NATURAL AND ARTIFICIAL HIBERNACULA
USE BY THREE SYMPATRIC SNAKE SPECIES
by

VERONICA MARTHA LOUISE MCKELVEY
BSc (Hons), University of Toronto, 2021

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
MASTER OF SCIENCE IN ENVIRONMENTAL SCIENCE

Thompson Rivers University
Kamloops, British Columbia
May 2024
Thesis examining committee:
Karl Larsen (PhD), Thesis Supervisor
Professor, Department of Natural Resource Sciences, Thompson Rivers University
Leigh Anne Isaac (PhD), Thesis Co-Supervisor
Small Mammal and Herpetofauna Specialist,
Ministry of Water, Land and Resource Stewardship
Darryl Carlyle-Moses (PhD), Committee Member
Associate Professor, Department of Geography & Environmental Studies,
Thompson Rivers University
Nicholas Cairns (PhD), External Examiner
Curator of Non-Avian Vertebrates, Royal Alberta Museum

ii

ABSTRACT
Species that persist at higher latitudes require specialized adaptations to survive
decreased temperatures, inclement weather, and reduced resource availability during winter.
As ectotherms, the internal temperature of snakes is determined by their external
environment. Thus, within temperate ranges, snakes rely on hibernacula that meet the
structural, thermal, and humidity conditions needed to survive overwinter.
The apparent scarcity of snake hibernacula in cooler climes makes understanding the
selection of this critical habitat feature significant for conservation. I used a community of
three sympatric snake species to explore how habitat features relate to the number of each
species at hibernacula and determine how these species vary in their selection of
overwintering habitat. Little to no measured habitat metrics effectively explained the detected
number of Great Basin Gophersnakes (Pituophis catenifer deserticola), Western Yellowbellied Racers (Coluber constrictor mormon) and Western Rattlesnakes (Crotalus oreganus
oreganus) at hibernacula. Further, I reveal that although these species share the same
landscape and, in some cases, hibernacula, they are not selecting habitat in the same way.
This variation in habitat selection aligned with differences in detected egress behaviour.
Given the apparent scarcity of denning sites, one can theorize the tremendous impact
on snake populations when they are destroyed. Constructing artificial hibernacula is one
proposed conservation or mitigation measure to combat this loss. In this study, I used the
creation of two artificial hibernacula following a disturbance event to (i) collect the internal
temperature and humidity of the hibernacula and the internal temperatures of snakes and (ii)
monitor the adoption of the main artificial hibernacula by the original cohort of displaced
snakes. Both artificial hibernacula provided a range of thermal and humidity conditions, with
temperature generally increasing with depth and distance to the hibernacula mouth. Internal
hibernacula readings buffered against and were largely independent of external temperature
fluctuations (including during a low of -30.4°C). Temperatures of overwintering Great Basin
Gophersnakes within the main artificial hibernacula were comparable to those in natural
hibernacula. In particular, the artificial hibernacula provided thermal microsites conducive to
the survival of gophersnakes at depths ranging from 2.6 to 3.1m. Despite relocation efforts
and six translocated snakes overwintering the first year, no translocated snake naturally

iii
adopted or returned to the main hibernacula two years post-disturbance. However, two newly
identified snakes overwintered in the main hibernacula in the second winter.
Overall, my thesis highlights distinct interspecific variations in behaviour and
hibernacula selection. Such variation indicates that conservation efforts and impact
assessments that use hibernacula ‘models’ from well-studied species (e.g., Western
Rattlesnakes) to extrapolate to other species may be ineffective. I recommend larger
protection at the landscape level to avoid future disturbance of these hard-to-identify critical
habitat features. In cases of disturbances, I show that artificial hibernacula can be built to be
conducive to survival. However, we caution against using these structures as mitigation
measures due to the low uptake of focal individuals, at least in the short term. Future research
is needed to understand the use of these artificial structures over a longer period of time.
Keywords: Hibernacula, Artificial Hibernacula, Habitat Use, Microclimate, Critical Habitat,
Communal Denning, Great Basin Gophersnake, Western Yellow-bellied Racer, Western
Rattlesnake

iv

ACKNOWLEDGEMENTS
I would like to acknowledge that my research took place on Tk'emlúps te Secwépemc
territory, situated within the unceded ancestral lands of the Secwépemc Nation. As a settler
from Ontario, I am so grateful to have experienced the raw beauty of the Lac du Bois
Grasslands Protected Area. As I tracked snakes, I traversed this undulating landscape through
the pulse of spring growth and ticks, through the smoke and heat of the summer, following
them into the yellow hues of fall, and trudging through fresh blankets of snow. Each season
was just as gruelling and magical as the last. I can attest that enduring a never-ending line of
ticks crawling up my neck, wildfires, the 2021 heat dome, and a -30°C cold snap made me a
stronger person. A special nod to the prickly pear cactus for always 'sticking' around, usually
in the most undesirable of places (i.e., legs, hands, and... yes, even my butt).
To my supervisors, Dr. Karl Larsen and Dr. Leigh Anne Isaac, your guidance,
wisdom, and constructive feedback shaped the trajectory of this thesis and my future. I am
grateful for the mentorship and inspiration to continue my academic pursuit as I start a PhD.
A special thanks to Karl for making edits just a little bit easier with all the curated character
comments. Leigh Anne, I always looked forward to having you in the field and hope one day
to have as good a field lunch as you do. Thank you to my committee member, Dr. Darryl
Carlyle-Moses, for providing time and valuable insight from outside the snake bubble. I
would also like to acknowledge my external examiner, Dr. Nicholas Cairns; thank you for all
your feedback and interesting questions.
None of this work would have been possible without the hard work, dedication and
immense flexibility of Dr. James Sudhoff, Dr. Fergus Alexander, Tracy Reynolds, and Paige
McKnight. Thank you for always having the time for a snake surgery!
To my snake-wrangling partner in crime, Lily Ragsdale, I could not have done this
without you. I am so grateful that we shared a snake farm and the rollercoaster ride in our
Ford Expeditions that came with it. No one will understand how we name snakes except us,
and I love that. Together, we supervised many amazing field assistants crazy enough to spend
their summers catching snakes with us. I am forever grateful for the dedication, hard work,
and love of snakes that Maria Collins, Hailey Wynnyk, Jade Spruyt, and Jake Veenman gave
to this project.

v
When I started this Master’s degree, I could have never imagined how many amazing
people would come into my life and whom I would like to thank. To all my fellow grad
students (and amazing friends), Jade Spruyt, Chloe Howarth, Calen Wong, Lindsay
Whitehead, Kara Atkinson, Larisa Murdoch, Shannon Werden, and Mae Frank, thank you all
for listening to me freakout out about my thesis, and for our weekly (sometimes daily)
updates and tea-dropping sessions. It is not easy to chat through cubicles, but we made it
work.
To my family, even though we are provinces apart, thank you for always being just a
phone away… and for still picking up, even when my rants about this thesis became a broken
record.
A very special thank you and immense gratitude goes to my personal chef (my
partner), Jacob Tecca. Thank you for always encouraging me, loving snakes almost as much
as I do, providing beautiful art (see in Dedication section below), and having a warm meal to
come home to every night.
As I write this it is starting to feel like the Oscars’ music is coming on. Thanks to
everyone who played a role, both big and small. It takes a village… or something like that.
This work was supported by Trans Mountain, MITACS and the Natural Sciences and
Engineering Research Council of Canada (NSERC). All snake handling, transportation, and
surgical protocols were reviewed and approved by the Thompson Rivers University Animal
Care Committee (File No: 102730). Data were collected under BC Wildlife Permit KA21620041.

NOTE: In Chapters 2 and 3 of this thesis, plural pronouns have been chosen to align
with journal publication conventions and to recognize the contributions of my supervisors
and the significant role played by Lily Ragsdale.

vi

DEDICATION
I would like to dedicate this thesis to “The Big Three”
To all the Great Basin Gophersnakes, Western Yellow-bellied Racers, and Western
Rattlesnakes that were observed, handled, or tracked. Each snake played a significant role in
making this research possible. I hope it can now be used for their conservation.

Illustration of (left to right) a Western Yellow-bellied Racer, Western Rattlesnakes, and a
Great Basin Gophersnake at hibernacula. Illustration by Jacob Tecca.

vii

TABLE OF CONTENTS
ABSTRACT ..................................................................................................................... II
ACKNOWLEDGEMENTS ................................................................................................ IV
DEDICATION ................................................................................................................ VI
TABLE OF CONTENTS ................................................................................................. VII
LIST OF FIGURES ........................................................................................................... X
LIST OF TABLES......................................................................................................... XIII
CHAPTER 1. INTRODUCTION ......................................................................................... 1
OVERWINTER SURVIVAL .............................................................................................. 1
OVERWINTERING IN REPTILES ..................................................................................... 1
SNAKE HIBERNACULA .................................................................................................. 2
COMMUNAL DENNING.................................................................................................. 3
ARTIFICIAL HIBERNACULA .......................................................................................... 4
STUDY ANIMALS .......................................................................................................... 5
STUDY SITE .................................................................................................................. 8
DISTURBANCE EVENT ................................................................................................ 15
THESIS OBJECTIVES.................................................................................................... 18
LITERATURE CITED .................................................................................................... 19
CHAPTER 2. AN INTRA AND INTERSPECIFIC APPROACH TO IDENTIFYING AND
PROTECTING CRITICAL OVERWINTERING HABITAT FOR SNAKES ......................... 25
INTRODUCTION ........................................................................................................... 25
METHODOLOGY.......................................................................................................... 27
STUDY SITE .............................................................................................................. 27
LOCATING HIBERNACULA ....................................................................................... 29
COUNT OF SNAKES AT HIBERNACULA .................................................................... 30
HABITAT MEASUREMENTS ...................................................................................... 31
STATISTICAL ANALYSIS ........................................................................................... 34

viii

RESULTS ..................................................................................................................... 36
NUMBER OF SNAKES AT HIBERNACULA ................................................................. 37
HABITAT SELECTION ............................................................................................... 41
DISCUSSION ................................................................................................................ 41
LITERATURE CITED .................................................................................................... 47
CHAPTER 3. ARTIFICIAL SNAKE HIBERNACULA PROVIDE SUITABLE
MICROCLIMATIC CONDITIONS BUT ARE NOT READILY ADOPTED......................... 53
INTRODUCTION ........................................................................................................... 53
METHODOLOGY.......................................................................................................... 55
STUDY SITE .............................................................................................................. 55
DISTURBANCE EVENT .............................................................................................. 56
ARTIFICIAL HIBERNACULA ...................................................................................... 56
SNAKE RELEASE ...................................................................................................... 57
RADIOTELEMETRY ................................................................................................... 59
HIBERNACULA MONITORING ................................................................................... 61
EXTERNAL CONDITIONS .......................................................................................... 61
STATISTICAL ANALYSIS ........................................................................................... 63
RESULTS ..................................................................................................................... 64
MICROCLIMATE OF ARTIFICIAL HIBERNACULA ...................................................... 64
EFFECT OF ENVIRONMENTAL VARIATION ............................................................... 65
TEMPERATURE OF SNAKES ...................................................................................... 68
SNAKE ADOPTION .................................................................................................... 69
DISCUSSION ................................................................................................................ 71
LITERATURE CITED .................................................................................................... 74
CHAPTER 4. CONCLUSION .......................................................................................... 80
SUMMARY OF THESIS................................................................................................. 80
MANAGEMENT IMPLICATIONS.................................................................................... 81
FUTURE RESEARCH CONSIDERATIONS ....................................................................... 85
CONCLUSION .............................................................................................................. 87

ix

LITERATURE CITED .................................................................................................... 88
APPENDIX A: DIRECTION OF RELATIONSHIP OF HABITAT VARIABLES
CORRESPONDING TO THE POISSON REGRESSION .......................................................... 91
APPENDIX B: DIRECTION OF RELATIONSHIP OF HABITAT VARIABLES
CORRESPONDING TO THE CONDITIONAL LOGISTIC REGRESSION ................................ 92
APPENDIX C: UNIVARIATE ANALYSIS RESULTS OF CANDIDATE VARIABLES FOR
CONDITIONAL LOGISTIC REGRESSION .......................................................................... 93
APPENDIX D: COMPARISON OF WEEKLY TEMPERATURE OSCILLATION AT TIMES WITH
SNOW AND WITHOUT SNOW......................................................................................... 94
APPENDIX E: POST HOC EXACT TEST RESULTS FOR THE COMPARISON OF AVERAGE
DAILY TEMPERATURE BETWEEN DIFFERENT POSITIONS IN ARTIFICIAL SNAKE
HIBERNACULUM 1 AND 2 AND THE AVERAGE, MAXIMUM, AND MINIMUM SNAKE
TEMPERATURE ............................................................................................................. 95

x

LIST OF FIGURES
Figure 1.1: The three study species: A) Great Basin Gophersnake (Pituophis catenifer
deserticola), B) Western Yellow-bellied Racer (Coluber constrictor mormon), and C)
Western Rattlesnake (Crotalus oreganus oreganus). Photos by author ........................7
Figure 1.2: Location of the study site in the Lac du Bois Grasslands Protected Area, ~10 km
from the City of Kamloops, BC (Google Earth 2023; GBC 2023). Also shown is the
Trans Mountain Expansion Project (Pipeline). ............................................................10
Figure 1.3: Mean monthly temperatures (°C) in Kamloops, British Columbia, from 20202023, compared to the historical 30-year mean (1991-2020). Data presented in order
of ingress (August to October), the overwintering period (October-March), and egress
(March-May). Values measured at the Kamloops A (50 ° N, 120 ° W) weather station
(49° N, 119°W) (ECCC 2023). Standard error bars are present on the historical 30year mean (1991-2020 .................................................................................................11
Figure 1.4: Total precipitation (mm) in Kamloops, British Columbia, during the study years
(2020-2023) compared to the historical 30-year mean (1991-2020). Data presented in
order of ingress (August to October), the overwintering period (October-March), and
egress (March-May). Values measured at the Kamloops A (50 ° N, 120 ° W) weather
station (49° N, 119°W) (ECCC 2023). Standard error bars are present on the
historical 30-year mean (1991-2020). ..........................................................................12
Figure 1.5: Representative photos of the study site in the Lac du Bois Grasslands Protected
Area: A) Displays the south-western view of the site towards the development in
Kamloops. B) Shows the north-eastern view of the site towards further park. Photos
by author. .....................................................................................................................13
Figure 1.6: Representative microhabitats for the Great Basin Gophersnake, Western
Rattlesnake, and Western Yellow-bellied Racer in Kamloops, British Columbia,
Canada, including: A) Rodent burrows, B) Lower grassland, C) Rocky outcrops and
D) Talus slopes. Photos by author. .............................................................................14
Figure 1.7: Pipeline construction within the Lac du Bois Grasslands Protected Area,
Kamloops, British Columbia. A) Displays active excavation for pipeline placement.
B) Displays the landscape just after construction and seeding. Photos by author. ......16
Figure 2.1: Numbers of snakes and species composition at 24 hibernacula distributed across
the study landscape. Snakes were surveyed during the 2023 emergence period. If no
snakes were found at a hibernaculum during this survey period, the next highest
survey period from 2022, ingress or egress, was used to estimate the number of
snakes at that hibernaculum. Bars are subdivided according to the number of snakes
of each species .............................................................................................................35
Figure 2.2: The number of hibernacula used and shared by the Western Yellow-bellied Racer
“WYR”, Western Rattlesnake “WR”, and Great Basin Gophersnake “GBG” in Lac du
Bois Grasslands Protected Area. ..................................................................................38

xi
Figure 2.3: Examples of snake hibernacula within Lac du Bois Grassland Protected Area
(2021-2023). A) Found in the blast rock off the side of the road. B) Found under a
rock. C) Found in an animal burrow. D) Found in a rocky outcrop. Red arrows
identify the location of the hibernaculum entrance. Photos by author. .......................39
Figure 3.1: Schematic of (A) artificial hibernaculum 1 and (B) artificial hibernaculum 2. The
figures in the top row provide an overhead view and shows the spatial relationship of
the temperature standpipes in reference to each other and the hibernaculum mouth.
The figures in the bottom row provide side profiles. ‘Back’ refers to temperature
standpipes in the same row furthest from the hibernaculum entrance, with the front
row being those closest to the hibernaculum mouth. All figures were adapted from
Trans Mountain (2021b,c). Not drawn to scale ...........................................................58
Figure 3.2: SB-2T radio transmitter (Holohil Systems Ltd., Carp, ON, Canada) and
DS1925L-F5 iButton (Maxim Integrated Products Inc., San Jose, CA, United States)
modification. A) The transmitter prior to modification. B) The iButton prior to
modification. C) The iButton and transmitter after one coat of rubber Plasti Dip and
beeswax. D) The final product with another layer of rubber Plasti Dip (note that the
tape held a magnet on the battery and was removed before implantation). In all photos
the antenna has been cropped (actual length was 22cm). Photos by author. ...............60
Figure 3.3: Funnels traps used to catch snakes at artificial snake hibernaculum 1 (A) Spring
emergence (egress) orientation to catch snakes as they exit, (B) Fall return (ingress)
orientation to catch snakes as they return. Arrows indicate the direction of snake
movement. Not drawn to scale.....................................................................................62
Figure 3.4: The average daily winter temperature from November 1, 2021, to March 31,
2022, for (A) each position within artificial snake hibernaculum 1 (ASH1) and
artificial snake hibernaculum 2 (ASH2) and (B) the external ambient air. ‘Position’
refers to the distance of loggers to the hibernaculum mouth, where ‘Back’ is the
furthest and ‘Front’ is the closest (Figure 3.1). Each line is the average of the loggers
(n=3 for each position in ASH1, n=2 for each position in ASH2, and n=6 for
external). The cloud around the ambient line represents the daily maximum and
minimum. The solid vertical lines indicate a severe cold interval from December 22,
2021, to January 3, 2022, with the dashed vertical line showcasing the minimum
external temperature (December 27, 2021). ................................................................66
Figure 3.5: The average daily relative humidity from November 1, 2021, to March 31, 2022,
for (A) each position within artificial snake hibernaculum 1 (ASH1) and artificial
snake hibernaculum 2 (ASH2) and (B) the external ambient air. ‘Position’ refers to
the distance of loggers to the hibernaculum mouth, where ‘Back’ is the furthest and
‘Front’ is the closest (Figure 3.1). Each line is the average of the loggers (n=3 for
each position in ASH1, n=2 for each position in ASH2). Ambient relative humidity
data was obtained from the Kamloops A (50 ° N, 120 ° W) weather station (ECCC
2023). The cloud around the ambient line represents the daily maximum and
minimum. .....................................................................................................................67

xii
Figure 3.6: The average daily winter temperature from November 1, 2022, to March 31, 2023
of (A) Great Basin Gophersnakes hibernating in natural hibernacula superimposed on
the temperature profiles of each position within artificial snake hibernaculum 1
(ASH1) and artificial snake hibernaculum 2 (ASH2) and (B) the external ambient air.
‘Position’ refers to the distance of loggers to the hibernaculum mouth, where ‘Back’
is the furthest and ‘Front’ is the closest (Figure 3.1). Each line is the average of the
loggers (2 from the front of ASH2, 3 from the middle of ASH1, 1 from the middle of
ASH2, 2 from the front of each hibernaculum, n=4 snakes, and 8 ambient temperature
loggers). The grey clouds represent the daily maximum and minimum for snakes and
ambient.........................................................................................................................70

xiii

LIST OF TABLES
Table 1.1: The salvage summary for snakes retrieved from three excavated hibernacula in the
fall of 2020 (Heinrich 2020). Snakes were either released at natural hibernacula,
moved to the BC Wildlife Park for artificial overwintering, or died ...........................17
Table 2.1: Habitat features measured around the entrances to hibernacula and random sites.
Habitat features were measured at the microscale (1m radius), macroscale (10m
radius), and within 30m of the site...............................................................................33
Table 2.2: Candidate models investigating the effect of habitat on the numbers of snakes
detected at hibernacula at a micro- and macroscale. Models for the Great Basin
Gophersnake, Western Rattlesnake, and Western Yellow-bellied Racer were ranked
based on the Akaike Information Criterion small sample equivalent (AICc) and
ΔAICc (difference of AICc of the best model (minimum AICc) to the model being
compared). This table also includes the models' weight (wi). The top models with ⃤
AICc value <2.0 (Burnham and Anderson 2002) are shown, along with the global
model (G) that contained all of the selected habitat variables. The relationship of
habitat variables in the top models is indicated with a superscript, + describes habitat
features associated with increased snakes, and - describes variables not associated
with them (Appendix A). Full names and detailed descriptions of explanatory
variables are found in Table 2.1 names and detailed descriptions of explanatory
variables are found in Table 1. ....................................................................................40
Table 2.3: Candidate models comparing habitat features of hibernacula and paired random
sites at the micro and macroscale. Models for the Great Basin Gophersnake, Western
Rattlesnake, and Western Yellow-bellied Racers were ranked based on Akaike
Information criterion (AIC) small sample equivalent (AICc) and ΔAICc (difference
of AICc of the best model (minimum AICc) to the model being compared). This table
also includes the models' weight (wi). The top models with ⃤ AICc value <2.0
(Burnham and Anderson 2002) are shown, along with the global model (G) that
contained all of the selected habitat variables. The relationship of habitat variables in
the top models is indicated with a superscript, where + describes habitat features
selected for, and - describe variables selected against (Appendix B). Full names and
detailed descriptions of explanatory variables are found in Table 2.1. The Western
Rattlesnake model at the microscale did not converge ................................................42

1

CHAPTER 1
INTRODUCTION TO THE THESIS
OVERWINTER SURVIVAL
Species shift in their response to external seasonal changes, leading to differing
behaviours, morphology, and physiology depending on the time of year (Nelson et al. 1990;
Lisovski et al. 2017). Such shifts are pronounced more for species in highly seasonal
environments, that increase in winter intensity towards the poles (Lisovski et al. 2017). These
latitudes are characterized by long winters and short summers (Ultsch 1989). Here, summers
provide access to increased light, temperature, and resource availability, resulting in
heightened animal activity (Williams et al. 2017). Conversely, winters are characterized by
shorter days, colder temperatures, and inclement weather, that limit the availability and
access of resources and ultimately reduce animal activity (Williams et al. 2017).
Winters are arguably one of the most stressful times for organisms, with conditions
constraining the northern range and resulting in the direct mortality of many individuals
(C.M. Williams et al. 2015). For example, White et al. (2011) found that mountain goat
(Oreamnos americanus) survival in coastal Alaska was influenced by snow conditions,
particularly in years with snow higher than the long-term median. In contrast, Shine and
Mason (2004) found that winters with lighter snow cover in central Manitoba led to major
freezing events of Red-sided Garter Snakes (Thamnophis sirtalis parietalis). Further,
Baccante and Woods (2010) revealed that higher mule deer (Odocoileus hemionus) fawn
survival was linked to years with milder winters within the Peace Region of British
Columbia. These strong selective forces, which in some regions persist for over half a year or
longer, necessitate animals to possess specialized adaptations to survive winter (C.M.
Williams et al. 2015).

