animal experiences a net catabolic state (negative N balance)
associated with lean muscle degradation(Fuller et al., 2005, Habran et al., 2010, Lee et al., 2012,
Funck et al., 2020). .Funck et al., (2020) found that wood bison (Bison bison athabascae)
experiencing nutritional stress from harsh winter conditions or significant movement across the
landscape experienced a distinct increase in 15N values in hair during that period. This pattern
was also observed in Western white-bearded wildebeest (Connochaetes taurinus) in the
Serengeti, where mean δ15N value in tail hairs of starved animals was greater than that of nonstarved animals (Rysava et al. 2016). Conversely, stress from predation, and disturbance may be
correlated with reduced δ15N levels in certain tissues such as hair because stress hormones such
as glucocorticoids inhibit dietary protein assimilation (De Feo 1996, Deschner et al. 2012, Found
et al. 2022). Reproductive state may also influence δ15N in body tissues. Protein requirements for
fetal development and lactation are high, and if forage protein is limited, maternal body stores
are allocated to reproduction (Parker et al. 2009, Gustine et al. 2014a). During gestation and
lactation the increased demand for energy and protein results in protein synthesis, or an excess of
anabolism over catabolism (positive protein balance; Fuller et al., 2004, 2005, Schmidt et al.,
2023)). The increased use of dietary and body N for tissue synthesis of the fetus results in a
decrease in δ15N values in hair during gestation (Fuller et al. 2004, 2005, Schmidt et al. 2023).
The transfer of maternal body proteins to offspring in the milk can further deplete 15N in nursing

19

mothers, regardless of maternal nutritional condition (Fuller et al. 2004, Endo and Kobayashi
2022).
The effects of nutritional and physiological stress, as well as reproductive state, on
nutritional condition can be further examined using indicators of endocrine activity.
Glucocorticoids, a class of steroid hormones, can be used to evaluate physiological responses to
a wide variety of environmental, social, and internal conditions (Karaer et al. 2023).
Glucocorticoids are released in response to both acute and chronic stressors, enabling an animal
to respond to a physical challenge or environmental change. The primary function of
glucocorticoids is to mobilize energy reserves by converting glycogen to glucose, which
provides an animal with the energy to respond to immediate environmental pressures (Reeder
and Kramer 2005, Baker et al. 2013). A chronic elevation in glucocorticoids in response to a
persistent stressor can, however, negatively impact an animal by supressing the immune system,
depleting body reserves, and reducing reproductive success (Sapolsky et al. 2000, Romero 2004,
Reeder and Kramer 2005, Karaer et al. 2023).
Variations in the release of glucocorticoids can result from numerous extrinsic and
intrinsic factors including time of day, season, food availability, predation, disease, social status,
habitat, body condition, reproductive status, age, and sex (Sheriff et al. 2011, Baker et al. 2013).
Because glucocorticoids are released in response to a variety of stressors, they can be examined
alongside other bioindicators to characterize physiological and reproductive responses to
environmental stressors.
Thyroid hormones, such as triiodothyronine (T3), can be examined with glucocorticoids
to distinguish nutritional stress from other physiological stressors (Wasser et al. 2010, Ayres et
al. 2012, Behringer et al. 2018). Thyroid hormones regulate energy and lipid metabolism,

20

allowing an animal to adapt their metabolism to changes in environmental conditions and
nutritional availability and requirements. T3 levels decrease when food is scarce and increase
when food is abundant. This mechanism decreases metabolic rate and conserves energy during
periods of limited food availability (Blake et al. 1991, Wasser et al. 2010, Pasciu et al. 2024).
T3 is generally unresponsive to disturbance stress but responds rapidly to nutritional
stress, so combined glucocorticoid measurements, can enable environmental disturbance impacts
to be separated into nutritional and non-nutritional stress (Wasser et al. 2010, 2017, Ayres et al.
2012). Thyroid hormones are also involved in the progression and timing of developmental
processes such as skeletal maturation, linear growth, and brain growth (Todini 2007, Behringer
et al. 2018). Young animals generally have higher thyroid hormone levels than older animals
(Todini 2007, Behringer et al. 2018). Additionally, thyroid hormones play an important role in
thermogenesis, as animals adjust these hormones to regulate basal metabolism and maintain a
constant body temperature (Todini 2007, Behringer et al. 2018, Hunninck et al. 2020). Thyroid
hormones generally decrease when temperatures rise above a species’ thermal neutral zone and
increase when temperatures fall below it (Silanikove 2000, Wasser et al. 2010, Behringer et al.
2018, Hunninck et al. 2020, Pasciu et al. 2024).
If nutritional quality of summer habitat is limited and/or if caribou trade off better
foraging sites for safety from predators, we reasoned that females with calves would be more
likely to experience nutritional stress compared with females without calves due to the
nutritional and energetic costs of lactation and calf rearing. Furthermore, we expected yearling
females to be more vulnerable to nutritional limitation compared with adult females because
their smaller size, limited body reserves, and relatively higher metabolic demands make younger
animals more sensitive to environmental and forage conditions (Parker et al. 2009). We therefore

21

expected females with calves and yearlings to exhibit lower body condition scores, higher δ15N
levels in guard hair, lower fecal T3 concentrations, and higher fecal glucocorticoid concentrations
compared with adult females without calves upon arrival at the feeders. If supplemental food
during the feeding period leads to improved nutritional condition, we predicted that fecal cortisol
metabolites would decrease and fecal T3 metabolite would increase over the feeding period. We
tested for variation in bioindicators between years that might reflect inter-annual variation in
habitat. Inter-annual variation environmental conditions such as snow depth and cover,
temperature, precipitation, forage characteristics, and the start of vegetation growing season have
been linked to changes in movement, recruitment, mortality, parturition date, and body condition
in caribou (Adams and Dale 1998, Solberg et al. 2001, Larter et al. 2002, Chen et al. 2013, 2018,
Bergerud 2014). We hypothesized that supplemental feeding would have the most beneficial
effect in years when food quantity or quality was low, and that this would be reflected in higher
cortisol and lower T3 and a more pronounced change in these indicators over the feeding season
in years with poorer conditions. Moreover, we predicted that these changes would be more
pronounced among individuals vulnerable to poor nutritional condition at the start of the feeding
period (i.e., females with calves, yearlings, and individuals with smaller body mass or poor body
condition scores). Overall, our study provides insight into the effects of nutritional resources in
the summer and autumn on caribou physiology and demonstrates the potential for supplemental
feeding to enhance caribou conservation.
Methods
Study site
The range of the Kennedy Siding caribou herd is in central British Columbia, within the
traditional territory of the McLeod Lake Tse’Khene (McLeod Lake Indian Band). The habitat

22

used by the herd is in the Sub-Boreal Spruce and Engelmann Spruce Sub-Alpine Fir
biogeoclimatic zones (Meidinger and British Columbia 1991). The terrain is characterized by
mountains and rolling hills ranging in elevation from 710m to 2010m above sea level (Jones et
al. 2007). Caribou typically spend most of the year at high elevations dispersed in relatively
undisturbed sub-alpine meadows and forests of Engelmann Spruce (Picea engelmannii) and
Subalpline Fir (Abies lasiocarpa). Each fall, the population descends from summer range at high
elevation to a small distinct area of lodgepole pine (Pinus contorta) forest at low elevation on the
southwest edge of their range, which has abundant terrestrial lichen (Cladonia spp). Movement
from high to low elevations is stimulated by the first snowfall events in the mountains. When on
their fall range, caribou forage on terrestrial lichens, arboreal lichens, forbs and shrubs. Snow
begins to accumulate on the ground in November and when it reaches a depth of about 1 m
(usually in mid-January; (Jones et al. 2007), caribou move back to high elevations, where they
feed on terrestrial lichens in wind-swept areas or arboreal lichens where snow is deep and
supports their weight. Most (2894 ha) of the fall range of the Kennedy Siding herd, including the
feeding site, was designated an Ungulate Winter Range under the British Columbia Forest and
Range Practices Act in 2002 (Arthur 2002).
The feeding area is located approximately 30 km south-east of the community of
Mackenzie. Feeding occurs in the vicinity of the Kennedy railway siding and Highway 97. The
area experiences frequent vehicle traffic from forestry, railway, hydro, and other industries as
well as by recreational users such as hunters. Other large mammals that inhabit the area include
moose (Alces alces), white-tailed deer (Odocoileus virginianus), elk (Cervus canadensis), lynx
(Lynx canadensis), wolf (Canis lupus), coyote (Canis latrans), grizzly bear (Ursus arctos), and
black bear (Ursus americanus; Heard and Zimmerman 2021).

23

Feeding, identification of individuals, and body condition measurements
We provided food and identified individuals that used the feeders following the
procedures described in Heard and Zimmerman (2021). Feeding began when caribou first arrived
in late September to early October. We contracted the McLeod Lake Indian Band to build the
feeders and deliver the food pellets. We initially provided small amounts of pellets in plastic
livestock buckets so that any remaining feed could be removed from the site when the caribou
were not present. Food removal was carried out to limit attractants to the bears that used the area.
We began ad-lib feeding in November when the bears were assumed to be in hibernation and
continued feeding until January 15th. Feeding was stopped at that date to avoid interfering with
the natural movement of the herd back into their alpine habitats. Caribou consumed
approximately 70 kg of food per animal over the feeding period (Heard and Zimmerman 2021).
We used 7–9 Reconyx Hyperfire motion-sensor trail cameras, focused on either a feeder,
a salt block, or along a trail, to obtain a continuous photographic record of individual caribou
during the feeding period. The sex, age class (calf or adult), and body condition of each caribou
were recorded. Animals were considered in poor body condition if ribs were visible, and good
body condition if ribs were not visible (figure 2). Individual caribou were identified from photos
using the number and position of antler points, radio collars or ear tags, or unique markings (e.g.
scars). Calves did not always have unique antlers, so we identified individuals based on their
close association with mothers of known identity upon arrival at the feeders. The caribou were
attracted onto a platform scale (using salt or pellets), with trail cameras positioned to photograph
the individual caribou and record the weight (kg) on the scale’s digital display. If an animal was
photographed on the scale more than once, we used the average of the weights recorded as the
final body mass measurement. Every animal used in the analysis was weighed one to three times.

24

The food pellets were produced by Hi-Pro Feeds/Trouw Nutrition Canada Inc. in Grand
Prairie, Alberta. The pellets were developed at the University of Alaska Fairbanks to provide
balanced and adequate levels of energy, nitrogen, and minerals for growth and reproduction of
caribou (Barboza and Parker 2006) and have been used in other nutritional studies (Gustine et al.
2011) and conservation projects such as the Klinse-za herd maternity pen (McNay et al., 2022).

Figure 2: Examples of female caribou (Rangifer tarandus caribou) at the Kennedy Siding feeding site considered in poor body
condition and female caribou considered in good body condition. The caribou in images A and B are considered in poor
condition based on their visible ribs, spine, and hip bones in the trail camera images. The caribou in images C and D are
considered in good condition based on the lack of visible ribs or other bones.

Animal care
Sample collection was conducted under research permits provided by the British
Columbia Ministry of Forests as well an Animal Care and Use Protocol approved by the
University of Northern British Columbia (BC Permit FJ21-618702; UNBC ACUC Protocol

25

Number: 2021-15). Some of the adult female caribou that used the feeding site (21 in 2021 and
19 in 2022) had collars and/or ear tags that were deployed as part of the Northern Mountain and
Central Mountain Caribou Capture Program run by the BC Provincial Government (Ministry of
Forests), under the BC Provincial Animal Care Protocol (Permit Number: FJ21-618702).
Sample collection
We conducted sample collection with research technicians from the McLeod Lake Indian
Band. We collected fecal pellet samples opportunistically by observing female caribou around
the feeders. When defecation occurred, the individual caribou was identified, and the pellets
were collected in a resealable plastic bag after the caribou moved away. The pellets were stored
in a cooler with an ice pack until they could be transferred to a freezer and stored at -20°C within
four hours of defecation.
We collected hair samples using a noninvasive snaring method. We affixed metal
alligator clips to a piece of wood which were held open using small pieces of wooden toothpicks.
The piece of wood was attached to one of the feeders and when caribou put their head down to
eat, their neck pressed down on the clips, knocking out the toothpick and causing them to close.
When the caribou pulled away, the clip caught a sample of guard hair from the neck, at least 10
hairs/clip. This method appeared to cause very little disturbance to the caribou, as they did not
avoid using the feeder where the clips were set up.
Identification of individuals and aging using DNA
We sampled feces for DNA by taking three surface rubs from each pellet sample using
cotton swabs. We stored samples of guard hairs with visible root bulbs in unwaxed paper
envelopes. DNA extraction, amplification, and sequencing of hair and fecal samples were
conducted by Wildlife Genetics International in Nelson, BC. The DNA was extracted using

26

QIAGEN DNeasy Blood and Tissue kits. For hair samples, the roots clipped from a target of 10
guard hairs was used. For fecal swabs ~ 5 mm clippings from the used end of one swab per
sample was used. Individual identity was established using 18 microsatellite markers established
by Serrouya et al., (2012) along with a ZFX/ZFY sex marker. The success rate for producing all
19 loci was 85% for hair samples and 91% for fecal swabs. DNA results were used to identify
unmarked females between years and to obtain an exact or estimated age of the animal.
Yearlings were classified as animals that had been sampled for DNA the year before as calves.
Adults were classified as animals that had been sampled in the year before as adults / yearlings,
or as a calf more than two years previously.
Hair preparation for stable isotope analysis
We prepared samples for stable isotope analysis following procedures outlined by
Hobson et al., (2000). In brief, the hair samples were washed with a solution of two parts
chloroform [Thermo Fisher Scientific] and one part methanol [Thermo Fisher Scientific].
Approximately 1 mg of hair (equal to 2-3 hairs) was weighed into a tin capsule. We sent the
weighed hair samples to the Stable Isotopes in Nature lab at the University of New Brunswick
for analysis of δ 15N using continuous flow isotope mass spectrometry. Instruments calibrated
against international reference standards from the International Atomic Energy Agency
(International Atomic Energy Agency 1995).
Fecal hormone enzyme linked immunosorbent assay (ELISA)
To extract hormones from the fecal samples we first thawed and homogenized each
sample. A portion of the homogenized sample was transferred to a plastic snap cap vial which
was covered with a low lint delicate task wipe secured with an elastic band. The vials were
lyophilized for ~ 26 hours. The dried feces were powdered and stored in the freezer at -20˚C.

27

Hormone metabolites were extracted from 0.2 g of dried feces in 5 mL of HPLC grade 80%
methanol [Thermo Fisher Scientific] / 20% water. Vials were sealed and rotated at 160 rpm
overnight at room temperature. A portion of the methanol extract was transferred to a 2 mL
centrifuge tube which had caps with o-rings to prevent evaporation, centrifuged for 30 minutes,
and stored at -20˚C until analysis.
Fecal glucocorticoid metabolite (FGM) ELISA
We followed the protocol described by Graham et al., (2024 ) to determine
concentrations of FGMs. We coated ninety-six well plates (Thermo Fisher Scientific) with
affinity-purified goat anti-rabbit gamma globulin (25 µg/plate [Sigma-Aldrich]) dissolved in
coating buffer (0.015 M Na2CO3 [VWR], 0.035 M NaHCO3 [Sigma-Aldrich]; pH 9.6) and
incubated them overnight at room temperature. The coated plates were washed (0.04% Tween 20
[Sigma-Aldrich]) and refilled with 200 µl Trizma assay buffer (0.02 M Trizma [Sigma-Aldrich],
0.300 M NaCl [VWR], 0.1% bovine serum albumin [VWR]; pH 7.5) and stored at room
temperature for at least 1 hour to block non-specific binding. We diluted fecal extracts (1:4 to
1:32) in Trizma assay buffer prior to the assay. The wells were emptied of the Trizma buffer and
50 µl each of diluted sample and standards were dispensed. Horse-radish peroxidase-labeled
cortisol was dispensed, followed by anti-cortisol antibody or (Anti-cortisol #R4866; CJM
Munro, UC Davis). The plates were incubated overnight at room temperature. Following
incubation, we washed the plates three times and then added 200 µl of substrate solution (0.5 mL
of 0.016 M tetramethylbenzidine [VWR] in dimethylsulphoxide [Thermo Fisher Scientific], and
100 mL of 0.175 M H2O2 [VWR] diluted in 24 mL of 0.01 M C2H3O2Na [Thermo Fisher
Scientific]; pH 5.0) to each well. The plates were incubated for 45 minutes at room temperature
and then the enzyme reaction was stopped by adding 50 µl of stop solution to each well (3 M

28

H2SO4 [Thermo Fisher Scientific]). The optical density was measured at 450 nm. If the sample
duplicates had a coefficient of variation (CV) greater than 15%, we re-assayed the sample. The
cross-reactivity of the cortisol antibody was reported for the following substances, including
cortisol (100%), prednisolone (9.9%), prednisone (6.3%), cortisone (5%) and <1% with
corticosterone, desoxycorticosterone, 21-desoxycortisone, testosterone, androstenedione,
androsterone, and 11-desoxycortisol (Young et al. 2004).
As an analytical validation, fecal extracts from different female caribou were combined
and assayed with the cortisol ELISAs. To establish parallelism, we tested serial two-fold
dilutions of the pooled sample for comparison displacement curves (Figure 3A). The intra-assay
coefficient of variability was determined by assaying 31 replicates of the pooled sample on the
same plate. The intra-assay CV was 5.36% at 73.5% binding. The inter-assay coefficient of
variation was determined by running a replicate of the pooled sample at a high and low
concentration as controls. The inter-assay CV was 8.21% at 55% binding and 5.21% at 87%
binding for 22 plates. As a physiological validation, fecal samples were assayed before and after
a known stressful event in two caribou to ensure that the cortisol ELISA could measure the
increase in FGMs associated with an increase in adrenal activity. The known stressful event was
the capture and transfer of two female caribou from the Klinse-Za herd to a maternal pen.
Samples were collected from the rectum at capture to represent baseline FGMs and from the pen
approximately 12 hours after capture to represent elevated FGMs following capture and transfer.
Fecal T3 metabolite (FTM) ELISA
We measured FTM concentrations using T3 Total ELISA kits (EIA-4569R: T3 Total
ELISA, DRG Instruments GmbH, Germany). Fecal samples were evaporated to dryness under a
steady stream of nitrogen. Samples were reconstituted using a solution of Trizma assay buffer

29

(0.02 M Trizma [Sigma-Aldrich], 0.300 M NaCl [VWR]; pH 7.5) and 100% HPLC grade
methanol [Thermo Fisher Scientific]. We dispensed 50 µl of either the diluted sample, standards,
or controls into each well. We added reagents and incubated plates following instructions
provided by the kit manufacturer. The optical density was measured at 450 nm and if sample
duplicates had a coefficient of variation (CV) greater than 15%, we reanalyzed the samples. The
cross-reactivity of the FTM assay to several compounds was provided with the immunoassay kit,
including I-Triiodothyronine (100%), I-Thyroxine 69 (0.37%), Reverse T3 (0.75%), DThyroxine (0.1%), 3,5-Diiodo-L-Thyrosine (0.2%), and 4-Phynoxyphenol (0.2%).
As an analytical validation of the FTM ELISA, pools of fecal extracts from different
female caribou were combined and assayed. To establish parallelism, we tested serial two-fold
dilutions of the pooled sample for comparison displacement curves (Figure 3B). The intra-assay
variability was determined by assaying 15 replicates of the pooled sample. The intra-assay CV
was 4.24% at 51.6% binding Inter-assay variation was determined by running a replicate of the
pooled sample at a high and low concentration as controls. The inter assay CV was 10.56% at
56% binding and 8.43% at 86% binding for 10 plates.

30

31

Figure 3: Parallelism of cortisol (A) and T3 (B) standards (solid line) and a serially diluted extract from caribou (Rangifer
tarandus caribou) fecal samples (dashed line).

Statistical analysis
All data analysis were conducted using R (R Core Team 2023). We classified each
caribou sampled according to their age and reproductive status (yearlings and adults [≥ 2 years
old] with and without a calf at heel).. We performed Kruskal-Wallis rank sum test to compare
body mass between 1-year olds, adults with a calf, and adults without a calf, and a Wilcoxon test
to compare body mass in animals in good and poor body condition (Kassambara 2019).
We used linear mixed effects regression models to assess the effects of multiple variables
on hormone and isotope concentrations in hair and feces (Bates et al. 2015). We log transformed
fecal cortisol and square root transformed fecal T3 to obtain normal distributions of the model
residuals. We included the effects of reproductive age class, sampling day, sampling year, body
mass, and body condition as fixed effects (Table 1). Individual was included as a random effect.
We included reproductive age class in all the models because of the high energetic costs
associated with calf rearing and lactation. Sampling day was expressed as day of the year and
standardized to a mean of 0 and a standard deviation of 1 prior to analysis. Body mass was
square root transformed to normalize the distribution, then standardized to a mean of 0 and a
standard deviation of 1 prior to analysis. To account for the effect that sampling year, body mass,
body condition, and reproductive age class might have on how fecal cortisol and T3 might
change over time, we tested for a significant interaction between these variables and sampling
date using Type III Analysis of Variance (ANOVAs). The only significant interaction was
between sampling year and date in the cortisol models (W(1) = 4.77, p = 0.03). We tested for
multicollinearity among predictor variables using pairwise correlation coefficients and variance
inflation factors (Fox et al., 2001).

32

We selected the top model for each response variable using Akaike’s Information
Criterion corrected for small size (AICc; (Bartoń 2010). We used AIC scores to identify the most
parsimonious (ΔAICc <2) top models within each set of candidate models. We assessed the
validity of the top models based on the normality of residuals and plots of residuals versus
predicted values for each response variable. We examined any models that had a ΔAICc within
0–2 units of the top model for uninformative parameters by assessing whether they differed from
the best model by one parameter and had similar maximum log-likelihood values as the top
model (Arnold 2010).
Table 1: Prediction category and structure of eight candidate models compared to predict fecal glucocorticoid, fecal
triiodothyronine (T3), and hair δ15N in female caribou (Rangifer tarandus caribou) sampled in 2021and 2022 at the Kennedy
Siding feeding site in British Columbia, Canada.

Prediction
Reproduction (null)
Reproduction + time at
feeders
Reproduction +
body condition

Reproduction + time at
feeders
+ body condition

Fecal glucocorticoid
metabolites
Reproductive age class
Reproductive age class +
sampling date + sampling
year + sampling
date*sampling year
Reproductive age class +
body condition

Fecal T3 metabolites

Hair δ15N

Reproductive age class
Reproductive age class +
sampling date + sampling
year

Reproductive age class
Reproductive age class +
sampling year

Reproductive age class +
body condition

Reproductive age class +
body condition

Reproductive age class +
body mass
Reproductive age class +
body condition + body
mass
Reproductive age class +
sampling date + sampling
year + sampling
date*sampling year + body
condition
Reproductive age class +
sampling date + sampling
year + sampling
date*sampling year + body
mass
Reproductive age class +
sampling date + sampling
year + sampling
date*sampling year + body
condition + body mass

Reproductive age class +
body mass
Reproductive age class +
body condition + body
mass
Reproductive age class +
sampling date + sampling
year + reproductive age +
body condition

Reproductive age class +
body mass
Reproductive age class +
body condition + body
mass
Reproductive age class +
sampling year + body
condition

Reproductive age class +
sampling date + sampling
year + body mass

Reproductive age class +
sampling year + body
mass

Reproductive age class +
sampling date + sampling
year + body condition +
body mass

Reproductive age class +
sampling year + body
condition + body mass

33

Results
We identified and sampled 20 individual animals in both years plus 5 others unique to
2021 and 20 others unique to 2022. In 2021, we classified 14 animals as adults with a calf, five
as adults without a calf, and six as yearlings. In 2022, we classified 21 animals as adults with a
calf, 13 as adults without a calf, and six as yearlings. In the 2021 field season, we collected 155
pellet samples from 37 females (1-14 samples / caribou). During the 2022 field season, we
collected 286 pellet samples from 43 females (1–16 samples / caribou). We collected hair
samples from 27 caribou in 2021 and 33 caribou in 2022.
Body mass and condition
The body mass of caribou sampled in 2021 and 2022 ranged from 95.0 kg–153.0 kg
(n=65, median = 128 kg). Body mass of adults with a calf at heel ranged from 105 kg–150.5 kg
(n=35, median = 128.5 kg). For adults without a calf at heel body mass ranged from 107.0 kg–
153.0 kg (n = 18, median = 138.0 kg) yearling caribou ranged in body mass from 95 kg–125.5 kg
(n=12, median = 106.0 kg). A Kruskal-Wallis test showed that there was a significant difference
of means between the age classes (H = 22.87, df = 2, p <0.001). Post hoc pairwise comparisons
found that adults with a calf at heel had smaller body masses compared to adults without a calf (p
= 0.036, Figure 4). Yearlings had smaller body masses compared to both adults with a calf (p <
0.001, Figure 4) and adults without a calf (p < 0.001). Of the caribou sampled, almost half were
classified as being in poor body condition based on the visibility of ribs (Table 2, Figure 4). Most
of the individuals in poor body condition were adults with a calf (Table 2); however, a Wilcoxon
signed-rank test showed that the difference in body mass between caribou in good and poor body
condition was not significant (z = -0.118, p = 0.906).

34

Figure 4: Body mass (kg) of adult female caribou (Rangifer tarandus caribou) without a calf at heel (A) and with a calf at heel
(AC) compared with yearling females (Y) using the feeding site at Kennedy Siding, British Columbia, Canada, in 2021 and 2022.
The centre line of each box represents the median and the outer limits represent the 25th and 75th percentiles. The lines extending
below and above each box represent 1.5 times the interquartile range for each percentile.
Table 2: Number of female caribou (Rangifer tarandus caribou) in each reproductive age class with visible ribs at the Kennedy
Siding feeding site used in our analysis. Most of the individuals with visible ribs were adults with a calf.