OVERWINTERING IN REPTILES
As ectotherms, the body temperature of reptiles is determined largely by their
external environment (Voituron et al. 2002). Thus, their persistence at higher latitudes
requires specialized responses to survive through or avoid exposure to sub-zero temperatures
during winter (Ultsch 1989; Voituron et al. 2002). As winter temperatures generally are

2
below the normal thermal activity thresholds for reptiles, this leads to a period of dormancy
called brumation (Holden et al. 2021). This period is controlled by temperature exposure and
is a metabolic adaptation characterized by lower energy demands, metabolism, and heart rate,
with periods of activity (Hoekstra et al. 2020; Holden et al. 2021). Depending on exposure,
some reptiles may utilize further physiological adaptations to reduce their exposure to lethal
temperatures (i.e. freeze tolerance or freeze avoidance) (White and Lasiewski 1971; Ultsch
1989; Costanzo et al. 1995; Moss and MacLeod 2022). With these adaptations, ‘freeze
tolerators’ can endure some of their body being frozen, while ‘freeze avoiders’ can supercool
by depressing their freezing point (Zachariassen 1985; Storey and Storey 1988, 1996;
Voituron et al. 2002). For example, hatchling painted turtles that overwinter in shallow
subterranean nests (5-10cm) can survive to -4°C as a ‘freeze tolerator’ or to -15°C as a
‘freeze avoider’ (Costanzo and Lee 2013). These strategies vary across species and ranges
and largely depend on the risk of thermal exposure (Tsuji 1988; Moss and MacLeod 2022).
Many reptiles, however, behaviourally avoid lethal temperatures by utilizing a shelter, also
known as a den or hibernaculum (White and Lasiewski 1971; Ultsch 1989; Costanzo et al.
1995; Michels-Boyce and Zani 2015; Moss and MacLeod 2022).

SNAKE HIBERNACULA
Snakes inhabiting temperate zones must find adequate overwintering habitat to
circumvent reduced food availability and extreme and variable temperatures over winter
(Hein and Guyer 2009). Rather than creating new features, most snakes rely on subterranean
habitats that superficially appear as crevices in a rockface, holes in the ground, or piles of
rocks (Costanzo 1986; Prior and Weatherhead 1996; Gienger and Beck 2011; K.E. Williams
et al. 2015). Previous studies have showcased that snakes select distinct habitat features for
hibernacula compared to the broader landscape (Prior and Weatherhead 1996; Gienger and
Beck 2011). For example, Western Rattlesnake (Crotalus oreganus oreganus) hibernacula in
Washington State were found along low-grade south-facing talus slopes with intermediatesized rocks compared to random sites (Gienger and Beck 2011). Further, Black Rat
Snake (Pantherophis obsoletus) hibernacula within eastern Ontario were found on rocky
south-facing slopes, with snakes preferring large, partially dead or hollow trees for basking
during emergence (Prior and Weatherhead 1996).

3
Hibernacula must provide suitable conditions to ensure survival for upwards of half
the year at northern range limits (Brown et al. 1974; Prior and Weatherhead 1996; Gienger
and Beck 2011). These sites must be structurally stable to ensure snakes can safely hibernate
without disturbances (Markle et al. 2020). Hibernacula vulnerable to collapse, flooding, or
landslides can have catastrophic impacts on populations (Shine and Mason 2004; Gardiner
and Sonmor 2011; Maida et al. 2017). Further, such sites are especially needed within
northern latitudes to insulate against extreme and variable temperatures (Gregory 1982; Hein
and Guyer 2009). Intuitively, the temperature within hibernacula must be within a tolerable
range to ensure snakes do not freeze and are not expending energy reserves (Gregory 1982;
Costanzo 1989a; K.E. Williams et al. 2015; Markle et al. 2020) while also providing
sufficient humidity to prevent desiccation (Costanzo 1986). Hibernacula that meet these
requirements may be a limiting resource in some environments, forcing some individuals to
travel kilometres in order to return to the same denning location, presenting strong yearly
fidelity to those sites (Prior et al. 2001; Gienger and Beck 2011). In fact, some Western
Rattlesnake hibernacula may be used for hundreds of years (COSEWIC 2015a; Dyer et al.
2016b; Macartney 1985). The potential limitation of these sites is reinforced by increased
instances of communal denning at higher latitudes, as numerous snakes must use a single
hibernaculum (Gregory 1984, 2009).

COMMUNAL DENNING
Communal denning in snakes, pronounced in higher latitudes and altitudes of the
northern hemisphere, is a phenomenon where seemingly solitary individuals congregate to
overwinter (Gregory 1974, 1984, 2009). The epitome of this behaviour is the aggregations of
thousands of Red-sided Garter Snakes at hibernacula in the Interlake region of Manitoba
(Gregory 1973, 1977, 1984). This phenomenon, albeit not to the same degree, is also
observed in other snakes: Western Rattlesnakes (Gienger and Beck 2011), Western Yellowbellied Racers (Coluber constrictor mormon), Great Basin Gophersnakes (Pituophis catenifer
deserticola) (Parker and Brown 1973), Black Rat Snakes (Prior and Weatherhead 1996), etc.
Communal hibernacula also may support multispecies assemblages (COSEWIC 2013; K.E.
Williams et al. 2015).

4
Although widely observed, the question of why snakes hibernate communally is
largely unanswered. There are many potential explanations for why communal denning
exists. A leading theory is that the scarcity of adequate hibernacula results in increased
instances of communal denning (Bruckerhoff et al. 2021; Gregory 1984, 2009). Within
northern latitudes, harsh winter weather may exacerbate this phenomenon by reducing sites
that meet the requirements for survival (Gregory 1984). This phenomenon is quoted as being
“undoubtedly the main cause of communal overwintering” (Gregory 1984) and although
many studies reference this theory (Prior and Weatherhead 1996; Harvey and Weatherhead
2006; Gardiner and Sonmor 2011; Gardiner et al. 2013; Bruckerhoff et al. 2021), few have
tried to assess the availability of hibernacula (Prior and Weatherhead 1996; Harvey and
Weatherhead 2006; Williams et al. 2014; K.E. Williams et al. 2015).

ARTIFICIAL HIBERNACULA
Given the apparent scarcity of natural denning sites, their destruction likely has a
tremendous impact on snake populations. Such loss of viable denning sites can result from
anthropogenic and natural factors (Gardiner & Sonmor 2011; Maida et al. 2017). For
instance, a rockslide at a talus slope hibernaculum in Osoyoos, British Columbia, was
reported to have killed at least ten Western Rattlesnakes (Maida et al. 2017). Further, the
slumping of an earthen hibernaculum in Grasslands National Park was suspected of killing at
least 50% of the Eastern Yellow-bellied Racers (Colubar constrictor flaviventris) that
overwintered at this site (Gardiner and Sonmor 2011; Dyer et al. 2016b). Besides natural
disasters, habitat loss can result from anthropogenic development (Zappalorti et al. 2014;
COSEWIC 2015a; Golder Associates Ltd 2022). One proposed conservation or mitigation
measure to combat this loss is the construction of artificial hibernacula (Zappalorti and
Reinert 1994; Zappalorti et al. 2014).
Artificial hibernacula are anthropogenic features intentionally or incidentally created
(i.e. human-created slopes, construction debris, abandoned mine features, etc.) (Costanzo
1986; Ravesi et al. 2015; K.E. Williams et al. 2015; Bruckerhoff et al. 2021; Choquette et al.
2024). The former generally are used for research, conservation, or mitigation (Gillingham
and Carpenter 1978; Zappalorti and Reinert 1994; Todd et al. 2009; Whiting and Booth
2012; Bryan 2015; Choquette et al. 2024). Previous research on artificial hibernacula

5
suggests this approach has merit. For example, Zappalorti and Reinert (1994) found that 68%
of overwintering refugia installed along sewer and electric line right-of-ways in New Jersey
were used by nine snake species at the end of their 3-year study. Another study found that the
occupancy rate of artificial hibernacula was similar to natural counterparts, however, natural
hibernacula were mainly selected by pine snakes (Pituophis m. melanoleucus) in the New
Jersey Pine Barrens (Zappalorti et al. 2014). More detailed research on incidental and
experimental artificial hibernacula indicates the potential of these sites to provide optimal
microclimatic conditions overwinter (Costanzo 1986; Choquette et al. 2024). However, little
published research has examined artificial hibernacula built as a mitigation response to
human-caused disturbance, an area of study that will only become more pertinent as habitat
loss continues (Bruckerhoff et al. 2021). With the extreme winters of northern latitudes, this
importance increases further due to the potentially limited availability of thermally-secure
hibernacula.

STUDY ANIMALS
My three study species were the Great Basin Gophersnake, Western Yellow-bellied
Racer, and Western Rattlesnake (Figure 1.1). As of 2024, each species is nationally listed
under the Species at Risk Act as threatened (COSEWIC 2013, 2015a, b) and provincially
blue-listed within British Columbia (BCCDC 2008, 2010, 2013). Threats to these species
include road mortality, direct or indirect persecution, agriculture, and habitat loss
(COSEWIC 2013, 2015a, b).
The largest snake in British Columbia, the Great Basin Gophersnake (Family
Colubridae), is an oviparous constrictor that preys upon small mammals and birds
(Shewchuk 1996; White 2008). The Western Yellow-bellied Racer (also Family Colubridae),
is a medium-sized oviparous snake with a long yet slender body, aptly named “racer” for its
fast movements (COSEWIC 2015b). This snake preys upon invertebrates and small animals
and is suspected of feeding primarily on grasshoppers (Sarell 2004a; Hall and Cavitt 2011).
Compared to the aforementioned two species, the northern populations of Western
Rattlesnake (Family Viperidae) have arguably garnered the most research (Macartney et al.
1987; Gomez et al. 2007; Lomas et al. 2015; Harvey and Larsen 2020; Schmidt et al. 2020;

6
Howarth et al. 2023, etc.). This snake is viviparous and generally preys upon rodent species
(COSEWIC 2015a).
All three species meet their northern extent in the southern interior of British
Columbia (COSEWIC 2013, 2015a,b) and occur coincidently over most of this range
(Matsuda et al. 2006). Within this range, these species can be found within shrub-steppe,
grassland, and rocky areas, utilizing microhabitats such as shrubs, rocks, rodent burrows,
talus slopes, and rocky outcrops (Shewchuk 1996; Bertram et al. 2001; Sarell 2004a; White
2008; COSEWIC 2015a,b). These species are active on this landscape from March to
October, following their emergence from hibernacula (egress, March to May) (pers. observ.).
During the active season, gophersnakes and racers mate in May after leaving hibernacula
(Shewchuk 1996; Sarell 2004a; White 2008; COSEWIC 2013), whereas rattlesnakes mate in
late summer (Sarell 2004b). Following the active season, these snakes return to hibernacula
(ingress) from August to October and overwinter from roughly October to March (pers.
observ.).
Western Rattlesnake hibernacula are well documented within British Columbia, with
over 380 known hibernacula (COSEWIC 2015a). They are known to communally den within
rocky outcrops and talus slopes (Macartney 1985; Macartney and Gregory 1988; Bertram et
al. 2001; Gienger and Beck 2011), while gophersnakes and racers overwinter in deep talus,
animal burrows and rock outcrops both communally or solitary (Haney and Sarell 2005;
COSEWIC 2015b). All three species also have been known to share hibernacula (COSEWIC
2013; K.E. Williams et al. 2015). Rattlesnakes and racers display fidelity to their
overwintering sites (Brown and Parker 1976; Macartney 1985; COSEWIC 2015b), however,
previous studies on gophersnakes have shown variation across regions for site fidelity within
the Okanagan (K.E. Williams et al. 2015). Through “Wildlife Habitat Areas” (WHAs), the
communal hibernacula of all three of these species can be protected in British Columbia
(COSEWIC 2015b; Dyer et al. 2016a,b). Typically, WHAs are small, protected areas around
sensitive habitats for species at-risk: for communal snake hibernacula, WHAs extend 200300 ha around hibernacula (MWLAP 2004; Williams et al. 2012).

7

A

B

B

C

Figure 1.1: The three study species: A) Great Basin Gophersnake (Pituophis catenifer
deserticola), B) Western Yellow-bellied Racer (Coluber constrictor mormon), and C)
Western Rattlesnake (Crotalus oreganus oreganus). Photos by author.

8

STUDY SITE
My primary study area lied within the Lac du Bois Grasslands Protected Area (51° N,
120.4° W) on the north-western side of the city of Kamloops, British Columbia, Canada
(Figure 1.2) (GCC 2009; BC Parks 2017). This extensive grassland encompasses
approximately 15,712 ha within the Thompson-Okanagan Ecoregion, supporting a variety of
resident and migratory species, including at least 17 considered at risk, three of which are my
study animals (BC Parks 2017). This ecoregion is known for having some of the hottest and
driest climates within British Columbia (BC Parks 2017). See Figures 1.3 and 1.4 for
representative average monthly temperature and precipitation during our study compared to a
30-year average (1991-2020). During my study period’s winters, mean monthly temperatures
during December 2021 and 2022 dropped to as low as -30°C, lower than the 30-year norm
(Figure 1.3) (ECCC 2023).
Although the second largest protected grassland in British Columbia, Lac du Bois is
isolated from further grassland habitats to the south and east (BC Parks 2017). This isolation
is, in part, the result of residential and commercial development that borders the park (Figure
1.5). Due to the proximity, Lac du Bois offers easily accessible recreational activities to the
over 98,000 residents of Kamloops in the form of biking, hiking, camping, hunting, fishing,
off-roading, and horseback riding (BC Parks 2017). This site also is used for cow grazing,
divided into two primary pastures that undergo grazing in a rotating three-year schedule
(GCC 2009). In addition to these activities, this site is heavily fragmented by trails, offroading tracks, barbed wire fencing, access roads, and, more recently, a pipeline right of way.
The study site is roughly 1100ha at the southeastern corner of the park (Figure 1.2)
(BC Parks 2017), primarily comprised of lower and middle grassland dominated by Big
Sagebrush (Artemisia tridentata), Bluebunch Wheatgrass (Pseudoroegneria spicata), and
spring-blooming plants (GCC 2009; BC Parks 2017). Other habitats include upper
grasslands, dry forests, rocky outcrops, talus slopes, ponds, wetlands, and lakes (BC Parks,
2017). This landscape affords the snakes environmental features important to their life
histories (i.e., hibernacula, basking sites, refuges, egg-laying sites, shedding sites, and
rookeries) (Figure 1.6).

9
Prior to my study, research on snakes within Lac du Bois had been restricted to only a
few studies. Bertram et al. (2001) conducted a radiotracking study on gophersnakes and
rattlesnakes, while Gomez (2007) focused on tracking rattlesnakes exclusively, both aiming
to characterize habitat utilization and movement patterns within the area. Further, Hobbs
(2007) studied the overwintering thermal conditions of rattlesnakes. From this past work, in
addition to the present study, a total of 28 hibernacula have been identified in the study site
within a variety of habitats (i.e., rocky outcrops, rodent burrows, holes under rocks, and blast
rocks on the side of roads).

Figure 1.2: Location of the study site in the Lac du Bois Grasslands Protected Area, ~10 km from the City of Kamloops, BC
(Google Earth 2023; GBC 2023). Also shown is the Trans Mountain Expansion Project (Pipeline).
10

Figure 1.3: Mean monthly temperatures (°C) in Kamloops, British Columbia, from 2020-2023, compared to the historical 30-year
mean (1991-2020). Data presented in order of ingress (August to October), the overwintering period (October-March), and egress
(March-May). Values measured at the Kamloops A (50 ° N, 120 ° W) weather station (49° N, 119°W) (ECCC 2023). Standard
error bars are present on the historical 30-year mean (1991-2020).
11

Figure 1.4: Total precipitation (mm) in Kamloops, British Columbia, during the study years (2020-2023) compared to the
historical 30-year mean (1991-2020). Data presented in order of ingress (August to October), the overwintering period (OctoberMarch), and egress (March-May). Values measured at the Kamloops A (50 ° N, 120 ° W) weather station (49° N, 119°W) (ECCC
2023). Standard error bars are present on the historical 30-year mean (1991-2020).
12

13

A

B

Figure 1.5: Representative photos of the study site in the Lac du Bois Grasslands Protected
Area: A) Displays the south-western view of the site towards the development in Kamloops.
B) Shows the north-eastern view of the site towards further park. Photos by author.

14

Figure 1.6: Representative microhabitats for the Great Basin Gophersnake, Western
Rattlesnake, and Western Yellow-bellied Racer in Kamloops, British Columbia, Canada,
including: A) Rodent burrows, B) Lower grassland, C) Rocky outcrops and D) Talus slopes.
Photos by author.

15

DISTURBANCE EVENT
The Trans Mountain Expansion Project (TMEP) was initiated to increase the
transportation of crude oil by twinning an existing pipeline built in 1951, running from
Edmonton, Alberta, to Burnaby, British Columbia (GC 2022; Trans Mountain 2023a,b). This
project commenced construction within Kamloops in June 2020 (Trans Mountain 2021c).
Within Lac du Bois, the TMEP pipeline right-of-way disturbance stretched 9.2 km and was,
on average, 30 m wide during construction (Figure 1.2, 1.7) (Trans Mountain Pipeline ULC
2017).
Before construction, TMEP was obligated by the Ministry of Forests, Lands, and
Natural Resource Operations to obtain required provincial permits and conduct a snake
inventory on the site (Heinrich 2020). From September 21, 2020, to October 16, 2020,
habitat assessments, reptile sweeps, and snake salvages were performed (Heinrich 2020).
During this timeframe, three communal snake hibernacula were excavated during
construction along the pipeline right of way within a 1.6 km straight-line distance (Heinrich
2020). The snakes salvaged from this disturbance included 51 gophersnakes, 71 racers, and
10 rattlesnakes (Table 1.1) (Heinrich 2020). Snakes retrieved before October 15 (10
rattlesnakes, 12 gophersnakes, and 13 racers) were moved to three natural hibernacula in the
area and were not marked for future identification as it is not a regulatory requirement (Table
1.1) (Heinrich 2020), a situation that later on complicated my work. Snakes recovered after
October 15 (including 37 gophersnakes and 56 racers) were transferred to the BC Wildlife
Park for artificial overwintering and were implanted with Passive Integrated Transponders
(PIT tags) to permit individual identification (Table 1.1) (Heinrich 2020).
Two artificial hibernacula were constructed between April and May 2021 to replace
the lost natural sites. The first artificial snake hibernacula (ASH1) was completed in late
April 2021, and all snakes artificially overwintered were released here between April 31 and
May 3, 2021 regardless of their point of origin. This release strategy was necessary as the
second artificial snake hibernacula (ASH2) remained incomplete until after the natural
emergence period for snakes in the region.

Figure 1.7: Pipeline construction within the Lac du Bois Grasslands Protected Area, Kamloops, British Columbia. A) Displays
active excavation for pipeline placement. B) Displays the landscape just after construction and seeding. Photos by author.

16

17

Table 1.1: The salvage summary for snakes retrieved from three excavated hibernacula
between September 21, 2020, to October 16, 2020 (Heinrich 2020). Snakes were either
released at natural hibernacula, moved to the BC Wildlife Park for artificial overwintering, or
died.
Hibernaculum
1

2

3

Species
Western Rattlesnake
Great Basin Gophersnake
Western Yellow-bellied Racer
Western Rattlesnake
Great Basin Gophersnake
Western Yellow-bellied Racer
Western Rattlesnake
Great Basin Gophersnake
Western Yellow-bellied Racer

Released
10
10
13
0
1
0
0
1
0

BC Wildlife Park
0
20
44
0
5
1
0
12
11

Died
0
0
2
0
1
0
0
1
0

18

THESIS OBJECTIVES
Given the significance of snake hibernacula and their apparent scarcity, it is crucial to
properly protect and avoid disturbance of these critical habitat features. Understanding the
habitat selection of a species or multiple species utilizing hibernacula in a given community
may help elucidate where to protect these sites. If disturbance cannot be avoided, ensuring
the efficacy of artificial hibernacula used as mitigation is crucial. In this thesis, I use the
community of three sympatric snake species, Great Basin Gophersnake, Western Yellowbellied Racer, and Western Rattlesnake, following the outlined disturbance event to pursue
the following objectives:
In Chapter 2, I explore the variation in hibernacula habitat use between the three
snake species by collecting habitat features at documented hibernacula and analyzing data
using a match case-control design. My objectives in this chapter were to:
a. Document hibernacula and their species assemblages on our study site.
b.

Investigate how the habitat features relate to the degree of communal occupation present
at hibernacula.

c. Explore how these three species vary in their selection of overwintering habitat.
In Chapter 3, I studied the efficacy of the newly built artificial hibernacula following
the disturbance event. To do so, I collected the internal temperature and humidity of two
artificial hibernacula and the internal temperatures of snakes via remote data loggers. My
objectives in this chapter were to:
d. Highlight the available microclimate within each hibernaculum and assess how well they
buffer against external environmental conditions.
e. Determine whether the internal thermal conditions of the artificial hibernacula are
comparable to those naturally experienced by snakes.
f. Monitor the adoption of the main artificial hibernaculum by the original cohort of
displaced snakes.
In the final Chapter (Ch. 4), I revisit the key findings of the previous chapters and
contextualize them within the realm of conservation, biology, management and the need for
future research.

19

LITERATURE CITED
Baccante D, Woods R. 2010. Relationship between winter severity and survival of mule deer
fawns in the Peace Region of British Columbia. BC JEM. 10(3):145-153.
[BCCDC] B.C. Conservation Data Centre. 2008. Species summary: Pituophis catenifer
deserticola. BC Ministry of Environment; [accessed 2024 Feb 2].
https://a100.gov.bc.ca/pub/eswp/
[BCCDC] B.C. Conservation Data Centre. 2010. Species summary: Coluber constrictor. BC
Ministry of Environment; [accessed 2024 Feb 2]. https://a100.gov.bc.ca/pub/eswp/.
[BCCDC] B.C. Conservation Data Centre. 2013. Species summary: Crotalus oreganus. BC
Ministry of Environment; [accessed 2024 Feb 2]. https://a100.gov.bc.ca/pub/eswp/.
BC Parks. 2017. Lac du Bois Grasslands Protected Area management plan. Kamloops.
Bertram N, Larsen KW, Surgenor J. 2001. Identification of critical habitats and conservation
issues for the Western Rattlesnake and Great Basin Gopher Snake within the
Thompson-Nicola region of British Columbia. Kamloops.
Brown WS, Parker WS, Elder JA. 1974. Thermal and spatial relationships of two species
of Colubrid snakes during hibernation. Herpetol. 30(1):32–38.
Brown WS, Parker WS. 1976. Movement Ecology of Coluber constrictor near communal
hibernacula. Copeia. 225-42.
Bruckerhoff LA, Kamees LK, Holycross AT, Painter CW. 2021. Patterns of survival of a
communally overwintering rattlesnake using an artificial hibernaculum. Ichthyol
Herpetol. 109(1):64–74.
Bryan DL. 2015.Survival and natural history of the Timber Rattlesnake (Crotalus Horridus)
at Center Hill Lake in Dekalb County, Tennessee [thesis]. [Cookville (TN)]:
Tennessee Technolgical University.
Choquette JD, Mokdad AI, Pitcher TE, Litzgus JD. 2024. Selection and validation of release
sites for conservation translocations of temperate-zone snakes. MethodsX. 12:1-11
COSEWIC. 2013. COSEWIC assessment and status report on the Great Basin Gopersnake
Pituophis catenifer deserticola in Canada. Ottawa.
COSEWIC. 2015a. COSEWIC assessment and status report on the Western Rattlesnake
Crotalus oreganus in Canada. Ottawa.
COSEWIC. 2015b. COSEWIC assessment and status report on the Eastern Yellow-bellied
Racer Coluber constrictor flaviventris and the Western Yellow-bellied Racer Coluber
constrictor mormon in Canada. Ottawa.
Costanzo JP. 1986. Influences of hibernaculum microenvironment on the winter life history
of the garter snake (Thamnophis sirtalis). OJS. 86(5):199–204.
Costanzo JP. 1989. Effects of humidity, temperature, and submergence behavior on
survivorship and energy use in hibernating garter snakes, Thamnophis sirtalis. Can J
Zool. 67:2486–2492.

20
Costanzo JP, Lee RE, Dcvries AL, Wang T, Layne JR. 1995. Survival mechanisms of
vertebrate ectotherms at subfreezing temperatures: applications in cryomedicine.
FASEB J. 9(5):295–459.
Costanzo JP, Lee RE. 2013. Avoidance and tolerance of freezing in ectothermic vertebrates.
J Exp Biol. 216(11):1961–1967.
Dyer O, Bishop C, Govindarajulu P, Anderson L, Cunnington D, Hobbs J, Huntley M,
Larsen K, Reudink R, Sadler K, Sarell M, Steciw J, et al. 2016a. British Columbia
recovery strategy series recovery plan for the Gopher Snake, deserticola subspecies
(Pituophis catenifer deserticola) in British Columbia.
Dyer O, Bishop C, Govindarajulu P, Anderson L, Cunnington D, Hobbs J, Huntley M,
Larsen K, Reudink R, Sadler K, Sarell M, Tedesco L, et al. 2016b. British Columbia
recovery strategy series recovery plan for the Western Rattlesnake (Crotalus
oreganus) in British Columbia.
[ECCC] Environment and Climate Change Canada. 2023. Historical climate data; [accessed
2023 Jul 18]. http://climate.weather.gc.ca/.
Gardiner LE, Sonmor KW. 2011. Reptiles major slump event at grasslands national park
snake pit in southwestern Saskatchewan. Blue Jay. 69(3):120–124.
[GCC] The Grasslands Conservation Council of British Columbia. 2009. Working together
for the conservation of BC’s grasslands. Kamloops.
Gienger CM, Beck DD. 2011. Northern Pacific Rattlesnakes (Crotalus oreganus) use thermal
and structural cues to choose overwintering hibernacula. Can J Zool. 89(11):1084–
1090.
Gillingham JC, Carpenter CC. 1978. Snake hibernation: construction of and observations
on a man-made hibernaculum (Reptilia, Serpentes). J Herpetol. 12(4):495–498.
Golder Associates Ltd. 2022. Site C Wildlife Mitigation Structures Snake Hibernacula
Monitoring 2021. Nanaimo.
Gomez LM. Habitat use and movement patterns of the Northern Pacific Rattlesnake
(Crotalus o. oreganus) in British Columbia [thesis]. [Victoria (BC)]: University of
Victoria.
Google Earth. 2023. Google Earth; [accessed 2024 Mar 2].
https://www.google.com/earth/index.html.
[GBC] Government of British Columbia. 2023. B.C. Conservation data centre; [accessed
2023 Jul 19]. https://www2.gov.bc.ca/gov/content/environment/plants-animalsecosystems/conservation-data-centre.
[GC] Government of Canada. 2022. What is the Trans Mountain Expansion Project?;
[accessed 2023 Nov 9]. https://www.canada.ca/en/campaign/trans-mountain/what-istmx.html.
Gregory PT. 1973. Life history parameters of a population of Red-Sided Garter Snakes
(Thamnophis sirtalis parietalis) adapted to a rigorous and fluctuating environment
[thesis]. [Winnipeg (MB)]: University of Manitoba.

21
Gregory PT. 1974. Patterns of spring emergence of the Red-sided Garter Snake (Thamnophis
sirtalis parietalis) in the Interlake region of Manitoba. Can J Zool. 52:1063–1069.
Gregory PT. 1977. Life-history parameters of the Red-sided Garter Snake (Thamnophis
Sirtalis Parietalis) in an extreme environment, the Interlake Region of Manitoba.
Ottawa (ON): National Museums of Canada.
Gregory P. 1982. Reptillian hibernation. In: Gans C, Pough FH, editors. Biology of the
Reptillia. Vol. 13. New York (USA): Academic Press. p. 53–154.
Gregory PT. 1984. Communal denning in snakes. In: Seigel R, Hunt L, Knight J, Malaret L,
Zuschlag N, editors. Vertebrate Ecology and Systematic A Tribute to Henry S. Fitch.
Lawrence: University of Kansas Publications. p. 57-76.
Gregory PT. 2009. Northern lights and seasonal sex: the reproductive ecology of coolclimate snakes. Herpetol. 65(1):1–13.
Hall LK, Cavitt J. 2011. Coluber constrictor mormon (Western Yellow-bellied Racer) and
Pituophis catenifer deserticola (Great Basin Gopher Snake). Herpetol Rev. 42(2).
Haney A, Sarell M. 2005. Conservation analysis for the Great Basin Gophersnake (Pituophis
catenifer deserticola) in British Columbia.
Harvey JA, Larsen KW. 2020. Rattlesnake migrations and the implications of thermal
landscapes. Mov Ecol. 8(1).
Hein AM, Guyer C. 2009. Body temperature of overwintering Cottonmouth snakes:
hibernaculum use and inter-individual variation. J Ala Acad Sci. 80(1): 35-44.
Heinrich R. 2020. Technical memorandum. Kamloops (BC): Triton Environmental
Consultants.
Hobbs J. 2007. Thermal factors in relation to the denning ecology of northern Pacific
rattlesnakes in British Columbia [thesis]. [Victoria (BC)]: Royal Roads University.
Hoekstra LA, Schwartz TS, Sparkman AM, Miller DAW, Bronikowski AM. 2020. The
untapped potential of reptile biodiversity for understanding how and why animals
age. Funct Ecol. 34(1):38–54.
Holden KG, Gangloff EJ, Gomez-Mancillas E, Hagerty K, Bronikowski AM. 2021.
Surviving winter: physiological regulation of energy balance in a temperate ectotherm
entering and exiting brumation. Gen Comp Endocrinol. 307: 1-10.
Howarth, Bishop CA, Larsen KW. 2023. Western Rattlesnake (Crotalus oreganus) spring
migration in British Columbia: a comparative study of juveniles and adults. Can J
Zool. 101(7): 530-540.
Lisovski S, Ramenofsky M, Wingfield JC. 2017. Defining the degree of seasonality and its
significance for future research. Integr Comp Biol. 57(5):934–942.
Lomas E, Larsen KW, Bishop CA. 2015. Persistence of Northern Pacific Rattlesnakes masks
the impact of human disturbance on weight and body condition. Anim Conserv.
18(6):548–556.
Macartney J. 1985. The ecology of the Northern Pacific Rattlesnake, Crotalus viridis
oreganus, in British Columbia [thesis]. [Victoria (BC)]: University of Victoria.

22
Macartney JM, Larsen KW, Gregory PT. 1987. Body temperatures and movements
of hibernating snakes (Crotalus and Thamnophis) and thermal gradients of natural
hibernacula. Can J Zool. 67:108–114.
Macartney JM, Gregory PT. 1988. Reproductive biology of female rattlesnakes (Crotalus
viridis) in British Columbia. Copeia. 5(1):47–57.
Maida J, Larsen K, Hooper C, Bishop C. 2017. Crotalus oreganus (Western Rattlesnake)
Mortality. Herpetol Rev. 48(2):447.
Markle CE, Moore PA, Waddington JM. 2020. Temporal variability of overwintering
conditions for a species-at-risk snake: Implications for climate change and habitat
management. Glob Ecol Conserv. 22: 1-9.
Matsuda BM, Green DM, Gregory PT. 2006. Amphibians and reptiles of British Columbia.
Victoria: Royal BC Museum.
Michels-Boyce M, Zani PA. 2015. Lack of supercooling evolution related to winter severity
in a lizard. J Therm Biol. 53:72–79.
[MWLAP] Ministry of Water Land and Air Protection. 2004. Identified wildlife
management strategy accounts and measures for managing identified wildlife
southern interior forest region.
Moss JB, MacLeod KJ. 2022. A quantitative synthesis of and predictive framework for
studying winter warming effects in reptiles. Oecologia. 200(1–2):259–271.
Nelson RJ, Badura LL, Goldman BD. 1990. Mechanisms of seasonal cycles of behavior.
Annu Rev Psychol. 41(1):81–108.
Parker WS, Brown WS. 1973. Species composition and population changes in two
complexes of snake hibernacula in northern Utah. Herpetol. 29(4):319–326.
Prior KA, Blouin-Demers G, Weatherhead PJ. 2001. Sampling biases in demographic
analyses of Black Rat Snakes (Elaphe obsoleta). Herpetol. 57(4):460–469.
Prior KA, Weatherhead PJ. 1996. Habitat features of Black Rat Snake hibernacula in
Ontario. J Herpetol. 30(2):211–218.
Ravesi M, Tetzlaff SJ, Kingsbury B. 2015. Sistrurus catenatus (Massasauga) overwintering
Site. Herpet Rev. 46(3):454.
Sarell M. 2004a. Racer. Victoria: Ministry of Water, Land and Air Protection; [accessed
2024 Mar 2]. https://www.env.gov.bc.ca/wld/frpa/iwms/accounts.html
Sarell M. 2004b. Western Rattlesnake. Victoria: Ministry of Water, Land and Air Protection;
[accessed 2024 Mar 2]. https://www.env.gov.bc.ca/wld/frpa/iwms/accounts.html
Schmidt DA, Govindarajulu P, Larsen KW, Russello MA. 2020. Genotyping-in-thousands
by sequencing reveals marked population structure in Western Rattlesnakes to inform
conservation status. Ecol Evol. 10(14):7157–7172.
Shewchuk CH. 1996. The natural history of reproduction and movement patterns in the
Gopher Snake (Pituophis melanoleucus) in southern British Columbia [thesis].
[Victoria (BC)]: University of Victoria.

23
Shine R, Mason RT. 2004. Patterns of mortality in a cold-climate population of garter snakes
(Thamnophis sirtalis parietalis). Biol Conserv. 120(2):201–210.
Storey KB, Storey JM. 1988. Freeze tolerance in animals. Physiol Rev. 68(1):27–84.
Storey KB, Storey JM. 1996. Natural freezing survival in animals. Annu Rev Ecol Syst.
27:365–386.
Todd J, Amiel J, Wassersug R. 2009. Factors influencing the emergence of a northern
population of Eastern Ribbon Snakes (Thamnophis sauritus) from artificial
hibernacula. Can J Zool.
Trans Mountain. 2021. Time flies: one year of construction in the Kamloops urban area;
[accessed 2024 Feb 3].
https://www.transmountain.com/news/2021/time-flies-one-year-of-construction-inthe-kamloops-urbanarea#:~:text=What%20a%20year%20it%20has,the%20Trans%20Mountain%20Expa
nsion%20Project.
Trans Mountain. 2023a. Expansion project. [accessed 2024 Mar 24].
https://www.transmountain.com/project-overview.
Trans Mountain. 2023b. Our history. [accessed 2024 Mar 24].
https://www.transmountain.com/history.
Trans Mountain Pipeline ULC. 2017. Lac du Bois Grasslands Protected Area reclamation
plan for the Trans Mountain pipeline ULC Trans Mountain Expansion Project.
Calgary (AB): Kinder Morgan Canada Inc.
Tsuji JS. 1988. Thermal acclimation of metabolism in sceloporus lizards from different
latitudes. Physiol Zool. 3:241–253.
Ultsch GR. 1989. Ecology and physiology of hibernation and overwintering among
freshwater fishes, turtles, and snakes. Biol Rev. 64:435–516.
Voituron Y, Mouquet N, De Mazancourt C, Clobert J. 2002. To freeze or not to freeze? An
evolutionary perspective on the cold-hardiness strategies of overwintering
ectotherms. Am Nat. 160(2): 255-270.
White FN, Lasiewski RC. 1971. Rattlesnake denning: theoretical considerations on winter
temperatures. J Theor Biol. 30:553–557.
White KE. 2008. Spatial ecology and life history of the Great Basin Gophersnakes (Pituophis
catenifer deserticola) in British Columbia’s Okanagan Valley [thesis]. [Kelowna
(BC)]: University of British Columbia.
White KS, Pendleton GW, Crowley D, Griese HJ, Hundertmark KJ, McDonough T, Nichols
L, Robus M, Smith CA, Schoen JW. 2011. Mountain goat survival in coastal Alaska:
Effects of age, sex, and climate. J Wildl Manag. 75(8):1731–1744.
Whiting C, Booth HJ. 2012. Adder Vipera berus hibernacula construction as part of a
mitigation scheme for flood defence works, Horsey Corner, Norfolk, England. Conser
Evid. 9: 9-16.

24
Williams CM, Henry HAL, Sinclair BJ. 2015. Cold truths: how winter drives responses of
terrestrial organisms to climate change. Biol Rev. 90(1):214–235.
Williams KE, Hodges KE, Bishop CA. 2012. Small reserves around hibernation sites may
not adequately protect mobile snakes: the example of great basin gophersnakes
(Pituophis catenifer deserticola) in British Columbia. Can J Zool. 90(3):304-312.
Williams KE, Hodges KE, Bishop CA. 2015. Hibernation and oviposition sites of Great
Basin Gophersnakes (Pituophis catenifer deserticola) near their northern range Limit.
J Herpetol. 49(2):207-216.
Williams CM, Ragland GJ, Betini G, Buckley LB, Cheviron ZA, Donohue K, Hereford J,
Humphries MM, Lisovski S, Marshall KE, et al. 2017. Understanding evolutionary
impacts of seasonality: an introduction to the symposium. Integr Comp Biol.
57(5):921–933.
Zachariassen KE. 1985. Physiological reviews physiology of cold tolerance in insects.
APS. 65(4):799-832.
Zappalorti RT, Reinert H. 1994. Artificial refugia as a habitat improvement strategy for
snake conservation. Contributions to Herpetology. 11:369–375.
Zappalorti RT, Burger J, Burkett DW, Schneider DW, McCort MP, Golden DM. 2014.
Fidelity of northern pine snakes (Pituophis m. melanoleucus) to natural and artificial
hibernation sites in the New Jersey Pine Barrens. J. Toxicol. Environ. Health - A:
Curr. Issues. 77(21):1285–1291.