Age class
Visible
ribs

No
Yes

Adult with a calf
10
25

Adult without a calf
15
3

Yearling
10
2

35

Hair nitrogen stable isotope (δ15N)
We analyzed δ15N in 52 hair samples collected from thirty-nine individual caribou using
the feeding site in 2021 and 2022. Concentrations of δ15N ranged from 1.712 ‰–6.921 ‰ (mean
= 3.814 ‰, standard error [SE] = 0.151). The mean δ15N concentration was 3.526 ‰ (n= 28, SE
= 0.221) for adults with a calf, 4.095‰ (n = 15, SE = 0.255) for adults without a calf, and 4.240
‰ (n = 9, SE = 0.257) for yearlings.
Of the 8 candidate models, 2 had a ΔAICc score <2 and carried 76.27% of the model set
weight. The model that received the strongest support included an effect of reproductive age
class and sampling year (ωi = 0.509, Table 3). Compared to adults with a calf, adults without a
calf (Top model: β = 0.783, SE = 0.242, CI = 0.297–1.268) and yearlings (Top model: β =0.769,
SE = 0.292, CI = 0.184–1.353) had higher δ15N concentrations (Figure 6A). Yearling females
had slightly lower δ15N concentrations than adults without a calf, however, the confidence
interval on the coefficient overlapped 0 (Figure 5, Figure 6A). Concentrations of δ15N were
lower in hair samples collected in 2022 compared to samples from 2021 (Top model: β = -0.823,
SE = 0.178, CI = -1.216 – -0.465; Figure 6B).
The second ranked model also contained an effect of body condition. δ15N concentration
was lower in animals in good condition compared to those in poor condition, however the
confidence interval overlapped 0 (Figure 4). Moreover, the model differed from the top model by
only one parameter, had an AICc score that differed by 1.4 and had a similar log-likelihood
value. This suggests that the top model was the most parsimonious and that the effect of body
condition on δ15N concentrations in the hair was negligible (Arnold 2010). The null model had a
ΔAICc score of 15.34 and a model weight of 2.377 x 10-4 (Table 3).

36

Table 3: Top 2 of 8 linear mixed-effects models used to predict δ15N concentrations in hair samples (n=52) from female caribou
(Rangifer tarandus caribou) sampled in fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in British Columbia,
Canada. Competing models within two ∆AICc are shown, as well as the null for reference.

Rank

Candidate model

df

logLik

AICc

ΔAICc

ωi

1

Reproductive age class + sampling year

6

-62.74

139.3

0

0.509

2

Reproductive age class + sampling year +
body condition

7

-62.09

140.7

1.387

0.254

5

Null

5

-71.69

154.67

15.34

0

Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model;
ωi = Akaike weight.

37

Figure 5: Coefficients and 95% and 85% confidence intervals (CI) for variables in the top three models (< 2 ∆AICc) predicting
hair δ15N concentrations in female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021 and 2022 at the
Kennedy Siding feeding site in British Columbia, Canada (n = 52). The baseline category for reproductive age class is adult
without a calf at heel. The baseline category for body condition is poor (ribs visible).

38

39

Figure 6: (A) The highest hair δ15N concentrations measured in caribou (Rangifer tarandus caribou) sampled at Kennedy
Siding, BC, Canada, were found in adults without a calf (A) and 1- year olds (Y). Adults with a calf (AC) had the lowest values.
The center point represents the adjusted prediction, and the error bars represent the adjusted predictions ± the standard error. (B)
Hair δ15N concentrations were higher in 2021 compared to 2022. The center point represents the adjusted prediction, and the
error bars represent the adjusted predictions ± the standard error.

Fecal glucocorticoid metabolites (FGM)
We analyzed FGMs in 432 pellet samples from forty-five female caribou that used the
feeding site in 2021 and 2022. Concentrations of FGMs ranged from 120.06 ng/g–2018.88 ng/g
(median = 457.04). Concentrations of FGMs ranged from 154.78 ng/g–1367.69 ng/g (median =
456.92 ng/g) in adults with a calf at heel, 176.34 ng/g–1013.06 ng/g (median = 436.31 ng/g) in
adults without a calf, and 120.06 ng/g–2018.88 ng/g (median = 489.61 ng/g) in yearlings.
Of the eight candidate models for FGMs, three had a ΔAICc score less than two and
carried 88.01% of the model weight (Table 4). The model with reproductive age class, sampling
day, and sampling year received the strongest support (ωi = 0.462, Table 4). There was an
interaction between sampling day and sampling year (F(1, 431.5) = 4.811, p = 0.029; Figure 7). The
interaction revealed that FGM concentrations declined over time in 2021 and 2022; however, the
decline was more pronounced in 2022 (Top model: β = -0.074, SE = 0.034, CI = -0.140 – -0.008;
Figure 7, Figure 8A). Compared to adults without a calf at heel, adults with a calf had higher
FGM concentrations however while the 85% confidence interval did not encompass zero, the
95% confidence interval did (Top model: β = 0.06, CI = -0.02–0.14, SE = 0.04, Figure 7, Figure
8B). Yearlings had greater FGM concentrations compared to adults without a calf (Top Model: β
= 0.12, CI = 0.02–0.23) SE = 0.06; Figure 8B). The second ranked model contained the same
variables as the top ranked model, plus an effect of body mass. FGM concentrations increased
with increasing body mass, however, the confidence interval encompassed 0 (Figure 7). The
third ranked model contained the same variables as the top model plus an effect of body
condition. FGM concentration was greater in caribou with good body condition, however, the

40

confidence interval on this variable also encompassed 0 (Figure 7). Both the second and third
ranked models differed from the top model by only one parameter, had ΔAICc scores that
differed by < 2 from the top model and had similar log-likelihood values. This suggests that the
top ranked model is the most parsimonious and that body mass and body condition have little
influence on the concentration of FGMs (Table 4; (Arnold 2010). The null model had a ΔAICc
score of 7.911 and a model weight of 0.009 (Table 4).
Table 4: Top 3 of 8 linear mixed-effects models used to predict fecal glucocorticoid metabolites in female caribou (Rangifer
tarandus caribou) sampled in fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in British Columbia, Canada.
Competing models within two ∆AICc are shown, as well as the null for reference.

Rank

Candidate model

df

logLik

AICc

ΔAICc

ωi

1

Reproductive age class + sampling date x sampling
year
Reproductive age class + sampling date x sampling
year + body mass
Reproductive age class + sampling date x sampling
year + body condition
Null

8

-124.8

265.9

0

0.462

9

-124.4

267.3

1.430

0.226

9

-124.6

267.6

1.762

0.192

5

-131.8

273.8

7.911

0.009

2
3
5

Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model; ω i
= Akaike weight.

41

Figure 7: Coefficients and 95% and 85% confidence intervals (CI) for variables in the top three models (< 2 ∆AICc) predicting
FGM concentrations in female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021 and 2022 at the Kennedy
Siding feeding site in British Columbia, Canada (n = 432). The baseline category for reproductive age class is adult without a calf
at heel. The baseline category for body condition is poor (ribs visible).

42

43

Figure 8: (A) Concentrations of FGMs decreased over time during the 2021 and 2022 feeding seasons in female caribou
(Rangifer tarandus caribou) using the Kennedy Siding site, however the decline was more significant in 2022. The relationship
between log transformed FGM concentrations and sampling date was predicted from the top model with other terms held
constant at their mean value. Sampling date was standardized to a mean of zero and a standard deviation of one prior to analysis.
(B) The highest FGM concentrations were found in yearlings (Y), followed by adults with calves (AC). Adults without calves
(A) had the lowest values. The center point represents the adjusted prediction value, and the error bars represent the adjusted
predictions ± the standard error.

Fecal T3 metabolites (FTM)
We analyzed FTMs in 385 fecal samples from forty-five female caribou that used the
feeding site in 2021 and 2022. FTM concentrations ranged from 14.53 ng/g–150.73 ng/g (median
= 59.47ng/g). The concentration of FTMs ranged from 25.10 ng/g–138.17 ng/g (median = 57.67
ng/g) in adults with a calf at heel, 14.53 ng/g–150.73 ng/g (median = 62.40, IQR = 25.00) in
adults without a calf, and 25.33 ng/g–150.40 ng/g (median = 60.90) in yearlings.
Of the 8 candidate models for FTM concentrations, two had a ΔAICc score <2 and
carried 78.35% of the model set weight (Table 5). The model containing reproductive age
category, sampling day, sampling year, and body mass received the strongest support (ωi =
0.537; Table 5; Figure 9). FTM concentrations declined over time during the feeding periods,
which occurred between October and January in 2021 and 2022 (Top model: β = -0.14, CI = 0.26 – -0.01, SE = 0.06; Figure 10A). In addition, larger caribou had lower FTM concentrations
(Top model: β = -0.29, CI = -0.54– -0.03, SE = 0.13; Figure 10B). The top model also included
an effect of reproductive age class. The concentration of FTMs was higher in yearlings compared
to adults without a calf at heel (Top model: β = 0.71, CI = 0.06–1.37, SE = 0.34; Figure 10C).
FTM concentrations were similar between in adults with a calf and without (Top model: β =
0.04, CI = -0.38–0.47, SE = 0.26; Figure 9, Figure 10C). The concentration of FTMs was higher
in 2022 than in 2021. This difference was significant at the 85% confidence level, but not at the
95% confidence level (Top model β = 0.28, CI = -0.01–0.58, SE = 0.15; Figure 9).

44

The second ranked model contained the same covariates as the top model, as well as an
effect of body condition. Caribou in good body condition had lower FTM concentrations
compared to those in poor body condition, however, the confidence interval on this variable
overlapped with 0 (Figure 9). The second ranked model only differed from the top model by 1
parameter, had a ΔAICc score that differed by 1.6 and had a similar log-likelihood value. This
suggests that the top model was the most parsimonious and that body condition has little
influence on FTM concentrations (Arnold 2010). The null model had an AICc of 10.75 and a
model weight of 0.002 (Table 5).
Table 5: Top 2 of 8 linear mixed-effects models used to predict female caribou (Rangifer tarandus caribou) FTM concentrations
in caribou sampled in fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in British Columbia, Canada.
Competing models within two ∆AICc are shown, as well as the null for reference.

Rank

Candidate model

1

Reproductive age class + sampling year + sampling
date + body mass

2

Reproductive age class + sampling year + sampling
date + body mass + body condition

df

logLik

AICc

ΔAICc

ωi

8

-610.62

1237.63

0

0.537

9

-610.36

1239.19

1.56

0.253

7
Null
5
-619.11 1248.38 10.75
0.002
Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model;
ωi = Akaike weight.

45

Figure 9: Coefficients and 95% and 85% confidence intervals (CI) for variables in the top three models (< 2 ∆AICc) predicting
FTM concentrations in female caribou (Rangifer tarandus caribou) sampled in fall and winter of 2021 and 2022 at the Kennedy
Siding feeding site in British Columbia, Canada (n = 385). The baseline category for reproductive age class is adult without a calf
at heel. The baseline category for body condition is ribs visible.

46

47

Figure 10: The relationships between sampling day, body mass, and reproductive age class on the concentration of fecal
triiodothyronine metabolites (FTM) in feces collected from female caribou (Rangifer tarandus caribou) at Kennedy Siding,
British Columbia. The distribution of FTMs was normalized with a square root transformation prior to analysis. The distribution
of body mass was normalized using a square root transformation and both sampling day and normalized body mass were
standardized to a mean of zero and a standard deviation of one prior to analysis. The concentration of FTMs was predicted from
the top model with all other terms held constant at their mean value. (A) The concentration of FTMs declined over time during
both the 2021 and 2022 feeding periods. (B). The concentration of FTMs declined with increasing body mass of caribou. (C)
Adults without a calf at heel had lower FTM concentrations than yearlings and similar concentrations to adults with a calf. The
center point represents the adjusted prediction, and the error bars represent the adjusted predictions ± the standard error.

Discussion
Our findings provide insight into how autumn nutrition and reproduction influence body
condition. Though further comparisons are required with unfed populations, these findings
indicate that supplemental feeding could contribute to population recovery by facilitating rapid
recovery of energy and protein stores before winter, particularly in adult females with a calf at
heel. We found that adult females with calves were in poor nutritional condition compared to
adults without calves and yearlings based on body mass, visual body scoring, and δ15N in guard
hair, suggesting that it may be difficult for caribou to fully compensate for both lactation and
maintenance of body reserves during the summer. We found differences in indicators of
condition between years, highlighting the need to account for environmental conditions when
considering nutrition. This is the first time we are aware of that repeated sampling of fecal
glucocorticoid or thyroid hormone metabolites has been conducted on woodland caribou during
the autumn. Our hormone analyses are the first of their kind for this subspecies in this region and
season; therefore, they should be viewed as a baseline upon which future studies can expand.
Nutritional limitation in females with calves
We predicted that if supplemental feeding has the potential to be effective in the Kennedy
Siding Herd, there would be evidence of nutritional limitation in the herd. Consistent with this
prediction, adult females with calves had smaller body masses and were more likely to be in
poorer body condition (i.e., have ribs showing) compared with those without calves. Combined,

48

these findings suggest that calf rearing can reduce the nutritional condition of adult females over
the summer and that summer habitat may not provide adequate nutrition to compensate for the
demands of lactation and calf-rearing. This finding is consistent with several studies that found
substantial differences in body fat between caribou that raised a calf and those that did not (Crête
and Huot 1993, Chan-McLeod et al. 1999, Cook et al. 2021). These differences are the result of
increasing nutritional requirements beginning in late gestation that double during lactation
compared with normal maintenance requirements (Gerhart et al. 1997b, a, Cook et al. 2004,
2021, Barboza and Parker 2006). Caribou often give birth before the onset of new vegetation
growth. The lichens that are still present at that time are high in digestible energy but may may
not satisfy the daily protein and dry matter needs of lactating caribou (Pulliainen 1971, Rominger
and Robbins 1996, Storeheier et al. 2002). Mothers of newborn calves will also sacrifice
foraging time and efficiency for increased vigilance and select habitats for predator avoidance
rather than forage (Gustine et al. 2006, Viejou et al. 2018). Combined, this means that lactating
females must rely on endogenous energy and protein stores from spring-early summer and
experience a minimum of nutritional condition by mid-summer (Crête and Huot 1993, ChanMcLeod et al. 1999, Cook et al. 2021). Reduced body condition may affect individual
performance by reducing fat accretion and ovulation in adults and survival and growth in calves
and juveniles (Crête and Huot 1993, Cook et al. 2004, Albon et al. 2017). A substantial loss of
condition during this time may also be life threatening by increasing susceptibility to predation,
disease, extreme climactic events, and starvation (Ratcliffe 1980, Mech and Delgiudice 1985,
Depperschmidt et al. 1987, Bender et al. 2008, Metz et al. 2012, Mattisson et al. 2016, Cook et
al. 2021). In British Columbia, most adult female mortality occurs between May and October,
with mortalities peaking in July and August (Seip 1992, Kinley and Apps 2001, McLoughlin et

49

al. 2003b, Wittmer et al. 2005). The animals most vulnerable to low nutritional condition are
those that have a calf or those that raised a calf the previous year, a carry over effect from one
summer to another (Cook et al. 2021). The period after mid to late summer when calves are more
mobile represents a critical time for caribou to recover depleted body reserves before the
breeding season and winter (Chan-McLeod et al. 1999). If good quality summer forage is
available and accessible, females with a calf at heel can reach a similar fatness to caribou without
a calf (Adamczewski et al. 1987, Cook et al. 2004). If summer forage is not fully adequate, fall
supplemental feeding could be beneficial, especially for females that raised a calf over the
summer, because improved nutrition at this time can result in higher body fat levels, and overall
greater survival (Crête and Huot 1993, Cook et al. 2004). Greater body fat levels would increase
the chances of a female breeding and would improve calf and juvenile survival, body size, and
growth rates (Crête and Huot 1993, Cameron and Hoef 1994, Cook et al. 2004, Dale et al. 2008).
In addition to lower body weights and condition, we predicted that females with calves
would have higher δ15N (negative nitrogen balance) compared to those without calves if dietary
protein inputs were inadequate due to their use of body tissues for both maintenance and milk
production in the neonatal period. In contrast with our prediction, caribou with calves had lower
hair δ15N than adults without a calf and yearlings. One explanation for these lower δ15N values is
that the synthesis of milk proteins results in a net anabolic state for the mother during lactation,
which coincides with the period when hair is growing and incorporating nitrogen. Caribou are
considered “capital breeders” that rely on body stores for reproduction (Taillon et al. 2013,
Williams et al. 2017). The primary source of N in milk proteins is maternal body protein (Taillon
et al. 2013). At calving, maternal body protein may contribute to 88–91% of milk N (Taillon et
al. 2013). Maternal stores remain important throughout lactation, even when more protein

50

becomes available in their diets (Gustine et al. 2011, Taillon et al. 2013). Nursing mothers
transfer 15N to their offspring when body protein reserves are used to produce milk in the
neonatal period , which results in maternal tissue, including hair, being depleted in δ15N, while
neonate tissue becomes enriched. (Barboza and Parker 2006, 2008, Taillon et al. 2013, Endo and
Kobayashi 2022).
Another possible mechanism leading to lower hair δ15N in females with calves is that the
demands of lactation combined with dietary restrictions during this period result in increased
nitrogen-use efficiency. Before spring green up, caribou subsist primarily on lichen, which is
high in digestible energy but low in N. This diet requires caribou to minimize N losses from the
body by limiting the oxidation of amino acid N to urea (Barboza and Parker 2008). During
periods of increased nutritional demand, such as pregnancy or lactation, the body can
preferentially reroute more dietary amino acids from oxidation and excretion toward deposition
at sites of tissue synthesis (Fuller et al. 2004, Taillon et al. 2013). In caribou, maternal stores are
used to ensure a supply of protein for milk production, while dietary income is more likely used
to rebuild maternal body stores (Taillon et al. 2013). Since dietary proteins have lower δ15N than
body tissues, this mechanism of protein allocation in lactating caribou could account for the
lower hair δ15N we observed.
Finally, female caribou with calves separate spatially from those without during
parturition, seeking sites that reduce predation risk (Parker et al. 2009, Viejou et al. 2018).
Caribou with calves may avoid areas of high forage quality or quantity if those areas are
associated with increased predation risk, which can influence their dietary protein intake since
different forage types have different δ15N values (Parker et al. 2009, Drucker et al. 2010). The
mass of calves at birth, which is correlated with calf survival, is impacted by maternal protein

51

stores. In red deer (Cervus elaphus), calf birth weight, milk production, and milk protein content
have significant allometric relationships with maternal body mass (Landete‐Castillejos et al.
2003). Milk protein content is correlated with calf birth mass and calf mass gain (Landete‐
Castillejos et al. 2003). The relationship between maternal condition, and calf growth makes the
replenishment of body protein during summer vital for future reproductive success (Landete‐
Castillejos et al. 2003, Adams 2005, Barboza and Parker 2008, Parker et al. 2009). Since the
body mass of adults with calves was significantly lower than those without, it suggests that
dietary protein and energy available in the Kennedy Siding summer range was insufficient to
recover mass in caribou that successfully weaned a calf before the autumn.
Though differences between females with calves and those without were opposite to
predicted, they probably reflect demands of lactation and calf-rearing. The δ15N measurements
represent the amount of protein that is assimilated into body tissues (Young 1991, Gustine et al.
2012). Lactation increases the protein requirements of caribou by 110–130%, and because early
lactation often occurs before spring green up, meeting these demands may require the use of
body protein stores. (Barboza and Parker 2008, Parker et al. 2009). Lactation may also increase
the energy requirements of ungulates by 65–215% (Hardesty et al. 1988, Robbins 1993, Parker et
al. 2009). In elk, fecal glucocorticoid levels have been linked to daily and seasonal energy
requirements, and increase in response to the energy demands of lactation (Newediuk et al.
2024). Increased energy requirements, use of nutritionally inferior habitats, and increased
vigilance and predator avoidance could account for the increased fecal glucocorticoid and hair
cortisol (see supplemental material for Chapter Two) in adults with calves compared to adults
without that we observed. Supplemental feeding could accelerate the recovery of body protein
and energy reserves before winter, especially for caribou that lactated, thereby increasing

52

ovulation rates and calf survival through improvements to maternal body condition (Albon et al.
2017).
Nutritional limitation in yearlings
The proportion of animals with ribs showing and the measurements of δ15N were similar
among yearlings compared with adult females without a calf, which suggests yearlings are not
experiencing increased nutritional stress. Results from FGM and FTM models provide further
support that yearlings, as a group, are not nutritionally stressed. Although yearlings had greater
FGM levels compared with adults without a calf, which is consistent with nutritional stress, an
alternative explanation more consistent with body condition scores relates to social dynamics
within the herd. Caribou, especially females, are gregarious, and coexistence in large groups
requires them to use behaviours to determine and maintain social rank (Barrette and Vandal
1986, Weckerly and Ricca 2014). Aggressive behaviour in female caribou is most often
associated with conflicts over food resources. Barrette and Vandal (1986) found that in a mixed
sex group of woodland caribou, aggression was correlated with access to resources, amount of
time spent feeding at snow craters, age, and antler size. Among females, adults initiated most
aggressive interactions with one-year olds and the outcome (gaining or maintaining access to a
snow crater) was most often determined by antler size. Females with larger antlers were more
successful in their interactions, regardless of age (Barrette and Vandal 1986). Within the
Kennedy Siding herd, antler size and number of points is positively correlated with age, and in
anecdotal observations larger females act aggressively towards smaller females (unpublished
data), suggesting that adult females outrank one-year olds and have better access to and spend
more time at the feeders. Female caribou in mixed sex groups, like the one formed at Kennedy
Siding in the autumn, also exhibit higher frequencies of aggressive behaviour than females in

53

same sex groups (Weckerly and Ricca 2014). The relationship between social status and stress
has been examined in a wide range of species. In social, non-cooperatively breeding species such
as caribou subordinate animals generally exhibit higher physiological stress than dominants,
though this can be complicated by environmental conditions, reproductive status, and sex (Creel
2001, Goymann and Wingfield 2004). If younger caribou are subordinate to older individuals,
this could explain the elevated FGM levels we observed. Future research could combine
glucocorticoid analysis with behavioural observations to determine how social rank, age, and
physical characteristics (antler size, body weight, etc.) influence physiological stress in caribou.
In contrast with our prediction that yearlings would have lower FTMs compared with
females without calves, we found that FTMs were greatest among yearlings, probably because
they are still growing. Elevated T3 levels in younger animals have been observed in several
ungulate species because of the role thyroid hormones have in processes such as skeletal
maturation, reproductive maturation, and brain development (Todini 2007, Pasciu et al. 2024).
Thyroid hormone levels are also positively correlated with growth hormone levels in domestic
sheep and goats (Todini 2007, Pasciu et al. 2024). T3 levels decline as an animal ages (Todini
2007, Pasciu et al. 2024). Age related differences in T3 could also explain the relationship
between FTMs and body mass we observed. Female caribou reach mature body mass around
three-years of age (Parker 1981) . We can then infer that smaller caribou are likely younger and
would therefore have higher FTM levels than larger (older) animals.
Combined, these findings suggest that yearlings are not nutritionally stressed relative to
older females without calves. Improved nutritional condition in yearlings may be the result of
improved nutrition in both the current year as well as the previous year, when they were calves.
Calves that develop with abundant resources will be physically larger and more fecund as adults

54

(Pigeon et al. 2017, 2019). Evidence of an effect of feeding on body condition is suggested by
the fact that yearlings have fecal progesterone values indicating ovulatory activity, and some
returned to Kennedy Siding the following year having produced a calf(see chapter 3 for more
detail).
Inter-annual variation in nutritional condition
Pronounced differences in physiological bioindicators between years suggest that
environmental stochasticity could lead to inter-annual variation in nutritional value of habitat that
could make supplemental feeding particularly important in some years relative to others. Hair
δ15N measurements were lower in 2022 than in 2021, while fecal glucocorticoid and T3
metabolites were higher across all reproductive age classes, which may reflect interannual
variation in summer environmental conditions that affected nutritional condition for all animals.
Climatic factors such as air temperature, precipitation, and sunlight can affect the quality and
quantity of summer forage available to caribou (Lenart et al. 2002, Albon et al. 2017). Climate
change is expected to alter precipitation and temperature patterns that are critical for plant
survival and growth, thereby affecting forage composition and quality (Lenart et al. 2002, Albon
et al. 2017, Zamin et al. 2017). In Svalbard reindeer (Rangifer tarandus platyrhynchus), changes
in winter and summer weather patterns were linked to seasonal fluctuations in body mass and
vital rates (Albon et al. 2017). Higher temperatures and low levels of soil moisture have been
linked to decreased nitrogen content of forage (Lenart et al. 2002, Zamin et al. 2017).
Additionally, changes in climate that affect snow cover can alter both the timing of spring greenup and forage nutrient content, potentially resulting in a trophic mismatch if peak nutrient
availability does not coincide with the period of greatest nutritional need (Cebrian et al. 2008,
Zamin et al. 2017). Caribou will increase digestible dry matter intake to restore tissues that were

55

mobilized when food intakes were low. If 2022 foraging conditions were poor relative to 2021,
caribou may have increased their intake at Kennedy Siding. This could account for the increased
fecal T3 we observed in 2022, as T3 is associated with increased dry matter intake. If summer
range quality is inadequate or reduced by ongoing climate change, supplemental feeding in the
autumn could help buffer caribou against adverse nutritional effects. The BC Rocky Mountains
have experienced more extreme high temperatures, such as the 2021 heat dome, and northern BC
has been experiencing drier winter conditions, such as record low snowpack of 2023, conditions
which are expected to impact the high elevation plant communities caribou rely on (Holmes et al.
2015, Haghshenas and Wood 2025). Further research is needed to determine the impact of
climatic conditions on the protein and energy balance of caribou in British Columbia.
Changes in nutritional condition over the feeding period
We predicted that if caribou experienced nutritional deficits during the summer, FGM
levels would decline over the fall feeding period as supplemental feeding would alleviate
nutritional stress. Our modeling framework identified an overall decline in FGMs over the course
of the feeding period from September to January, which is consistent with our prediction. This
overall decline could reflect improved nutritional condition. An association between higher
glucocorticoids and reduced food availability, quality, and nutrients has been found for a variety
of species (Delgiudice et al. 1990, Stetz et al. 2013, Gregorio et al. 2019, Christianson et al.
2021, Touitou et al. 2021) including caribou and reindeer (Barboza et al. 2004, Wasser et al.
2011).
In contrast with our predictions, however, the magnitude of the decrease in FGMs was
similar among adults with calves, adults without calves, and yearlings. Furthermore, there were
no interactions between date and body condition or body mass, which we predicted would occur

56

if supplemental feeding were particularly important for individuals in poor nutritional condition.
These findings suggest that the magnitude of the change in FGMs over time was not strongly
affected by differences in nutritional condition among individuals.
An alternative or complementary explanation for the observed decrease in FGMs relates
to seasonal changes in photoperiod, with production of melatonin increasing with decreasing day
length. Melatonin inhibits the production of glucocorticoids by inhibiting the secretion of
corticotropin-releasing hormone leading to a decrease in the production of adrenocorticotropic
hormone, and, consequently a reduction cortisol or corticosterone levels (Konakchieva et al.
1997, Lightman et al. 2020). In Svalbard reindeer, Nilssen et al., (1985) found that serum cortisol
levels were lower in autumn and winter than summer, regardless of age or sex. Glucocorticoids
increase the sensation of hunger by binding to receptor sites that promote muscle catabolism,
limiting the secretion of appetite-suppressing hormones, and increasing blood glucose (Bini et al.
2022). A decrease in glucocorticoids in fall could contribute to the reduction in voluntary food
intake and basal metabolic rates seen in reindeer during the winter (Ryg 1984). Since
glucocorticoids are involved in glucose metabolism, a decrease in production could facilitate fat
deposition in the autumn and the conservation of body stores over the winter. Although
photoperiod could play a role in the decrease in FGMs we observed, the interaction between date
and year indicated that the magnitude of the decrease in FGMs differed between years. This
finding suggests that despite a potential seasonal effect, environmental and/or nutritional
differences between years probably also influence FGMs in the Kennedy Siding herd.
Similar to FGMs, we predicted that changes in FTMs during the feeding period would
reflect reduced nutritional stress (i.e., greater FTMs). In contrast with this prediction, we
observed that FTMs decreased during the feeding period. One explanation for this finding is that,

57

similarly as for glucocorticoids, seasonal changes in photoperiod may influence production of
thyroid hormones. In studies on domestic ruminants, blood thyroid hormone concentrations
increased with increasing daylength in spring and decreased with decreasing daylength in
autumn, and that plasma T3 can change rapidly in response to artificial photoperiodic cycles
(Todini et al. 2006, Todini 2007). Yasuo et al., (2006) also found that the expression of the type
II deiodinase gene, which catalyzes the conversion of thyroxine into T3 decreases with increased
photoperiod in Saanen goats (Capra hircus). In reindeer, serum T3 levels vary seasonally,
peaking in summer, and declining in winter (Ryg 1984, Nilssen et al. 1985). Although T3 levels
in cervids are thought to fluctuate with food intake, Ryg (1984) found that serum T3 in semi
domestic reindeer was greater in April and July than August and November, even when dry
matter intake was similar. This suggests that a less pronounced effect of food intake on T3 levels
could facilitate the deposition of fat in late summer and autumn, which is needed for
reproduction and over winter survival.
One limitation of our study is the lack of an unfed population for comparison. The
Kennedy Siding herd is unique in that it consistently uses the same small, easily accessible area
every year, making repeated sampling much easier than it would be for other Central Mountain
herds. Very little fecal hormone data exists for woodland caribou in British Columbia, and the
existing data is either from a single sample per individual and/or collected at a different time of
year, making comparisons to our data challenging (Freeman 2008, Lamb et al. 2024a). Repeated
sampling of an unfed population during the autumn would help to clarify our FGM and FTM
results and make a more explicit link between these indicators and supplemental feeding. For
example, Forristal et al., (2012) found that elk wintering on supplemental feed grounds in
Wyoming had higher FGM levels then elk wintering in native habitats. Bahnak et al., (1981)

58

found that serum T3 in white tailed deer does was higher in winter and early spring when fed a
high protein, medium energy diet compared to does fed a low protein, low energy diet. The
increase in female body mass since the start of feeding in 2014 along with continued population
growth suggests that supplemental feeding has a beneficial effect on population health (Heard
and Zimmerman 2021). While feeding couldn’t be directly linked to fecal hormone levels, the
decline in FGMs over time suggests that caribou are not experiencing an increase in
physiological stress from time spent at the feeders, except possibly yearlings, which may
experience stress from social dynamics in the herd.
Management implications
Our research adds to a growing body of literature on the importance of summer and
autumn nutrition to caribou in northeastern BC (Dale et al. 2008, Couturier et al. 2009, Kelly
2020, Cook et al. 2021, Heard and Zimmerman 2021, Denryter et al. 2022). We found that
caribou that raised a calf had lower body mass, were in poorer body condition and had higher
glucocorticoid measurements in hair and feces, which indicated that they arrived at Kennedy
Siding from their summer ranges in reduced physiological and nutritional condition compared to
those that did not have a calf. Caribou with calves had lower δ15N which, combined with lower
body masses, suggests that they relied on body stores of protein throughout lactation, and had not
fully recovered body stores by the autumn. These findings suggests that summer habitat may be
limiting for lactating caribou (Cook et al. 2004, Denryter et al. 2022). Lactating females are more
sensitive to variations in forage quality and quantity than non lactating females. Non lactating
females maintained on a low nutrition diet are able accrete fat and muscle tissues at nearly the
same rate as lactating cows maintained on a high nutrition diet (Cook et al. 2004). It is possible

59

for lactating females to accrete as much fat as non lactating cows if summer forage is accessible
and quality is adequate (Adamczewski et al. 1987, Cook et al. 2004).
Caribou that enter winter in poor body condition experience reduced survival and
reproductive success, which could have serious consequences for conservation and recovery
outcomes for small populations such as the Kennedy Siding herd. Autumn supplemental feeding
may help compensate for summer nutritional limitations by allowing caribou, especially those
that reared a calf, to rapidly recover body stores lost before the onset of winter. We found that
yearlings did not appear to be nutritionally limited, which suggests that improved nutrition in the
previous year could carry over and benefit caribou in the next. Interpreting the changes in FGMs
and FTMs over time requires further research, however these hormones did vary between years,
suggesting that environmental conditions influence these bioindicators in addition seasonal
conditions such as photoperiod. Ultimately, we could not link the bioindicators we measured
directly to supplemental feeding, however, the decline in FGMs we observed suggests that
feeding is minimally invasive and unlikely to have adverse health consequences, though
continued monitoring is required to ensure that remains the case. Finally, we conducted some of
the first longitudinal monitoring of endocrine activity in woodland caribou that reveals
information about caribou physiology and highlights gaps in understanding around links between
habitat and physiology. Our interpretation was limited by lack of data from other caribou
populations, so more studies are needed to address some the information gaps we identified,
ideally using a control population or pre or post-rut implementation of a feeding program.
Autumn supplemental feeding may be an effective management action for herds with
inadequate summer range conditions, predictable movement patterns, or prior exposure to
pelleted food, such as in maternity pens. The cost of supplemental feeding is around

60

$7000/caribou (CDN), considerably less costly than other management actions such as wolf
reduction ($26,000/caribou) or maternity penning ($148,000/caribou; Johnson et al., 2019, Heard
and Zimmerman 2021). More broadly, summer and autumn nutrition should be considered when
developing habitat management strategies for caribou. Widespread habitat conservation and
restoration is needed if caribou are to remain on the landscape in the long term. Future land use
planning should account for the nutritional landscape of the environment and promote and
protect diverse plant community types (Cook et al. 2021, Denryter et al. 2022). Fostering a
landscape that supports caribou nutritional needs will help maintain self-sustaining populations
by improving body condition, reproduction, and survival.

61

Chapter 3: Reproductive rates among female caribou with fall supplemental feeding
Introduction
Habitat loss from human activities is the primary cause of population decline for most
threatened and endangered species worldwide (Myers et al. 2000). In North America, the once
abundant woodland caribou (Rangifer tarandus caribou) have experienced dramatic declines
throughout their range over the past few decades (Festa-Bianchet et al. 2011, Environment
Canada 2014). In British Columbia (BC), the Central and Southern Mountain populations have
experienced steep declines of up to 64%, and 12 of 24 subpopulations have been extirpated
(Environment Canada 2014, Environment and Climate Change Canada 2018). The primary cause
of these declines has been attributed to human resource extraction activities such as clearcut
logging (Festa-Bianchet et al. 2011, Wasser et al. 2011, Johnson et al. 2015). The proximate
cause is thought to be unsustainable predation caused by habitat changes that increase the
abundance, access, and mobility of predators such as wolves (Canis lupus; Wasser et al., 2011,
Hervieux et al., 2014, DeMars and Boutin 2018). Threats to caribou are further exacerbated by
climate change and the associated changes in snowpack, fire regimes, and forage availability and
quality (Kinley et al. 2007, Cebrian et al. 2008, Joly et al. 2010, Tyler 2010, Gustine et al.
2014b).
Forage resources influence most aspects of ungulate ecology. Body condition is the state
of body components (e.g., fat and protein) which influence an animal's ability to survive and
reproduce (Harder and Kirkpatrick, 1994). The deposition and mobilization of body fat and
protein depends on an animal’s physiological state and environmental conditions (Parker et al.
2009). Body condition is often important in determining whether or not an animal dies, lives

62

without reproducing, or lives and reproduces (Parker et al. 2009). The maintenance and growth
of a population is in part dependent on the environment having sufficient nutritional resources
year-round (Crête and Huot 1993, Cook et al. 2004, Dale et al. 2008, Parker et al. 2009).
Most studies of nutritional ecology in northern ungulates, including caribou, have
focused on winter foraging conditions, as winter is the season when food is assumed to be most
limiting. There is, however, increasing evidence to suggest that late spring to early autumn may
be when nutritional limitation occurs for caribou (Crête and Huot 1993, Dale et al. 2008,
Couturier et al. 2009, Cook et al. 2013, 2021, Kelly 2020, Heard and Zimmerman 2021, Denryter
et al. 2022). During the summer, the energy costs of lactation plus those required to recover from
the weight loss in the previous winter can more than double daily energy requirements compared
to those of non-lactating females in the summer and autumn (Barboza and Parker 2008, Parker et
al. 2009). Female caribou lose body mass for approximately three weeks following calving
because of the high costs of lactation (Parker et al. 2009). Adequate nutrition is needed for a
caribou to both lactate and replenish body reserves before winter (Cook et al. 2004, 2013, Dale et
al. 2008, Parker et al. 2009). Despite this requirement, many habitats in BC may be inadequate to
support the daily energy and protein requirements needed for both lactation and replenishment of
body reserves, and those that can are not highly available (Denryter et al. 2022). Widespread
nutritional inadequacies in habitat may result in a gradual reduction in population productivity
that varies in magnitude along with the degree of nutritional inadequacy (Crête and Huot 1993).
Furthermore, changes in predation risk associated with disturbances such as linear features
within caribou habitat can alter the ability of females to access high-quality forage. Caribou are
generally predator-sensitive foragers and may trade off high-quality foraging opportunities as
vegetation greens up at lower elevations to reduce predation risk, especially when they have a

63

calf (Gustine et al. 2006, Jones et al. 2007, Parker et al. 2009, Viejou et al. 2018). The combined
effects of poor nutritional quality and predation risk both limit potential for population growth of
imperilled caribou populations in BC. For an already imperilled population, any additional
limitations on population growth rate could compromise recovery and conservation actions.
The body reserves accumulated by female caribou during the summer are closely linked
with reproductive rates, including the probability of an animal ovulating, carrying a fetus to term,
the growth of offspring in utero and post partum, and over-winter survival (Crête and Huot 1993,
Cameron and Hoef 1994, Chan-McLeod et al. 1994, Adams and Dale 1998, Barboza and Parker
2006, Parker et al. 2009, Albon et al. 2017). If caribou cannot recover body stores lost to
lactation and calf rearing before the breeding season, they may forgo reproduction that year in
favour of investing energy into self maintenance (Cameron 1994, Chan-McLeod et al. 1994,
Parker et al. 2009). Nutrition also plays a major role in determining the age of primiparity;
animals often begin reproducing at a younger age when nutritional conditions are favourable
(Festa-Bianchet et al. 1994, Gaillard et al. 2000). Furthermore, mothers in good condition may
invest more energy into their offspring, which can increase survival (Landete‐Castillejos et al.
2003). In caribou, mothers in better body condition produce calves with greater neonatal birth
weights, which in turn increases calf viability and reduces susceptibility of calves to predation
(Cameron et al. 1993, Chan-McLeod et al. 1994, Adams 2005, Taillon et al. 2012).
In BC, The Central Group of Mountain caribou occur on the east side of the Rocky
Mountains, north of Kakwa Provincial Park in the south, to the southern shore of the east arm of
Williston Lake in the north (Environment Canada 2014, Environment and Climate Change
Canada 2020). There are currently six populations defined in the Central Group, the Quintette,
Narraway, Burnt Pine, Scott West, Klinse-Za (also known as Moberly Scott), and Kennedy

64

Siding herds (COSEWIC 2014, Environment and Climate Change Canada 2018, 2020). The
Kennedy Siding herd declined from a high of 120 in 2007 to 41 in 2012, the Quintette herd from
265 in 2002 to 106 in 2014 and the Klinse-Za and Scott East herds combined from 191 in 1997
to 36 in 2013 (Seip and Jones 2018, Serrouya et al. 2019). The Burnt Pine subpopulation was
confirmed extirpated in 2014 (COSEWIC 2014). In response to these declines, several
management strategies have been applied to herds in the Central Group. The province of British
Columbia began a wolf reduction program in the area in 2015, and over five years the density of
wolves was reduced by over 80% (Bridger 2019). A maternal penning program was implemented
for the Klinse-Za herd starting in 2014 with the goal of reducing predation on caribou calves
(McNay et al. 2022). A supplemental feeding program was implemented for the Kennedy Siding
herd beginning in 2014 (Heard and Zimmerman 2021).
Supplemental feeding is a management tool that uses nutrition to maintain or increase a
population by improving body condition, reproduction and survival (Milner et al. 2014, Murray
et al. 2016, Horstkotte et al. 2022). If summer nutrition is inadequate for caribou in BC, then fall
supplemental feeding could compensate for that limitation and contribute to population growth.
Supplemental feeding does appear to have an additional effect on population growth rate over the
effect of wolf reduction alone, with the Kennedy Siding herd having a faster population growth
rate (λ = 1.16) than the Quintette herd, which only received wolf reduction (λ = 1.08; Heard and
Zimmerman 2021). One possible mechanism by which feeding improves population growth is by
increasing the proportion of females who become pregnant each year by increasing body weight
and fat reserves, factors that can impact the onset of cyclic ovarian activity and pregnancy
(Cameron et al. 1993, Gerhart et al. 1997a). Specifically, feeding could reduce skipped years of
reproduction among adult females and/or increase the proportion of yearlings that breed.

65

Alternatively, improved maternal body condition going into winter could enable females to
invest in higher quality calves, leading to increased calf survival.
An understanding of basic reproductive processes and components of individual fitness
can aid species management and conservation because they influence population dynamics in
wild animals (Schwarzenberger 2007, Sontakke 2018). Common methods for assessing
reproduction in caribou include analysis of hormone levels and movement patterns of GPS
collared animals. Progesterone is a steroid hormone that increases after ovulation and is the
principal hormone responsible for the maintenance of pregnancy in mammals (Smith 2009,
Crowe 2016, Shrestha 2021). The mammalian estrous cycle consists of a follicular and luteal
phase (Crowe 2016). During the follicular phase an ovarian follicle and its enclosed oocyte
mature in preparation for ovulation and fertilization and progesterone levels are low (O’Connor
et al. 2013, Crowe 2016). The luteal phase is the period following ovulation when the corpus
luteum is formed and is associated with increased secretion of progesterone (Asher 2011,
O’Connor et al. 2013, Shrestha 2021). If pregnancy does not occur, the corpus luteum
degenerates and progesterone levels decline, allowing for another opportunity to establish a
pregnancy (Asher 2011, Shrestha 2021). If pregnancy is established the corpus luteum continues
to produce progesterone through the gestation period or until the placenta takes over
progesterone production, depending on the species (Asher 2011). The effect of nutrition on
progesterone secretion and ovulation has been extensively studied in domestic ruminants.
Improved pre-ovulatory nutrition increases the size of the ovulatory follicle and the
progesterone-secreting ability of the corpus luteum (Robinson et al. 2006). For cattle, postpartum nutrition plays a major role in the timing of the onset of oestrous cyclicity after calving,
the normality of its expression, serum progesterone levels, and, conception rate (Robinson et al.

66

2006). Cattle in good body condition at calving and those receiving high food intakes after
calving have shorter anoestrous periods than thinner and less well-nourished animals (Robinson
et al. 2006). In sheep, undernutrition reduces the number of follicles that develop and therefore
the number available to ovulate and may also impact the secretion of gonadotropins such as
follicle-stimulating hormone and luteinizing hormone (Robinson et al. 2002, 2006). Flushing, the
practice of providing improved nutrition at the beginning of the mating season, is commonly
used to increase ovulation rates in domestic ewes (Robinson et al. 2002, 2006). Flushing in ewes
in low body condition has a larger effect on ovulation than in ewes in higher body condition
(Robinson et al. 2002, 2006).
In caribou, serum progesterone tests on blood samples taken from captured animals have
commonly been used to assess pregnancy status (Rettie and Messier 1998, Wittmer et al. 2005,
DeMars et al. 2013). The stress and hazards associated with capture and restraint can be
detrimental to the animals' health (Schwarzenberger 2007, Morden et al. 2011). This means that
repeated blood sampling on a regular basis is often not possible. Fecal hormone analysis has
been developed as an alternative, non-invasive technique that is now widely employed for
studies of stress and reproductive physiology in wildlife (Schwarzenberger 2007). Hormones like
progesterone are metabolized by the liver into several metabolites which are excreted in feces
and can then be measured using immunoassay techniques (Brown 2018). Due to the pooling of
metabolites during excretion, fecal samples present cumulative hormone secretion hours or days,
depending on the species (Schwarzenberger et al. 1996, Möstl and Palme 2002, Palme et al.
2005). Fecal sampling also enables frequent data collection without disturbance of the animal,
allowing for longitudinal studies of hormones that can be used to assess changes in reproductive
activity (Schwarzenberger 2007). The concentration of progesterone metabolites in feces has

67

been used to study reproductive components such as seasonality, estrus and ovulation,
pregnancy, and fertility status in a variety of wildlife species (Schwarzenberger 2007, Brown
2018, Sontakke 2018). In caribou, analysis of single fecal samples has been used to determine
pregnancy in late winter, but repeated sampling has not been used to monitor ovarian activity,
likely due to difficulties obtaining repeated samples from wild populations (Messier et al. 1990a,
Morden et al. 2011, Joly et al. 2015).
Previous methods for assessing parturition in wild ungulates have included aerial surveys
during the calving season (Whiting et al., 2012), serum progesterone tests on blood samples
taken from captured animals (Wittmer et al., 2005), and vaginal implant transmitters. Rates of
neonate calf survival are usually determined by late spring aerial surveys or by radio-collaring
newborn calves. These methods are invasive and stressful to the animal, costly to researchers,
and may miss early post natal deaths (DeMars et al. 2013). Analysis of movement patterns using
GPS radio-telemetry data provides an alternative method of inferring parturition and survival of
neonatal offspring (DeMars et al. 2013). Parturition in caribou corresponds with an abrupt
decrease in movement rates. Over the first four to five weeks following parturition, the female’s
movement rate will gradually increase as the calf grows and becomes more mobile. If the female
loses her calf, her movement will quickly return to pre calving rates. This change in movement
pattern is detectable from GPS collar data when a location is acquired every one to three hours,
though false negatives can occur if the calf is still born or dies shortly after birth (DeMars et al.
2013, Mueller et al. 2022). GPS movements can assess calf survival for the first four weeks of
life, after which calf mobility begins to approach adult movement rates (DeMars et al. 2013).
Understanding parturition and neonate survival will better inform conservation measures and aid
recovery actions such as the conservation of specific habitat types used for calving.

68

We used a combination of population counts, body condition measurements, and fecal
progesterone measurements collected during autumn from the Kennedy Siding supplemental
feeding site and GPS-collar data from female caribou during May and June to characterize rates
of ovulation, pregnancy, parturition, and calf survival in the Kennedy Siding herd. We discuss
these rates in the context of reproductive parameters reported in caribou herds elsewhere.
Specifically, we predicted that the Kennedy Siding herd would have greater rates of pregnancy
among adults and yearlings, higher parturition rates, earlier parturition dates, and improved calf
survival compared with values reported in the literature for unfed caribou herds and closer to
values for herds known to have access to high amounts of summer food (Ouellet et al. 1997,
Prichard et al. 1999). Further, we speculated what these comparisons indicate about potential
mechanisms linking supplemental feeding and population growth. Overall, our results contribute
to an increased understanding of reproduction in woodland caribou and help to inform
conservation strategies and land-use plans for central mountain caribou.
Methods
Study site
The range of the Kennedy Siding caribou herd is in central British Columbia, within the
traditional territory of the McLeod Lake Tse’Khene (McLeod Lake Indian Band). The habitat
used by the herd is within the Sub-Boreal Spruce and Engelmann Spruce Sub-Alpine Fir
biogeoclimatic zones (Meidinger and British Columbia 1991). The terrain is characterized by
mountains and rolling hills ranging in elevation from 710m to 2010m above sea level (Jones et
al. 2007). Caribou typically spend most of the year at high elevations dispersed in relatively
undisturbed sub-alpine meadows and forests of Engelmann Spruce (Picea engelmannii) and
Subalpline Fir (Abies lasiocarpa). Each fall the population descends to a small distinct area of

69

lodgepole pine (Pinus contorta) forest at low elevation on the southwest edge of their range,
which has abundant terrestrial lichen. Movement from high to low elevations is stimulated by the
first snowfall events in the mountains. When on their fall range, caribou forage on terrestrial
lichens, arboreal lichens, forbs and shrubs. Snow begins to accumulate on the ground in
November and when it reaches a depth of about 1 m (usually in mid-January; (Jones et al. 2007),
the herd moves back to high elevations, where they feed on terrestrial lichens in wind-swept
areas or arboreal lichens where snow is deep and supports their weight. Most (2894 ha) of the
fall range of the Kennedy Siding herd, including the feeding site, was designated an Ungulate
Winter Range under the British Columbia Forest and Range Practices Act in 2002 (Arthur 2002).
The feeding area is located approximately 30 km south-east of the community of
Mackenzie. Feeding occurs in the vicinity of the Kennedy (BCR) railway siding and Highway
97. The area experiences frequent vehicle traffic from forestry, railway, hydro, and other
industries as well as by recreational users such as hunters. Other large mammals that inhabit the
area include moose (Alces alces), white-tailed deer (Odocoileus virginianus), elk (Cervus
canadensis), lynx (Lynx canadensis), wolf (Canis lupus), coyote (Canis latrans), grizzly bear
(Ursus arctos), and black bear (Ursus americanus; (Heard and Zimmerman 2021).
Feeding, identification of individuals, and body condition measurements
We conducted supplemental feeding and identification of individuals based on the
procedures described in Heard and Zimmerman (2021). The food pellets were produced by HiPro Feeds/Trouw Nutrition Canada Inc. in Grand Prairie, Alberta. The formula for the pellets
was developed at the University of Alaska to provide balanced and adequate levels of energy,
nitrogen, and minerals for growth and reproduction of caribou (Barboza and Parker 2006) and

70

have been used in other nutritional studies (Gustine et al. 2011) and conservation projects such as
the Chisana Herd maternity pen (Adams et al., 2019).
Feeding began when caribou first arrived in late September and early October. We
contracted the McLeod Lake Indian Band to build the feeders and deliver the food pellets.
Initially, we provided small amounts of pellets in plastic livestock buckets so that any remaining
feed could be removed from the site when the caribou were not present. Food was removed to
limit attractants to the bears that used the area. We began ad-lib feeding in November when the
bears were assumed to be in hibernation and continued feeding until January 15th. We stopped
feeding on that date to avoid interfering with the natural movement of the herd back into their
alpine habitats. Caribou consumed approximately 70kg of food per animal over the feeding
period.
We used 7–9 Reconyx Hyperfire motion-sensor trail cameras, focused on either a feeder,
a salt block, or along a trail, to obtain a continuous photographic record of individual caribou
during the feeding period. The sex, age class (calf or adult), and the relative body condition of
each caribou were recorded. Animals were considered in poor body condition if ribs were
visible, and good body condition if ribs were not visible (figure 11). Individual caribou were
identified from photos using the number and position of antler points, radio collars or ear tags, or
unique markings (e.g., scars). Calves did not always have unique antlers, so we identified
individuals based on their close association with mothers of known identity upon arrival at the
feeders. The caribou were attracted onto a platform scale (using salt or pellets), with trail
cameras positioned to photograph the individual caribou and record the weight (kg) on the
scale’s digital display. If an animal was photographed on the scale more than once, we used the
average of the weights recorded as our final body mass measurement.

71

Figure 11: Examples of female caribou (Rangifer tarandus caribou) at the Kennedy Siding feeding site considered in poor body
condition and female caribou considered in good body condition. The caribou in images A and B are considered in poor
condition based on their visible ribs, spine, and hip bones in the trail camera images. The caribou in images C and D are
considered in good condition based on the lack of visible ribs or other bones.