25

CHAPTER 2
AN INTRA- AND INTERSPECIFIC APPROACH TO IDENTIFYING AND PROTECTING
CRITICAL OVERWINTERING HABITAT FOR SNAKES
INTRODUCTION
By definition, designated critical habitat represents those features on the landscape
essential for the survival, recovery, persistence, and conservation of species-at-risk (Murphy
and Noon 1991; Hall et al. 1997; Camaclang et al. 2015; GC 2023). In turn, determining
where and why critical habitat features exist allows for a deeper understanding of the habitat
requirements of these species (Mayor et al. 2009; Boyce et al. 2016; Northrup et al. 2022).
These habitats may be broadly distributed over space and time or tied to specific spatial
features or temporal periods (Mayor et al. 2009), such as migration, dispersal, reproduction,
or overwintering (Krausman 1999; Goldberg et al. 2020).
Critical habitat for northern wildlife may include distinct features that enable survival
through winter (Burger et al. 1988; Prior and Weatherhead 1996; Gienger and Beck 2011;
Goldberg et al. 2020). This period of time is particularly stressful for animals as it reduces
the availability of resources and potentially exposes them to thermal extremes, thus placing
emphasis on the selection of habitat (Huey 1991; Krausman 1999; Mayor et al. 2009;
Dubiner et al. 2023). Species may respond to winter with migration (Winger et al. 2019),
morphological or physiological changes (Zimova et al. 2018), hibernation (Prior and
Weatherhead 1996; Gienger and Beck 2011; Goldberg et al. 2020), or a combination thereof
(Goldberg et al. 2020; Howarth et al. 2023). In particular, animals hibernating may require
distinct habitats to meet their differing requirements over winter, such as a specific structure,
temperature, or humidity (Brown et al. 1974; Prior and Weatherhead 1996; Hein and Guyer
2009; Goldberg et al. 2020). Due to the use of spatially and temporally distinct critical
habitats, such sites warrant the need for identification to conserve species in colder climates
outside their active season.
The use of hibernacula as a habitat feature that enables overwinter survival in
northern snakes has received considerable attention, in part because they warrant designation
as ‘critical habitat’ but also because large communal hibernacula provide valuable research
opportunities to study behaviour, demography, and habitat selection (Shine et al. 2001; Shine

26
and Mason 2004; Gienger and Beck 2011). As a result, several well-studied systems [ex.
Red-sided garter snake (Thamnophis sirtalis parietalis - Gregory 1974; Shine et al. 2001;
Shine and Mason 2004; Gregory 2011), Western Rattlesnake (Crotalus oreganus oreganus Lomas et al. 2015; Maida et al. 2018; Howarth et al. 2023)] have become iconic and almost
synonymous with the phenomenon of ‘communal denning’ in snakes, even serving as tourist
destinations (Manitoba 2024). However, not all northern snakes are associated with large
communal hibernacula; they instead may be found denning alone or in low numbers (Parker
and Brown 1973; Williams et al. 2015). Due to this variation, it is important to consider how
species shift in their habitat selection when the degree of communal occupation changes. For
our study, we define communal denning as an aggregation of individuals of the same species,
where the degree of communal occupation would refer to how many individuals of a given
species use a site. By way of example, Williams et al. (2014; 2015) assessed eight communal
and 25 solitary hibernacula for the Great Basin Gophersnake (Pituophis catenifer
deserticola) in western Canada, finding that rocky areas were used within the southern part
of the study, where one-third of snakes were denning communally; in the north, the snakes
tended to hibernate in grassy hillsides, with only one-fifth denning communally (Williams et
al. 2015). No research, save for the aforementioned study, has provided a comparison of
habitat as it relates to communal occupation of snakes at hibernacula. Communal denning
makes populations vulnerable to mass mortality events through persecution, natural disasters,
and habitat loss (Shine and Mason 2004; Gardiner and Sonmor 2011; COSEWIC 2015a;
Maida et al. 2017). Therefore, understanding the factors influencing the occurrence of
species-specific communal denning is essential for the conservation of these critical habitats.
No studies have explored both the common selection and species-specific preferences
of denning habitats for sympatric snake species that, in some cases, share the same
overwintering sites. Many studies have looked at hibernacula habitat characteristics for single
snake species (Prior and Weatherhead 1996; Harvey and Weatherhead 2006), such as the
Great Basin Gophersnake (Williams et al. 2015) and Western Rattlesnake (Gienger and Beck
2011). However, some sites have interspecific importance, with multiple species known to
share hibernacula (Parker and Brown 1973; Prior and Weatherhead 1996; COSEWIC 2013;
Williams et al. 2015). This provides opportune conditions for studying how and why
hibernacula selection varies between sympatric species, particularly within an environment

27
where natural selection should be relatively strong for the use of effective hibernacula.
Variation between species will highlight unique qualities for species-specific management
(including monitoring and inventory) and protection of hibernacula. Finally, through
comparison, this approach has the potential to provide further context for habitat selection of
little-known and hard-to-study species.
In British Columbia, Canada, three threatened species of snakes occur coincidentally
over much of their respective northern range extremes. The Great Basin Gophersnake and the
Western Yellow-bellied Racer (Coluber constrictor mormon) each hibernate communally or
solitarily in rocky outcrops, deep talus, or animal burrows (Haney and Sarell 2005;
COSEWIC 2015b; Williams et al. 2015), while the Western Rattlesnake largely dens
communally within rocky outcrops and talus slopes (Gienger and Beck 2011; COSEWIC
2015a). Within southern British Columbia, these species routinely are found denning with
each other (COSEWIC, 2013; Williams et al., 2015). In this study, we use a community of
these three snake species occurring at coincidental northern limits to compare hibernacula
selection and communal denning. We (i) document the hibernacula and their species
assemblages on our study site; from these sites, we (ii) investigate how the habitat features
relate to the degree of communal occupation present at hibernacula and (iii) explore how
these three species vary in their selection of overwintering habitat.

METHODOLOGY
Study Site
This study was conducted within and around the Lac du Bois Grasslands Protected
Area (51° N, 120.4° W) on the northern edge of the City of Kamloops, in south-central
British Columbia, Canada (Figure 1.2). The area is dominated by lower and middle
grasslands and, to a lesser extent, upper grassland, rocky outcrops, talus slopes, and wetlands
(BC Parks 2017). The major plants of this study site consist of Big Sagebrush
(Artemisia tridentata), Bluebunch Wheatgrass (Pseudoroegneria spicata), and a variety of
spring-blooming plants (BC Parks 2017). Although protected, this area is not immune to
anthropogenic practices such as recreational use, livestock grazing, bordering residential
development, and pipeline construction (GCC 2009; BC Parks 2017; Trans Mountain 2021a).

28
Refer to Chapter 1 for further details on the climatic and biogeographic conditions of the
study area.
Our site lies within the coincidental northern limits of the Great Basin Gophersnake,
Western Yellow-bellied Racer, and Western Rattlesnake (Matsuda et al. 2006). The study
area has critical habitat features supporting basking, oviposition, gestation, birthing, and
overwintering of our study species. The overwintering period commences when snakes return
to their hibernaculum (ingress) between August and October and is completed when snakes
leave their hibernaculum (egress) between March and May (pers. observ.). Snakes are active
on the study site from when they emerge until returning to their hibernacula (April to
September) (pers. observ.). Data collection in this study was centered around the emergence
and return of snakes to hibernacula.
This project was established as part of the mitigation efforts following the unearthing
of several snake hibernacula during construction of the Trans Mountain Expansion Project
(Heinrich 2020). Three hibernacula, containing variable numbers of the three aforementioned
species, were displaced late in Autumn 2020 through active pipeline construction (Heinrich
2020; Trans Mountain 2021b – Figure 1.2). A total of 10 rattlesnakes, 49 gophersnakes, and
69 racers were salvaged from the excavation (Heinrich 2020). Prior to the involvement of
the authors and this research project, a portion of these snakes (10 rattlesnakes, 12
gophersnakes, and 13 racers) were relocated to natural hibernacula in the immediate area
(Heinrich 2020). These snakes were not marked for future identification as it is not a
regulatory requirement. The remaining 37 gophersnakes and 56 racers were overwintered in
captivity at the BC Wildlife Park, providing an opportunity to study the response of the
animals to the attempted mitigation. In the spring of 2021, the artificially-overwintered
snakes were released at an artificial hibernaculum built to compensate for lost habitat
(Heinrich 2020). All released individuals were implanted with a passive integrated
transponder (PIT tags) (Biomark Inc., Boise, ID, United States) to allow for future
identification and a subset of these snakes were tracked to newly identified hibernacula (see
Locating Hibernacula below). Refer to Chapter 1 for further details on the disturbance event.

29
Locating Hibernacula
We located existing hibernacula using historical records, hibernacula surveys, and
radiotelemetry. Historical records of hibernacula were obtained through the Province of
British Columbia’s Conservation Data Centre (Bertram et al. 2001; GBC 2023).
We conducted routine surveys of known hibernacula to ensure sites were still active
and to obtain count data (see Count of Snakes at Hibernacula below). Surveys were
conducted at a subset of identified hibernacula during periods when snakes emerged or
returned to hibernacula. When additional hibernacula were identified, they were added to the
surveys. Between April 6 to May 6, 2022, 17 hibernacula were surveyed each at least 10
times; from August 25 to October 17, 2022, 19 hibernacula were surveyed each at least 19
times; and between April 1 to May 10, 2023, 24 hibernacula were surveyed each at least 16
times. Surveys began well in advance of the first observed snakes and continued until snakes
were no longer observed or the same individuals were recaught over more than one visit
across multiple sites. To survey hibernacula, we would initially survey the hibernacula
entrance, followed by thoroughly searching for snakes within a ~5m radius around the
entrances. Special attention was given to checking the underneath of cover objects in the
vicinity (i.e., shrubs, rocks, etc.). These searches would, at times, be limited to a smaller area
depending on the topography of hibernacula (i.e. increased slopes/cliffs). New hibernacula
were discovered opportunistically when a snake or snakes were found using a particular
habitat during these windows and subsequent surveys.
In addition to the subset of translocated gophersnakes and racers (see Study Site
above), we also tracked a sample of undisturbed, free-ranging snakes to collect comparative
data on movement behaviour (Ragsdale, MSc. Thesis, in prep.). Racers tracked in the study
were generally too small to carry transmitters large enough to enable tracking throughout the
active season to hibernacula. Rattlesnakes were not tracked as (i) described above,
rattlesnakes from the disturbed hibernacula were relocated to natural hibernacula without
permanent marking, and (ii) considerable telemetry on that species has been conducted in BC
(Bertram et al. 2001; Gomez et al. 2007; COSEWIC 2015a; Lomas et al. 2015; Harvey and
Larsen 2020; Schmidt et al. 2020; Atkins 2021; Eye 2022; Howarth et al. 2023, etc.),
including in the Lac du Bois area (Bertram et al. 2001; Gomez 2007; Hobbs 2007), such that

30
communal denning is known to be virtually ubiquitous in this region, and (iii) as a result, a
number of rattlesnake hibernacula in the study area already were known.
We selected gophersnakes for intracoelomic radio transmitter implantation when the
transmitters (SB-2T 5.2g or SB-2T and ibutton modification 9.5g with 12-month battery, or
SB-2T 3.8g with six month battery; Holohil Systems Ltd., Carp, ON, Canada) did not exceed
4% of snake body mass (Bryant et al. 2010). Surgeries were performed at nearby veterinary
clinics and followed similar methodologies to Reinert and Cundall (1982) with minor
modifications; the transmitter implants occurred on the left side of the body at ~55% of the
SVL as measured caudally from the snout. The antennae were fed caudally under the skin
rather than cranially. Animals were tracked every 2-3 days throughout the two active seasons
until each hibernaculum location was determined. Hibernaculum location was confirmed
when a telemetered snake remained in a location through the fall and into winter. The
hibernaculum mouth was identified as the closest opening to the point of strongest signal
from radiotelemetry (Harvey and Weatherhead 2006). Radio transmitters were surgically
removed at the end of the battery life following the same procedure.
Precise locations of the hibernacula identified and monitored in this study cannot be
revealed, given restrictions on the publication of sensitive/secured data. None of the
displaced snakes naturally adopted the newly-created hibernaculum in the subsequent fall
(see Study Site above and Chapter 3). Instead, many found and used other hibernacula (all
used with other snakes) that would have been extremely difficult to locate otherwise. We do
not use the newly created artificial hibernacula in our sampling of denning sites.
Counts of Snakes at Hibernacula
Following the previously described survey methodology (see Locating Hibernacula
above), 2023 egress hibernaculum surveys (April 3, 2023, to May 10, 2023) were used to
gauge the number of snakes occurring at the hibernacula. If no snakes were found at a known
site during the 2023 egress hibernacula survey period, the highest number from the 2022
egress (April 6, 2022, to May 6, 2022) or ingress hibernacula survey period (August 25,
2022, to October 17, 2022) was used. The overall data from these previous survey periods
were not used in combination with 2023 egress data, as not all hibernacula had been
identified until the end of the 2022 ingress period. During surveys, individual snakes were

31
caught and implanted with a PIT tag (Biomark Inc., Boise, ID, United States) to permit
identification in subsequent surveys. The displaced snakes were included in this count of
snakes. Because snakes were transplanted directly into new hibernacula by contractors (see
Study Site above), we could not unequivocally eliminate them from our samples. Snakes
seen but not caught were only included in the count if we could confirm they were unique
individuals (i.e., caught and tagged two gophersnakes, but four were seen). The count of
snakes at a given hibernacula was the total number of unique individuals found over the
entire survey period. Hibernacula were only included in this inventory if they could be
surveyed with effort similar to the other sites. Two hibernacula were omitted due to distance
from other sites as they could not be evenly surveyed with the other hibernacula. An
additional two hibernacula were omitted as they were not evenly surveyed along with the
other hibernacula during this period. Overall, 24 hibernacula out of 28 known hibernacula
were surveyed to estimate the number of snakes occupying them.
Habitat Measurements
Habitat data were collected at each recorded hibernaculum and a paired random site,
using a matched case-control design (Harvey and Weatherhead 2006; Thomas and Taylor
2006; Eye 2022). From the hibernaculum entrance, a random direction and distance (40 - 300
m) were chosen to identify the random site (Prior and Weatherhead 1996; Harvey and
Weatherhead 2006; Gienger and Beck 2011; Eye 2022). The range of distances used to find
the random sites was selected based on previous studies to ensure hibernacula and the
random site did not overlap and to represent the greater environment available to snakes
(Harvey and Weatherhead 2006; Eye 2022). The entrance of each hibernaculum was used as
the epicenter for data collection. If a site had multiple entrances, the entrance with the largest
area was used for measurements. Following Prior & Weatherhead (1996), measurements at
the random site were taken around the closest feature within 5 m of the site found from the
random distance and direction, representing an underground opening (i.e., rodent burrow,
hole under rock, crack in a rockface, etc.). At each known hibernaculum and random site,
three types of habitat features were measured: microscale features (1m radius), macroscale
features (10m radius), and objects within 30m of the site (Table 2.1) (Howarth 2023). Habitat
features were selected based on similar studies (Prior and Weatherhead 1996; Harvey and
Weatherhead 2006; Gienger and Beck 2011; Williams et al. 2015; Eye 2022; Howarth et al.

32
2023) (Table 2.1). The percent cover of shrub, grass, dirt, and rock were visually estimated in
1m and 10m radius plots. At the microscale, slope, aspect, elevation, number of openings,
and the average area of those openings were measured. The distance to the nearest shrub and
rock over 20cm in length was recorded (if found within 30m) (Williams et al. 2015). The
height of the nearest shrub and the volume of the nearest rock over 20cm in length were
measured (if found within 30m). Distances and size metrics were obtained using a tape
measure.

Table 2.1: Habitat features measured around the entrances to hibernacula and random sites. Habitat features were measured at the
microscale (1m radius), macroscale (10m radius), and within 30m of the site.

Abbreviation
OPEN

Description
Number of openings or mouths

Scale
micro

AREA

Area (cm2) of largest opening

micro

SLOPE

Average of three slope (%) measures

micro

ELE

Elevation (m)

micro

ASP

Aspect (°) converted to radians

micro

S1

%Shrub cover at 1m

micro

G1

%Grass cover at 1m

micro

R1

%Rock cover at 1m

micro

D1

%Exposed dirt at 1m

micro

AVGRV

micro

S10

Average volume of 10 randomly selected surface rocks
volumes (cm3)
%Shrub cover at 10m

macro

G10

%Grass cover at 10m

macro

R10

%Rock cover at 10m

macro

D10

%Exposed dirt at 10m

macro

RD

Distance (cm) of nearest rock >20cm in length within 30m

30m

RV

Volume of nearest rock (cm3) >20cm in length within 30m

30m

SD

Distance to nearest shrub (cm) within 30m

30m

SH

Height of nearest shrub (cm) within 30m

30m
33

34
Statistical analysis
All statistical analysis was performed using RStudio (Posit Software 2023). The
effects of season were not accounted for in this analysis (Howarth 2023). For analysis,
habitat variables were modelled in two separate groups: microscale was modelled alone,
while macroscale features and objects measured within 30m were modelled together
(Howarth 2023). Rarely, habitat plots on neighbouring hibernacula overlapped at the
macroscale. When this occurred, the two plots were analyzed as one at that scale by taking
the average value of the habitat metrics across both plots; also, in these situations, the
number of snakes at those sites was pooled, and a single random site was chosen randomly.
Aspect was converted from degrees into radians (Mardia and Jupp 2000; Eye 2022).
Number of Snakes at Hibernacula
The relationship of hibernacula features on the count of snakes was assessed
separately for gophersnakes, racers, and rattlesnakes, at both the micro and
macroscale (Figure 2.1). Data within these groups were assessed for correlation using
Spearman rank correlation tests (Hauke & Kossowski, 2011). When a correlation coefficient
between a pair of metrics was ≥0.7, only one of the variables was retained if deemed more
biologically relevant than the other from the habitat selection analysis (see Habitat Selection
below). Using the ‘glm’ function in base R (Posit Software 2023) and ‘glm.nb’ function from
the ‘MASS’ package (Ripley et al. 2023) in RStudio (Posit Software 2023), Poisson and
negative binomial regressions were fit for each species at both scales. The negative binomial
model preferentially was selected when overdispersion of the data was significant, and AICc
values were lower than the Poisson regression model (Ismail and Jemain 2007). Top models
were determined using the difference in AICc values obtained from the ‘dredge’ function in
the ‘MuMIn’ package (Bartoń 2023) in RStudio (Posit Software 2023). Models with ⃤ AICc
value <2.0 were considered equally supported (Burnham and Anderson 2002).

Figure 2.1: Numbers of snakes and species composition at 24 hibernacula distributed across the study landscape. Snakes were
surveyed during the 2023 emergence period. If no snakes were found at a hibernaculum during this survey period, the next highest
survey period from 2022, ingress or egress, was used to estimate the number of snakes at that hibernaculum. Bars are subdivided
according to the number of snakes of each species.
35

36
Habitat Selection
We compared the habitat features at random sites to hibernacula used by
gophersnakes, racers, and rattlesnakes at both micro- and macro-scales. Univariate analysis
was performed to reduce the number of variables imputed into the final models (Hosmer and
Lemeshow 2000; Howarth 2023). This was done by conducting a Wilcoxon signed-rank test
(paired sample) or paired t-test, depending on normality, for each habitat variable between
hibernacula and random sites. Habitat variables were retained for modelling when P<0.25
(Hosmer and Lemeshow 2000; Howarth 2023). The selected habitat variables were then
assessed for correlation using Spearman rank correlation tests (Hauke and Kossowski 2011).
When a correlation coefficient was ≥0.7, the variable with the least variation, as determined
from the univariate analysis, was removed from further analysis (Dormann et al. 2013).
Following univariate analysis and checking for correlation, we compared the habitat at
hibernacula to paired random plots using conditional logistic regression. These models were
created at the two scales for each of the three snake species. We used the ‘clogit’ function in
the ‘survival’ package (Therneau et al. 2023) in RStudio (Posit Software 2023) to fit these
models. Top models were determined as described in Number of Snakes at Hibernacula. If a
global model did not converge, the variable with the next lowest variation, as determined
from the univariate analysis, was removed. Following this, if the model still did not
converge, interpretations were carried out using the initial univariate analysis following the
removal of correlated variables (Ballantyne and Nol 2011).
RESULTS
We identified 28 hibernacula within the study site: Of these, 7 were known
historically, 6 were identified through our hibernacula surveys, and 15 were located through
telemetry on 44 gophersnakes (19 from the displaced population and 25 reference). Among
the 15 hibernacula discovered through radiotelemetry, 5 were found from displaced snakes,
and 10 were found from reference individuals. Of the 28 hibernacula, gophersnakes were
detected at 21, with racers and rattlesnakes occurring in 15 hibernacula each (Figure 2.2). All
told, 5 hibernacula (18%) supported all three species (Figure 2.2). Habitat data were collected
at each hibernaculum and 28 paired random sites; 46 sites were surveyed following egress in
2022, and the remaining 10 were collected during ingress in 2022. Hibernacula were located
in rodent burrows (18%, n=5), at the base of rocky outcrops (61%, n=17), within

37
anthropogenically-placed boulders on the side of roads (blast rock) (11%, n=3), and in holes
found under rocks (11%, n=3) (Figure 2.3). Racers and gophersnakes were found in
hibernacula across all of these four categories, while rattlesnakes overwintered exclusively in
rocky outcrops and blast rock.
Number of Snakes at Hibernacula
A total of 24 hibernacula (19, 13, and 12 for gophersnakes, racers, and rattlesnakes,
respectively) were used to examine the effect of habitat on the number of snakes detected at
each location for each species. Overall, the top model suggested that no habitat features
explained the number of racers and gophersnakes at hibernacula at the microscale, with the
top models for these species being the null models (i.e. the models without any habitat
features) (Table 2.2; Appendix A). However, for gophersnakes, models with ⃤ AICc <2.0
suggested the number of those snakes detected at hibernacula was negatively associated with
the number of openings and shrub cover, and racers were positively associated with rock
cover (Burnham and Anderson 2002) (Table 2.2; Appendix A). The number of rattlesnakes
detected at a hibernaculum was negatively associated with average rock volume and dirt
cover, and positively associated with the number of openings and aspect at this scale (Table
2.2.; Appendix A).
The top model suggests that no habitat features using macroscale measurements
explained the number of detected racers and gophersnakes, with the top models for these
species being the null models (Table 2.2; Appendix A). However, models with a ⃤ AICc
value <2.0 suggest gophersnakes are negatively associated with rock cover and volume of the
nearest rock (Table 2.2; Appendix A). At this scale, dirt cover and volume of the nearest rock
were negatively associated with the number of rattlesnakes present, while shrub cover was
positively associated (Table 2.2; Appendix A).