Animal care
Sample collection was conducted under research permits provided by the British
Columbia Ministry of Forests as well an Animal Care and Use Protocol approved by the
University of Northern British Columbia (ACUC Protocol Number: 2021-15). Some of the adult
female caribou that used the feeding site (21 in 2021 and 19 in 2022) had collars and/or ear tags
that were deployed as part of the Northern Mountain and Central Mountain Caribou Capture
Program run by the BC Provincial Government, under the BC Provincial Animal Care Protocol
(Permit Number: FJ21-618702).

72

Sample collection
We conducted sample collection with research technicians from the McLeod Lake Indian
Band. We collected fecal pellet samples opportunistically by observing female caribou around
the feeders. When defecation occurred, the individual caribou was identified, and the pellets
were collected in a resealable plastic bag after the caribou moved away. The pellets were stored
in a cooler with an ice pack until they could be transferred to a freezer and stored at -20°C within
four hours of defecation. In 2021, we collected 155 pellet samples from 37 females (1-14
samples / caribou). During the 2022 field season, we collected 286 pellet samples from 43
females (1–16 samples / caribou).
We collected hair samples using a noninvasive snaring method. We affixed metal
alligator clips to a piece of wood which were held open using small pieces of wooden toothpicks.
The piece of wood was attached to one of the feeders and when caribou put their head down to
eat, their neck pressed down on the clips, knocking out the toothpick and causing them to close.
When the caribou moved away, the clip caught a sample of guard hair from the neck, typically
with at least 10 hairs/clip. This method appeared to cause little disturbance to the caribou, as they
did not avoid using the feeder where the clips were set up. We collected hair samples from 27
caribou in 2021 and 33 caribou in 2022, which were used for DNA analysis.
DNA analysis
We sampled feces for DNA by taking three surface rubs from each pellet sample using
cotton swabs. We stored samples of guard hairs with visible root bulbs in unwaxed paper
envelopes. DNA extraction, amplification, and sequencing of hair and fecal samples were
conducted by Wildlife Genetics International in Nelson, BC. The DNA was extracted using
QIAGEN DNeasy Blood and Tissue kits. For hair samples, the roots clipped from a target of ten

73

guard hairs was used. For fecal swabs ~ 5 mm clippings from the used end of one swab per
sample was used. Individual identity was established using 18 microsatellite markers established
by Serrouya et al., (2012) along with a ZFX/ZFY sex marker. The success rate for producing all
19 loci was 85% for hair samples and 91% for fecal swabs. DNA results were used to identify
unmarked females between years and to obtain an exact or estimated age of the animal.
Fecal progesterone enzyme linked immunosorbent assay (ELISA)
To extract hormones from the fecal samples, we first thawed and homogenized each
sample. A portion of the homogenized sample was transferred to a plastic snap cap vial which
was covered with a low lint delicate task wipe secured with an elastic band. The vials were
lyophilized for ~ 26 hours. The dried feces were powdered and stored in the freezer at -20˚C.
Hormone metabolites were extracted from 0.2g of dried feces in 5ml of 80% HPLC grade
methanol (Thermo Fisher Scientific)/20% water. Vials were sealed and rotated at 160 rpm
overnight at room temperature. A portion of the methanol extract was transferred to a 2ml
centrifuge tube with a cap that included an o-ring to prevent evaporation. Extracts were
centrifuged for 30 minutes, and the supernatant was removed and stored at -20˚C until analysis.
We followed a modified version of the protocol described by Graham et al. (2001) to
determine concentrations of fecal progesterone metabolites. We coated ninety-six well plates
with affinity-purified rabbit anti mouse gamma globulin (25 µg/plate [Sigma-Aldrich] dissolved
in coating buffer (0.015 M Na2CO3 [VWR], 0.035 M NaHCO3 [Sigma-Aldrich]; pH 9.6) and
incubated them overnight at room temperature. The coated plates were washed (0.04% Tween 20
[Sigma-Aldrich]) and refilled with 200 µl Trizma assay buffer (0.02 M Trizma [Sigma-Aldrich],
0.300 M NaCl [VWR], 0.1% bovine serum albumin [VWR]; pH 7.5) and stored at room
temperature for at least 1 hour to block non-specific binding. We diluted fecal extracts (1:4 to

74

1:32) in Trizma assay buffer prior to the assay. The wells were emptied of the Trizma buffer and
50 µl each of diluted sample and standards were dispensed. Horse-radish peroxidase-labeled
cortisol was dispensed, followed by anti-cortisol antibody or (Anti-progesterone CL425; CJM
Munro, UC Davis). The plates were incubated overnight at room temperature. Following
incubation, we washed the plates three times and then added 200 µl of substrate solution (0.5 mL
of 0.016 M tetramethylbenzidine [VWR] in dimethylsulphoxide [Thermo Fisher Scientific], and
100 mL of 0.175 M H2O2 [VWR] diluted in 24 mL of 0.01 M C2H3O2Na [Thermo Fisher
Scientific]; pH 5.0) to each well. The plates were incubated for 45 minutes at room temperature
and then the enzyme reaction was stopped by adding 50 µl of stop solution to each well (3 M
H2SO4 [Thermo Fisher Scientific]). The optical density was measured at 450 nm. If the sample
duplicates had a coefficient of variation (CV) greater than 15%, we re-assayed the sample. The
cross-reactivity of the progesterone antibody was reported by Graham et al. (2001), including
progesterone (100%), and pregnanediol, androstenedione, and corticosterone (<0.1%).
As an analytical validation, we first created a pooled sample of fecal extracts from
different female caribou. To establish parallelism, we tested serial two-fold dilutions of the
pooled sample for comparison displacement curves (Figure 12). The intra-assay coefficient of
variability was determined by assaying replicates of the pooled sample on the same plate. The
intra-assay CV for 35 samples was 6.9% at 59.3% binding. The inter-assay coefficient of
variation was determined by running a replicate of the pooled sample at a high and low
concentration as controls on each plate. The inter-assay CV was 16% at 21.3% binding and
10.8% at 71.8% binding over 12 plates. As a physiological validation, we compared the
concentration of fecal progesterone metabolites between the pooled extracts of female caribou
and a combined extract from fecal samples taken from two adult male caribou at the feeders. We

75

used this comparison to ensure that the progesterone ELISA could measure changes in fecal
progesterone metabolites associated with the female reproductive system with the assumption
that fecal progesterone metabolite concentrations would be similarly low in males and
anovulatory females.

Figure 12: Parallelism of progesterone standards (solid line) and a serially diluted pooled sample extract from female caribou
(Rangifer tarandus caribou) fecal samples (dashed line) collected in 2021 and 2022 at the Kennedy Siding feeding site in British
Columbia, Canada.

76

Statistical analysis
All data analysis were conducted using R (R Core Team 2023). We classified each
caribou sampled according to their age and reproductive status (yearlings and adults [≥ 2 years
old] with and without a calf at heel). Yearlings were identified using DNA collected from calves
the previous year. Adults without calves were identified using DNA collected in previous years
or unique radio collars and/or ear tags. We performed Kruskal-Wallis rank sum tests followed by
Wilcoxon post-hoc pairwise comparisons to compare body mass between one -year olds, adults
with a calf, and adults without a calf, and a Wilcoxon test to compare body mass in animals in
good and poor body condition (Kassambara 2019).
We considered an animal to be ovulatory based on whether they had at least one fecal
sample with a FPM concentration above the baseline determined from male samples. We used
general linear mixed effects regression models to assess the impact of age, reproductive status,
and body condition on fecal progesterone concentrations (Bates et al. 2015). We applied a
square-root transformation of fecal progesterone prior to analysis to obtain normal distributions
of the model residuals. We included the effects of reproductive age class, body mass, and body
condition body condition as fixed effects. Body mass was square root transformed to normalize
the distribution, then standardized to a mean of 0 and a standard deviation of 1 prior to analysis.
Individual identity was included as a random effect in all models. We examined any models that
had a ΔAICc within 0–2 units of the top model for uninformative parameters by assessing
whether they differed from the best model by one parameter and had similar log-likelihood
values as the top model (Arnold 2010).

77

We estimated parturition rates based on the radio collar movements of 19 individual
female caribou in the Kennedy Siding herd between 2020 and 2024 using data and analysis
provided by M. Klaczek (personal communication, 2024). Parturition status was estimated using
methods from DeMars et al. (2013). The movement models represent three states of females
during calving season; 1) did not calve, 2) calved and the calf survived to four weeks, and 3)
calved and calf was lost by four weeks of age. For the model representing females that did not
calve, the movement rate remained constant over time. For the model representing females that
calved, and the calf survived to four weeks, there is a single break point where movement rate
abruptly drops, indicating calving, before increasing again without a clear breakpoint. For the
model representing females that calved and then subsequently lost their calf, there are two break
points indicating when calving and calf loss occurred. We compared the model results to
observations made in the fall to confirm that females did or did not have a calf. Parturition dates
and dates of calf mortality were also estimated by radio collar movement models. We estimated
calf survival by comparing the estimated parturition rate to the number of adult females with a
calf at heel when they arrived at the feeders.
Results
Body mass and nutritional condition
We collected body condition and hormone data from 45 individual female caribou during
the 2021 and 2022 feeding seasons. We identified and sampled 20 of the total 45 animals in both
years. In 2021, we classified 14 animals as adults with a calf, five as adults without a calf, and
six as yearlings. In 2022, we classified 21 animals as adults with a calf, 13 as adults without a
calf, and six as yearlings.

78

The body mass of all caribou sampled in 2021 and 2022 ranged from 95 –153 kg overall
(n=65, median = 128 kg). Among reproductive age classes, body mass ranged from 105 –150.5
kg (n=35, median = 128.5 kg) for adults with a calf at heel and 107 –153 kg (n = 18, median =
138.0 kg) for adults without a calf at heel. Yearling caribou ranged in body mass from 95 –125.5
kg (n=12, median = 106 kg). The distribution of body masses differed among reproductive age
classes (Kruskal-Wallis Test; H = 22.87, df = 2, p <0.001). Post-hoc pairwise comparisons
revealed that adults with a calf at heel had smaller body masses compared to adults without a calf
(p = 0.036). As expected, yearlings had smaller body masses compared to adults with a calf (p <
0.001) and adults without a calf (p < 0.001). All yearling females had body mass measurements
that were greater than 70% (92.75 kg) of the median adult female body mass (132.5 kg; Figure
13). Of the caribou sampled, almost half were classified as being in poor body condition based
on the visibility of ribs (Table 6). Most (83%) of the individuals in poor body condition were
adults with a calf (Table 6); however, a Wilcoxon signed-rank test showed that the difference in
body mass between caribou in good and poor body condition was not significant (z = -0.118, p =
0.906).

79

Figure 13: Body mass (kg) of adult female caribou (Rangifer tarandus caribou) without a calf at heel (A) and with a calf at heel
(AC) compared with yearling females (Y) using the feeding site at Kennedy Siding, British Columbia, Canada, in 2021 and 2022.
The centre line of each box represents the median and the outer limits represent the 25th and 75th percentiles. The lines extending
below and above each box represent 1.5 times the interquartile range for each percentile. The dashed line represents the median
adult female body mass (132.5 kg). The dotted line represents 70% of the median adult female body mass (92.75 kg). Sample
sizes are shown at the bottom of the panel.
Table 6: Body condition (assessed as visibility of ribs) of female caribou (Rangifer tarandus caribou) in each reproductive age
class at the Kennedy Siding feeding site in British Columbia, Canada, in 2021 and 2022.

Age class
Ribs
visible

No
Yes

Adult with a calf
10
25

Adult without a calf
15
3

Yearling
10
2

80

Fecal progesterone metabolites
We analyzed progesterone metabolites (FPM) in 387 fecal samples from 45 individual
female caribou. The number of samples collected per individual ranged from 1 –18.
Concentrations of FPMs ranged from 141– 2612.2 ng/g with an overall median of 897.35 ng/g.
The median and range of FPM concentrations were very similar among age classes. Yearlings
had a median FPM concentration of 942.1 ng/g (range = 169.4 –2316.7 ng/g), adults with a calf
at heel had a median of 877.6 ng/g (range = 228.6 ng/g –2269.9 ng/g), and adults without a calf
had a median of 925.1 ng/g (range = 141– 2612.2 ng/g; Figure 14). The baseline progesterone
concentration determined by quantifying FPMs in feces from five adult male caribou was 162.8
± 28.9 ng/g. All the female caribou sampled had at least one FPM value above the baseline value
determined in males, and therefore all were considered ovulatory (Figure 14).

81

Figure 14: Fecal progesterone metabolite concentrations adult female caribou (Rangifer tarandus caribou) without a calf at heel
(A) and with a calf at heel (AC) compared with yearling females (Y) using the feeding site at Kennedy Siding, British Columbia,
Canada, in 2021 and 2022. The centre line of each box represents the median and the outer limits represent the 25th and 75th
percentiles. The lines extending below and above each box represent 1.5 times the interquartile range for each percentile. The
dotted line represents the baseline progesterone metabolite concentration derived from a pooled male sample mass (162.8 ng/g).
Sample sizes are shown at the top of the panel.

Of four candidate models used to predict FPM concentrations, the null model ranked
highest, indicating that none of the variables tested (i.e., reproductive age class, body condition,

82

and body mass) explained more variation in FPMs than would be expected due to chance alone
(Table 7).
Table 7: Ranking of four candidate general linear mixed-effects models used to predict fecal progesterone metabolite
concentrations in fecal samples (n = 397) from female caribou (Rangifer tarandus caribou) sampled between fall and winter of
2021 and 2022 at the Kennedy Siding feeding site in British Columbia, Canada.

Rank
1
2
3
4

Candidate model
df
logLik
AICc
ΔAICc
ωi
Null
3
-1313.54
2633.15
0
0.543
Reproductive age class
5
-1312.17
2634.51
1.355
0.276
Body condition + body mass
5
-1312.89
2635.93
2.779
0.135
Reproductive age class + body
7
-1311.89
2638.07
4.921
0.046
condition + body mass
Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model; ω i
= Akaike weight.

Parturition and calf survival rates
The total annual parturition rate estimated using radio collar data from 40 females
monitored between 2020 – 2024 was 87.5% with 12.5% of females not producing calves (Table
3). Calving date ranged from May 11 to June 6. Of females that gave birth, 25.7% lost their
calves before the fall (Table 8). Estimated parturition dates were available for 20 animals
between 2021 and 2024. Parturition dates ranged from May 11th to June 7th (mean date = May
25th ± 7 days). Estimated dates of calf loss were available for five of the nine animals identified
as having lost their calf. All recorded calf mortalities occurred within 30 days of parturition,
between June 12th and June 21st.
Table 8: Parturition and calf survival status of 40 female caribou (Rangifer tarandus caribou) with radio collars in the Kennedy
Siding herd, central British Columbia, Canada, between 2020 and 2024. Parturition and calf mortality was estimated using a
model developed by DeMars et al. (2013).

Year

N

Calved (%)

Did not calve (%)

2020

9

9 (100)

0 (0)

Lost calf
(%)
1 (11)

2021

10

8 (80)

2 (20)

2 (25)

2022

7

7 (100)

0 (0)

1 (14)

2023

10

7 (70)

3 (30)

2 (29)

2024

4

4 (100)

0 (0)

3 (75)

83

Total

40

35 (87.5 ± 5)

5 (12.5 ± 5)

9 (25.7 ±
7)

Discussion
We used a combination of body condition metrics, hormone analysis, and demographic
rates to characterize reproductive parameters in the Kennedy Siding caribou herd. We found that
all females, including yearlings, had fecal progesterone concentrations reflective of ovulation,
indicating that all females had the capacity to become pregnant. In contrast, some females,
especially those that produced a calf the year sampling occurred and yearlings, would be
expected to exhibit suppressed ovarian activity if amount or quality of nutrition limits pregnancy
rates in this population (Thomas 1982, Mani et al. 1992, Webb et al. 2004, Chaves et al. 2024).
The rates of ovulation we observed could be associated at least in part to supplemental feeding,
though further comparisons with unfed populations would be required to confirm the influence
of supplemental feeding on ovulation. Compared with caribou herds elsewhere, caribou at
Kennedy Siding had relatively high rates of parturition, pregnancy among yearlings, and calf
survival. These findings provide preliminary evidence that supplemental feeding could lead to
population growth in this herd via multiple mechanisms.
Body condition, ovulation, and primiparity
All females sampled had at least one FPM value above the male baseline, indicating that
all females in the herd were ovulatory and therefore capable of becoming pregnant. The fecal
samples, however, were collected during the breeding season and early gestation and were still
fluctuating; therefore we were unable to use FPMs to determine which females became pregnant
following ovulation and fertilization. Body mass, body condition, and reproductive age class
were not identified by the modeling framework as strong predictors of FPM values. These

84

findings suggest that the body mass threshold for ovulation was met, and all animals sampled
had the capacity to become pregnant. By contrast, findings in caribou and other large herbivore
species indicate that nutritional limitation can suppress ovulation cycles, leading to years of
skipped reproduction in adult females or later age at primiparity (Thomas 1982, Festa-Bianchet
et al. 1994, Cook et al. 2013, Newby and DeCesare 2020).
Our FPM results combined with the minimum body weight among females in this study
(95 kg) suggest that females weighing 72% of median mature female body mass (132.5 kg) can
ovulate and become pregnant. We cannot determine a lower body mass threshold for ovulation
because all females in this study were ovulatory. However, a body mass threshold between 65
and 84% of the mean population body mass has been reported as the threshold for ovulation in
other ungulate populations. Specifically, in woodland caribou in Labrador, Canada, Parker
(1981) reported a mean (± standard deviation) spring body mass of 78.6 ± 4.5 kg in pregnant
two-year olds (one-year old at breeding), 84% of the mean body mass of females older than two
years. In the same population, Crête et al. (1993) suggested a live autumn body mass threshold of
approximately 112 kg for parturient females, 75% of the mean body mass of all females
sampled. Even lower threshold masses have been reported in elk (65–70%; Hudson et al., 1991)
and domestic sheep (50–70%; Hafez 1952, Dýrmundsson, 1973). Overall, the body mass
measurements of females in this study—all of which were ovulatory— fell above 70% of the
body mass of the population median, which is consistent with reports of body mass thresholds
for ovulation in other ungulate populations. In Svalbard reindeer, Albon et al. (2017) found that
adult ovulation rate was positively related to variation in mean October adult body mass. These
findings provide further support that nutritional condition does not limit ovulation rates in this
population.

85

Nutritional conditions in early life, including maternal investment, can have a strong
effect on growth and age at primiparity in many species (Festa-Bianchet et al. 1994, Gaillard et
al. 2000). We found that 100% (n = 12) of yearling females at Kennedy Siding had FMP values
indicating ovulation (i.e., capacity to become pregnant) and all had body mass measurements
greater than 70% of median adult body mass, which is an approximate threshold for
reproduction. Moreover, based on observations at the feeders in 2021 and 2022, and radio collar
movements in 2023, at least four of seven yearlings re-identified the following year as two-year
olds had calved, for a minimum pregnancy rate among yearlings of 57%. Combined, these
findings suggest that the average age of primiparity is one year in the Kennedy Siding herd.
Caribou are known to breed as one-year olds; however, the proportion varies depending on the
population and environmental conditions with conception in yearling caribou occurring only
when nutrition is especially good (Parker 1981, Ouellet et al. 1997, Rettie and Messier 1998). In
Svalbard reindeer, the variation in yearling ovulation rates is positively correlated to their mean
body mass (Albon et al. 2017). In arctic caribou populations, females generally first give birth at
three years old, and the pregnancy rate of animals that were one-years old at breeding is often
less than 50% (Thomas 1982, Cuyler and Østergaard 2005), however it may be as high as 100%
in years with high summer and winter range quality . Rates of yearling pregnancy in wild
woodland caribou populations outside of BC have been estimated at 43% (n = 21; Parker 1981),
33% (n = 3; Crête et al., 1993), and 100% (n = 5; Rettie and Messier 1998). Notably, our finding
that 57% of yearling females bred is comparable to data from the Chisana herd in southern
Yukon and Alaska, where 57% (n = 67) of females produced their first calf at two-years old
(one-year old at conception) during 2003 to 2006 when maternal penning was applied as a
management strategy (Adams et al. 2019). Caribou in maternity pens have daily access to

86

pelleted feed and supplementary lichen that could improve nutrition and growth, resulting in
earlier age of puberty (Adams et al. 2019, Lamb et al. 2024a). Although comparable data would
be required from an unfed control population, our findings of reproductive activity in a high
proportion of yearlings among females at Kennedy Siding relative to caribou elsewhere could
conceivably be influenced by supplemental food. Such an effect could occur if females that use
feeders invest greater energy into offspring nutrition (i.e., silver spoon effects) leading to faster
growth and earlier primiparity, if feeding contributes to accelerated growth and puberty among
females that use feeders as calves, or if a pulse of nutrients provided to 1.5-year-olds promotes
ovulation cycles and subsequent pregnancies. We found that the median mass of adults without a
calf at heel was 9.5 kg heavier than those with a calf, further demonstrating the impact that calf
rearing has on body condition. The length of time that females spent nursing a calf influences
their body condition in autumn. Since all females were ovulatory, most of the adults that arrived
at Kennedy Siding without a calf likely had given birth but lost their calf some time before
autumn. Most calf mortality occurs within summer shortly after parturition; therefore, caribou
that lost a calf may have up to three extra months to regain body stores lost to lactation and
would likely enter winter in better condition. In caribou, the average difference in body fat
between females that raised a calf and those that didn’t is 3 percentage points (Crête and Huot
1993, Chan-McLeod et al. 1999, Cook et al. 2021). In elk, Cook et al. (2013) found that
individuals that did not give birth (i.e., not pregnant the previous winter) averaged 8 percentage
points more body fat than those that raised a calf. When summer conditions are inadequate
females with a calf accrue smaller fat reserves than non-reproductive animals, resulting in
decreased fecundity and survival (Crête and Huot 1993).

87

Supplemental feeding may also increase ovulation rates through the effects of nutritional
flushing. Flushing is the practice of providing improved nutrition at the beginning of the mating
season and is commonly used to increase ovulation rates in domestic ruminants like sheep
(Robinson et al., 2002, 2006). Adequate protein and energy intake prior to breeding is essential
for reproduction, as it regulates reproductive hormones and supports normal ovulation and
fertility in females (Robinson et al. 2006, Ashworth et al. 2009, Assan et al. 2025). Flushing
increases the release of gonadotropic hormones, including luteinizing hormone and follicle
stimulating hormone, which improves the ovulation rate (Ashworth et al. 2009, Assan et al.
2025). Flushing increases the number and size of preovulatory follicles in ewes (Robinson et al.
2006, Ashworth et al. 2009). Flushing in ewes in low body condition has a larger effect on
ovulation than in ewes in higher body condition (Robinson et al., 2002, 2006). Further research
on the effects of nutrition on ovulation in caribou is needed to assess whether autumn
supplemental feeding could have similar effects as flushing. One limitation of our study is the
lack of an unfed population to which our FPM and body mass results could be compared. The
Kennedy Siding herd is unique in that it consistently uses the same small, easily accessible area
every year, making repeated sampling easier than it would be for other herds. Very little fecal
hormone data exists for woodland caribou in British Columbia, and the existing data is either
from a single sample per individual and/or collected at a different time of year, making
comparisons to our data challenging (Flasko et al. 2017, Mueller et al. 2022). Body mass is also
difficult to measure noninvasively in other populations. Repeated sampling of an unfed
population during the autumn would help to clarify our FPM results and make a more explicit
link supplemental feeding and the survival of adult females, their ability to conceive and carry a
fetus to term over multiple years, and calf survival.

88

Demographic rates
Based on the radio collar movement data, we estimated a total parturition rate of 87.5 % for the Kennedy Siding herd.
This estimate is greater than total parturition rates reported in other herds in BC (82–59%; Table 9), and the standard error for
the estimates did not overlap with the standard error of our estimate. Like the other herds our estimate varied from year to year.
These results support our prediction that supplemental feeding would increase parturition rates relative to other nearby herds
that do not experience supplemental feeding. Parturition rates in caribou are closely tied to autumn body mass, and are sensitive
to even small shifts in mass distribution (Cameron and Hoef 1994). Barboza et al., (2024) found that captive caribou
fed a restricted diet during the summer then transitioned onto a high energy diet for the autumn breeding period gained
approximately 0.09 kg/day, whereas caribou at Kennedy Siding may gain approximately 0.103 kg/day over the feeding period
(Heard and Zimmerman 2021). Since parturition rates are positively correlated with body mass during the autumn, this rapid
mass gain could improve the probability of pregnancy and parturition, especially for smaller animals (Cameron et al.

1993, Cameron and Hoef 1994, Heard and Zimmerman 2021). The effects of nutritional flushing may also
influence parturition rates. Improved pre-mating body condition and diets in domestic ruminants can affect embryo survival by
altering the quality of the oocyte (Abecia et al. 1999, Lozano et al. 2003, Ashworth et al. 2009). In sheep,
ewes that were underfed had poorer quality oocytes, lower rates of cleavage and blastocyst formation, and lower pregnancy
rates compared to ewes that received adequate or increased nutrition (Abecia et al. 1999, Lozano et al. 2003,

Borowczyk et al. 2006). Table 9: Parturition rates of caribou (Rangifer tarandus caribou) in five populations of woodland
caribou in British Columbia and Alberta, Canada. Standard error (SE) was calculated as the standard error of the proportion.