38

Figure 2.2: The number of hibernacula used and shared by the Western Yellow-bellied
Racer “WYR”, Western Rattlesnake “WR”, and Great Basin Gophersnake “GBG” in Lac du
Bois Grasslands Protected Area.

39

A

C

B

D

Figure 2.3: Examples of snake hibernacula within Lac du Bois Grassland Protected Area
(2021-2023). A) Found in the blast rock off the side of the road. B) Found under a rock. C)
Found in an animal burrow. D) Found in a rocky outcrop. Red arrows identify the location of
the hibernaculum entrance. Photos by author.

Table 2.2: Candidate models investigating the effect of habitat on the numbers of snakes detected at hibernacula at a micro- and
macroscale. Models for the Great Basin Gophersnake, Western Rattlesnake, and Western Yellow-bellied Racer were ranked based
on the Akaike Information Criterion small sample equivalent (AICc) and ΔAICc (difference of AICc of the best model (minimum
AICc) to the model being compared). This table also includes the models' weight (wi). The top models with ⃤ AICc value <2.0
(Burnham and Anderson 2002) are shown, along with the global model (G) that contained all of the selected habitat variables. The
relationship of habitat variables in the top models is indicated with a superscript, + describes habitat features associated with
increased snakes, and - describes variables not associated with them (Appendix A). Full names and detailed descriptions of
explanatory variables are found in Table 2.1.
Scale

Group
Gophersnake

Micro

Racer

Rattlesnake

Gophersnake
Macro

Racer
Rattlesnake

# Model
1 Null Model
2 OPEN3 S1G ASP+ELE+AREA+R1+D1+OPEN+S1+G1
1 Null Model
2 R1+
G ASP+ELE+OPEN+AREA+R1+D1+AVGRV+S1+G1
1 AVGRV-+D1-+OPEN+
2 ASP++AVGRV-+D1-+OPEN+
G ASP+OPEN+AREA+R1+D1+AVGRV+S1+G1
1 Null Model
2 R103 RVG RV+R10+D10+SD+SH+S10
1 Null model
G RD+RV+R10+G10+SD+SH+S10
1 D102 D10-+RV3 D10-+S10+
G RD+RV+R10+D10+S10+G10

AICc
56.9
58.5
58.8
89.5
39
40.6
161.9
59.8
60.9
145.8
55.7
57.6
57.6
76.7
40.4
87.0
63.3
65.1
65.1
96.2

ΔAICc
0
1.6
1.9
32.6
0
1.6
122.9
0
1.1
86
0
1.9
1.9
21.0
0
46.6
0
1.8
1.8
32.9

wi
0.12
0.05
0.05
0.00
0.15
0.07
0.00
0.40
0.24
0.00
0.21
0.08
0.08
0.00
0.21
0.00
0.35
0.14
0.14
0.00
40

41
Habitat Selection
Habitat information was collected at each of the 28 hibernacula and a paired random
site. At the microscale, gophersnake hibernacula were associated with an increased area of
the main hibernaculum mouth, number of openings, shrub cover, and slope (Tables 2.3;
Appendix B). At this scale, the racer model did not initially converge; following the removal
of the variable with the next lowest variation as determined from the univariate analysis,
ASP, convergence was achieved (Appendix C). Racers selected for steeper slopes, an
increased area of the main hibernaculum mouth, and rock and shrub cover (Tables 2.3;
Appendix B). At this scale, the rattlesnake model did not converge regardless of variable
removal; thus, univariate analysis following correlation testing was used to compare known
hibernacula to random sites. Known rattlesnake hibernacula had significantly larger openings
(Wilcoxon signed-rank tests, P=0.0001), were found on steeper slopes (Paired ttest=3.72, P=0.0023), and had less grass cover than random sites (Wilcoxon signed-rank
tests, P=0.0007) (Appendix C).
At the macroscale, gophersnakes selected for rock cover, and increased height of the
nearest shrub (Tables 2.3; Appendix B). At this scale, racers and rattlesnakes selected against
grass cover (Tables 2.3; Appendix B). All three species selected larger distances to the
nearest shrub and smaller distances to the nearest rock (Tables 2.3; Appendix B).

DISCUSSION
This study provides the first direct comparison of surface habitat at hibernacula used
by multiple snake species within the same community. At these hibernacula, we observed
variation in the extent of communal occupation for all three species. However, our habitat
metrics around these sites failed to effectively explain the numbers of gophersnakes, racers,
and rattlesnakes at the hibernacula. Further, we reveal that although these three species share
the same landscape and, in some cases, hibernacula, they do not appear to be selecting habitat
in the same fashion. Overall, this research provides compelling evidence to reconsider our
approach in identifying and protecting these critical habitats.

Table 2.3: Candidate models comparing habitat features of hibernacula and paired random sites at the micro and macroscale.
Models for the Great Basin Gophersnake, Western Rattlesnake, and Western Yellow-bellied Racers were ranked based on Akaike
Information criterion (AIC) small sample equivalent (AICc) and ΔAICc (difference of AICc of the best model (minimum AICc) to
the model being compared). This table also includes the models' weight (wi). The top models with ⃤ AICc value <2.0 (Burnham
and Anderson 2002) are shown, along with the global model (G) that contained all of the selected habitat variables. The
relationship of habitat variables in the top models is indicated with a superscript, where + describes habitat features selected for,
and - describes variables selected against (Appendix B). Full names and detailed descriptions of explanatory variables are found in
Table 2.1. The Western Rattlesnake model at the microscale did not converge.
Scale

Micro

Group
Gophersnake

Racer

Gophersnake

Macro

Racer

Rattlesnake

#
1
2
3
4
G
1
2
3
4
5
G
1
2
3
G
1
2
3
G
1
2
G

Model
+

+

SLOPE +AREA
SLOPE++AREA++S1++OPEN+
SLOPE++AREA++S1+
SLOPE++OPEN+
SLOPE+AREA+S1+OPEN+G1
R1+
R1++SLOPE+
SLOPE++AREA+
R1++S1+
R1++SLOPE++S1+
SLOPE+AREA+S1+OPEN+AVGRV +R1
SH++R10+
SH++SD+ +R10+
SH+ +RD-+R10+
RD+SD+SH+R10
G10G10-+SD+
G10-+RDG10+RD+SD
G10-+RDG10-+SD++RDG10+SD+RD+S10

AICc
21.5
21.9
21.9
22.4
25.3
13.6
14.4
14.8
15.4
15.4
27.5
13.4
14.2
15
17.3
25.8
26.4
26.7
28.5
13
14.9
19.1

ΔAICc
0
0.4
0.4
0.9
3.8
0
0.8
1.2
1.8
1.8
13.9
0
0.8
1.6
3.9
0
0.6
0.9
2.7
0
1.9
6.1

wi
0.18
0.15
0.15
0.12
0.03
0.14
0.10
0.08
0.06
0.06
0.00
0.43
0.29
0.19
0.06
0.30
0.23
0.19
0.11
0.45
0.18
0.02

42

43
We now know that Western Yellow-bellied Racers and Great Basin Gophersnakes
appear to be more generalist in their selection of habitat, overwintering in rodent burrows,
holes under rocks, rocky outcrops, and blast rock features in comparison to rattlesnakes
present at rocky outcrops and blast rock on this study site. Gophersnakes, racers, and
Western Rattlesnakes all are known to overwinter together (COSEWIC 2013,2015a,b).
While we observed this phenomenon, there were cases where animals overwintered
individually or in alternate groupings of the species. Coincidental occupation of hibernacula
occurred the most between gophersnakes and racers at six sites. The generalist nature of
these species may provide more opportunities for racers and gophersnakes to overlap.
Overall, the variation between assemblages and the difference in the counts of snakes at
hibernacula suggests that these species are selecting differently and may coincidentally
overlap at select hibernacula.
Rattlesnakes were found increasingly at hibernacula with more entrances on
southwestern and southern aspects. Rattlesnakes typically congregate during ingress or
egress in the immediate entrance area to hibernacula (Brown and Parker 1982; Gienger and
Beck 2011); presumably, more entranceways would allow increased snakes to bask and
provide adequate refuge should individuals need to evade. As with this study, previous
research has shown Western Rattlesnake hibernacula occur on south-facing slopes in
Washington State (Gienger and Beck 2011), receiving increased solar radiation in
comparison to north-facing slopes (Gates 1980; Bonon 2002; Geiger et al. 2003; Gienger and
Beck 2011). Presumably, heightened solar radiation increases opportunities for
thermoregulation in the form of basking.
Gophersnakes and racers demonstrate a broad range of habitat preferences for
hibernation sites. This variability might help to explain why none of the habitat features (of
those measured) effectively explain their count at hibernacula, as the null models were the
top models at the micro and macroscale. Further, the potential underrepresentation of racers
and gophersnakes due to detection bias may contribute to these results (Kéry 2002; Biro and
Dingemanse 2009; Johnstone et al. 2021). For example, Parker and Brown (1973) found that
the number of detected racers increased from 5 to 149 when their hibernaculum was
fenced. This generalist behaviour and low detectability at hibernacula compared to
rattlesnakes may contribute to the absence of habitat features that differentiate their count.

44
Regardless, gophersnakes and racers are not using hibernacula like rattlesnakes on our study
site. Western Rattlesnakes conspicuously congregate at hibernacula (Brown and Parker 1982;
Gienger and Beck 2011), whereas racers and gophersnakes are more cryptic and generally
found away from the hibernaculum entrance (Parker and Brown 1973; Brown and Parker
1982; Kéry 2002). This behavioural heterogeneity may contribute to the lack of effect habitat
has on their numbers (Williams et al. 2015).
Certain features were commonly selected among the species. Our study demonstrated
that gophersnake, racer, and rattlesnake hibernacula were situated on steeper slopes with
relatively larger openings of the main hibernaculum entrance compared to random sites.
Previous research on Western Rattlesnake hibernacula in Washington States, USA, found
56+/-12° slopes compared to random sites 62+/-7° (Gienger and Beck 2011). Stability and
access decrease as slopes increase due to potential rockfall, landslides, and near-vertical
climbs (Giani 1992; Gienger and Beck 2011). In contrast, as slope increases, solar exposure
increases, providing increased soil temperature and basking opportunities (Barry and Chorley
1976; Brady and Weil 2002; Hamilton and Nowak 2009). Presumably, these midrange slopes
would provide the optimal balance between stability, access, and solar radiation. One of the
most vulnerable times for snakes is during egress when they are cold, less responsive to
potential risk, and thus seeking to bask (Angilletta et al. 2002; Harvey and Weatherhead
2006; Gienger and Beck 2011). Larger entrances would help to provide adequate refuge to
evade predation and room to bask.
The variation in the selection of cover and open habitat features at the micro and
macroscale aligns with observed behavioural differences between these species (pers.
observ). During egress, racers and gophersnakes are more cryptic and generally found away
from hibernacula entrances (Parker and Brown 1973; Brown and Parker 1982). In contrast,
Western Rattlesnakes conspicuously bask at the hibernacula (Brown and Parker 1982;
Gienger and Beck 2011). We found that rattlesnake hibernacula were associated with open
habitat, with those sites having significantly less grass cover than random sites at both the
micro and macroscale. Such habitat could provide more basking opportunities with increased
solar exposure and better visibility to potential predators. Open habitat surrounding
hibernacula is crucial for various snake species, including the Timber Rattlesnake (Crotalus
horridus) (Brown 1993), where the lifespan of hibernacula may be constrained by the

45
shading caused by nearby trees (Brown 1993). This limitation is also speculated to affect
Western Rattlesnakes in forested regions (COSEWIC 2015b). Although our study site is an
open grassland, and trees are not a dominant feature this example showcases the potential
dependence of some snakes to open habitat at their hibernacula. In contrast, we found that
gophersnakes and racers were associated with further cover features at the microscale, with
gophersnakes selecting for more openings and shrub cover and racers selecting for increased
rock and shrub cover. Gophersnakes are thought to rely on cover objects for
thermoregulation and security around their hibernacula (Parker and Brown 1980; COSEWIC
2013). Their selection of further cover features (i.e. increased rock cover) at the macroscale
would support this behaviour. However, such behavioural differences may not capture the
whole story, as racers selected against grass cover, and all three snakes selected for larger
distances to the nearest shrubs and smaller distances to the nearest rock at the macroscale.
Such habitat would provide open conditions conducive to basking and cover rocks in the
greater vicinity.
Why do rattlesnakes exclusively overwinter in rocky outcrops and blast rock features
on our study site, whereas the other two species appear much more catholic in hibernacula
site selection? One obvious factor distinguishing racers and gophersnakes from rattlesnakes
is the reproductive mode. Racers and gophersnakes are oviparous, retaining eggs until
oviposition occurs (COSEWIC 2013; COSEWIC 2015b), whereas rattlesnakes are viviparous
(COSEWIC 2015a). Rookeries, distinct sites where parturition occurs, are generally found
far closer to hibernacula than active season foraging sites (as reviewed in Graves and Duvall
1995). For example, Western Rattlesnakes within the Okanagan were found to utilize
hibernacula or the general area (on average less than <140m from hibernacula) outside of
winter as rookeries (Macartney and Gregory 1988; Eye 2022). Sites physically closer to
hibernacula benefit gravid females by reducing the distance needed to migrate while gravid
(Duvall et al. 1985; Graves and Duvall 1995). Additionally, proximity to hibernacula would
aid neonates by making it easier to locate these sites and reducing the potential energy
expenditure associated with searching for a suitable site (Graves and Duvall 1995). Thus, the
selection of rocky outcrops and blast rock features as hibernacula for rattlesnakes may
facilitate or be influenced by the selection of habitat for these reproductive behaviours. In the

46
Okanagan, gravid Western Rattlesnakes selected for rock cover at their rookeries across
multiple scales and sites (Eye 2022).
We did not track rattlesnakes or racers for this study, and we acknowledge that this
may bias results by not providing the full breadth of sites selected. However, rattlesnakes are
one of the most heavily studied reptiles within the province of British Columbia, with over
380 known hibernacula and multiple research studies (Bertram et al. 2001; Gomez et al.
2007; COSEWIC 2015a; Lomas et al. 2015; Harvey and Larsen 2020; Schmidt et al. 2020;
Atkins 2021; Eye 2022; Howarth et al. 2023, etc.). Further, rattlesnakes were previously
tracked within the study site (Bertram et al. 2001; Gomez 2007; Hobbs 2007). Given this, we
believed that additional tracking of rattlesnakes would not have revealed anything outside the
scope of our results. Racers, however, are understudied, and any data regarding their
overwintering behaviour is valuable. Although we may not have represented the full scope of
their potential selection, their presence at all habitat types (i.e., rodent burrows, rocky
outcrops, holes under rocks, and blast rock) is similar to gophersnakes in this study.
Our work has implications for future research studies, impact assessments, and
protection of these critical habitat features. We have identified key habitat characteristics
related to the number of snakes at hibernacula and distinguished these sites from the broader
environment. In practice, the ability to readily distinguish a hibernaculum, communal or not,
with these features is not likely (Prior and Weatherhead 1996). Instead, these features only
provide a general guide to personnel conducting surveys. Unfortunately, these features may
not encompass all of the attributes and environmental cues these species are selecting for
(Gienger & Beck, 2011), instead portraying only “average hibernacula” features. These
features also are restricted to surface habitat, arguably pertaining more to shoulder season
use. Due to the limited access, underground hibernacula habitat has largely been unexplored.
We also showcase variation of habitat selection between these unique species that, in some
cases, share these habitats. Thus, we caution against using hibernacula ‘models’ from certain
iconic, tractable species (e.g., Western Rattlesnakes) to extrapolate to other species with lesswell known denning habits, even within the same ecosystem.
Within British Columbia, protection mechanisms currently apply to 200-300 ha
around communal hibernacula (MWLAP 2004; Williams et al. 2012; COSEWIC 2015a;
Dyer et al. 2016a,b). However, protection only is effective if this habitat can be identified.

47
Prior to our study, only seven out of 28 hibernacula were known on this landscape, six of
which were rocky outcrop hibernacula with a strong presence of rattlesnakes. To ensure the
protection of cryptic hibernacula, especially for racers and gophersnakes, protection must be
extended to the broader landscape where these hibernacula occur. As per Prior and
Weatherhead (1996), radiotelemetry may be the most effective way to identify hibernacula in
a new area, being responsible for the location of 15 of 28 hibernacula (54%) in this study.
Such results support our findings that these sites are not easily visually identified through
surveys (Prior and Weatherhead 1996), a point worth considering during future research and
impact assessments. Overall, this study offers argument that we need to rethink how we
identify and protect critical habitats for even species that share habitat within a community of
snakes.
LITERATURE CITED
Angilletta MJ, Niewiarowski PH, Navas CA. 2002. The evolution of thermal physiology in
ectotherms. J Therm Biol. 27:249–268.
Atkins MCP. 2021. Temporal and spatial changes in a Western Rattlesnake (Crotalus
oreganus) population in British Columbia [thesis]. [Kamloops (BC)]: Thompson
Rivers University.
Ballantyne K, Nol E. 2011. Nesting habitat selection and hatching success of whimbrels near
Churchill, Manitoba, Canada. Waterbirds. 34(2):151–159.
Barry RG, Chorley RJ. 1976. Atmosphere, weather, and climate. 3rd ed. London: Methuen.
Bartoń K. 2023. MuMIn: Multi-model inference; [accessed 2023 Apr 13]. https://cran.rproject.org/web/packages/MuMIn/index.html.
BC Parks. 2017. Lac du Bois Grasslands Protected Area management plan. Kamloops.
Bertram N, Larsen KW, Surgenor J. 2001. Identification of critical habitats and conservation
issues for the Western Rattlesnake and Great Basin Gopher Snake within the
Thompson-Nicola region of British Columbia. Kamloops.
Biro PA, Dingemanse NJ. 2009. Sampling bias resulting from animal personality. Trends
Ecol Evol. 24(2):66–67.
Bonon GB. 2002. Ecological climatology concepts and applications. Cambridge: Cambridge
University Press.
Boyce MS, Johnson CJ, Merrill EH, Nielsen SE, Solberg EJ, van Moorter B. 2016. Can
habitat selection predict abundance? Jour Anim Ecol. 85(1):11–20.
Brady NC, Weil RR. 2002. Elements of the nature of properties of soil. Upper Saddle River:
Pearson Prentice Hall (NJ).

48
Brown WS, Parker WS, Elder JA. 1974. Thermal and spatial relationships of two species
of Colubrid snakes during hibernation. Herpetol. 30(1):32–38.
Brown WS, Parker WS. 1976. Movement Ecology of Coluber constrictor near Communal
Hibernacula. Copeia. 1976(2):225-242.
Brown WS, Parker WS. 1982. Niche dimensions and resource partitioning in a Great Basin
Desert snake community. In: Scott, NJ. Herpetological Communities. A Symposium
of the Society for the Study of Amphibians and Reptiles and the Herpotologist
League; 1977 Aug; Washington (DC): United Stated Departments of the Interior Fish
and Wildlife Service. p. 59–82.
Brown WS. 1993. Biology, status, and management of the Timber Rattlesnake (Crotalus
horridus): A guide for conservation. Lawrence (KS): Society for the Study of
Amphibians and Reptiles.
Bryant GL, Eden P, De Tores P, Warren K. 2010. Improved procedure for implanting
radiotransmitters in the coelomic cavity of snakes. Aust Vet J. 88(11):443–448.
Burger J, Zappalorti RT, Gochfeld M, Boarman WI, Caffrey M, Doig V, Garber SD, Lauro
B, Mikovsky M, Safina C, et al. 1988. Hibernacula and summer den sites of pine
snakes (Pituophis melanoleucus) in the New Jersey Pine Barrens. J Herpetol.
22(4):425–433.
Burnham KP, Anderson D. 2002. Model selection and multimodel inference: a practical
information-theoretic approach. 2nd ed. New York (NY): Springer.
Camaclang AE, Maron M, Martin TG, Possingham HP. 2015. Current practices in the
identification of critical habitat for threatened species. Conserv Biol. 29(2):482–492.
COSEWIC. 2013. COSEWIC Assessment and status report on the Great Basin Gopersnake
Pituophis catenifer deserticola in Canada. Ottawa.
COSEWIC. 2015a. COSEWIC Assessment and status report on the Western Rattlesnake
Crotalus oreganus in Canada. Ottawa.
COSEWIC. 2015b. COSEWIC Assessment and status report on the Eastern Yellow-bellied
Racer Coluber constrictor flaviventris and the Western Yellow-bellied Racer Coluber
constrictor mormon in Canada. Ottawa.
Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carré G, Marquéz JRG, Gruber B,
Lafourcade B, Leitão PJ, et al. 2013. Collinearity: a review of methods to deal with it
and a simulation study evaluating their performance. Ecography. 36(1):27–46.
Dubiner S, Jamison S, Meiri S, Levin E. 2023. Squamate metabolic rates decrease in winter
beyond the effect of temperature. J Anim Ecol.92(11): 2163-2174.
Duvall D, King MB, Gutzwiller KJ. 1985. Behavioral ecology and ethology of the prairie
rattlesnake. Natl Geogr Res. 1:80–111.
Dyer O, Bishop C, Govindarajulu P, Anderson L, Cunnington D, Hobbs J, Huntley M,
Larsen K, Reudink R, Sadler K, Sarell M, Steciw J, et al. 2016a. British Columbia
recovery strategy series recovery plan for the Gopher Snake, deserticola subspecies
(Pituophis catenifer deserticola) in British Columbia.