Reference

Population / region

Years

n

Parturition (%)

SE

Gustine et al. 2006 a

Besa Prophet

2002
2003

9
13

55.6
76.9

17.6
12.2

Total

22

68.2

10.2

2004
2011
2012

10
24
15

90
79
80

Total

49

81.6

5

Narraway

1998 - 2014

48

80

6

Redrock-Prairie Creek

1998 - 2014

35

61

8

Klinse-Za (penned) c

2018
2019
2020

12
16
13

75
81
69

Total

41

76

7

2018
2019
2020
Total

4
10
13
27

75
50
62
59

9

DeMars et al. 2013 b

Nobert et al. 2016 b

Mueller et al. 2022 b

Fort Nelson

Klinse-Za (free ranging)

a

Parturition rate determined using observations from fixed-wing aircraft twice daily
Parturition rate determined using GPS movement model developed by DeMars et al. 2013
c
Received supplemental feeding
b

89

The effect of autumn supplemental feeding did not appear to influence parturition dates
for the Kennedy Siding herd as dates were similar to those reported in other herds in BC. Calving
occurred between May 11th and June 7th in the Kennedy Siding herd, April 29th and July 4th in the
Klinse-za herd (Mueller et al. 2022), May 18th and June 19th in the Narraway and Redrock Prairie
Creek herds (Nobert et al. 2016), and May 21st and June 10th in herds in the Besa Prophet region
(Gustine et al. 2006). Parturition date has been related to maternal condition in late gestation and
at time of conception, with animals in better condition giving birth earlier (Cameron et al. 1993,
Crête et al. 1993). An earlier calving date may enhance fat accretion and increase the chance that
a female will successfully reproduce because she is able to escape the high nutritional demands
of lactation earlier (Cook et al. 2004). Parturition dates may also be influenced by variables not
considered in this study, such as latitude, spring and summer climatic conditions and vegetation
characteristics. For example, Post et al. (2003) found that caribou calving seasons are highly
synchronized to the progression of forage plant phenology.
Total calf survival, based on females with a recorded calving event returning to Kennedy
Siding with a calf at heel, was 74.3%. This rate is slightly higher than previously reported total
calf survival estimates for caribou that did not receive supplemental feeding, which range from
58 – 63% (Table 10). The standard error of our estimate was smaller and did not overlap with the
standard errors of the other herd estimates. The higher calving rate lends support to our
hypothesis that improved maternal nutrition from supplemental feeding might lead to
improvements in calf survival. Calf survival can, however, be highly variable between years
since it can be caused by a wide range of factors including predation, accidents, emaciation,
congenital defects, disease, and stillbirths, with predation being the main cause of mortality in
BC (Seip and Cichowski 1996, Gustine et al. 2006). Notably, calf survival observed in 2003 in

90

the Besa Prophet was 79% even though the total calf survival over two years was 66% (Gustine
et al., 2006). Estimates of calf mortality may also be limited by sample size. In 2024 collar
locations were only available for four caribou in the Kennedy Siding herd, which may have
inflated the estimates of mortality that year. All of mortalities recorded in the Kennedy Siding
herd occurred within 30 days of parturition, consistent with calf mortality patterns in other herds
in BC (Gustine et al. 2006, DeMars et al. 2013, Nobert et al. 2016).
Table 10: Caribou (Rangifer tarandus caribou) calf survival in four populations of woodland caribou in British Columbia and
Alberta, Canada. Standard error (SE) was calculated as the standard error of the proportion.

Reference

Population / region

Years

n

Calf survival (%)

SE

Gustine et al. 2006 a

Besa Prophet

2002
2003

25
25

54
79

11
8

Total

50

66

7

2004
2011
2012

9
12
6

55.6
66.7
66.7

Total

27

63

9

Narraway

1998 – 2014

28

60

9

Redrock-Prairie Creek

1998 – 2014

21

58

11

Klinse-Za (penned)d

2014–2020

65

95.4

DeMars et al. 2013
b

Nobert et al. 2016 b

McNay et al. 2022c

Fort Nelson

a

Calf survival determined using observations from aerial surveys
Calf survival determined using GPS movement model developed by DeMars et al. 2013
c
Calf survival determined from direct observations in maternal pen
d
Received supplemental feeding
b

The data from the movement models suggested that approximately 10% of caribou did
not calve. Since all the females we sampled had FPM values indicating they were able to become
pregnant, this implies that some caribou either failed to become pregnant, despite ovulating, or
lost their pregnancy before spring. Early embryonic mortality in caribou occurs close to the
breeding season and is inversely correlated to maternal body fat and mass (Russell et al. 1998).
Previous lactation status during the summer may impact the probability of early embryonic

91

mortality because caribou in poor condition in autumn may sacrifice pregnancy to allocate
nutrients to the growth and survival of their current calf or their own survival. All of the
incidences of early embryonic mortality recorded by Russell et al. (1998) occurred in caribou
that extended lactation into the fall. Late gestation in-utero reabsorption of a pregnancy is rare in
caribou and reindeer but can occur when a female is in very poor condition (Ringberg and
Aakvaag 1982). The prevalence of early and late term pregnancy loss in woodland caribou is not
well documented, though Mueller et al. (2022) reported that out of 34 pregnant caribou held in
the Klinse-Za maternity pen, two aborted births and one still birth occurred (9% of pregnancies).
In Svalbard reindeer, the loss of reproductive potential between ovulation and parturition was
associated with the severity of the winter and there were significantly greater losses in winters
with low April body mass (Albon et al. 2017). The increase in nutritional resources provided by
supplemental feeding could reduce the number of pregnancies lost due to inadequate nutritional
condition. Notably, two of the females that did not give birth in 2023 were observed to be in
good body condition and did not have a calf at heel at the feeders in 2022. Non nutritional
causes, such as disease or congenital issues may be more likely to cause pregnancy loss. The
movement model used may not detect parturition if the calf is stillborn, non-viable, abandoned,
or killed immediately and the female immediately resumes a normal daily step length (DeMars et
al. 2013, Mueller et al. 2022). The difference between ovulation and parturition in our study
could be due to undetected parturition events that were masked by early calf mortality.
Comparison of population growth between supplemental feeding and maternal penning
Applying more than one management action simultaneously can improve population
growth in caribou (Serrouya et al. 2019). Assessing the response of caribou populations to
different recovery actions independently can help to identify the mechanisms by which an action

92

influences population growth and helps inform managers on the number and intensity of
recovery actions needed for population growth. We compared the demographic and population
growth rates of the Kennedy Siding herd to the nearby Klinse-Za herd. The ranges of the two
herds are adjacent and have similar habitat characteristics (Heard and Zimmerman 2021, McNay
et al. 2022). Both herds experienced significant declines around the same time; the Klinse-Za
herd declined from ~200 animals in the 1990s to 36 in 2013 and the Kennedy Siding herd
declined from ~120 in 2007 to 49 in 2014 (Heard and Zimmerman 2021, McNay et al. 2022).
Maternal penning began for the Klinse-Za herd in 2014, the same year supplemental feeding
began for Kennedy Siding and an aerial-based program wolf removal program was initiated by
the BC Government in 2015 in both ranges (Heard and Zimmerman 2021, McNay et al. 2022).
Since then, both populations have increased at similar rates. The average population growth rate
from 2014 to 2018 was 1.14 in the Klinse-Za herd and 1.16 in the Kennedy Siding herd.
Although feeding and maternal penning appear to have had similar effects on recruitment, the
mechanism underlying population growth may differ. Specifically, parturition rates were greater
in the Kennedy Siding herd by 11.5–28.5 % (Table 9), whereas rates of calf survival in the 30
days following birth were greater in the Klinse-Za herd by 21 % (Table 10; McNay et al., 2022).
This difference suggests that the contribution to recruitment from supplemental feeding comes
mostly from an increase in the proportion of females that breed and parturition rates (Table 9),
maternal penning primarily affects calf survival (Table 10).
In addition to improving calf survival, maternal penning may also benefit females
through improved nutrition as Mueller et al. (2022) found that penned caribou had a higher total
parturition rate (76 %) than free ranging animals 59 %; table 9). One explanation for this effect is
that improved nutrition during late gestation influences parturition rates. While in the pen,

93

caribou are supplemented with lichen and the same pelleted food used at Kennedy Siding
(McNay et al. 2022). The pelleted food contains essential micronutrients such as iodine and
selenium, which are critical to fetal development and survival at birth. The production of the
hormones T3 and Thyroxine (T4) by pregnant animals are dependent on adequate selenium and
iodine intake (Robinson et al. 2001). The thyroid hormones T3 and T4 play important roles in
fetal cell and tissue differentiation, lung maturation, and development of brown adipose tissue,
which are important to neonatal thermogenesis and survival (Robinson et al. 2001). Selenium
and iodine deficiencies are often a cause of abortions in domestic sheep and goats. Deficencies in
copper and zinc have also been implicated in abortions in domestic livestock. Caribou in the
Klinse-Za herd have lower selenium, copper, and zinc compared to nearby Northern Mountain
caribou and boreal caribou subpopulations (Lamb et al. 2024a). Selenium levels were greater in
penned animals compared to free ranging animals (Lamb et al. 2024a). The increase in
micronutrients provided by feed during penning may help caribou overcome any deficiencies
later in gestation, thereby reducing pregnancy loss. Further comparison of trace nutrients in
Kennedy Siding and Klinse-Za caribou could help to delineate the effect of nutrient
supplementation in early or late gestation on parturition and calf survival.
.
Management implications
The successful management of wildlife requires an understanding of species'
reproductive biology. Information about reproductive physiology (ovulation, estrous-cycle
length, pregnancy, lactational anovulation, age at the onset of puberty, birth rates etc.) can
provide useful insights into the influence of factors such as population density, nutrition,
genetics, climate, and anthropogenic disturbances on animal health and reproduction

94

(Schwarzenberger 2007, Ghosal et al. 2023). Fecal hormone sampling allows for frequent
sampling needed for longitudinal studies of wildlife without subjecting animals to undue stress
(Holt 2003). That all animals, regardless of age, body condition, or reproductive status, had FPM
values indicating reproductive activity suggests that caribou may use the nutritional increase
provided by feeding to compensate for lactation costs and insufficient summer forage that would
otherwise limit reproduction. This result is similar to the effects of flushing in domestic sheep,
where improved nutrition increases ovulation rates during the breeding season, especially for
ewes with low body condition scores (Robinson et al. 2002, 2006). Improved body condition in
the autumn could increase the probability of caribou carrying a fetus to term and decrease age at
first reproduction. Although our results are consistent with an increase in reproductive success
via improved nutritional condition, a control population for comparison is needed to definitively
link supplemental feeding to improved pregnancy, parturition, and calf survival.
Supplemental feeding could be applied in other small, declining herds or herds with low
pregnancy rates to improve female body condition and increase calf production. At the landscape
scale, summer and autumn nutrition should be considered when developing management
strategies for caribou. Protection and restoration of quality summer habitats would promote
population growth and support caribou populations that are self-sustaining via effects on body
condition, reproduction, and survival (Cook et al., 2021, Denryter et al., 2022).

95

Chapter 4: Conclusion

Research summary
Averting the extirpation and increasing the abundance of woodland caribou in BC
represents a significant conservation challenge that intersects with multiple ecological, cultural,
and economic values (Bixler 2013, Serrouya et al. 2019, Lamb et al. 2022). Caribou are a key
umbrella species for boreal biodiversity due to their selection for intact, mature forest habitats
forests (Serrouya et al. 2019, Campbell et al. 2020, Falconer and Ford 2020). Indigenous Peoples
throughout the northern hemisphere have long relied on caribou for subsistence, ceremonial, and
community purposes (Muir and Booth 2012, Lamb et al. 2022). In BC, declining caribou
populations threaten fundamental cultural connections to the land (Muir and Booth 2012, Lamb
et al. 2022). Activities related to the exploration and extraction of natural resources have
degraded the quantity and quality of habitat for a large proportion of woodland caribou (Vors et
al. 2007, Wasser et al. 2011, Johnson et al. 2015). The cumulative effects of multiple disturbance
types have resulted in increased predation, altered movement patterns, which has affected the
availability and quality of forage resources for caribou (Wasser et al. 2011, Johnson et al. 2015,
McGreer et al. 2015, Dickie et al. 2020).
The conservation of caribou requires a combination of both short and long-term measures
and has been most successful where more than one recovery action has been implemented
(Serrouya et al. 2019, Heard and Zimmerman 2021, McNay et al. 2022, Lamb et al. 2024b).
Understanding the role that nutrition plays in individual survival and reproduction and overall
population dynamics can be used to enhance conservation efforts for caribou. Improved
nutrition, through short term strategies such as supplemental feeding and long term habitat

96

management, can result in increased juvenile growth rates, adult mass gain, body condition,
probability of pregnancy, over-winter survival, and calf survival (Gerhart et al. 1997b, Parker et
al. 2009, Cook et al. 2021, Heard and Zimmerman 2021, Denryter et al. 2022).
I used multiple indicators of physiological, nutritional, and reproductive condition to
examine the effect that nutrition and supplemental feeding could have on female caribou in the
Kennedy Siding herd. Specifically, I used fecal hormone metabolites to evaluate changes in
nutritional condition over the course of the feeding period. I used a combination of fecal
hormone metabolites and GPS collar data to assess rates of pregnancy, parturition, and calf
survival in the Kennedy Siding herd and compared them with measures of demographic rates
reported in the literature for other herds.
Age and reproductive status had the largest effect on measurements of body condition. I
found that adult females that arrived at Kennedy Siding with a calf at heel were more likely to
been in poor body condition based on visual scoring and had lower body masses than adults
without a calf at heel. Adults with a calf had significantly lower δ15N values in guard hair than
both adults without a calf and yearlings. These results indicate the profound effect of lactation
and calf rearing on body condition and protein / fat reserves (Chan-McLeod et al. 1994, Gerhart
et al. 1997b, a, Taillon et al. 2013). My results corroborate past findings that while the nutritional
landscape available to caribou in BC may be adequate to support pregnancy, it appears to be
inadequate to support both lactation and the recovery and maintenance of body reserves (Dale et
al. 2008, Cook et al. 2021, Denryter et al. 2022).
I found that fecal glucocorticoid metabolites (FGM) differed between reproductive age
classes. Yearlings had the greatest FGM concentrations, perhaps reflecting the effects of social
dynamics in a group with mixed sexes and ages (Barrette and Vandal 1986, Creel 2001,

97

Weckerly and Ricca 2014). Adults with a calf at heel had slightly greater FGM measurements
than adults without a calf, possibly representing the increased energetic demands of calf rearing
(Cook et al. 2021, Newediuk et al. 2024). Fecal T3 metabolites (FTM) also differed between age
classes, with yearlings having much higher concentrations than the adults in general. This
difference likely reflects the role that T3 plays in growth and development, as caribou are still
growing and have not yet reached adult size at one-years old (Todini 2007, Behringer et al.
2018).
I found differences in δ15N, FGMs and FTMs between the two study years. δ15N was
lower whereas FGMs and FTMs were higher in samples collected in 2022 compared to samples
from 2021. These differences could be attributed to the difference in environmental conditions
between the two years. Variables such as the timing of vegetation emergence, temperature,
precipitation, wildfires, and snowpack can all influence the availability and quality of the forage
available for caribou (Zamin et al. 2017, Mallory and Boyce 2018, DeMars et al. 2021, Twining
et al. 2022, Russell et al. 2025). The concentration of FGMs declined over the course of the
feeding seasons overall, with a more pronounced decline in 2022 than 2021. This result suggests
that supplemental feeding may have the most beneficial effect on physiological condition in
years where environmental conditions limit the quality and quantity of forage resources. Further
investigation into inter annual differences in condition indicators is needed to understand the role
that environmental conditions play in the nutritional condition of caribou in the Kennedy Siding
herd, especially as climate change is expected to alter boreal and mountain ecosystems. The
concentration of FTMs also declined over the course of the feeding season. This was contrary to
my initial prediction that FTMs would increase in response to increased food availability. The
decrease in FTMs instead likely represents the metabolic changes that occur as caribou transition

98

from summer to winter conditions (Ryg 1984, Nilssen et al. 1985). Lower T3 could facilitate
body fat gains before winter and conserve energy during the winter (Ryg 1984).
Every caribou that I sampled had fecal progesterone metabolite (FPM) values indicative
of ovarian activity, regardless of age, body condition, or whether they had a calf at heel or not.
This indicated that caribou in the Kennedy Siding herd were not forgoing reproduction in
response to poor condition and all females had the capacity to become pregnant during the fall
breeding season (Cameron and Hoef 1994, Gerhart et al. 1997b, a). Based on the FPM values,
the age of puberty in Kennedy Siding caribou is one-years old, which further supports
supplemental feeding having a beneficial impact on body condition, as yearlings have to accrue
enough body fat and achieve approximately 70% of their mature mass in order to undergo
puberty and conceive (Parker 1981, Rettie and Messier 1998, Cook et al. 2004).
Using a combination of FPMs, GPS collar data, and fall counts conducted at the feeding
site, I estimated that in the Kennedy Siding herd, the rate of ovulation and potential pregnancy
was 100%, the parturition rate was 87%, and the calf survival rate was 74%. Parturition rate was
greater than previously reported rates in other caribou herds in BC (Gustine et al. 2006, DeMars
et al. 2013, Nobert et al. 2016, Mueller et al. 2022). Calf survival was slightly higher than in
other herds, but was variable between years (Gustine et al., 2006, DeMars et al., 2013, Nobert et
al., 2016). These results suggest that fall supplemental feeding could contribute to population
growth by both increasing the proportion of females with successful pregnancies and increasing
the proportion of calves that survive until the fall.
My interpretation is limited by the lack of data from an unfed control population to which
I could compare my results to. Repeated sampling of an unfed population during the autumn
would help to clarify the hormonal results and make a more explicit link between these

99

indicators and supplemental feeding. Although I was unable to make definitive links to
supplemental feeding, the decline in FGMs we observed suggests that feeding is unlikely to be
negatively impacting caribou physiology and the FPM and movement model results are
consistent with an increase in reproductive rates expected with improved nutrition. The increases
in body mass observed by Heard and Zimmerman (2021), along with the high rates of ovulation,
parturition, calf survival, and yearling reproduction suggests that supplemental feeding could
provide a meaningful improvement to female body condition, resulting in cascading beneficial
demographic impacts such as increased parturition and calf survival rates.
Management implications
Autumn supplemental feeding appears to be improving population growth in the
Kennedy Siding herd by increasing parturition and calf survival rates via improved female body
condition. My research emphasizes the importance of summer and autumn nutrition on survival
and reproduction of caribou in BC (Dale et al., 2008, Couturier et al., 2009, Kelly 2020, Cook et
al., 2021, Heard and Zimmerman 2021, Denryter et al., 2022). In populations where summer
habitat is insufficient to meet the needs of both lactation and recovery of body reserves, fall
supplemental feeding could be used to facilitate the rapid recovery of body fat before winter.
Improved body condition in the autumn increases the probability of caribou conceiving and
carrying a fetus to term. Supplemental feeding could be applied in conjunction with other
conservation measures, such as habitat restoration or predator reduction, to more rapidly enhance
small populations (Lamb et al. 2024b).
Widespread habitat conservation and restoration is needed if caribou are to remain on the
landscape in the long term, and summer and autumn nutrition should be considered when
developing habitat management strategies for caribou. Protection and restoration of quality

100

summer habitats would promote population growth and support caribou populations that are selfsustaining via effects on body condition, reproduction, and survival (Cook et al., 2021, Denryter
et al., 2022).
Finally, my research represents some of the first longitudinal monitoring of endocrine
activity in woodland caribou. Analysis of single fecal samples have been used to examine
reproduction and physiological/nutritional stress in caribou, however repeated sampling has
rarely occurred due to the difficulties obtaining repeated samples from wild populations (Messier
et al. 1990a, Morden et al. 2011, Joly et al. 2015, Lamb et al. 2024a). The Kennedy Siding herd
consistently uses the same small, easily accessible area every year, providing a unique
opportunity to collect repeated samples from known individuals to address questions surrounding
the links between habitat conditions, physiology, and reproduction. My results demonstrate the
potential of fecal hormone measurement for health monitoring and provide a baseline upon
which future studies can expand.

101

References
Abecia, J. A., F. Forcada, and J. M. Lozano. 1999. A preliminary report on the effect of dietary
energy on prostaglandin f2α production in vitro, interferon-tau synthesis by the
conceptus, endometrial progesterone concentration on days 9 and 15 of pregnancy and
associated rates of embryo wastage in ewes. Theriogenology 52:1203–1213.
Adamczewski, J. Z., C. C. Gates, R. J. Hudson, and M. A. Price. 1987. Seasonal changes in body
composition of mature female caribou and calves (Rangifer tarandus groenlandicus) on
an arctic island with limited winter resources. Canadian Journal of Zoology 65:1149–
1157.
Adams, L. G. 2005. Effects of maternal characteristics and climatic variation on birth masses of
Alaskan caribou. Journal of Mammalogy 86:506–513.
Adams, L. G., and B. W. Dale. 1998. Reproductive performance of female Alaskan caribou. The
Journal of Wildlife Management 62:1184.
Adams, L. G., R. Farnell, M. P. Oakley, T. S. Jung, L. L. Larocque, G. M. Lortie, J. Mclelland,
M. E. Reid, G. H. Roffler, and D. E. Russell. 2019. Evaluation of maternal penning to
improve calf survival in the Chisana caribou herd. Wildlife Monographs 204:5–46.
Albon, S. D., R. Justin. Irvine, O. Halvorsen, R. Langvatn, L. E. Loe, E. Ropstad, V. Veiberg, R.
Van Der Wal, E. M. Bjørkvoll, E. I. Duff, B. B. Hansen, A. M. Lee, T. Tveraa, and A.
Stien. 2017. Contrasting effects of summer and winter warming on body mass explain
population dynamics in a food‐limited Arctic herbivore. Global Change Biology
23:1374–1389.
Arnold, T. W. 2010. Uninformative parameters and model selection using Akaike’s information
criterion. Journal of Wildlife Management 74:1175–1178.
Arthur, B. 2002. Kennedy Siding ungulate winter range (UWR) (U-7-001) report. Ministry of
Water, Land and Air Protection, British Columbia.
Asher, G. W. 2011. Reproductive cycles of deer. Animal Reproduction Science 124:170–175.
Ashworth, C. J., L. M. Toma, and M. G. Hunter. 2009. Nutritional effects on oocyte and embryo
development in mammals: implications for reproductive efficiency and environmental
sustainability. Philosophical Transactions of the Royal Society B: Biological Sciences
364:3351–3361.
Assan, N., C. Bhakat, P. Chisoro, and E. Muteyo. 2025. The role of feed resources in optimizing
reproductive efficiency in goats and sheep. International Journal of Multidisciplinary
Research and Growth Evaluation. 6:213–233.
Ayres, K. L., R. K. Booth, J. A. Hempelmann, K. L. Koski, C. K. Emmons, R. W. Baird, K.
Balcomb-Bartok, M. B. Hanson, M. J. Ford, and S. K. Wasser. 2012. Distinguishing the
impacts of inadequate prey and vessel traffic on an endangered killer whale (Orcinus
orca) Population. B. Fenton, editor. PLoS ONE 7:e36842.
Bahnak, B. R., J. C. Holland, L. J. Verme, and J. J. Ozoga. 1981. Seasonal and nutritional
influences on growth hormone and thyroid activity in white-tailed deer. The Journal of
Wildlife Management 45:140.
Baker, M. R., K. S. Gobush, and C. H. Vynne. 2013. Review of factors influencing stress
hormones in fish and wildlife. Journal for Nature Conservation 21:309–318.

102

Barboza, P. S., D. W. Hartbauer, W. E. Hauer, and J. E. Blake. 2004. Polygynous mating impairs
body condition and homeostasis in male reindeer (Rangifer tarandus tarandus). Journal
of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology
174:309–317.
Barboza, P. S., and K. L. Parker. 2006. Body protein stores and isotopic indicators of N balance
in female reindeer (Rangifer tarandus) during winter. Physiological and Biochemical
Zoology 79:628–644.
Barboza, P. S., and K. L. Parker. 2008. Allocating protein to reproduction in arctic reindeer and
caribou. Physiological and Biochemical Zoology 81:835–855.
Barboza, P. S., R. D. Shively, D. D. Gustine, and J. A. Addison. 2020. Winter is coming:
conserving body protein in female reindeer, caribou, and muskoxen. Frontiers in Ecology
and Evolution 8:150.
Barboza, P. S., R. D. Shively, and D. P. Thompson. 2024. Robust responses of female caribou to
changes in food supply. Ecological and Evolutionary Physiology 97:29–52.
Barrette, C., and D. Vandal. 1986. Social rank, dominance, antler size, and access to food in
snow-bound wild woodland caribou. Behaviour 97:118–146.
Bartoń, K. 2010. MuMIn: Multi-Model Inference.
Bates, D., M. Mächler, B. Bolker, and S. Walker. 2015. Fitting linear mixed-effects models using
lme4. Journal of Statistical Software 67.
Behringer, V., C. Deimel, G. Hohmann, J. Negrey, F. S. Schaebs, and T. Deschner. 2018.
Applications for non-invasive thyroid hormone measurements in mammalian ecology,
growth, and maintenance. Hormones and Behavior 105:66–85.
Bender, L. C., J. G. Cook, R. C. Cook, and P. B. Hall. 2008. Relations between nutritional
condition and survival of North American elk Cervus elaphus. Wildlife Biology 14:70–
80.
Bergerud, A. T. 1975. The reproductive season of Newfoundland caribou. Canadian Journal of
Zoology 53:1213–1221.
Bergerud, A. T. 2014. Return of caribou to Ungava. McGill-Queen’s Native and Northern Series
v. 50, McGill-Queen’s University Press, Montreal.
Bergerud, A. T., and J. P. Elliott. 1998. Wolf predation in a multiple-ungulate system in northern
British Columbia. Canadian Journal of Zoology 76:1551–1569.
Bini, J., L. Parikh, C. Lacadie, J. J. Hwang, S. Shah, S. B. Rosenberg, D. Seo, K. Lam, M.
Hamza, R. B. De Aguiar, T. Constable, R. S. Sherwin, R. Sinha, and A. M. Jastreboff.
2022. Stress-level glucocorticoids increase fasting hunger and decrease cerebral blood
flow in regions regulating eating. NeuroImage: Clinical 36:103202.
Bishop, C. J., G. C. White, D. J. Freddy, B. E. Watkins, and T. R. Stephenson. 2009. Effect of
enhanced nutrition on mule deer population rate of change. Wildlife Monographs 172:1–
28.
Bixler, R. 2013. The political ecology of local environmental narratives: power, knowledge, and
mountain caribou conservation. Journal of Political Ecology 20.
Blake, N. G., D. J. A. Eckland, O. J. F. Foster, and S. L. Lightman. 1991. Inhibition of
hypothalamic thyrotropin-releasing hormone messenger ribonucleic acid during food
deprivation. Endocrinology 129:2714–2718.