49
Dyer O, Bishop C, Govindarajulu P, Anderson L, Cunnington D, Hobbs J, Huntley M,
Larsen K, Reudink R, Sadler K, Sarell M, Tedesco L, et al. 2016b. British Columbia
recovery strategy series recovery plan for the Western Rattlesnake (Crotalus
oreganus) in British Columbia.
Eye DM. 2022. Reproductive ecology of female Western Rattlesnakes (Crotalus Oreganus)
in Canada [thesis]. [Kamloops (BC)]: Thompson Rivers University.
Gardiner LE, Sonmor KW. 2011. Reptiles major slump event at Grasslands National Park
snake pit in southwestern Saskatchewan. Blue Jay. 69(3):120–124.
Gates DM. 1980. Biophysical ecology. New York(NY): Springer.
[GBC] Government of British Columbia. 2023. B.C. Conservation Data Centre; [accessed
2023 Jul 19]. https://www2.gov.bc.ca/gov/content/environment/plants-animalsecosystems/conservation-data-centre.
[GC] Government of Canada. 2023. Critical habitat for species at risk national dataset –
Canada; [accessed 2024 Mar 2]. https://datadonnees.az.ec.gc.ca/data/species/protectrestore/critical-habitat-species-at-risk-canada/
[GCC] The Grasslands Conservation Council of British Columbia. 2009. Working together
for the conservation of BC’s grasslands. Kamloops.
Geiger R, Aron RH, Todhunter P. 2003. The climate near the ground. 6th ed. Oxford:
Rowman and Littlefield Publishers.
Gienger CM, Beck DD. 2011. Northern pacific rattlesnakes (Crotalus oreganus) use thermal
and structural cues to choose overwintering hibernacula. Can J Zool. 89(11):1084–
1090.
Giani G. 1992. Rock slope stability analysis. Rotterdam: Balkema Publishers.
Goldberg AR, Conway CJ, Mack DE, Burak G. 2020. Winter versus summer habitat
selection in a threatened ground squirrel. J Wildl Manag. 84(8):1548–1559.
Gomez LM. Habitat use and movement patterns of the Northern Pacific Rattlesnake
(Crotalus o. oreganus) in British Columbia [thesis]. [Victoria (BC)]: University of
Victoria.
Graves BM, Duvall D. 1995. Aggregation of Squamate reptiles associated with gestation,
oviposition, and parturition. Herpetol Monogr. 9:102–119.
Gregory PT. 1974. Patterns of spring emergence of the red-sided garter snake (Thamnophis
sirtalis parietalis) in the Interlake region of Manitoba. Can J Zool. 52:1063–1069.
Gregory PT. 2011. Temporal dynamics of relative-mass variation of red-sided garter snakes
(thamnophis sirtalis parietalis) at a communal hibernaculum in manitoba. Écoscience.
18(1):1–8.
Hall LS, Krausman PR, Morrison ML. 1997. The habitat concept and a plea for standard
terminology. Wildl Soc Bull. 25(1):173–182.
Hamilton BT, Nowak EM. 2009. Relationships between insolation and rattlesnake
hibernacula. West N Am Nat. 69(3):319–328.

50
Haney A, Sarell M. 2005. Conservation analysis for the Great Basin Gophersnake (Pituophis
catenifer deserticola) in British Columbia.
Harvey DS, Weatherhead PJ. 2006. Hibernation site selection by Eastern Massasauga
Rattlesnakes (Sistrurus catenatus catenatus) near their northern range limit. J
Herpetol.
Harvey JA, Larsen KW. 2020. Rattlesnake migrations and the implications of thermal
landscapes. Mov Ecol. 8(1).
Hauke J, Kossowski T. 2011. Comparison of values of pearson’s and spearman’s correlation
coefficients on the same sets of data. Quaest Geogr. 30(2):87–93.
Hein AM, Guyer C. 2009. Body temperature of overwintering Cottonmouth snakes:
hibernaculum use and inter-individual variation. J Ala Acad Sci. 80(1): 35-44.
Hobbs J. 2007. Thermal factors in relation to the denning ecology of northern Pacific
rattlesnakes in British Columbia [thesis]. [Victoria (BC)]: Royal Roads University.
Hosmer D, Lemeshow S. 2000. Applied logistic regression. 2nd ed. New York (NY):
Wiley-Interscience.
Howarth CR., Bishop CA, Larsen KW. 2023. Western Rattlesnake (Crotalus oreganus)
spring migration in British Columbia: a comparative study of juveniles and adults.
Can J Zool. 101(7):530–540.
Howarth CR. 2023. Western Rattlesnake (Crotalus oreganus) migration and habitat use in
British Columbia [thesis]. [Kamloops (BC)]: Thompson Rivers University.
Huey RB. 1991. Physiological consequences of habitat selection. Am Nat. 137:91–115.
Ismail N, Jemain AA. 2007. Handling overdispersion with negative binomial and
generalized poisson regression models. In: Casualty Actuarial Society E Forum. p.
103–158.
Johnstone KC, McArthur C, Banks PB. 2021. Behavioural drivers of survey bias: interactive
effects of personality, the perceived risk and device properties. Oecologia.
197(1):117–127.
Kéry M. 2002. Inferring the absence of a species: a case study of snakes. J Wildl Manag.
66(2):330-338.
Krausman PR. 1999. Some basic principles of habitat use. In: Launchbaugh KL, Sanders
KD, Mosley JL, editors. Grazing Behaviour of Livestock and Wildlife. Moscow:
University of Idaho. p. 85–90.
Lomas E, Larsen KW, Bishop CA. 2015. Persistence of northern Pacific Rattlesnakes masks
the impact of human disturbance on weight and body condition. Anim Conserv.
18(6):548–556.
Maida J, Larsen K, Hooper C, Bishop C. 2017. Crotalus oreganus (Western Rattlesnake)
Mortality. Herpetol Rev. 48(2):447.

51
Maida JR, Kirk DA, Mckibbin O, Row JR, Larsen KW, Stringam C, Bishop CA. 2018.
Population estimate, survivorship, and generation time of the northern pacific
rattlesnake (Crotalus o. oreganus) at its northern-most range limits. Herpetol Conserv
Biol. 13(3):662–672.
Macartney JM, Gregory PT. 1988. Reproductive biology of Female Rattlesnakes (Crotalus
viridis) in British Columbia. Copeia. 5(1):47–57.
Manitoba. 2024. Snakes of Narcisse; [accessed 2024 Feb 3].
https://www.gov.mb.ca/nrnd/fish-wildlife/snakes_narcisse/index.html.
Mardia K, Jupp P. 2000. Directional statistics. Barnett V, Cressie N, Fisher N, Johnstone I,
Kadane J, Kendall D, Scott D, Silverman B, Smith A, Teugel J, editors. West Sussex:
John Wiley and Sons Ltd.
Matsuda BM, Green DM, Gregory PT. 2006. Amphibians and reptiles of British Columbia.
Victoria (BC): Royal BC Museum.
Mayor SJ, Schneider DC, Schaefer JA, Mahoney SP. 2009. Habitat selection at multiple
scales. Écoscience. 16(2):238–247.
[MWLAP] Ministry of Water Land and Air Protection. 2004. Identified wildlife
management strategy accounts and measures for managing identified wildlife
southern interior forest region.
Murphy DD, Noon BR. 1991. Coping with uncertainty in wildlife biology. J Wildl Manag.
55(4):773–782.
Northrup JM, Vander Wal E, Bonar M, Fieberg J, Laforge MP, Leclerc M, Prokopenko CM,
Gerber BD. 2022. Conceptual and methodological advances in habitat-selection
modeling: guidelines for ecology and evolution. Ecol Appl. 32(1).
Parker WS, Brown WS. 1973. Species composition and population changes in two
complexes of snake hibernacula in northern Utah. Herpetol. 29(4):319–326.
Parker WS, Brown WS. 1980. Comparative ecology of two Colubrid snakes, Masticophis t.
taeniatus and Pituophis melanoleucus deserticola, in northern Utah. Milwaukee (WI):
Milwaukee Public Museum.
Posit Software. 2023. RStudio IDE; [accessed 2024 Feb 3].
https://posit.co/products/open-source/rstudio/.
Prior KA, Weatherhead PJ. 1996. Habitat features of Black Rat Snake hibernacula in
Ontario. J Herpetol. 30(2):211–218.
Reinert HK, Cundall D. 1982. An improved surgical implantation method for radiotracking snakes. Source: Copeia. 1982(3):702–705.
Ripley B, Venable B, Bates DM, Hornik K, Gebhardt A, Firth D. 2023. MASS: support
functions and datasets for venables and Ripley’s MASS; [accessed 2023 Apr 7].
https://cran.r-project.org/web/packages/MuMIn/index.html.
Schmidt DA, Govindarajulu P, Larsen KW, Russello MA. 2020. Genotyping-in-thousands
by sequencing reveals marked population structure in Western Rattlesnakes to inform
conservation status. Ecol Evol. 10(14):7157–7172.

52
Shine R, LeMaster MP, Moore IT, Olsson MM, Mason RT. 2001. Bumpus in the snake den:
effects of sex, size, and body condition on mortality of red-sided garter snakes. Evol.
55(3):598–604.
Shine R, Mason RT. 2004. Patterns of mortality in a cold-climate population of garter
snakes (Thamnophis sirtalis parietalis). Biol Conserv. 120(2):201–210.
Therneau TM, Lumley T, Elizabeth A, Cynthia C. 2023. survival: Survival analysis;
[accessed 2024 Feb 3].https://cran.r-project.org/web/packages/survival/index.html
Thomas DL, Taylor EJ. 2006. Study designs and tests for comparing resource use and
availability II. Source: J Wildl Manag. 70(2):324–336.
Trans Mountain. 2021a. Time flies: one year of construction in the Kamloops urban area;
[accessed 2024 Feb 3].
https://www.transmountain.com/news/2021/time-flies-one-year-of-construction-inthe-kamloops-urbanarea#:~:text=What%20a%20year%20it%20has,the%20Trans%20Mountain%20Expa
nsion%20Project.
Trans Mountain. 2021b. Snakes on a plain – salvaging snakes in the BC Interior;
[accessed 2024 Feb 3].
https://www.transmountain.com/news/2021/snakes-on-a-plain-salvaging-snakes-inthe-bc-interior.
Williams KE, Hodges KE, Bishop CA. 2012. Small reserves around hibernation sites may
not adequately protect mobile snakes: the example of great basin gophersnakes
(Pituophis catenifer deserticola) in British Columbia. Can J Zool. 90(3):304–312.
Williams KE, Hodges KE, Bishop CA. 2014. Phenology and demography of great basin
gophersnakes (Pituophis catenifer deserticola) at the northern edge of their range.
Herpetol Conserv Biol. 9(2):246–256.
Williams KE, Hodges KE, Bishop CA. 2015. Hibernation and oviposition sites of Great
Basin Gophersnakes (Pituophis catenifer deserticola) near their northern range Limit.
J Herpetol. 49(2):207–216.
Winger BM, Auteri GG, Pegan TM, Weeks BC. 2019. A long winter for the red queen:
rethinking the evolution of seasonal migration. Biol Rev. 94(3):737–752.
Zappalorti RT, Burger J, Burkett DW, Schneider DW, McCort MP, Golden DM. 2014.
Fidelity of northern pine snakes (Pituophis m. melanoleucus) to natural and artificial
hibernation sites in the New Jersey Pine Barrens. J. Toxicol. Environ. Health - A:
Curr. Issues. 77(21):1285–1291.
Zimova M, Hackländer K, Good JM, Melo-Ferreira J, Alves PC, Mills LS. 2018. Function
and underlying mechanisms of seasonal colour moulting in mammals and birds: what
keeps them changing in a warming world? Biol Rev. 93(3):1478–1498.

53

CHAPTER 3
ARTIFICIAL SNAKE HIBERNACULA PROVIDE SUITABLE MICROCLIMATIC
CONDITIONS BUT ARE NOT READILY ADOPTED
INTRODUCTION
The irrevocable impact of habitat loss on global biodiversity demands innovative
solutions to mitigate the effect (Hanski 2011; Banks-Leite et al. 2020; Gonçalves-Souza et al.
2020). To address this loss, artificial structures have gained popularity to compensate for lost
habitat features (Rueegger 2016; Rueegger 2019; Fontaine et al. 2021; Tillman et al. 2021;
McComb et al. 2022). These structures can include habitats such as nest boxes for birds
(Fontaine et al. 2021; Sudyka et al. 2023), dens for mammals (McComb et al. 2022), ponds
for amphibians (Cook et al. 2023), and refugia for reptiles (Arida and Bull 2008).
Habitat plays a crucial role in the thermoregulatory behaviour of organisms (Long et
al. 2005; Scheffers et al. 2014; Milling et al. 2017). Ideally, the microclimatic conditions of
artificial refugia should model natural habitat, enabling the target species to thermoregulate
and, consequently, improving the overall success of the offset (Rueegger, 2019). However,
studies have shown considerable variation in the quality of the microclimate of artificial
structures compared to their natural counterparts (Griffiths et al. 2018; Fontaine et al. 2021;
McComb et al. 2022; Sudyka et al. 2023). Such research underscores the need to understand
the microclimate of artificial structures before their use or reliance as a conservation or
mitigation measure.
As ectotherms, snakes are unable to metabolically regulate their internal temperature
(Huey and Stevenson 1979; Hein and Guyer 2009). Thus, their behaviour, ecology, and
physiology are largely dictated by available environmental temperatures (Huey and
Stevenson 1979; Hein and Guyer 2009). In temperate regions overwinter, snakes either
succumb to the elements or find a hibernaculum that meets the structural, thermal, and
humidity conditions needed to survive (Brown et al. 1974; Prior and Weatherhead 1996;
Hein and Guyer 2009). Such sites should sufficiently buffer against the external
overwintering conditions to ensure snakes do not freeze and are not expending energy
reserves (Gregory 1982; Costanzo 1989a; Williams et al. 2015; Markle et al. 2020) while
also providing sufficient humidity to prevent desiccation (Costanzo 1986). For instance, the

54
potential dependency on snow insulation has been shown at natural hibernacula for Redsided Garter Snakes (Thamnophis sirtalis parietalis) in Manitoba (Shine and Mason 2004),
where years with lighter snow cover result in major mortality events (Shine and Mason
2004). This dependency showcases the importance of understanding environmental factors
and the specific seasonal requirements of the focal species when building artificial habitats.
The microclimate within artificial snake hibernacula has been investigated relatively
little: research only has investigated the thermal requirements of snakes within incidental
hibernacula created as a by-product of anthropogenic practices (i.e. human-created slopes,
construction debris, abandoned mine features, etc.) or those intentionally created for research
purposes, as opposed to those built for conservation or as mitigation measures (Gillingham
and Carpenter 1978; Costanzo 1986; Bryan 2015; Choquette et al. 2024). For example,
Costanzo (1986) monitored an incidental artificial snake hibernaculum within a buried
cylindrical cistern that provided access to a stable thermal environment within water for
Common Garter Snakes (Thamnophis sirtalis). More recently, unoccupied artificial
hibernacula built out of plumbing hardware buffered against external environmental
variation, while providing low mean temperatures, and high humidity (Choquette et al.
2024). These studies showcase the potential of artificial sites to provide optimal
microclimatic conditions. However, due to differences in application, such results provide
limited insight into the microclimate and its influence on the colonization and efficacy of
artificial snake hibernacula built as mitigation measures.
The northern extremes for snake species (Matsuda et al. 2006) provide optimal
locations to test the thermal suitability of artificial snake hibernacula. Within the Canadian
province of British Columbia, artificial hibernacula are increasingly being turned to as a
means to compensate for the loss (intentional or unintentional) of natural hibernacula for a
number of snake species (MFLNRO 2016; TranBC n.d.). Given this trend, detailed work
comparing the microhabitat afforded by natural and anthropogenic features is becoming
increasingly important. This is imperative given the extreme winters of northern latitudes,
where overwinter survival depends heavily on animals finding and using appropriate
hibernacula. However, data of this nature remain scant.
The recent construction of artificial snake hibernacula following the excavation of
hibernacula for three threatened snake species in British Columbia, Canada, provided a

55
unique opportunity to monitor the thermodynamics of the structures. Our objectives were to
(i) highlight the resulting microclimate within each hibernaculum and assess how well it
buffered against external environmental conditions, (ii) determine whether the internal
thermal conditions of the hibernacula were comparable to those naturally experienced by
snakes, and (iii) monitor the adoption of the main artificial hibernaculum by the original
cohort of displaced snakes.

METHODOLOGY
Study Site
This study was conducted near the city of Kamloops, British Columbia, Canada, in
the Lac du Bois Grasslands Protected Area (51° N, 120.4° W) (BC Parks 2017) (Figure 1.2).
This region experiences some of the driest and warmest conditions in the province, with short
winters and hot summers (BC Parks 2017). The study area features a mosaic of Bluebunch
Wheatgrass (Pseudoroegneria spicata), Big Sagebrush (Artemisia tridentata), and springblooming plants found primarily in lower and middle grasslands (BC Parks 2017). The park
is impacted by anthropogenic activities, such as recreational use, bordering residential
neighbourhoods, livestock grazing, and pipeline construction (GCC 2009; BC Parks 2017;
Trans Mountain 2021a). Refer to Chapter 1 for further details on the climatic and
biogeographic conditions of the study area.
Three threatened snake species, the Great Basin Gophersnake (Pituophis catenifer
deserticola), Western Yellow-bellied Racer (Coluber constrictor mormon), and Western
Rattlesnake (Crotalus oreganus oreganus), all meet their northern range limit near our study
site (Matsuda et al. 2006). Following spring emergence (egress, March-May) and an active
season (April-September) spent foraging and reproducing, the snakes return to their
hibernacula (ingress, August-October) (pers. observ.) to avoid inclement weather and subzero temperatures of winter (October-March) (Costanzo, 1986; Gienger & Beck, 2011; pers.
observ.). Snakes in the study area overwinter in rocky outcrops, talus slopes, rodent burrows,
and anthropogenic features (see Chapter 2).

56
Disturbance Event
Disturbance to natural snake hibernacula occurred during the Trans Mountain
Expansion Project (TMEP), a venture to expand oil transportation in western Canada by
constructing a twin pipeline parallel to the existing one built in 1951 (Trans Mountain
2023a,b). One portion of this project required a linear excavation ~9.2 km long (~30 m wide)
within the boundaries of the Lac du Bois Grasslands Protected Area, starting in 2020 (Trans
Mountain Pipeline ULC 2017; Trans Mountain 2021c).
Wildlife biologists contracted by TMEP conducted reptile sweeps, habitat
assessments, and snake salvages from September 21 to October 16, 2020 (Heinrich 2020).
During construction, three snake hibernacula were excavated within a 1.6km straight-line
distance along the pipeline right-of-way (Heinrich 2020). Prior to October 15, salvaged
snakes were relocated to neighbouring natural hibernacula; these animals included 10
rattlesnakes, 12 gophersnakes, and 13 racers (Heinrich 2020). Unfortunately, as it is not a
regulatory requirement these snakes were not marked for future identification prior to
release. Snakes recovered after October 15 were transferred to the BC Wildlife Park for
artificial overwintering; this sample included 37 gophersnakes and 56 racers (Heinrich 2020).
Snakes were overwintered in containers within fridges maintained at ~4°C (ranging from
2°C to 8°C). The bottom of the container was filled with coconut fibre substrate, and each
container was outfitted with a container of water to provide humidity.
Artificial Hibernacula
Mitigation efforts in the field began in late April to mid-May 2021, when TMEP
designed and constructed two artificial hibernacula to compensate for the lost overwintering
habitat (Figure 1) (Trans Mountain 2021a; Trans Mountain 2021b). Both artificial
hibernacula were offset roughly 50 m south from the original hibernacula locations (Heinrich
2020; Trans Mountain 2021a,b). These locations were selected for their proximity to the
original hibernacula and distance from future disturbance (i.e., pipeline maintenance and
recreational trails) while ensuring a similar microhabitat, aspect, and slope (Heinrich 2020;
Trans Mountain 2020).
Rock and soil salvaged during the excavation of the original hibernacula were
strategically placed at the base of the artificial hibernacula to create interstitial spaces (Trans
Mountain 2020, 2021). Small to medium-sized rocks (0.1-0.3m) were used towards the front

57
of the hibernacula to reduce access to predators, with larger rocks being used towards the rear
(0.3-0.75m) (Heinrich 2020; Trans Mountain 2021b). The rocks were installed at a 2.5%
grade upwards from the front to promote drainage and reduce cold air (Heinrich 2020; Trans
Mountain 2021a,b). A layer of geotextile cloth was placed over the rocks to prevent the 2.53.0m layer of soil placed on top from filling the interstitial spaces (Heinrich 2020; Trans
Mountain 2021a,b). In total, the interior floor area of the first artificial snake hibernacula
(ASH1) was ~ 262m2, and the second (ASH2) was ~ 150m2 (as estimated during
construction). ASH1 was constructed at an elevation of ~520m, and ASH2 at ~480m.
During the construction of the artificial hibernacula, PVC standpipes were added to
allow the monitoring of internal temperature and humidity data (Figure 3.1). Nine standpipes
were added to ASH1, and eight to ASH2 (Figure 3.1). For each hibernaculum, temperature
standpipes were arranged into three rows (Back, Middle, Front – see Figure 3.1). The back
standpipes of ASH1 reach an average depth of 4.0m (ASH2 2.8m), the middle 3.6m (ASH2
3.0m), and the front 3.1m (ASH2 2.6m). HOBO U23 Pro v2 Temperature and Relative
Humidity Data Loggers (ONSET, Bourne, MA, USA) were deployed into each standpipe to
monitor the internal conditions every hour.
Snake Release
All snakes overwintering at the BC Wildlife Park were released at ASH1 between
April 31 and May 3, 2021. No snakes were released at ASH2 as construction on that site was
not completed until mid-May, after the natural egress of snakes in the region had been
completed (McKelvey and Ragsdale, pers. obs.). All snakes released at ASH1 were
implanted with passive integrated transponders (PIT tags) (Biomark Inc., Boise, ID, United
States) to permit individual identification, and relatively larger animals were selected for
radiotelemetry (see Radiotelemetry below). A ‘soft release’ approach was used to reintroduce
snakes to the wild (Mitchell et al. 2011): To do this, a 1.8 m tall fence, created with
overlapped chicken wire (outside) and window screen mesh (inside), was erected around the
hibernacula mouth. To prevent escape, the foot of the fence was buried roughly 10 cm
underground and the top was equipped with a 10 cm fold (90°) to hinder climbing.

Figure 3.1: Schematic of (A) artificial hibernaculum 1 and (B) artificial hibernaculum 2. The figures in the top row provide an
overhead view and shows the spatial relationship of the temperature standpipes in reference to each other and the hibernaculum
mouth. The figures in the bottom row provide side profiles. ‘Back’ refers to temperature standpipes in the same row furthest from
the hibernaculum entrance, with the front row being those closest to the hibernaculum mouth. All figures were adapted from Trans
Mountain (2021b,c). Not drawn to scale.
58

59
Radiotelemetry
Radiotelemetry was employed to facilitate a concurrent study on the movement
ecology of the translocated gophersnakes and racers (Ragsdale MSc in prep) but also helped
this study record the overwintering temperatures of snakes. In 2021, telemetered snakes
included a subset released at ASH1 plus ‘reference snakes’. Reference snakes in 2021 and all
telemetered snakes in 2022 were captured during hibernacula surveys or opportunistically
during the active season. Snakes released at natural hibernacula and not marked following the
disturbance event (see Disturbance Event above) could be included in our sample of
reference snakes, as there was no way to differentiate these individuals reliably. Telemetered
racers were excluded from this study. Due to their smaller size, racers often could not carry
transmitters large enough to facilitate tracking to hibernacula. No rattlesnakes were tracked,
as all displaced rattlesnakes were moved to natural hibernacula without prior identification.
Surgically-implanted transmitters did not weigh more than 4% of the body weight of
the captured snakes (SB-2T 3.8g six-month battery, or SB-2T 5.2g 12-month battery; Holohil
Systems Ltd., Carp, ON, Canada) (Bryant et al. 2010). Select gophersnakes in 2022 were
implanted with SB-2T (12-month battery) transmitters with a temperature data logger
(iButton DS1925L-F5 3g; Maxim Integrated Products Inc., San Jose, CA, United States)
modification (9.5g, see Figure 3.2). After programming to record hourly temperature, the
iButton was glued to the antenna-less end of the radio transmitter (Figure 3.2) (Hobbs 2007;
Nordberg and Cobb 2017; Crowell 2019). Beeswax was added to streamline the connection,
and the entire transmitter+iButton unit was coated twice in rubber Plasti Dip (Plasti Dip
International, Inc, Blaine, MN, USA) to waterproof it (Figure 3.2) (Taylor et al. 2004; Hobbs
2007; Crowell 2019). Implantations were performed by licensed veterinarians following
modified methodologies similar to Reinert and Cundall (1982). Implantations occurred just
over midway between the distance from the cloacal vent to the snout of each animal. The
antennae were fed caudally under the skin rather than cranially. Individuals were tracked
every 2-3 days throughout the active season, from April to October. Monitoring continued
until the animals selected hibernacula, which occurred when an individual persisted in a
location from fall into winter. In winter 2021-2022, snakes with temperature-sensitive radio
transmitters were tracked biweekly from ingress to egress to obtain temperature. Individuals
were caught for transmitter removal in the ensuing egress.

60

A)

C)

D)

B)

Figure 3.2: SB-2T radio transmitter (Holohil Systems Ltd., Carp, ON, Canada) and
DS1925L-F5 iButton (Maxim Integrated Products Inc., San Jose, CA, United States)
modification. A) The transmitter prior to modification. B) The iButton prior to modification.
C) The iButton and transmitter after one coat of rubber Plasti Dip and beeswax. D) The final
product with another layer of rubber Plasti Dip (note that the tape held a magnet on the
battery and was removed before implantation). In all photos the antenna has been cropped
(actual length was 22cm). Photos by author.

61
Hibernacula Monitoring
During egress and ingress, hibernacula surveys took place daily at ASH1. Natural
hibernacula were coincidentally monitored during these periods to mark when snakes had
started naturally returning or leaving. During these periods, the fence was erected and
equipped with funnel traps (Figure 3.3). The direction of the funnels and the coupling of
mesh holding boxes (traps) allowed us to intercept all snakes attempting to leave (egress) or
enter (ingress) the hibernaculum. Funnels equipped with traps were monitored twice daily (in
the morning when entering the field and late afternoon when leaving the field) and closed if
(i) overnight temperatures dropped below 5°C or (ii) the monitoring schedule could not be
adhered to. Cover boards provided the traps with shade from the elements, and a synthetic
upholstery covered the exposed trap floors (Burgdorf et al. 2005). The fencing apparatus on
ASH1 was erected well before the first observed snakes at natural hibernacula and kept in
place until the emergence of snakes at natural hibernacula was complete (i.e. no snakes were
detected and/or the same individuals were recaptured across multiple visits).
To establish the artificial hibernacula and ensure displaced snakes reached
overwintering habitat, snakes found in this area previously impacted by the disturbance were
actively moved into the ASH1 fence line during ingress 2021. Coincident to our own work,
Trans Mountain had the responsibility to move any snakes (both displaced and reference)
found in the pipeline construction footprint to the closest hibernacula (by default ASH1)
during ingress 2021. The following ingress, we monitored for the natural return of snakes,
and since construction was over, no snakes were forcefully moved to ASH1 in 2022. Any
captured individuals not already tagged were subcutaneously implanted with a PIT tag.
External Conditions
In conjunction to monitoring the internal artificial hibernacula conditions, HOBO
TidbiT v2 Water Temperature Data Loggers (ONSET, Bourne, MA, USA) were placed in the
shade around the study site to monitor external air temperature. Wildlife cameras (Reconyx,
Holmen, WI, United States) assessed snow depth on top of and in front of both artificial
hibernacula mouths by taking pictures of mounted meter sticks every hour throughout winter.

Figure 3.3: Funnels traps used to catch snakes at artificial snake hibernaculum 1 (A) Spring emergence (egress) orientation to catch
snakes as they exit, (B) Fall return (ingress) orientation to catch snakes as they return. Arrows indicate the direction of snake
movement. Not drawn to scale.

62

63
Statistical Analysis
Daily average temperature and humidity readings were used for all subsequent
analyses as the mean of the 24 consecutive hourly data points recorded on a given calendar
day (Amat-Valero et al. 2014). For each artificial hibernaculum, temperature and humidity
recorded by loggers within the same row were averaged together (Back, Middle, Front – see
Figure 3.1). Weekly or monthly oscillation was calculated as the difference between the
maximum and the minimum value in that period (Amat-Valero et al. 2014). All statistical
analyses were performed using RStudio (Posit Software 2023).
Environmental Variation
Data from November 1st, 2021, to March 31st, 2022, were used to determine the
impact of external conditions on the internal microclimate. External air temperature profiles
collected across the site were averaged together (Griffiths et al. 2018). Relative humidity of
the broader environment was measured at the Kamloops A (50 ° N, 120 ° W) weather station
(ECCC 2023). We assessed the temporal coherence of the average daily temperature and
humidity between outside and inside conditions for both hibernacula using Spearman’s
coefficient (Hachem et al. 2012; Nordberg and Cobb 2017). To determine the importance of
snow insulation on internal hibernacula conditions, we compared weekly temperature
oscillation for each hibernaculum at every position for time with snow and without snow
utilizing either an unpaired t-test or a Wilcoxon rank sum test depending on normality.
Temperature of Snakes
The temperatures of telemetered snakes that overwintered in ASH1 in 2021-2022
were compared to those of reference and displaced snakes that overwintered elsewhere in
natural hibernacula. This comparison was conducted using either an ANOVA or a KruskalWallis test, depending on normality.
To determine if and where the artificial hibernacula met the thermal requirements for
overwintering snakes, the average daily temperature at the different positions within the
artificial hibernacula was compared against the average, maximum, and minimum daily
snake body temperature obtained through transmitter+iButton implantation from November
1st, 2022, to March 31st, 2023. All comparisons were made using a repeated measures
ANOVA or non-parametric Friedman rank sum test depending on normality (Thierry et al.

64
2009; Roznik and Alford 2012). In all cases, temperature or humidity was the dependent
variable, position (or snake) was the factor, and time between measures was the repeated
measure (Roznik and Alford 2012). If significant, a pairwise t-test or exact p-values post hoc
test was performed (Eisinga et al. 2017; Pohlert 2023). To determine what region of the
artificial hibernacula best corresponded to the temperatures being experienced by snakes in
natural hibernacula, we superimposed the snakes’ temperature profiles with the temperature
profiles of the artificial hibernacula over time (Forget-Klein and Green 2021).
RESULTS
Microclimate of Artificial Hibernacula
Data from November 1st, 2021 - March 31st, 2022 were used to compare the
overwintering microclimate between and within the artificial hibernacula. All nine deployed
loggers were recovered from ASH1, three in each position (Figure 3.1). At ASH2, a failure in
the ropes used to retrieve the monitors enabled only six of the eight deployed loggers to be
recovered, two in each position (Figure 3.1). Overall, the artificial hibernacula experienced
average interior winter temperatures of 9.3°C (ASH1) and 7.9°C (ASH2), ranging from
3.3°C to 16.6°C (ASH1) and 2.5°C to 14.2 °C (ASH2) (Figure 3.4). The back of ASH1 was
warmer on average (10.7±1.9°C) than the middle (9.6±2.5°C) or front (7.9±2.5°C), with
lower average monthly oscillation (1.3±0.5°C) than the middle (1.7±0.7°C) or front
(2.3±0.7°C). Similarly, the back of ASH2 (8.8±2.4°C) was warmer on average than the
middle (8.1±2.2°C) and the front (6.9±2.2°C), with lower average monthly oscillation
(2.0±1.1°C) than the middle (1.7±0.9) or front (2.4±0.7). The artificial hibernacula
experienced average relative humidity of 99.5% (ASH1) and 97.4% (ASH2), ranging from
84.7% (ASH1) and 77.4% (ASH2) to 100% (Figure 3.5). No obvious trend was found in the
average humidity between positions, with similar average humidity recordings in the back
(ASH1 99.9±0.2%; ASH2 99.9±0.02%), middle (ASH1 100.0±0.0%; ASH2 93.7±4.9%) and
front (ASH1 98.7±1.7%; ASH2 98.5±2.5%). However, the front of ASH1 had the most
unstable humidity (3.9±3.6%) with higher average monthly oscillation than the middle
(0.0±0.0%) or back (0.3±0.3%). While the middle (12.4±6.5%) of ASH2 had the most
unstable monthly humidity compared to the back (1.4±3.2%) and front (7.5±6.1%).

65
Effect of Environmental Variation
From November 1, 2021, to March 31, 2022, six loggers recorded ambient air
temperature across the study site, averaging -0.4°C and ranging from 27.4°C to -30.4 °C
(Figure 3.4). During this time, relative humidity measured at the Kamloops A (50 ° N, 120 °
W) weather station (ECCC 2023) was, on average, 72.8%, ranging from 99.0% to 16.0%
(Figure 3.5). All correlations between external ambient and internal hibernacula temperatures
were found to be negative, weak correlations with Spearman’s ρ ranging from -0.04 to -0.25
(ASH1 back and middle, ρ= -0.25, P=<0.01; ASH1 front, ρ= -0.22, P=<0.01; ASH2 Back, ρ=
-0.21, P=<0.01; ASH2 middle, ρ= -0.16, P=0.06; ASH2 front, ρ= -0.04, P=0.65). Similarly,
correlation between external ambient and internal hibernacula humidity were found to be
weak correlations ranging from -0.05 to 0.29 (ASH1 back, ρ= 0.29, P=<0.01; ASH1 front, ρ=
-0.26, P=<0.01; ASH2 back, ρ= 0.07, P=0.43; ASH1 middle, ρ= -0.1, P=0.24; ASH1 front ρ=
-0.05, P=0.56). No Spearman correlation coefficient could be determined for the middle of
ASH1 as it sat consistently at 100% relative humidity throughout this period. Although some
of these correlations were significant, the relationships were still weak or negative, indicating
low and sometimes inverse correlation with external conditions. The lack of temporal
coherence was further supported during a cold snap between December 22, 2021, and
January 3, 2022, as internal hibernacula conditions remained relatively stable (Figure
3.4). When the external temperature dropped to a minimum of -30.4°C on December 27,
2021, internal conditions ranged from 8.3°C to 11.7°C (ASH1) and 6.1°C to 9.4°C (ASH2)
(Figure 3.4). External humidity reached a minimum of 16% on March 8, 2022, and internal
conditions in ASH1 ranged from 98.4.0 to 100%, and ASH2 ranged from 94.5% to 100.0%.
Due to the failure of two cameras overwinter (top and bottom of ASH2), the average
reading of snow depth was retrieved from the wildlife cameras positioned on the top and
bottom of ASH1. No significant variance was seen in the fluctuation of weekly temperatures
between periods with snow and without snow from November 1, 2021, to March 31, 2022,
across the various datalogger positions and hibernacula (Appendix D).

66

Figure 3.4: The average daily winter temperature from November 1, 2021, to March 31, 2022, for (A) each position within
artificial snake hibernaculum 1 (ASH1) and artificial snake hibernaculum 2 (ASH2) and (B) the external ambient air. ‘Position’
refers to the distance of loggers to the hibernaculum mouth, where ‘Back’ is the furthest and ‘Front’ is the closest (Figure 3.1).
Each line is the average of the loggers (n=3 for each position in ASH1, n=2 for each position in ASH2, and n=6 for external). The
cloud around the ambient line represents the daily maximum and minimum. The solid vertical lines indicate a severe cold interval
from December 22, 2021, to January 3, 2022, with the dashed vertical line showcasing the minimum external temperature
(December 27, 2021).

Figure 3.5: The average daily relative humidity from November 1, 2021, to March 31, 2022, for (A) each position within artificial
snake hibernaculum 1 (ASH1) and artificial snake hibernaculum 2 (ASH2) and (B) the external ambient air. ‘Position’ refers to the
distance of loggers to the hibernaculum mouth, where ‘Back’ is the furthest and ‘Front’ is the closest (Figure 3.1). Each line is the
average of the loggers (n=3 for each position in ASH1, n=2 for each position in ASH2). Ambient relative humidity data was
obtained from the Kamloops A (50 ° N, 120 ° W) weather station (ECCC 2023). The cloud around the ambient line represents the
daily maximum and minimum.
67

68
Temperature of Snakes
Comparison of snakes in artificial and natural hibernacula
Overall, 13 telemetered gophersnakes had their body temperatures recorded remotely
on five occasions from December 6, 2021, to February 1, 2022. One snake in a natural
hibernaculum was omitted from the comparison due to missing values during this period.
Temperatures taken after February 1, 2022, were omitted when many telemetry batteries had
expired. During this period, the two snakes overwintering in ASH1, on average, 12.5°C,
ranging from 9.6°C to 14.9°C. Snakes previously displaced and overwintering in natural
hibernacula (n= 6) were, on average, 11.4°C, ranging from 7.1°C to 15.3°C. Reference
snakes in natural hibernacula (n=4) were, on average, 13.0°C, ranging from 7.0°C to 18.9°C.
Overall, gophersnakes overwintering in ASH1 did not experience significantly different
overwintering temperature from snakes naturally overwintering on the same landscape
(ANOVA, F2,57=2.90, P=0.06).
Artificial hibernacula conditions versus naturally overwintering snakes
Four of 7 gophersnakes implanted with radio transmitters+iButtons were successfully
retrieved following the 2022-2023 winter. All individuals overwintered in different natural
hibernacula. The overwintering body temperatures of these snakes were, on average, 7.8 °C,
ranging from 4.5°C to 17.0°C from November 1st, 2022, to March 31st, 2023 (Figure
3.6). At the same time, only five loggers were retrieved from each artificial hibernacula: two
from the front of ASH2, three from the middle of ASH1, one from the middle of ASH2, and
two from the front of each hibernacula. Internal temperatures were, on average, 9.2°C
(ASH1) and 7.9°C (ASH2), ranging from 4.2°C to 16.3°C (ASH1) and 2.4°C to 16.1°C
(ASH2) (Figure 3.6).
The overall difference between average daily temperature profiles at different
positions within the artificial hibernacula and the maximum, minimum, and average
temperature of snakes was significant (Friedman rank sum test, χ 2 = 950.35 (7), P< 2.2e-16)
(Figure 3.6; Appendix E). All positions were significantly different from naturally
overwintering snakes except the average daily temperature of snakes with the front of ASH1
(Post hoc exact test, P= 0.3674) and the middle of ASH2 (Post hoc exact test, P= 0.1107), the
back of ASH2 and the maximum daily snake temperature (Post hoc exact test, P= 0.5414),

69
and the front of ASH2 and the minimum daily snake temperature (Post hoc exact test, P=
1.0000) (Figure 3.6; Appendix E). This is reflected in a comparison of the thermal profile of
the average, overall maximum and overall minimum of these gophersnakes with the thermal
profiles of the artificial hibernacula (Figure 3.6). In particular, the average daily temperature
of these snakes mirrored the temperature profile of the front of ASH1 and the middle of
ASH2, found at depths of 3.1m and 3m, respectively. The maximum snake temperatures
overwinter mirrored the temperature profile of the back of ASH2 found at a depth of 2.8m,
and the minimum snake temperatures mirrored the front of ASH2 at a depth of 2.6m.
Snake Adoption of ASH1
Overall, 41 snakes (9 gophersnakes, 29 racers, and 3 rattlesnakes) were placed in
ASH1 in the fall of 2021. Of these animals, 19 (4 gophersnakes, 15 racers) were from the
original displaced population. All of these snakes were introduced by the research team or
consultants, except for 3 individuals found trying to enter the fence or funnel traps on their
own (1 gophersnake, 2 racers), with the gophersnake and one racer being from the original
displaced population. Only 7 of these 41 snakes (3 gophersnakes, 4 racers), 6 being from the
displaced population (3 gophersnakes, 3 racers), were confirmed to have overwintered when
caught the following spring. We cannot comment on the overall survival of snakes, as during
ingress (Fall) 2021, we observed at least 15 individuals, previously placed inside the fenced
enclosure, outside of ASH1 (i.e. escaping the fence line, or finding an alternative exit).
Although the snakes were placed back into the ASH1 enclosure if found, it is unclear how
many left during this time. Three newly identified snakes naturally came to ASH1 in the fall
of 2022 (2 rattlesnakes and 1 racer), and the two rattlesnakes were confirmed to have
overwintered in the spring of 2023.