103

Bleke, C. A., E. M. Gese, and S. S. French. 2021. Variations, validations, degradations, and
noninvasive determination of pregnancy using fecal steroid metabolites in free-ranging
pronghorn. General and Comparative Endocrinology 312:113841.
Bookhout, T. A. and Wildlife Society, editors. 1994. Research and management techniques for
wildlife and habitats. 5th ed. Wildlife Society, Bethesda, Md.
Borowczyk, E., J. S. Caton, D. A. Redmer, J. J. Bilski, R. M. Weigl, K. A. Vonnahme, P. P.
Borowicz, J. D. Kirsch, K. C. Kraft, L. P. Reynolds, and A. T. Grazul-Bilska. 2006.
Effects of plane of nutrition on in vitro fertilization and early embryonic development in
sheep1. Journal of Animal Science 84:1593–1599.
Bridger, M. 2019. South Peace caribou recovery following five years of experimental wolf
reduction. BC Ministry of Forests, Lands, Natural Resource Operations and Rural
Development.
Brown, J. L. 2018. Comparative ovarian function and reproductive monitoring of endangered
mammals. Theriogenology 109:2–13.
Bryan, H. M., A. G. Adams, R. M. Invik, K. E. Wynne-Edwards, and J. E. G. Smits. 2013a. Hair
as a meaningful measure of baseline cortisol levels over time in dogs. Journal of the
American Association for Laboratory Animal Science: JAALAS 52:189–196.
Bryan, H. M., C. T. Darimont, P. C. Paquet, K. E. Wynne-Edwards, and J. E. G. Smits. 2013b.
Stress and reproductive hormones in grizzly bears reflect nutritional benefits and social
consequences of a salmon foraging niche. N. Moreira, editor. PLoS ONE 8:e80537.
Butler, W. R. 2003. Energy balance relationships with follicular development, ovulation and
fertility in postpartum dairy cows. Livestock Production Science 83:211–218.
Cameron, R. D. 1994. Reproductive Pauses by Female Caribou. Journal of Mammalogy 75:10–
13.
Cameron, R. D., and J. M. V. Hoef. 1994. Predicting parturition rate of caribou from autumn
body mass. The Journal of Wildlife Management 58:674.
Cameron, R. D., W. T. Smith, S. G. Fancy, K. L. Gerhart, and R. G. White. 1993. Calving
success of female caribou in relation to body weight. Canadian Journal of Zoology
71:480–486.
Campbell, M. A., B. Kopach, P. E. Komers, and A. T. Ford. 2020. Quantifying the impacts of oil
sands development on wildlife: perspectives from impact assessments. Environmental
Reviews 28:129–137.
Cao, L., F. Liu, L. Guo, Q. Sun, S. Ge, B. Li, Y. Zhou, X. Fang, Y. Geng, W. Song, C. C.
Grueter, Y. Li, L. Cui, and W. Xiao. 2025. Continuous provisioning increases breeding
success of black‐and‐white snub‐nosed monkeys (Rhinopithecus bieti) in the Baima
Snow Mountain Nature Reserve, Yunnan, China. American Journal of Primatology
87:e70006.
Cebrian, M. R., K. Kielland, and G. Finstad. 2008. Forage quality and reindeer productivity:
multiplier effects amplified by climate change. Arctic, Antarctic, and Alpine Research
40:48–54.
Chan-McLeod, A. C. A., R. G. White, and D. F. Holleman. 1994. Effects of protein and energy
intake, body condition, and season on nutrient partitioning and milk production in caribou
and reindeer. Canadian Journal of Zoology 72:938–947.

104

Chan-McLeod, A. C. A., R. G. White, and D. E. Russell. 1999. Comparative body composition
strategies of breeding and nonbreeding female caribou. Canadian Journal of Zoology
77:1901–1907.
Chaves, A. S., F. Silva, R. Valentim, and H. Quintas. 2024. Body Condition in small
ruminants—effects of nutrition on the hypothalamic–pituitary–gonad axis and ovarian
activity that controls reproduction. Physiologia 4:213–225.
Chen, W., J. Z. Adamczewski, L. White, B. Croft, A. Gunn, A. Football, S. G. Leblanc, D. E.
Russell, and B. Tracz. 2018. Impacts of climate-driven habitat change on the peak
calving date of the Bathurst caribou in Arctic Canada. Polar Biology 41:953–967.
Chen, W., D. E. Russell, A. Gunn, B. Croft, W. R. Chen, R. Fernandes, H. Zhao, J. Li, Y. Zhang,
K. Koehler, I. Olthof, R. H. Fraser, S. G. Leblanc, G. R. Henry, R. G. White, and G. L.
Finstad. 2013. Monitoring habitat condition changes during winter and pre-calving
migration for Bathurst Caribou in northern Canada. Biodiversity 14:36–44.
Chinn, S. M., P. E. Schlichting, T. J. Smyser, C. F. Bowden, and J. C. Beasley. 2022. Factors
influencing pregnancy, litter size, and reproductive parameters of invasive wild pigs. The
Journal of Wildlife Management 86:e22304.
Christianson, D., T. H. Coleman, Q. Doan, and M. A. Haroldson. 2021. Physiological
consequences of consuming low-energy foods: herbivory coincides with a stress response
in Yellowstone bears. S. Cooke, editor. Conservation Physiology 9:coab029.
Cichowski, D., G. D. Sutherland, R. S. McNay, and R. Sulyma. 2022. Direct and indirect effects
of habitat disturbances on caribou terrestrial forage lichens in montane forests of British
Columbia. Forests 13:251.
Cook, J. G., B. K. Johnson, R. C. Cook, R. A. Riggs, T. Delcurto, L. D. Bryant, and L. L. Irwin.
2004. Effects of summer‐autumn nutrition and parturition date on reproduction and
survival of elk. Wildlife Monographs 155:1–61.
Cook, J. G., A. P. Kelly, R. C. Cook, B. Culling, D. Culling, A. McLaren, N. C. Larter, and M.
Watters. 2021. Seasonal patterns in nutritional condition of caribou (Rangifer tarandus)
in the southern Northwest Territories and northeastern British Columbia, Canada.
Canadian Journal of Zoology 99:845–858.
Cook, R. C., J. G. Cook, D. J. Vales, B. K. Johnson, S. M. Mccorquodale, L. A. Shipley, R. A.
Riggs, L. L. Irwin, S. L. Murphie, B. L. Murphie, K. A. Schoenecker, F. Geyer, P. B.
Hall, R. D. Spencer, D. A. Immell, D. H. Jackson, B. L. Tiller, P. J. Miller, and L.
Schmitz. 2013. Regional and seasonal patterns of nutritional condition and reproduction
in elk. Wildlife Monographs 184:1–45.
Cortés-Avizanda, A., G. Blanco, T. L. DeVault, A. Markandya, M. Z. Virani, J. Brandt, and J. A.
Donázar. 2016. Supplementary feeding and endangered avian scavengers: benefits,
caveats, and controversies. Frontiers in Ecology and the Environment 14:191–199.
COSEWIC. 2014. COSEWIC assessment and status report on the Caribou Rangifer tarandus,
Northern Mountain population, Central Mountain population and Southern Mountain
population in Canada. Committee on the Status of Endangered Wildlife in Canada,
Ottawa.
Couturier, S., S. D. Côté, J. Huot, and R. D. Otto. 2009. Body-condition dynamics in a northern
ungulate gaining fat in winter. Canadian Journal of Zoology 87:367–378.
Creel, S. 2001. Social dominance and stress hormones. Trends in Ecology & Evolution 16:491–
497.

105

Crête, M., and J. Huot. 1993. Regulation of a large herd of migratory caribou: summer nutrition
affects calf growth and body reserves of dams. Canadian Journal of Zoology 71:2291–
2296.
Crête, M., J. Huot, R. Nault, and R. Patenaude. 1993. Reproduction, growth and body
composition of Rivière George caribou in captivity. ARCTIC 46:189–196.
Crowe, M. A. 2016. Reproduction, events and management: estrous cycles: characteristics. Page
B9780081005965010398 in. Reference Module in Food Science. Elsevier.
Cuyler, C., and J. B. Østergaard. 2005. Fertility in two West Greenland caribou Rangifer
tarandus groenlandicus populations during 1996/97: potential for rapid growth. Wildlife
Biology 11:221–227.
Dale, B. W., L. G. Adams, W. B. Collins, K. Joly, P. Valkenburg, and R. Tobey. 2008.
Stochastic and compensatory effects limit persistence of variation in body mass of young
caribou. Journal of Mammalogy 89:1130–1135.
De Feo, P. 1996. Hormonal regulation of human protein metabolism. European Journal of
Endocrinology 135:7–18.
Delgiudice, G. D., L. D. Mech, and U. S. Seal. 1990. Effects of winter undernutrition on body
composition and physiological profiles of white-tailed deer. The Journal of Wildlife
Management 54:539.
DeMars, C. A., M. Auger‐Méthé, U. E. Schlägel, and S. Boutin. 2013. Inferring parturition and
neonate survival from movement patterns of female ungulates: a case study using
woodland caribou. Ecology and Evolution 3:4149–4160.
DeMars, C. A., and S. Boutin. 2018. Nowhere to hide: Effects of linear features on predator–prey
dynamics in a large mammal system. L. Börger, editor. Journal of Animal Ecology
87:274–284.
DeMars, C. A., S. Gilbert, R. Serrouya, A. P. Kelly, N. C. Larter, D. Hervieux, and S. Boutin.
2021. Demographic responses of a threatened, low-density ungulate to annual variation in
meteorological and phenological conditions. M. Apollonio, editor. PLOS ONE
16:e0258136.
Denryter, K., R. C. Cook, J. G. Cook, and K. L. Parker. 2022. Animal‐defined resources reveal
nutritional inadequacies for woodland caribou during summer–autumn. The Journal of
Wildlife Management 86.
Depperschmidt, J. D., S. C. Torbit, A. W. Alldredge, and R. D. Deblinger. 1987. Body condition
indices for starved pronghorns. The Journal of Wildlife Management 51:675.
Deschner, T., B. T. Fuller, V. M. Oelze, C. Boesch, J.-J. Hublin, R. Mundry, M. P. Richards, S.
Ortmann, and G. Hohmann. 2012. Identification of energy consumption and nutritional
stress by isotopic and elemental analysis of urine in bonobos (Pan paniscus): Isotopic
analysis of nutritional stress in bonobo urine. Rapid Communications in Mass
Spectrometry 26:69–77.
Dickie, M., S. R. McNay, G. D. Sutherland, M. Cody, and T. Avgar. 2020. Corridors or risk?
Movement along, and use of, linear features varies predictably among large mammal
predator and prey species. A. Loison, editor. Journal of Animal Ecology 89:623–634.
Drucker, D. G., K. A. Hobson, J.-P. Ouellet, and R. Courtois. 2010. Influence of forage
preferences and habitat use on 13 C and 15 N abundance in wild caribou (Rangifer
tarandus caribou) and moose (Alces alces) from Canada. Isotopes in Environmental and
Health Studies 46:107–121.

106

Dubois, S., and D. Fraser. 2013. A framework to evaluate wildlife feeding in research, wildlife
management, tourism and recreation. Animals 3:978–994.
Dunkley, L., and M. R. L. Cattet. 2003. A comprehensive review of the ecological and human
social effects of artificial feeding and baiting of wildlife. Canadian Cooperative Wildlife
Health Centre: Newsletters & Publications.
Dýrmundsson, Ó. R. 1973. Puberty and early reproduction performance in sheep. 1. Ewe lambs.
Animal Breeding Abstracts 41:273–289.
Endo, T., and M. Kobayashi. 2022. Typical changes in carbon and nitrogen stable isotope ratios
and mercury concentration during the lactation of marine mammals. H. Abdelhay
Essayed Kaoud, editor. Marine Mammals. IntechOpen.
Environment and Climate Change Canada. 2018. Imminent threat assessment for southern
mountain caribou.
Environment and Climate Change Canada. 2020. Amended recovery strategy for the woodland
caribou (Rangifer tarandus caribou), boreal population, in Canada.
Environment Canada. 2014. Recovery strategy for the woodland caribou, southern mountain
population (Rangifer tarandus caribou) in Canada.
Falconer, S., and A. T. Ford. 2020. Evaluating policy-relevant surrogate taxa for biodiversity
conservation: a case study from British Columbia, Canada. Canadian Journal of Zoology
98:279–286.
Fanson, K. V., T. Keeley, and B. G. Fanson. 2014. Cyclic changes in cortisol across the estrous
cycle in parous and nulliparous Asian elephants. Endocrine Connections 3:57–66.
Festa-Bianchet, M., J. C. Ray, S. Boutin, S. D. Côté, and A. Gunn. 2011. Conservation of
caribou (Rangifer tarandus) in Canada: an uncertain future 1 This review is part of the
virtual symposium “Flagship Species – Flagship Problems” that deals with ecology,
biodiversity and management issues, and climate impacts on species at risk and of
Canadian importance, including the polar bear (Ursus maritimus), Atlantic cod (Gadus
morhua), Piping Plover (Charadrius melodus), and caribou (Rangifer tarandus).
Canadian Journal of Zoology 89:419–434.
Festa-Bianchet, M., M. Urquhart, and K. G. Smith. 1994. Mountain goat recruitment: kid
production and survival to breeding age. Canadian Journal of Zoology 72:22–27.
Flasko, A., M. Manseau, G. Mastromonaco, M. Bradley, L. Neufeld, and P. Wilson. 2017. Fecal
DNA, hormones, and pellet morphometrics as a noninvasive method to estimate age
class: an application to wild populations of Central Mountain and Boreal woodland
caribou (Rangifer tarandus caribou). Canadian Journal of Zoology 95:311–321.
Forristal, V. E., S. Creel, M. L. Taper, B. M. Scurlock, and P. C. Cross. 2012. Effects of
supplemental feeding and aggregation on fecal glucocorticoid metabolite concentrations
in elk: Elk Feedgrounds and Stress. The Journal of Wildlife Management 76:694–702.
Fortin, D., F. Barnier, P. Drapeau, T. Duchesne, C. Dussault, S. Heppell, M.-C. Prima, M.-H. StLaurent, and G. Szor. 2017. Forest productivity mitigates human disturbance effects on
late-seral prey exposed to apparent competitors and predators. Scientific Reports 7:6370.
Fortin, D., R. Courtois, P. Etcheverry, C. Dussault, and A. Gingras. 2008. Winter selection of
landscapes by woodland caribou: behavioural response to geographical gradients in
habitat attributes. Journal of Applied Ecology 45:1392–1400.

107

Found, R., J. A. Baker, J. M. Fryxell, A. A. D. McLaren, A. R. Rodgers, J. Shuter, I. Thompson,
and B. R. Patterson. 2022. Stable isotopes indicate reduced body condition of caribou in
disturbed areas. Animal Biology 72:415–433.
Fox, J., S. Weisberg, and B. Price. 2001. car: Companion to Applied Regression.
Fraser, B. C., K. Eyre, J. B. Gaughan, G. Wijffels, and L. F. P. Silva. 2025. Use of 15N
abundance in tail hair to predict feed efficiency and response to a urea supplement in
young cattle fed a tropical grass forage. W. Bryden, editor. Animal Production Science
65.
Freeman, N. L. 2008. Motorized backcountry recreation and stress response in Mountain Caribou
(Rangifer tarandus caribou).
Fuller, B. T., J. L. Fuller, N. E. Sage, D. A. Harris, T. C. O’Connell, and R. E. M. Hedges. 2004.
Nitrogen balance and δ15 N: why you’re not what you eat during pregnancy. Rapid
Communications in Mass Spectrometry 18:2889–2896.
Fuller, B. T., J. L. Fuller, N. E. Sage, D. A. Harris, T. C. O’Connell, and R. E. M. Hedges. 2005.
Nitrogen balance and δ15 N: why you’re not what you eat during nutritional stress. Rapid
Communications in Mass Spectrometry 19:2497–2506.
Funck, J., C. Kellam, C. T. Seaton, and M. J. Wooller. 2020. Stable isotopic signatures in
modern wood bison (Bison bison athabascae) hairs as telltale biomarkers of nutritional
stress. Canadian Journal of Zoology 98:505–514.
Gaillard, J.-M., M. Festa-Bianchet, N. G. Yoccoz, A. Loison, and C. Toigo. 2000. Temporal
variation in fitness components and population dynamics of large herbivores. Annual
Review of Ecology and Systematics 31:367–393.
Gerhart, K. L., D. E. Ruussell, D. V. D. Wetering, R. G. White, and R. D. Cameron. 1997a.
Pregnancy of adult caribou (Rangifer tarandus): evidence for lactational infertility.
Journal of Zoology 242:17–30.
Gerhart, K. L., R. G. White, R. D. Cameron, D. E. Russell, and D. Van De Wetering. 1997b.
Pregnancy rate as an indicator of nutritional status in Rangifer. implications of lactational
infertility. Rangifer 17:21.
Ghosal, R., K. L. Edwards, T. L. Chiarelli, K. V. Fanson, A. Ganswindt, T. Keeley, D. C.
Koester, B. Roberts, T. L. Majelantle, J. Wauters, and A. E. Newell-Fugate. 2023.
Biomarkers of reproductive health in wildlife and techniques for their assessment.
Theriogenology Wild 3:100052.
Gilbert, M., R. T. Watson, S. Ahmed, M. Asim, and J. A. Johnson. 2007. Vulture restaurants and
their role in reducing diclofenac exposure in Asian vultures. Bird Conservation
International 17:63–77.
Gort‐Esteve, A., A. Carbajal, M. López, X. Manteca, J. Ruiz‐Olmo, and J. L. Riera. 2024. Faecal
cortisol levels in a wild Iberian red deer population are best explained by prior weather
conditions. Journal of Zoology 322:375–385.
Goymann, W., and J. C. Wingfield. 2004. Allostatic load, social status and stress hormones: the
costs of social status matter. Animal Behaviour 67:591–602.
Graham, L. H., E. M. Leishman, K. Demers, D. P. Whiteside, and M. McAdie. 2024. Factors
associated with reproductive success in captive Vancouver Island marmots (Marmota
vancouverensis). Animals 14:387.
Graham, L. H., K. Reid, T. Webster, M. Richards, and S. Joseph. 2002. Endocrine patterns
associated with reproduction in the Nile hippopotamus (Hippopotamus amphibius) as

108

assessed by fecal progestagen analysis. General and Comparative Endocrinology 128:74–
81.
Graham, L., F. Schwarzenberger, E. Möstl, W. Galama, and A. Savage. 2001. A versatile
enzyme immunoassay for the determination of progestogens in feces and serum. Zoo
Biology 20:227–236.
Gregorio, P. F., A. Panebianco, R. Ovejero Aguilar, P. A. Taraborelli, P. G. Moreno, N. M.
Schroeder, L. R. Leggieri, A. A. Marozzi, and P. D. Carmanchahi. 2019. Linking diet
quality and energy demand in free‐living guanacos: an eco‐physiological innovative
approach. Journal of Zoology 308:243–252.
Gustine, D. D., P. S. Barboza, L. G. Adams, R. G. Farnell, and K. L. Parker. 2011. An isotopic
approach to measuring nitrogen balance in caribou: Estimating nitrogen balance in
caribou. The Journal of Wildlife Management 75:178–188.
Gustine, D. D., P. S. Barboza, L. G. Adams, and N. B. Wolf. 2014a. Environmental and
physiological influences to isotopic ratios of N and protein status in a montane ungulate
in winter. C. Sueur, editor. PLoS ONE 9:e103471.
Gustine, D. D., P. S. Barboza, J. P. Lawler, L. G. Adams, K. L. Parker, S. M. Arthur, and B. S.
Shults. 2012. Diversity of nitrogen isotopes and protein status in caribou: implications for
monitoring northern ungulates. Journal of Mammalogy 93:778–790.
Gustine, D. D., T. J. Brinkman, M. A. Lindgren, J. I. Schmidt, T. S. Rupp, and L. G. Adams.
2014b. Climate-driven effects of fire on winter habitat for caribou in the Alaskan-Yukon
arctic. B. Bond-Lamberty, editor. PLoS ONE 9:e100588.
Gustine, D. D., K. L. Parker, R. J. Lay, M. P. Gillingham, and D. C. Heard. 2006. Calf survival
of woodland caribou in a multi-predator ecosystem. Wildlife Monographs 165:1–32.
Habran, S., C. Debier, D. E. Crocker, D. S. Houser, G. Lepoint, J.-M. Bouquegneau, and K. Das.
2010. Assessment of gestation, lactation and fasting on stable isotope ratios in northern
elephant seals (Mirounga angustirostris). Marine Mammal Science 26:880–895.
Hafez, E. S. E. 1952. Studies on the breeding season and reproduction of the ewe Part I. The
breeding season in different environments Part II. The breeding season in one locality.
The Journal of Agricultural Science 42:189–231.
Haghshenas, M., and L. J. Wood. 2025. Subalpine fir response to drought in northern British
Columbia, Canada. F. Fassnacht, editor. Forestry: An International Journal of Forest
Research 98:426–436.
Hardesty, L. H., R. J. Hudson, and R. G. White. 1988. Bioenergetics of wild herbivores. Journal
of Range Management 41:270.
Hata, A., R. Nakashita, T. Anezaki, M. Minami, Y. Fukue, N. Higuchi, H. Uno, Y. Nakajima, M.
Saeki, C. Kozakai, and M. B. Takada. 2021. Agricultural crop consumption induces
precocious maturity in deer by improving physical and reproductive performance.
Ecosphere 12:e03464.
Heard, D. C., and K. L. Zimmerman. 2021. Fall supplemental feeding increases population
growth rate of an endangered caribou herd. PeerJ 9:e10708.
Heffelfinger, L. J., D. G. Hewitt, R. W. DeYoung, T. E. Fulbright, L. A. Harveson, W. C.
Conway, and S. S. Gray. 2023. Shifting agriculture and a depleting aquifer: implications
of row-crop farming on mule deer population performance. G. Dryden, editor. Animal
Production Science 63:1633–1647.

109

Hervieux, D., M. Hebblewhite, D. Stepnisky, M. Bacon, and S. Boutin. 2014. Managing wolves
(Canis lupus) to recover threatened woodland caribou (Rangifer tarandus caribou) in
Alberta. Canadian Journal of Zoology 92:1029–1037.
Hobson, K. A., B. N. McLellan, and J. G. Woods. 2000. Using stable carbon (δ 13 C) and
nitrogen (δ 15 N) isotopes to infer trophic relationships among black and grizzly bears in
the upper Columbia River basin, British Columbia. Canadian Journal of Zoology
78:1332–1339.
Holmes, K. R., N. C. Coops, T. A. Nelson, F. M. A. Fontana, and M. A. Wulder. 2015.
Indicators of vegetation productivity under a changing climate in British Columbia,
Canada. Applied Geography 56:135–144.
Holt, R. D. 1977. Predation, apparent competition, and the structure of prey communities.
Theoretical Population Biology 12:197–229.
Holt, W. V., editor. 2003. Reproductive science and integrated conservation. Conservation
biology 8, Cambridge University Press, Cambridge.
Horstkotte, T., Ø. Holand, and J. Kumpula, editors. 2022. Reindeer husbandry and global
environmental change: pastoralism in Fennoscandia. Earthscan studies in natural resource
management, Routledge, Milton Park, Abingdon, Oxon New York, NY.
Hudson, R. J., H. M. Kozak, J. Z. Adamczewski, and C. D. Olsen. 1991. Reproductive
performance of farmed wapiti (Cervus elaphus nelsoni). Small Ruminant Research 4:19–
28.
Hunninck, L., C. R. Jackson, R. May, E. Røskaft, R. Palme, and M. J. Sheriff. 2020.
Triiodothyronine (T3) levels fluctuate in response to ambient temperature rather than
nutritional status in a wild tropical ungulate. S. Cooke, editor. Conservation Physiology
8:coaa105.
Hunt, K. E., J. Robbins, C. L. Buck, M. Bérubé, and R. M. Rolland. 2019. Evaluation of fecal
hormones for noninvasive research on reproduction and stress in humpback whales
(Megaptera novaeangliae). General and Comparative Endocrinology 280:24–34.
International Atomic Energy Agency. 1995. Reference and Intercomparison Materials for Stable
Isotopes of Light Elements. TECDOC Series 825, International Atomic Energy Agency,
Vienna.
IUCN. 2025. The IUCN Red List of Threatened Species. Version 2025-1.
James, A. R. C., S. Boutin, D. M. Hebert, and A. B. Rippin. 2004. Spatial separation of caribou
from moose and its relation to predation by wolves. Jr. White, editor. Journal of Wildlife
Management 68:799–809.
Jenkins, S. G., S. T. Partridge, T. R. Stephenson, S. D. Farley, and C. T. Robbins. 2001. Nitrogen
and carbon isotope fractionation between mothers, neonates, and nursing offspring.
Oecologia 129:336–341.
Jimeno, B., and S. Verhulst. 2023. Meta-analysis reveals glucocorticoid levels reflect variation in
metabolic rate, not ‘stress.’ eLife 12:RP88205.
Johnson, C. A., G. D. Sutherland, E. Neave, M. Leblond, P. Kirby, C. Superbie, and P. D.
McLoughlin. 2020. Science to inform policy: Linking population dynamics to habitat for
a threatened species in Canada. K. Suryawanshi, editor. Journal of Applied Ecology
57:1314–1327.

110

Johnson, C. J., L. P. W. Ehlers, and D. R. Seip. 2015. Witnessing extinction – Cumulative
impacts across landscapes and the future loss of an evolutionarily significant unit of
woodland caribou in Canada. Biological Conservation 186:176–186.
Johnson, C. J., M. A. Mumma, and M. St‐Laurent. 2019. Modeling multispecies predator–prey
dynamics: predicting the outcomes of conservation actions for woodland caribou.
Ecosphere 10:e02622.
Joly, K., F. S. Chapin, and D. R. Klein. 2010. Winter habitat selection by caribou in relation to
lichen abundance, wildfires, grazing, and landscape characteristics in northwest Alaska.
Ecoscience 17:321–333.
Joly, K., S. K. Wasser, and R. Booth. 2015. Non-Invasive assessment of the interrelationships of
diet, pregnancy rate, group composition, and physiological and nutritional stress of
barren-ground caribou in late winter. C. Sueur, editor. PLOS ONE 10:e0127586.
Jones, E. S., M. P. Gillingham, D. R. Seip, and D. C. Heard. 2007. Comparison of seasonal
habitat selection between threatened woodland caribou ecotypes in central British
Columbia. Rangifer 27:111.
Karaer, M. C., N. Čebulj-Kadunc, and T. Snoj. 2023. Stress in wildlife: comparison of the stress
response among domestic, captive, and free-ranging animals. Frontiers in Veterinary
Science 10:1167016.
Kassambara, A. 2019. rstatix: Pipe-Friendly Framework for Basic Statistical Tests.
Kelly, A. 2020. Seasonal patterns of mortality for boreal caribou (Rangifer tarandus caribou) in
an intact environment.
Kelly, J. F. 2000. Stable isotopes of carbon and nitrogen in the study of avian and mammalian
trophic ecology. Canadian Journal of Zoology 78:1–27.
Kinley, T. A., and C. D. Apps. 2001. Mortality patterns in a subpopulation of endangered
mountain caribou. Wildlife Society Bulletin (1973-2006) 29:158–164.
Kinley, T. A., T. Goward, B. N. McLellan, and R. Serrouya. 2007. The influence of variable
snowpacks on habitat use by mountain caribou. Rangifer 27:93.
Kirkpatrick, J. F., D. F. Gudermuth, R. L. Flagan, J. C. McCarthy, and B. L. Lasley. 1993.
Remote monitoring of ovulation and pregnancy of Yellowstone bison. The Journal of
Wildlife Management 57:407.
Kitaysky, A. S., M. D. Romano, J. F. Piatt, J. C. Wingfield, and M. Kikuchi. 2005. The
adrenocortical response of tufted puffin chicks to nutritional deficits. Hormones and
Behavior 47:606–619.
Konakchieva, R., Y. Mitev, O. F. Almeida, and V. K. Patchev. 1997. Chronic melatonin
treatment and the hypothalamo‐pituitary‐adrenal axis in the rat: Attenuation of the
secretory response to stress and effects on hypothalamic neuropeptide content and
release. Biology of the Cell 89:587–596.
Lamb, C. T., E. Dubman, R. S. McNay, L. Giguere, Y. Majchrzak, C. Thacker, O. Slater, B.
Macbeth, N. Owens-Beek, B. Muir, and A. T. Ford. 2024a. Assessing the health-fitness
dynamics of endangered mountain caribou and the influence of maternal penning.
Canadian Journal of Zoology cjz-2023-0032.
Lamb, C. T., S. Williams, S. Boutin, M. Bridger, D. Cichowski, K. Cornhill, C. DeMars, M.
Dickie, B. Ernst, A. Ford, M. P. Gillingham, L. Greene, D. C. Heard, M. Hebblewhite, D.
Hervieux, M. Klaczek, B. N. McLellan, R. S. McNay, L. Neufeld, B. Nobert, J. J.
Nowak, A. Pelletier, A. Reid, A. Roberts, M. Russell, D. Seip, C. Seip, C. Shores, R.

111

Steenweg, S. White, H. U. Wittmer, M. Wong, K. L. Zimmerman, and R. Serrouya.
2024b. Effectiveness of population‐based recovery actions for threatened southern
mountain caribou. Ecological Applications e2965.
Lamb, C. T., R. Willson, C. Richter, N. Owens‐Beek, J. Napoleon, B. Muir, R. S. McNay, E.
Lavis, M. Hebblewhite, L. Giguere, T. Dokkie, S. Boutin, and A. T. Ford. 2022.
Indigenous‐led conservation: Pathways to recovery for the nearly extirpated KLINSE‐ZA
mountain caribou. Ecological Applications.
Landete‐Castillejos, T., A. García, J. Á. Gómez, A. Molina, and L. Gallego. 2003. Subspecies
and Body Size Allometry Affect Milk Production and Composition, and Calf Growth in
Red Deer: Comparison of Cervus elaphus hispanicus and Cervus elaphus scoticus.
Physiological and Biochemical Zoology 76:594–602.
Larter, N. C., J. A. Nagy, and D. S. Hik. 2002. Does seasonal variation in forage quality
influence the potential for resource competition between muskoxen and Peary caribou on
Banks Island? Rangifer 22:143.
Lavery, A., and C. Ferris. 2021. Proxy measures and novel strategies for estimating nitrogen
utilisation efficiency in dairy cattle. Animals 11:343.
Lee, T. N., C. L. Buck, B. M. Barnes, and D. M. O’Brien. 2012. A test of alternate models for
increased tissue nitrogen isotope ratios during fasting in hibernating arctic ground
squirrels. Journal of Experimental Biology.
Leech, H., D. E. Jelinski, L. DeGroot, and G. Kuzyk. 2017. The temporal niche and seasonal
differences in predation risk to translocated and resident woodland caribou (Rangifer
tarandus caribou). Canadian Journal of Zoology 95:809–820.
Lenart, E. A., R. T. Bowyer, J. V. Hoef, and R. W. Ruess. 2002. Climate change and caribou:
effects of summer weather on forage. Canadian Journal of Zoology 80:664–678.
Lightman, S. L., M. T. Birnie, and B. L. Conway-Campbell. 2020. Dynamics of ACTH and
cortisol secretion and implications for disease. Endocrine Reviews 41:bnaa002.
Lozano, J., P. Lonergan, M. Boland, and D. O’Callaghan. 2003. Influence of nutrition on the
effectiveness of superovulation programmes in ewes: effect on oocyte quality and postfertilization development. Reproduction (Cambridge) 125:543–553.
Mallory, C. D., and M. S. Boyce. 2018. Observed and predicted effects of climate change on
Arctic caribou and reindeer. Environmental Reviews 26:13–25.
Mani, A. U., W. A. C. McKelvey, and E. D. Watson. 1992. The effects of low level of feeding
on response to synchronization of estrus, ovulation rate and embryo loss in goats.
Theriogenology 38:1013–1022.
Mattisson, J., G. R. Rauset, J. Odden, H. Andrén, J. D. C. Linnell, and J. Persson. 2016.
Predation or scavenging? Prey body condition influences decision‐making in a facultative
predator, the wolverine. Ecosphere 7:e01407.
McEwan, E. H., and P. E. Whitehead. 1972. Reproduction in female reindeer and caribou.
Canadian Journal of Zoology 50:43–46.
McGreer, M. T., E. E. Mallon, L. M. Vander Vennen, P. A. Wiebe, J. A. Baker, G. S. Brown, T.
Avgar, J. Hagens, A. M. Kittle, A. Mosser, G. M. Street, D. E. B. Reid, A. R. Rodgers, J.
Shuter, I. D. Thompson, M. J. Turetsky, S. G. Newmaster, B. R. Patterson, and J. M.
Fryxell. 2015. Selection for forage and avoidance of risk by woodland caribou (Rangifer
tarandus caribou) at coarse and local scales. Ecosphere 6:1–11.

112

McLeod Lake Indian Band. n.d. Traditional Stewardship. <https://mlibwildlife.ca/traditionalstewardship/>. Accessed 14 Aug 2025.
McLoughlin, P. D., E. Dzus, B. Wynes, and S. Boutin. 2003a. Declines in populations of
woodland caribou. The Journal of Wildlife Management 67:755–761.
McNay, R. S., C. T. Lamb, L. Giguere, S. H. Williams, H. Martin, G. D. Sutherland, and M.
Hebblewhite. 2022. Demographic responses of nearly extirpated endangered mountain
caribou to recovery actions in Central British Columbia. Ecological Applications 32.
Mech, L. D., and G. D. Delgiudice. 1985. Limitations of the marrow-fat technique as an
indicator of body condition. Wildlife Society Bulletin (1973-2006) 13:204–206.
Meidinger, D. and British Columbia, editors. 1991. Ecosystems of British Columbia. Special
report series; 6, British Columbia Ministry of Forests, Victoria.
Messier, F., D. M. Desaulniers, A. K. Goff, R. Nault, R. Patenaude, and M. Crete. 1990a.
Caribou pregnancy diagnosis from immunoreactive progestins and estrogens excreted in
feces. The Journal of Wildlife Management 54:279.
Metz, M. C., D. W. Smith, J. A. Vucetich, D. R. Stahler, and R. O. Peterson. 2012. Seasonal
patterns of predation for gray wolves in the multi‐prey system of Yellowstone National
Park. Journal of Animal Ecology 81:553–563.
Milner, J. M., F. M. Van Beest, K. T. Schmidt, R. K. Brook, and T. Storaas. 2014. To feed or not
to feed? Evidence of the intended and unintended effects of feeding wild ungulates:
Effects of Feeding Ungulates. The Journal of Wildlife Management 78:1322–1334.
Minagawa, M., and E. Wada. 1984. Stepwise enrichment of 15N along food chains: Further
evidence and the relation between δ15N and animal age. Geochimica et Cosmochimica
Acta 48:1135–1140.
Mohammed, O. B., D. I. Green, and W. V. Holt. 2011. Fecal progesterone metabolites and
ovarian activity in cycling and pregnant mountain gazelles (Gazella gazella).
Theriogenology 75:542–548.
Morden, C.-J. C., R. B. Weladji, E. Ropstad, E. Dahl, Ø. Holand, G. Mastromonaco, and M.
Nieminen. 2011. Fecal hormones as a non-invasive population monitoring method for
reindeer: Fecal Hormones for Monitoring of Reindeer. The Journal of Wildlife
Management 75:1426–1435.
Moskovitz, D. 2019. The contiguous United States just lost its last wild caribou. Science.
https://doi.org/10.1126/science.aaw7110.
Möstl, E., and R. Palme. 2002. Hormones as indicators of stress. Domestic Animal
Endocrinology 23:67–74.
Mueller, M., C. J. Johnson, and R. S. McNay. 2022. Influence of maternity penning on the
success and timing of parturition by mountain caribou (Rangifer tarandus caribou).
Canadian Journal of Zoology 100:548–560.
Muir, B. R., and A. L. Booth. 2012. An environmental justice analysis of caribou recovery
planning, protection of an Indigenous culture, and coal mining development in northeast
British Columbia, Canada. Environment, Development and Sustainability 14:455–476.
Murray, M. H., D. J. Becker, R. J. Hall, and S. M. Hernandez. 2016. Wildlife health and
supplemental feeding: A review and management recommendations. Biological
Conservation 204:163–174.
Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. Da Fonseca, and J. Kent. 2000.
Biodiversity hotspots for conservation priorities. Nature 403:853–858.

113

Nak’azdli Whut’en. 2021. Nak’azdli Whut’en land stewardship planning report version 1.0.
Planning Report.
Newbold, T., L. N. Hudson, S. L. L. Hill, S. Contu, I. Lysenko, R. A. Senior, L. Börger, D. J.
Bennett, A. Choimes, B. Collen, J. Day, A. De Palma, S. Díaz, S. Echeverria-Londoño,
M. J. Edgar, A. Feldman, M. Garon, M. L. K. Harrison, T. Alhusseini, D. J. Ingram, Y.
Itescu, J. Kattge, V. Kemp, L. Kirkpatrick, M. Kleyer, D. L. P. Correia, C. D. Martin, S.
Meiri, M. Novosolov, Y. Pan, H. R. P. Phillips, D. W. Purves, A. Robinson, J. Simpson,
S. L. Tuck, E. Weiher, H. J. White, R. M. Ewers, G. M. Mace, J. P. W. Scharlemann, and
A. Purvis. 2015. Global effects of land use on local terrestrial biodiversity. Nature
520:45–50.
Newby, J. R., and N. J. DeCesare. 2020. Multiple nutritional currencies shape pregnancy in a
large herbivore. Canadian Journal of Zoology 98:307–315.
Newediuk, L., G. F. Mastromonaco, and E. Vander Wal. 2024. Associations between
glucocorticoids and habitat selection reflect daily and seasonal energy requirements.
Movement Ecology 12:30.
Nilssen, K. J., K. Bye, J. A. Sundsfjord, and A. Schytte Blix. 1985. Seasonal changes in T3, FT4,
and cortisol in free-ranging Svalbard reindeer (Rangifer tarandus platyrhynchus).
General and Comparative Endocrinology 59:210–213.
Nobert, B. R., S. Milligan, G. B. Stenhouse, and L. Finnegan. 2016. Seeking sanctuary: the
neonatal calving period among central mountain woodland caribou (Rangifer tarandus
caribou). Canadian Journal of Zoology 94:837–851.
O’Connor, T. P., J. Lawrence, P. Andersen, V. Leathers, and E. Workman. 2013. Immunoassay
applications in veterinary diagnostics. Pages 623–645 in. The Immunoassay Handbook.
Elsevier.
Ossi, F., J.-M. Gaillard, M. Hebblewhite, N. Morellet, N. Ranc, R. Sandfort, M. Kroeschel, P.
Kjellander, A. Mysterud, J. D. C. Linnell, M. Heurich, L. Soennichsen, P. Sustr, A.
Berger, M. Rocca, F. Urbano, and F. Cagnacci. 2017. Plastic response by a small cervid
to supplemental feeding in winter across a wide environmental gradient. Ecosphere
8:e01629.
Ouellet, J.-P., D. C. Heard, S. Boutin, and R. Mulders. 1997. A comparison of body condition
and reproduction of caribou on two predator-free arctic islands. Canadian Journal of
Zoology 75:11–17.
Palme, R., S. Rettenbacher, C. Touma, S. M. El‐Bahr, and E. Möstl. 2005. Stress hormones in
mammals and birds: comparative aspects regarding metabolism, excretion, and
noninvasive measurement in fecal samples. Annals of the New York Academy of
Sciences 1040:162–171.
Parker, G. R. 1981. Physical and reproductive characteristics of an expanding woodland caribou
population (Rangifer tarandus caribou) in northern Labrador. Canadian Journal of
Zoology 59:1929–1940.
Parker, K. L., P. S. Barboza, and M. P. Gillingham. 2009. Nutrition integrates environmental
responses of ungulates. Functional Ecology 23:57–69.
Parker, K. L., P. S. Barboza, and T. R. Stephenson. 2005. Protein conservation in female caribou
(Rangifer tarandus): effects of decreasing diet quality during winter. Journal of
Mammalogy 86:610–622.

114

Pasciu, V., M. Nieddu, F. D. Sotgiu, E. Baralla, and F. Berlinguer. 2024. Fecal thyroid hormone
metabolites in wild ungulates: a mini-review. Frontiers in Veterinary Science
11:1407479.
Pérez-Flores, J., S. Calmé, and R. Reyna-Hurtado. 2016. Scoring body condition in wild Baird’s
Tapir (Tapirus bairdii) using camera traps and opportunistic photographic material.
Tropical Conservation Science 9:1940082916676128.
Pigeon, G., M. Festa-Bianchet, and F. Pelletier. 2017. Long-term fitness consequences of early
environment in a long-lived ungulate. Proceedings of the Royal Society B: Biological
Sciences 284:20170222.
Pigeon, G., L. E. Loe, R. Bischof, C. Bonenfant, M. Forchhammer, R. J. Irvine, E. Ropstad, A.
Stien, V. Veiberg, and S. Albon. 2019. Silver spoon effects are constrained under
extreme adult environmental conditions. Ecology 100:e02886.
Post, E., P. S. Bøving, C. Pedersen, and M. A. MacArthur. 2003. Synchrony between caribou
calving and plant phenology in depredated and non-depredated populations. Canadian
Journal of Zoology 81:1709–1714.
Prichard, A. K., G. L. Finstad, and D. H. Shain. 1999. Lactation in yearling Alaskan reindeer:
Implications for growth, reproduction, and survival. Rangifer 19:77.
Pulliainen, E. 1971. Nutritive values of some lichens used as food by reindeer in northeastern
Lapland. Annales Zoologici Fennici 8:385–389.
R Core Team. 2023. R: A Language and Environment for Statistical Computing. R Foundation
for Statistical Computing, Vienna, Austria}.
Ratcliffe, P. R. 1980. Bone marrow fat as an indicator of condition in roe deer. 25:333–340.
Reeder, D. M., and K. M. Kramer. 2005. Stress in free-ranging mammals: integrating
physiology, ecology, and natural history. Journal of Mammalogy 86:225–235.
Rettie, W. J., and F. Messier. 1998. Dynamics of woodland caribou populations at the southern
limit of their range in Saskatchewan. Canadian Journal of Zoology 76:251–259.
Ringberg, T., and A. Aakvaag. 1982. The diagnosis of early pregnancy and missed abortion in
European and Svalbard reindeer (Rangifer tarandus tarandus and Rangifer tarandus
platyrhyncus. Rangifer 2:26.
Robb, G. N., R. A. McDonald, D. E. Chamberlain, and S. Bearhop. 2008. Food for thought:
supplementary feeding as a driver of ecological change in avian populations. Frontiers in
Ecology and the Environment 6:476–484.
Robbins, C. T. 1993. Wildlife feeding and nutrition. 2nd ed. Animal feeding and nutrition,
Academic Press, San Diego.
Robinson, J. J., C. J. Ashworth, J. A. Rooke, L. M. Mitchell, and T. G. McEvoy. 2006. Nutrition
and fertility in ruminant livestock. Animal Feed Science and Technology 126:259–276.
Robinson, J. J., J. A. Rooke, and T. G. McEvoy. 2002. Nutrition for conception and pregnancy.
Pages 189–211 in M. Freer and H. Dove, editors. Sheep Nutrition. First edition. CABI
Publishing, UK.
Robinson, J., T. McEvoy, and C. Ashworth. 2001. Nutrition in the expression of reproductive
potential. Journal of Animal and Feed Sciences 10:15–27.
Romero, L. M. 2004. Physiological stress in ecology: lessons from biomedical research. Trends
in Ecology & Evolution 19:249–255.
Rominger, E. M., and C. T. Robbins. 1996. Generic preference and in-vivo digestibility of
alectorioid arboreal lichens by woodland caribou. Rangifer 16:379.

115

Ropstad, E., M. Forsberg, J. E. Sire, H. Kindahl, T. Nilsen, O. Pedersen, and L.-E. Edqvist.
1995a. Plasma concentrations of progesterone, oestradiol, LH and 15-ketodihydro-PGF2
in Norwegian semi-domestic reindeer (Rangifer tarandus tarandus) during their first
reproductive season. Reproduction 105:307–314.
Ropstad, E., V. Veiberg, H. Säkkinen, E. Dahl, H. Kindahl, Ø. Holand, J. F. Beckers, and E.
Eloranta. 2005. Endocrinology of pregnancy and early pregnancy detection by
reproductive hormones in reindeer (Rangifer tarandus tarandus). Theriogenology
63:1775–1788.
Russell, D. E., K. L. Gerhart, R. G. White, and D. V. D. Wetering. 1998. Detection of early
pregnancy in caribou: evidence for embryonic mortality. The Journal of Wildlife
Management 62:1066.
Russell, K. L. M., C. J. Johnson, and T. M. Hegel. 2025. Importance of scale, season, and forage
availability for understanding the use of recent burns by woodland caribou during winter.
Canadian Journal of Forest Research 55:1–18.
Ryg, M. 1984. Seasonal changes in the relationship between food intake and serum
triiodothyronine in reindeer. Comparative Biochemistry and Physiology Part A:
Physiology 78:427–429.
Rysava, K., R. A. R. McGill, J. Matthiopoulos, and J. G. C. Hopcraft. 2016. Re‐constructing
nutritional history of Serengeti wildebeest from stable isotopes in tail hair: seasonal
starvation patterns in an obligate grazer. Rapid Communications in Mass Spectrometry
30:1461–1468.
Sapolsky, R. M., L. M. Romero, and A. U. Munck. 2000. How do glucocorticoids influence
stress responses? integrating permissive, suppressive, stimulatory, and preparative
actions. Endocrine Reviews 21:55–89.
Schiffmann, C., M. Clauss, S. Hoby, and J.-M. Hatt. 2017. Visual body condition scoring in zoo
animals – composite, algorithm and overview approaches. Journal of Zoo and Aquarium
Research 5:1–10.
Schmidt, N. M., A. Michelsen, L. H. Hansen, M. R. Aggerbeck, M. Stelvig, S. Kutz, and J. B.
Mosbacher. 2023. Sequential analysis of δ 15 N in guard hair suggests late gestation is the
most critical period for muskox calf recruitment. Rapid Communications in Mass
Spectrometry 37.
Schwarzenberger, F. 2007. The many uses of non‐invasive faecal steroid monitoring in zoo and
wildlife species. International Zoo Yearbook 41:52–74.
Schwarzenberger, F., W. Rietschel, J. Vahala, D. Holeckova, P. Thomas, J. Maltzan, K.
Baumgartner, and W. Schaftenaar. 2000. Fecal progesterone, estrogen, and androgen
metabolites for noninvasive monitoring of reproductive function in the female Indian
rhinoceros, Rhinoceros unicornis. General and Comparative Endocrinology 119:300–
307.
Schwarzenberger, F., C. H. Son, R. Pretting, and K. Arbeiter. 1996. Use of group-specific
antibodies to detect fecal progesterone metabolites during the estrous cycle of cows.
Theriogenology 46:23–32.
Seip, D., and E. Jones. 2013. Population status of caribou herds in the central mountain
designatable unit within British Columbia, 2013. Government of British Columbia.
Seip, D., and E. Jones. 2016. Population status of central mountain caribou herds within British
Columbia, 2015. Government of British Columbia.

116

Seip, D., and E. Jones. 2018. Population status of central mountain caribou herds in British
Columbia and response to recovery management actions, 2018. Government of British
Columbia.
Seip, D. R. 1992. Factors limiting woodland caribou populations and their interrelationships with
wolves and moose in southeastern British Columbia. Canadian Journal of Zoology
70:1494–1503.
Seip, D. R., and D. B. Cichowski. 1996. Population ecology of caribou in British Columbia.
Rangifer 16:73.
Sergio, F., T. Caro, D. Brown, B. Clucas, J. Hunter, J. Ketchum, K. McHugh, and F. Hiraldo.
2008. Top predators as conservation tools: ecological rationale, assumptions, and
efficacy. Annual Review of Ecology, Evolution, and Systematics 39:1–19.
Serrouya, R., D. Paetkau, B. N. McLellan, S. Boutin, M. Campbell, and D. A. Jenkins. 2012.
Population size and major valleys explain microsatellite variation better than taxonomic
units for caribou in western Canada. Molecular Ecology 21:2588–2601.
Serrouya, R., D. R. Seip, D. Hervieux, B. N. McLellan, R. S. McNay, R. Steenweg, D. C. Heard,
M. Hebblewhite, M. Gillingham, and S. Boutin. 2019. Saving endangered species using
adaptive management. Proceedings of the National Academy of Sciences 116:6181–
6186.
Sheriff, M. J., B. Dantzer, B. Delehanty, R. Palme, and R. Boonstra. 2011. Measuring stress in
wildlife: techniques for quantifying glucocorticoids. Oecologia 166:869–887.
Shrestha, H. K. 2021. Loss of luteal sensitivity to luteinizing hormone underlies luteolysis in
cattle: A hypothesis. Reproductive Biology 21:100570.
Silanikove, N. 2000. Effects of heat stress on the welfare of extensively managed domestic
ruminants. Livestock Production Science 67:1–18.
Smiley, R. A., C. D. Rittenhouse, T. W. Mong, and K. L. Monteith. 2020. Assessing nutritional
condition of mule deer using a photographic index. Wildlife Society Bulletin 44:208–
213.
Smith, M. S. 2009. Estrus and Menstrual Cycles: Neuroendocrine Control. Pages 1–5 in.
Encyclopedia of Neuroscience. Elsevier.
Solberg, E. J., P. Jordhøy, O. Strand, R. Aanes, A. Loison, B. ‐E. Sæther, and J. D. C. Linnell.
2001. Effects of density‐dependence and climate on the dynamics of a Svalbard reindeer
population. Ecography 24:441–451.
Sontakke, S. D. 2018. Monitoring and controlling ovarian activities in wild ungulates.
Theriogenology 109:31–41.
Stanton, D. E., A. Ormond, N. M. Koch, and C. Colesie. 2023. Lichen ecophysiology in a
changing climate. American Journal of Botany 110:e16131.
Stetz, J., K. Hunt, K. C. Kendall, and S. K. Wasser. 2013. Effects of exposure, diet, and
thermoregulation on fecal glucocorticoid measures in wild bears. F. Seebacher, editor.
PLoS ONE 8:e55967.
Storeheier, P. V., S. D. Mathiesen, N. J. C. Tyler, and M. A. Olsen. 2002. Nutritive value of
terricolous lichens for reindeer in winter. The Lichenologist 34:247–257.
Taillon, J., P. S. Barboza, and S. D. Côté. 2013. Nitrogen allocation to offspring and milk
production in a capital breeder. Ecology 94:1815–1827.

117

Taillon, J., V. Brodeur, M. Festa-Bianchet, and S. D. Côté. 2011. Variation in body condition of
migratory caribou at calving and weaning: Which measures should we use? Écoscience
18:295–303.
Taillon, J., V. Brodeur, M. Festa-Bianchet, and S. D. Côté. 2012. Is mother condition related to
offspring condition in migratory caribou (Rangifer tarandus) at calving and weaning?
Canadian Journal of Zoology 90:393–402.
Thomas, D. C. 1982. The relationship between fertility and fat reserves of Peary caribou.
Canadian Journal of Zoology 60:597–602.
Todini, L. 2007. Thyroid hormones in small ruminants: effects of endogenous, environmental
and nutritional factors. Animal 1:997–1008.
Todini, L., J. A. Delgadillo, A. Debenedetti, and P. Chemineau. 2006. Plasma total T3 and T4
concentrations in bucks as affected by photoperiod. Small Ruminant Research 65:8–13.
Touitou, S., M. Heistermann, O. Schülke, and J. Ostner. 2021. Triiodothyronine and cortisol
levels in the face of energetic challenges from reproduction, thermoregulation and food
intake in female macaques. Hormones and Behavior 131:104968.
Tŝilhqot’in National Government. 2019. Tŝilhqot’in and Ulkatcho announce emergency no
caribou hunting ban in their traditional territories. Media Resease, Tŝilhqot’in National
Government, Williams Lake, BC.
Twining, C. W., J. R. Shipley, and B. Matthews. 2022. Climate change creates nutritional
phenological mismatches. Trends in Ecology & Evolution 37:736–739.
Tyler, N. J. C. 2010. Climate, snow, ice, crashes, and declines in populations of reindeer and
caribou (Rangifer tarandus L.). Ecological Monographs 80:197–219.
Viejou, R., T. Avgar, G. S. Brown, B. R. Patterson, D. E. B. Reid, A. R. Rodgers, J. Shuter, I. D.
Thompson, and J. M. Fryxell. 2018. Woodland caribou habitat selection patterns in
relation to predation risk and forage abundance depend on reproductive state. Ecology
and Evolution 8:5863–5872.
Vors, L. S., J. A. Schaefer, B. A. Pond, A. R. Rodgers, and B. R. Patterson. 2007. Woodland
caribou extirpation and anthropogenic landscape disturbance in Ontario. Journal of
Wildlife Management 71:1249–1256.
Wasser, S. K., J. C. Azkarate, R. K. Booth, L. Hayward, K. Hunt, K. Ayres, C. Vynne, K.
Gobush, D. Canales-Espinosa, and E. Rodríguez-Luna. 2010. Non-invasive measurement
of thyroid hormone in feces of a diverse array of avian and mammalian species. General
and Comparative Endocrinology 168:1–7.
Wasser, S. K., J. L. Keim, M. L. Taper, and S. R. Lele. 2011. The influences of wolf predation,
habitat loss, and human activity on caribou and moose in the Alberta oil sands. Frontiers
in Ecology and the Environment 9:546–551.
Wasser, S. K., J. I. Lundin, K. Ayres, E. Seely, D. Giles, K. Balcomb, J. Hempelmann, K.
Parsons, and R. Booth. 2017. Population growth is limited by nutritional impacts on
pregnancy success in endangered Southern Resident killer whales (Orcinus orca). Z. D.
Deng, editor. PLOS ONE 12:e0179824.
Wathes, D. C., M. Fenwick, Z. Cheng, N. Bourne, S. Llewellyn, D. G. Morris, D. Kenny, J.
Murphy, and R. Fitzpatrick. 2007. Influence of negative energy balance on cyclicity and
fertility in the high producing dairy cow. Theriogenology 68:S232–S241.

118

Webb, R., P. C. Garnsworthy, J.-G. Gong, and D. G. Armstrong. 2004. Control of follicular
growth: Local interactions and nutritional influences1,2. Journal of Animal Science
82:E63–E74.
Webber, Q. M. R., K. M. Ferraro, J. G. Hendrix, and E. Vander Wal. 2022. What do caribou eat?
A review of the literature on caribou diet. Canadian Journal of Zoology 100:197–207.
Weckerly, F. W., and M. A. Ricca. 2014. Aggression and Coexistence in Female Caribou. Arctic
67:189–195.
Wilbur, S. R., W. D. Carrier, and J. C. Borneman. 1974. Supplemental feeding program for
California condors. The Journal of Wildlife Management 38:343.
Williams, C. T., M. Klaassen, B. M. Barnes, C. L. Buck, W. Arnold, S. Giroud, S. G. Vetter, and
T. Ruf. 2017. Seasonal reproductive tactics: annual timing and the capital-to-income
breeder continuum. Philosophical Transactions of the Royal Society B: Biological
Sciences 372:20160250.
Wittmer, H. U., B. N. Mclellan, R. Serrouya, and C. D. Apps. 2007. Changes in landscape
composition influence the decline of a threatened woodland caribou population. Journal
of Animal Ecology 76:568–579.
Wittmer, H. U., R. Serrouya, L. M. Elbroch, and A. J. Marshall. 2013. Conservation strategies
for species affected by apparent competition: conservation biology. Conservation Biology
27:254–260.
Wittmer, H. U., A. R. E. Sinclair, and B. N. McLellan. 2005. The role of predation in the decline
and extirpation of woodland caribou. Oecologia 144:257–267.
Yasuo, S., N. Nakao, S. Ohkura, M. Iigo, S. Hagiwara, A. Goto, H. Ando, T. Yamamura, M.
Watanabe, T. Watanabe, S. Oda, K. Maeda, G. A. Lincoln, H. Okamura, S. Ebihara, and
T. Yoshimura. 2006. Long-Day Suppressed expression of type 2 deiodinase gene in the
mediobasal hypothalamus of the Saanen goat, a short-day breeder: implication for
seasonal window of thyroid hormone action on reproductive neuroendocrine axis.
Endocrinology 147:432–440.
Young, K. M., S. L. Walker, C. Lanthier, W. T. Waddell, S. L. Monfort, and J. L. Brown. 2004.
Noninvasive monitoring of adrenocortical activity in carnivores by fecal glucocorticoid
analyses. General and Comparative Endocrinology 137:148–165.
Young, V. R. 1991. Nutrient interactions with reference to amino acid and protein metabolism in
non-ruminants; particular emphasis on protein-energy relations in man. Zeitschrift für
Ernährungswissenschaft 30:239–267.
Zamin, T. J., S. D. Côté, J. Tremblay, and P. Grogan. 2017. Experimental warming alters
migratory caribou forage quality. Ecological Applications 27:2061–2073.
Zielke, L., N. Wrage-Mönnig, and J. Müller. 2018. Development and assessment of a body
condition score scheme for European bison (Bison bonasus). Animals 8:163.
Zimmerman, K., R. Munro, and B. Ernst. 2025. Summary of B.C. caribou population estimates,
trends, and recovery actions up to Mar. 31, 2023. B.C. Caribou Recovery Program.

119

Appendices
Supplemental material for Chapter Two
Linear mixed effects models
Table 11: The eight linear mixed-effects models used to predict fecal glucocorticoid concentrations in female caribou

(Rangifer tarandus caribou) sampled in fall - winter of 2021 and 2022 at the Kennedy Siding feeding project site in
British Columbia, Canada.
Fecal cortisol candidate model

df

logLik

AICc

ΔAICc

ωi

Min
VIF
1.060

Max
VIF
2.160

Reproductive age class + sampling
8
-124.8
265.9
0
0.462
date * sampling year
Reproductive age class + sampling
9
-124.4
267.3
1.430
0.226 1.220
2.170
date * sampling year + body mass
Reproductive age class + sampling
9
-124.6
267.6
1.762
0.192 1.200
2.170
date * sampling year + body condition
Reproductive age class + sampling
10
-124.2
268.9
3.063
0.010 1.220
2.180
date * sampling year + body condition
+ body mass
Reproductive age class (null)
5
-131.8
273.8
7.911
0.009 N/A
N/A
Reproductive age class + body
6
-131.4
274.9
9.022
0.005 1.550
1.550
condition
Reproductive age class + body mass
6
-131.7
275.5
9.635
0.004 1.410
1.410
Reproductive age class + body
7
-131.1
276.5
10.58
0.002 1.430
2.200
condition + body mass
Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model;
ωi = Akaike weight, VIF = variance inflation factor
Table 12: The eight linear mixed-effects models used to predict fecal triiodothyronine concentrations in female caribou (Rangifer
tarandus caribou) sampled in fall - winter of 2021 and 2022 at the Kennedy Siding feeding project site in British Columbia,
Canada.

Fecal T3 candidate model

df

logLik

AICc

ΔAICc

Reproductive age class + sampling date +
sampling year + body mass
Reproductive age class + sampling date +
sampling year + body mass + body
condition
Reproductive age class + sampling day +
sampling year
Reproductive age class + sampling date +
sampling year + body condition
Reproductive age class + body mass

8

-610.62

1237.63

0

9

-610.36

1239.19

7

-613.04

8

ωi
0.538

Min
VIF
1.13

Max
VIF
1.53

1.56

0.246

1.13

2.59

1240.38

2.76

0.136

1.10

1.24

-612.80

1241.99

4.36

0.061

1.10

1.92

6

-616.58

1245.38

7.76

0.011

1.32

1.32

Reproductive age class + body mass +
body condition
Reproductive age class (null)

7

-616.26

1246.82

9.19

0.005

1.31

2.52

5

-619.11

1248.38

10.75

0.002

N/A

N/A

Reproductive age class + body condition

6

-618. 83

1249.88

12.25

0.001

1.70

1.70

Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model;
ωi = Akaike weight, VIF = variance inflation factor

120

Table 13: The eight linear mixed-effects models used to predict hair δ15N concentrations in female caribou (Rangifer tarandus
caribou) sampled in fall - winter of 2021 and 2022 at the Kennedy Siding feeding project site in British Columbia, Canada.

Hair δ15N
df
logLik
AICc
ΔAICc
ωi
Min VIF
Max VIF
candidate
models
Reproducti
6
-62.74
139.3
0 0.509
1.060
1.060
ve age
category +
sampling
year
Reproducti
7
-62.09
140.7
1.387 0.254
1.080
2.980
ve age
class +
sampling
year +
body
condition
Reproducti
7
-62.61
141.8
2.422 0.152
1.140
1.510
ve age
class +
sampling
year +
body mass
Reproducti
8
-61.78
142.9
3.570 0.085
1.200
4.500
ve age
class +
sampling
year +
body
condition +
body mass
Reproducti
5
-71.69
154.7
15.34 2.000 x 10- N/A
N/A
4
ve age
class (null)
Reproducti
6
-71.48
156.8
17.48 1.000 x 10- 1.350
1.350
4
ve age
class +
body mass
Reproducti
6
-71.59
157.1
17.72 1.000 x 10- 2.650
2.650
4
ve age
class +
body
condition
Reproducti
7
-71.42
159.4
20.05 2.000 x 10- 1.390
2.730
5
ve age
class +
body
condition +
body mass
Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model;
ωi = Akaike weight, VIF = variance inflation factor

121

Figure 15: Model validations from each top model set for caribou (Rangifer tarandus caribou fecal glucocorticoids, fecal T3, and
hair δ15N. The histograms denote the frequency of residuals, and the Normal Q-Q plots denote standardized residuals against the
values that would be expected under normality from each top model.

Hair cortisol
Methods
To extract hormones from hair samples, we followed the protocol described by (Bryan et
al. 2013a). The hair was washed twice with approximately 40 mL of water for three minutes

122

each time by rotating on a plate rotator (approx. 120 rpm). This washing step was repeated using
100% isopropanol [Sigma - Aldrich] for three minutes on a plate rotator. Samples were
thoroughly dried between the water and isopropanol washes. The hair was powdered in a ball
mill (Retch Mixer Mill 200) at 25 Hz and ~30 mg was weighed into a 20 ml glass scintillation
vial. We added 100µl of 100% % HPLC methanol (Thermo Fisher Scientific) /mg of hair
powder and sealed the lids of the scintillation vials with parafilm to prevent evaporation. The
samples were sonicated for 30 minutes then placed in an incubator at 50˚C and rotated at 160
rpm overnight. The methanol was then centrifuged for 30 minutes to separate the hair particles
from the extract. The supernatant was stored at -20 ˚C until analysis.
We used an Arbour Assays DetectX® Cortisol ELISA Kit to analyse hair cortisol
concentrations. Hair samples were evaporated to dryness under a steady stream of nitrogen.
Samples were reconstituted using a solution of assay buffer provided by the kit and 100% HPLC
methanol [Thermo Fisher Scientific]. We dispensed 50 µl of either the diluted sample, standards,
or controls into each well. Assay reagents were added, and the plate was incubated following
instructions provided by the kit manufacturer. The optical density was measured at 450 nm. The
cross reactivity of the cortisol assay to several compounds was provided with the immunoassay
kit, including cortisol (100%), dexamethasone (18.8%), prednisolone (7.8%), corticosterone
(1.2%), cortisone (1.2%), and progesterone, estradiol, cortisol 21-glucuronide, 1αhydroxycorticosterone, and testosterone (< 0.1%).
As an analytical validation, a pooled hair extract from different female caribou was
assayed with the hair cortisol ELISA. To establish parallelism, we tested serial two-fold dilutions
of the pooled sample for comparison displacement curves (Figure 16). The intra assay variability
was determined by assaying replicates of the pooled sample. The intra assay CV was <15% for

123

the hair cortisol ELISA. Because we only ran one plate for hair cortisol, inter assay variation was
not calculated.

Figure 16: Parallelism of cortisol standards with a serially diluted extract from caribou (Rangifer tarandus caribou) guard hair
(dashed line).

Results
We analyzed hair cortisol concentrations in twenty-three samples from eighteen
individual caribou sampled in 2021 and 2022. Hair cortisol concentrations ranged from 2.145
pg/mg–8.570 pg/mg (mean = 4.887 pg/mg, SE = 0.377). The mean hair cortisol concentration for

124

adults with a calf was 5.010 pg/mg (n= 14, SE = 0.505), 4.321 pg/mg (n = 6, SE = 0.696) for
adults without a calf, and 5.027 pg/mg (n=3, SE = 1.179) for one-year olds. Student's t-Tests
revealed that hair cortisol concentrations were higher in adults with a calf compared to adults
without a calf, however this difference was not significant (t(10.52) = -0.905, p = 0.386; Figure
17). Hair cortisol was also higher in adults without a calf compared to one-year olds, however
the difference was also not significant (t(3.468) = -0.515, p = 0.638; Figure 17). Adults with a calf
and yearlings had similar hair cortisol concentrations (t(2.785) = 0.057, p = 0.959; Figure 17).

125

Figure 17: Cortisol concentrations in hair samples collected from female caribou (Rangifer tarandus caribou) at Kennedy Siding
in 2021 and 2022 (n = 23). The boxes represent the median value (center line) and the 25th and 75th percentiles. The highest hair
cortisol concentrations were found in yearlings (Y) and adults with calves (AC). Adults without calves (A) had the lowest values.
The lines represent 1.5 times the interquartile range for each percentile. Sample sizes are shown at the bottom of the panel.

126

Supplemental material for Chapter Three
Linear mixed effects models
Table 14: The eight linear mixed-effects models used to predict fecal progesterone metabolite concentrations in female caribou
(Rangifer tarandus caribou) sampled in fall - winter of 2021 and 2022 at the Kennedy Siding feeding project site in British
Columbia, Canada.

Candidate model

df

logLik

AICc

ΔAICc

ωi

Null
Reproductive age class

3
5

-1308.4
-1306.9

2622.8
2624.0

0
1.209

0.531
0.290

Min
VIF
N/A
N/A

Max
VIF
N/A
N/A

Body condition
5
-1307.7 2625.6
2.799
0.131 1.002
1.002
Reproductive age class + body
7
-1306.6 2627.6
4.772
0.049 1.336
2.192
condition
Loglik = Loglikelihood, AICc = AICc score, ΔAICc = difference in corrected AIC between model and top model;
ωi = Akaike weight, VIF = variance inflation factor

Individual fecal progesterone profiles
As an extension of the fecal hormone analysis, we also wanted to characterize changes in
progesterone concentrations during the breeding season for woodland caribou. An understanding
of basic reproductive processes and components of individual fitness can aid species
management and conservation because they are drivers of population dynamics in wild animals
(Schwarzenberger 2007, Sontakke 2018). Common methods for assessing reproduction in
caribou include analysis of hormone levels and movement patterns of GPS collared animals.
Progesterone is a steroid hormone that regulates the estrous cycle and is the principal hormone
responsible for the maintenance of pregnancy in mammals (Smith 2009, Crowe 2016, Shrestha
2021). Caribou are seasonally polyestrous and will experience recurring estrous cycles during the
breeding season (McEwan and Whitehead 1972, Bergerud 1975, Ropstad et al., 1995a). The
mammalian estrous cycle consists of a follicular and luteal phase (Crowe 2016). During the
follicular phase an ovarian follicle and its enclosed oocyte mature in preparation for ovulation
and fertilization (Crowe 2016). Progesterone levels are low during this phase (O’Connor et al.,
2013). The luteal phase is the period following ovulation when the corpus luteum; a transient

127

endocrine organ formed from the remnants of a mature follicle after ovulation; is formed and is
associated with increased secretion of progesterone (Asher 2011, O’Connor et al., 2013, Shrestha
2021). If pregnancy does not occur, the corpus luteum degenerates and progesterone levels
decline, allowing for maturation and ovulation of a new follicle and another opportunity to
establish a pregnancy (Asher 2011, Shrestha 2021). If pregnancy is established, the luteal
lifespan is prolonged and the corpus luteum continues to produce progesterone through the
gestation period (Asher 2011). In caribou, serum progesterone tests on blood samples taken from
captured animals have commonly been used to assess pregnancy status (Rettie and Messier 1998,
Wittmer et al., 2005, DeMars et al. 2013). The stress and hazards associated with capture and
restraint can be detrimental to the animals' health and may bias the results of hormone analysis
(Schwarzenberger 2007, Morden et al. 2011). This means that repeated blood sampling on a
regular basis is often not possible. Fecal hormone analysis has been developed as an alternative,
non-invasive technique that is now widely employed for studies of stress and reproductive
physiology in wildlife (Schwarzenberger 2007). Hormones like progesterone are metabolized by
the liver into several metabolites which are excreted in feces and can then be measured using
immunoassay techniques (Brown 2018). Due to the pooling of metabolites during excretion,
fecal samples present cumulative hormone secretion hours or days, depending on the species,
which provides a more representative measure of hormone concentrations over time
(Schwarzenberger et al. 1996, Möstl and Palme 2002, Palme et al. 2005). Fecal sampling also
enables frequent data collection without disturbance of the animal, allowing for longitudinal
studies of hormones that can be used to assess changes in reproductive activity
(Schwarzenberger 2007). The concentration of progesterone metabolites in feces has been used
to study reproductive components such as seasonality, estrus and ovulation, pregnancy, and

128

fertility status in a variety of wildlife species and has produced detailed basic information about
reproductive biology such as the length of estrous cycle and ovulation pattern in many wild
ungulates (Schwarzenberger 2007, Brown 2018, Sontakke 2018). In caribou, analysis of single
fecal samples has been used to determine pregnancy, but repeated sampling has not been used to
monitor of ovarian activity, likely due to difficulties obtaining repeated samples from wild
populations (Messier et al. 1990a, Morden et al. 2011, Joly et al. 2015).
Methods
We considered an animal to be ovulatory based on whether they had at least one fecal
sample with a FPM concentration above the baseline determined from male samples. We created
fecal progesterone profiles for caribou that had been sampled at least 10 times during a 50-day
period (October 24thtoDecember 17th) where caribou visited the feeding site daily. We
determined FPM concentrations during the luteal phases for each individual using an iterative
process described by Graham et al. (2002). The average concentration of all samples from an
individual was calculated and the values greater than the average plus 1.75 standard deviations
(SD) were considered significant elevations and temporarily removed from the data set. The
average was then recalculated and the process repeated until no values exceeded the mean + 1.75
SD. Remaining fecal progesterone values were reported as baseline (follicular phase)
concentrations. If the data was more than the baseline concentration, we considered it to be the
luteal phase, while if the value was less than the baseline concentration, we considered it to be
the follicular phase. The FPM concentrations of the luteal and follicular phases of individuals
were compared with one-way ANOVA. We estimated the length of an estrous cycle based on the
methods used by (Mohammed et al. 2011). One estrous cycle was defined as the period from the

129

first increase in FPMs above baseline to the corresponding rise in the following cycle. The day of
sampling before the first increase in FPMs was considered day 0.
Results
The baseline concentrations (mean ± standard error) of FPMs for individual females were
488.1 ± 27.6 - 2059.2 ± 119.5 ng/g and averaged 957.7 ± 91.3 ng/g (Table 13). Mean luteal
concentrations per female, calculated as the average of all values that exceeded baseline ranged
from 578.9 ± 26.8 - 2351.9 ± 100.4 ng/g and averaged 1185.1 ± 98.7 (Table 13). The FPM
concentrations during the luteal phase were higher within individuals and on average (P <0.05)
than those observed during the follicular phase. All females showed a fluctuating pattern of FPM
levels around their individual baselines, and there was great variation in the height of peaks
among individuals (Figure 18).
Table 15: Mean fecal progesterone metabolite concentrations (± standard error) of the follicular (baseline) and luteal phases from
14 individual female caribou (Rangifer tarandus caribou) sampled between fall and winter of 2021 and 2022 at the Kennedy
Siding feeding site in British Columbia, Canada. The reproductive age classes are adult without a calf at heel (A), adult with a
calf at heel (AC), and yearling females (Y).

ID

Reproductive age class

Baseline (ng/g) ± SE

Luteal (ng/g) ± SE

21.045
21.047
21.079
22.001
22.007
22.016
22.017
22.03
22.032
22.038
22.048
22.058
22.066
22.077

AC
AC
AC
A
AC
A
A
AC
A
Y
AC
AC
A
Y

869.4 ± 82.2
857.4 ± 84.3
2059.2 ± 119.5
826.4 ± 49.7
488.1 ± 27.6
1111.5 ± 44.9
832.8 ± 57.6
891.8 ± 121.4
989.1 ± 71.1
989.1 ± 59.9
747.9 ± 48.5
1102.6 ± 91.2
913.6 ± 49.9
728.9 ± 105.4

1080.0 ± 31.9
1042.9 ± 71.6
2351.9 ± 100.4
994.2 ± 77.2
578.9 ± 26.8
1362.0 ± 90.1
1133.8 ± 80.9
1270.3 ± 128.6
1192.8 ± 54.8
1116.7 + 35.3
1027.8 ± 121.1
1097.4 ± 84.2
1037.8 ± 34.8
1304.8 ± 49.6

957.7 ± 91.3

1185.1 ± 98.7

Mean

130

131

Figure 18: Longitudinal fecal progesterone metabolite profiles for 14 female caribou (Rangifer tarandus caribou) sampled
between fall and winter of 2021 and 2022 at the Kennedy Siding feeding site in British Columbia, Canada. Individual identities
are displayed at the top of each plot. The dashed line indicates an individual’s baseline value used to differentiate between the
follicular and luteal phases.

Discussion
This is, to our knowledge, the first report of changes in progesterone concentrations over
the breeding season for wild woodland caribou. The characteristics and endocrinology of the
estrous cycle and breeding season of caribou and their interactions with external factors remains
understudied and has only been described in a few reports under captive conditions. Most studies
have also examined hormones in serum samples, the invasive nature of which is impractical for
repeated sampling of wild populations. Previous research suggests that the reproductive season
for Rangifer females begins in September to October, and that cyclic ovarian activity can occur
until February. The length of estrous can vary greatly. Ropstad et al. (1995) reported that in
Norwegian reindeer (R.t tarandus), estrous lasted on average (± SD) 19.4 ± 5.7 days (range =
13–33 days). McEwan and Whitehead (1972) described two types of estrous cycles in caribou
and reindeer, a short 10- to 12-day, and a long 24-day cycle depending on the functional life of
the corpus luteum. In woodland caribou in Newfoundland, Canada, Bergerud (1975) reported
that the length of the estrous cycle was approximately 10-12 days. Because sampling was limited
to the period most caribou used the feeders (mid October to mid December), we were only able
to detect 1–2 possible cycles per animal and could not determine the total length of the breeding
season. The mean FPM baseline of 957.7 ng/g we found was similar to the mean progestin (±
SD) level of 903 ± 162 ng/g reported by Messier et al. (1990) for the early gestation period (0–50
days) of woodland caribou in Quebec, Canada. The patterns of fluctuations were variable among
individual animals. Differences could be explained by individual (phenotypic, genotypic, or
physiologic) and environmental attributes (Chinn et al. 2022). One female (21.079), had a

132

baseline FPM concentration of 2059.2 ng/g, over two times the mean baseline and higher than
the 2025.9 ng/g threshold used by Morden et al. (2011) to identify pregnant reindeer in February,
suggesting she may already have been pregnant at the time of sampling.
Due to the opportunistic nature of our sampling design, fecal samples were collected at
irregular intervals. Daily sample collections would produce more detailed profiles and allow for
further analysis of estrous characteristics. Time series analyses, which becomes more powerful
with increasing length of the dataset, could be used to identify cycles and determine cycle
frequency with more accuracy (Fanson et al. 2014). More frequent sampling combined with a
supplementary method such as behavioural observations could also help to establish more
accurate estimates of when ovulation occurred. We only examined progesterone metabolites,
however other indicators of reproductive activity such as estrogens, androgens, luteinizing
hormone, follicle stimulating hormone, and pregnancy-associated glycoproteins could be
examined to better describe the reproductive biology of woodland caribou during the breeding
season and possibly detect early pregnancy (Kirkpatrick et al. 1993, Ropstad et al. 2005). An
estimate of early pregnancy would provide more information on the occurrence of calf mortality
between pregnancy and parturition, as there is currently very little data on this phenomenon in
caribou (Ringberg and Aakvaag 1982, Russell et al. 1998, Mueller et al. 2022). Our results
demonstrate the potential of fecal progesterone measurement for monitoring reproductive
activity in caribou and provide a preliminary description of progesterone changes associated with
estrous in free ranging woodland caribou.

133