A)

B)

Figure 3.6: The average daily winter temperature from November 1, 2022, to March 31, 2023 of (A) Great Basin Gophersnakes
hibernating in natural hibernacula superimposed on the temperature profiles of each position within artificial snake hibernaculum 1
(ASH1) and artificial snake hibernaculum 2 (ASH2) and (B) the external ambient air. ‘Position’ refers to the distance of loggers to
the hibernaculum mouth, where ‘Back’ is the furthest and ‘Front’ is the closest (Figure 3.1). Each line is the average of the loggers
(2 from the front of ASH2, 3 from the middle of ASH1, 1 from the middle of ASH2, 2 from the front of each hibernaculum, n=4
snakes, and 8 ambient temperature loggers). The grey clouds represent the daily maximum and minimum for snakes and ambient.
70

71

DISCUSSION
To the best of our knowledge, this study provides the first detailed analysis of the
internal microclimate of artificial snake hibernacula purposely built as mitigation measures.
Both hibernacula provided various thermal and humidity microenvironments, with
temperatures generally increasing and becoming more stable with greater depth and distance
from the hibernaculum mouth, while humidity did not appear to relate to depth. These
microsites were not strongly impacted by external environmental fluctuation and sufficiently
buffered against ambient thermal and humidity minima. Snow did not play a role in the
insulation of the hibernacula, suggesting the sites would be thermally secure in winters with
less snowfall. Temperatures of overwintering Great Basin Gophersnakes within ASH1 were
comparable to those in natural hibernacula. In particular, the artificial hibernacula provided
thermal microsites conducive to the survival of gophersnakes at depths ranging from 2.6 to
3.1m. Over the two study years, nine snakes were confirmed to have overwintered
successfully within ASH1, providing additional evidence of favourable microenvironments
for survival. Although conducive to survival, the initial adoption rate was poor, and will
require additional research to understand the long-term use.
The artificial hibernacula offer a range of temperature that snakes could potentially
exploit over winter. All measured positions within the artificial hibernacula remained well
above freezing throughout winter, with the rear of each hibernacula providing the warmest
and most thermally-stable conditions. In general, as the depth of each thermal datalogger
increased, the temperature increased, and temperature oscillation decreased. Further, no
position within either hibernaculum positively tracked external ambient temperature changes
over time. A severe cold interval occurred from December 22, 2021, to January 3, 2022, with
temperatures outside reaching values as low as -30.4°C, and during this event, temperatures
within both hibernacula remained above a minimum reading of 6.1°C. Such results indicate
the thermal security of these sites against the most extreme temperatures during these
northern winters. Overall, thermal dynamics will be partially attributable to the lagged
response of soil temperature at increased depths to external conditions (Al-Kaisi et al. 2017).
Soil temperature lags with increasing depth due to a damping effect where heat is absorbed
or released through the soil (Al-Kaisi et al. 2017). The distance away from the hibernacula
entrances also may influence temperature. Drafts and airflow may be more apparent towards

72
the front of the hibernacula and depend on the overall structure (Nordberg and Cobb 2017).
Building these structures with a suite of depths and distances from the hibernaculum mouth
may facilitate hibernation (Macartney et al. 1987; Lutterschmidt et al. 2006). As snakes enter
the hibernaculum in the fall, they may shift to deeper depths as shallower depths cool and
move with the reverse thermal gradient as they emerge to shallower and warmer depths in the
spring (Sexton and Hunt 1980; Sexton and Marion 1981; Macartney et al. 1987;
Lutterschmidt et al. 2006). Further, thermal variation may benefit snakes by providing a
selection of potential microsites. This is an important consideration when building habitat, as
individual snakes and species may vary in their selection and/or move between microsites
during winter (Macartney et al. 1987; Lutterschmidt et al. 2006; Nordberg and Cobb 2017).
Although snow insulates soil (Shine and Mason 2004), it had no apparent impact on
temperature oscillation within our artificial hibernacula. This may have resulted from
sufficient insolation being provided by the soil at these depths, paired with the short and
relatively mild Canadian winter experienced in this region (Environment Canada 2011; BC
Parks 2017): the nearby city of Kamloops, British Columbia, is ranked thirteenth (out of 100
cities assessed) for the mildest winter in Canada (Environment Canada 2011). Elsewhere,
years with lighter snow cover have resulted in major mortality events of Red-sided Garter
Snakes in natural hibernacula in Manitoba, Canada (Shine and Mason 2004). In New York,
decreased snow cover was suggested to potentially impact the overwintering temperature of
Timber Rattlesnakes (Crotalus horridus) (Brown 1982). Although our hibernacula do not
appear to be impacted by snow, future artificial snake hibernacula should factor this into the
design in similar or colder climes. This is especially important given the ongoing and
increasing impacts of climate change. For example, snow cover duration in Canada has been
decreasing (GC 2019), and although the overall increasing temperature may compensate for
this loss, less snow paired with variable and extreme cold events could be catastrophic (GC
2023).
Great Basin Gophersnakes overwintering in ASH1 in 2021-2022 experienced similar
temperatures to those overwintering in natural hibernacula. This indicates that ASH1
provided conditions that were not only thermally conducive to survival (as these snakes
emerged) but were within a range naturally experienced by snakes overwintering on the same
landscape. Hibernacula need to afford conditions above some critical thermal minimum (to

73
ensure snakes do not freeze) and maximum (not be too warm to minimize energy
expenditures) (Gregory 1982; Costanzo 1989a; Williams et al. 2015; Markle et al. 2020;
Huey et al. 2021). Although all positions within both hibernacula afforded temperatures well
above freezing overwinter, only some appeared to be biologically relevant to the
gophersnakes. The average daily temperature of naturally overwintering gophersnakes in
2022-2023 mirrored the temperature profiles in the artificial dens at depths from 2.6 to 3.1m.
Overall, the overwintering body temperatures of these gophersnakes ranged from 4.5°C to
17.0°C and were, on average, 7.8 °C. However, temperatures that are potentially damaging
appear to be subject to change with geography and the focal species. For example, Red-sided
Garter Snakes (Thamnophis sirtalis parietalis) in Manitoba reached a thermal minimum of
1.1 ± 0.16°C overwinter (Lutterschmidt et al., 2006), while Macartney et al. (1987) found the
average overwintering temperature of this species in the Northwest Territories to be 3.9°,
hitting a minimum of 1.8°C. Further, the lowest Timber Rattlesnake (Crotalus horridus)
temperature in Tennessee recorded by Nordberg & Cobb (2017) in one year was 5.9°C and
the next 1.1°C. However, Hobbs (2007) found Western Rattlesnakes overwintered on
average at 9.6 °C and ranged between 17.8 to 5.1°C on our study site. These results are very
similar to those experienced by our gophersnakes, which in some cases share the same
hibernacula as Western Rattlesnakes (COSEWIC 2013; Williams et al. 2015). Given this
variation, when building artificial snake hibernacula the temperature requirements of the
focal species in that location must be considered. Thus, having reference data from naturally
hibernating snakes in a given area provides context and guidelines for the construction and
assessment of artificial den performance.
The natural weight loss of snakes overwinter is believed to be mostly a result of water
loss (Costanzo 1989a). Thus, humidity is another important factor to consider when
constructing an artificial snake hibernaculum to ensure there is enough moisture to prevent
the desiccation of snakes (Costanzo 1986; Costanzo 1989a). Humidity, as we recorded inside
the artificial hibernacula, did not appear to relate to depth and for unknown reasons, the
middle section of ASH2 was variable compared to other positions. Regardless of this
variation, no position co-varied strongly in time with external relative humidity. Instead, the
hibernacula buffered against external humidity changes: even when ambient, above-ground
measurements reached a low of 16%, conditions inside the artificial hibernacula never

74
dropped below 77%. The adverse effects of water loss due to the low humidity appears to be
more of a concern than overly high humidity, given some snakes are known to overwinter
submerged in water (Costanzo 1989a). Since the hibernacula stayed above 77.4%, with many
regions remaining at or around 100%, the artificial hibernacula appeared conducive to
overwinter survival by the snakes.
Despite affording apparent suitable hibernating conditions, the ASH1 failed to serve
as replacement habitat for the 93 displaced snakes, as documented during the first- and
second-year post-disturbance. This occurred even with our intervention and monitoring. For
example, during ingress 2021, ASH1 was actively seeded by 19 displaced snakes. However,
only six of these snakes were determined to have successfully overwintered. In year 2, only
three snakes in total were detected at the hibernacula: two overwintered, none were displaced
snakes, and none were returning from the first year. These results suggest that costly artificial
dens will not necessarily achieve mitigation targets, at least in the short term. Without a
concerted monitoring program (as conducted herein) this modicum of success will not be
easily achieved or quantified. Unfortunately, it is unclear how long it takes for snakes to
adopt a hibernaculum naturally, and other mechanisms beyond our knowledge and control
may be at work (e.g., scent trails, visual cues, etc.) (Parker and Brown 1980; Costanzo 1986;
Costanzo 1989b; Gienger and Beck 2011; Zappalorti et al. 2014). Thus, the value of artificial
hibernacula and similar structures requires further research assessing their viability in the
long term, perhaps decades.
LITERATURE CITED
Al-Kaisi MM, Lal R, Olson KR, Lowery B. 2017. Chapter 1 - fundamentals and functions
of Soil Environment. In: Al-Kaisi MM, Lowery B, editors. Soil Health and
Intensification of Agroecosytems. Massachusetts (USA): Academic Press. p. 1–23.
Amat-Valero M, Calero-Torralbo MA, Václav R, Valera F. 2014. Cavity types and
microclimate: implications for ecological, evolutionary, and conservation studies. Int
J Biometeorol. 58(9):1983–1994.
Arida EA, Bull MC. 2008. Optimising the design of artificial refuges for the Australian
skink, Egernia stokesii. Appl Herpetol 5(2):161–172.
Banks-Leite C, Ewers RM, Folkard-Tapp H, Fraser A. 2020. Countering the effects of habitat
loss, fragmentation, and degradation through habitat restoration. One Earth. 3(6):672–
676.
BC Parks. 2017. Lac du Bois Grasslands Protected Area management plan. Kamloops.

75
Brown WS. 1982. Overwintering body temperatures of Timber Rattlesnakes (Crotalus
horridus) in northeastern New York. Source: J Herpetol. 16(2):145–150.
Brown WS, Parker WS, Elder JA. 1974. Thermal and spatial relationships of two species
of Colubrid snakes during hibernation. Herpetol. 30(1):32–38.
Bryan DL. 2015.Survival and natural history of the Timber Rattlesnake (Crotalus Horridus)
at Center Hill Lake in Dekalb County, Tennessee [thesis]. [Cookville (TN)]:
Tennessee Technolgical University.
Bryant GL, Eden P, De Tores P, Warren K. 2010. Improved procedure for implanting
radiotransmitters in the coelomic cavity of snakes. Aust Vet J. 88(11):443–448.
Burgdorf S, Rudolph D, Conner R, Saenz D, Schaefer R. 2005. A successful trap design for
capturing large terrestrial snakes. Herpetol Rev. 36(4):421–424.
Burger J, Zappalorti RT, Gochfeld M, Boarman WI, Caffrey M, Doig V, Garber SD, Lauro
B, Mikovsky M, Safina C, et al. 1988. Hibernacula and summer den sites of Pine
Snakes (Pituophis melanoleucus) in the New Jersey Pine Barrens. J Herpetol.
22(4):425–433.
Choquette JD, Mokdad AI, Pitcher TE, Litzgus JD. 2024. Selection and validation of release
sites for conservation translocations of temperate-zone snakes. MethodsX. 12:1-11.
Cook DG, Stemle LR, Stokes DL, Messerman AF, Meisler JA, Searcy CA. 2023. Habitat
value of constructed breeding pools for the endangered Sonoma population of
California tiger salamander. J of Wildl Manag. 87(3): 1-18.
COSEWIC. 2013. COSEWIC Assessment and status report on the Great Basin Gopersnake
Pituophis catenifer deserticola in Canada. Ottawa.
Costanzo JP. 1986. Influences of hibernaculum microenvironment on the winter life history
of the garter snake (Thamnophis sirtalis). OJS. 86(5):199–204.
Costanzo JP. 1989a. Effects of humidity, temperature, and submergence behavior on
survivorship and energy use in hibernating garter snakes, Tharnnophis sirtdis. Can J
Zool. 67:2486–2492.
Costanzo JP. 1989b. Conspecific scent trailing by garter snakes (Thamnophis sirtalis) during
autumn further evidence for use of pheromones in den location. J Chem Ecol. 15(11):
2531-2538.
Crowell HL. 2019. Comparative thermal ecology of coastal and inland Pacific Rattlesnakes
(Crotalus Oreganus) [thesis]. [San Luis Obispo (CA)]: California Polytechnic State.
Dong H, Zhao J, Zhao W, Si M, Liu J. 2019. Study on the thermal characteristics of crude oil
pipeline during its consecutive process from shutdown to restart. Case Stud Therm
Eng. 14: 1-20
Edkins TL, Somers CM, Vanderwel MC, Sadar MJ, Poulin RG. 2018. Variable habitat
selection and movement patterns among Bullsnake (Pituophis catenifer sayi)
populations in Saskatchewan. Can Field-Nat. 132(2):126–139.

76
Eisinga R, Heskes T, Pelzer B, Te Grotenhuis M. 2017. Exact p-values for pairwise
comparison of Friedman rank sums, with application to comparing classifiers. BMC
Bioinform. 18(1): 1-18.
Environment Canada. 2011. Mildest winter. National Climate Data and Information Archive;
[accessed 2024 Mar 22].
https://web.archive.org/web/20111125151335/http://climate.weatheroffice.gc.ca/winn
ers/categorydata_e.html?SelectedCategory=4.
[ECCC] Environment and Climate Change Canada. 2023. Historical climate data; [accessed
2023 Jul 18]. http://climate.weather.gc.ca/.
Eye DM. 2022. Reproductive ecology of female Western Rattlesnakes (Crotalus Oreganus)
in Canada [thesis]. [Kamloops (BC)]: Thompson Rivers University.
Fontaine A, Simard A, Dubois B, Dutel J, Elliott KH. 2021. Using mounting, orientation, and
design to improve bat box thermodynamics in a northern temperate environment. Sci
Rep. 11(1): 1-15.
Forget-Klein É, Green DM. 2021. Toads use the subsurface thermal gradient for temperature
regulation underground. J Therm Biol. 99: 1-9.
Gonçalves-Souza D, Verburg PH, Dobrovolski R. 2020. Habitat loss, extinction
predictability and conservation efforts in the terrestrial ecoregions. Biol Conserv. 246:
1-8.
[GC] Government of Canada. 2019. Changes in snow;[accessed 2024 Feb 6].
https://www.canada.ca/en/environment-climate-change/services/climatechange/canadian-centre-climate-services/basics/trends-projections/changessnow.html.
[GC] Government of Canada. 2023. Climate change. Canadian Centre for Occupation Health
and Safety; [accessed 2024 Mar 22].
https://www.ccohs.ca/oshanswers/safety_haz/climate/climate-change-extremeweather-cold.html.
[GCC] The Grasslands Conservation Council of British Columbia. 2009. Working together
for the conservation of BC’s grasslands. Kamloops.
Gillingham JC, Carpenter CC. 1978. Snake Hibernation: Construction of and observations
on a man-made hibernaculum (Reptilia, Serpentes). J Herpetol. 12(4):495–498.
Gienger CM, Beck DD. 2011. Northern pacific rattlesnakes (Crotalus oreganus) use thermal
and structural cues to choose overwintering hibernacula. Can J Zool. 89(11):1084–
1090.
Gregory P. 1982. Reptillian hibernation. In: Gans C, Pough FH, editors. Biology of the
Reptillia. Vol. 13. New York (USA): Academic Press. p. 53–154.
Griffiths SR, Lentini PE, Semmens K, Watson SJ, Lumsden LF, Robert KA. 2018.
Chainsaw-carved cavities better mimic the thermal properties of natural tree hollows
than nest boxes and log hollows. Forests. 9(5): 1-27.

77
Hachem S, Duguay CR, Allard M. 2012. Comparison of MODIS-derived land surface
temperatures with ground surface and air temperature measurements in continuous
permafrost terrain. TC. 6(1):51–69.
Hanski I. 2011. Habitat loss, the dynamics of biodiversity, and a perspective on conservation.
Ambio. 40(3):248–255.
Hein AM, Guyer C. 2009. Body temperature of overwintering Cottonmouth snakes:
hibernaculum use and inter-individual variation. J Ala Acad Sci. 80(1): 35-44.
Heinrich R. 2020. Technical memorandum. Kamloops (BC): Triton Environmental
Consultants.
Hobbs J. 2007. Thermal factors in relation to the denning ecology of northern Pacific
rattlesnakes in British Columbia [thesis]. [Victoria (BC)]: Royal Roads University.
Huey RB, Stevenson RD. 1979. Integrating thermal physiology and ecology of ectotherms:
a discussion of approaches. Am Zool. 19(1):357–366.
Huey RB, Ma L, Levy O, Kearney MR. 2021. Three questions about the eco-physiology of
overwintering underground. Ecol Lett. 24(2):170–185.
Larsen KW. 1986. Ecology of the common garter snake, Thamnophis sirtalis, at the northern
limit of its range. [Victoria]: University of Victoria.
Long RA, Martin TJ, Barnes BM. 2005. Body temperature and activity patterns in free-living
arctic ground squirrels. J Mammal. 86(2):314–322.
Lutterschmidt DI, LeMaster MP, Mason RT. 2006. Minimal overwintering temperatures of
red-sided garter snakes (Thamnophis sirtalis parietalis): A possible cue for
emergence? Can J Zool. 84(5):771–777.
Macartney JM, Larsen KW, Gregory PT. 1987. Body temperatures and movements
of hibernating snakes (Crotalus and Thamnophis) and thermal gradients of natural
hibernacula. Can J Zool. 67:108–114.
Markle CE, Moore PA, Waddington JM. 2020. Temporal variability of overwintering
conditions for a species-at-risk snake: Implications for climate change and habitat
management. Glob Ecol Conserv. 22: 1-9.
Matsuda BM, Green DM, Gregory PT. 2006. Amphibians and reptiles of British Columbia.
Victoria: Royal BC Museum.
McComb LB, Lentini PE, Harley DKP, Lumsden LF, Eyre AC, Briscoe NJ. 2022. Climate
and behaviour influence thermal suitability of artificial hollows for a critically
endangered mammal. Anim Conserv. 25(3):401–413.
Mitchell AM, Wellicome TI, Brodie D, Cheng KM. 2011. Captive-reared burrowing owls
show higher site-affinity, survival, and reproductive performance when reintroduced
using a soft-release. Biol Conserv. 144(5):1382–1391.
Milling CR, Rachlow JL, Johnson TR, Forbey JS, Shipley LA. 2017. Seasonal variation in
behavioral thermoregulation and predator avoidance in a small mammal. Behav Ecol.
28(5):1236–1247.

78
[MFLNRO] Ministry of Forest Lands and Natural Resource Operations. 2016. Best
management practice for amphibian and reptile salvages in British Columbia.
Naeth M, Chan Yk DA, McGILL W, Bailey A. 1993. Soil temperature regime in mixed
prairie rangeland after pipeline construction and operation. Can Agric Eng. 35(2):89–
95.
Nordberg EJ, Cobb VA. 2017. Body temperatures and winter activity in overwintering
Timber Rattlesnakes (Crotalus horridus) in Tennessee, USA. Herpetol Conserv Biol.
12(3): 606-615.
Parker WS, Brown WS. 1980. Comparative ecology of two Colubrid snakes, Masticophis t.
taeniatus and Pituophis melanoleucus deserticola, in northern Utah. Milwaukee (WI):
Milwaukee Public Museum.
Pohlert T. 2023. Calculate pairwise multiple comparisons of mean rank sums extended;
[accessed 2024 Feb 3]. https://CRAN.R-project.org/package=PMCMRplus.
Posit Software. 2023. RStudio IDE; [accessed 2024 Feb 3].
https://posit.co/products/open-source/rstudio/.
Prior KA, Weatherhead PJ. 1996. Habitat features of Black Rat Snake hibernacula in
Ontario. J Herpetol. 30(2):211–218.
Reinert HK, Cundall D. 1982. An improved surgical implantation method for radiotracking snakes. Source: Copeia. 1982(3):702–705.
Roznik EA, Alford RA. 2012. Does waterproofing thermochron iButton dataloggers
influence temperature readings? J Therm Biol. 37(4):260–264.
Rueegger N. 2016. Bat boxes-a review of their use and application, past, present and
future. Acta Chiropt. 18(1):279–299.
Rueegger N. 2019. Variation in summer and winter microclimate in multi-chambered bat
boxes in Eastern Australia: Potential eco-physiological implications for bats.
Environments. 6(2): 1-19.
Scheffers BR, Edwards DP, Diesmos A, Williams SE, Evans TA. 2014. Microhabitats reduce
animal’s exposure to climate extremes. 20(2):495–503.
Shine R, Mason RT. 2004. Patterns of mortality in a cold-climate population of garter snakes
(Thamnophis sirtalis parietalis). Biol Conserv. 120(2):201–210.
Sexton OJ, Hunt SR. 1980. Temperature relationships and movements of snakes (Elaphe
obsoleta, Coluber constrictor) in a cave hibernaculum. Herpetol. 36:20–26.
Sexton OJ, Marion KR. 1981. Experimental analysis of movements by Prairie Rattlesnakes,
Crotalus viridis, During Hibernation. Oecologia. 51:37–41.
Sudyka J, Di Lecce I, Szulkin M. 2023. Microclimate shifts in nest-boxes and natural cavities
throughout reproduction. J Avian Biol. 2023(1–2):1-17.
Taylor EN, DeNardo DF, Malawy MA. 2004. A comparison between point- and semicontinuous sampling for assessing body temperature in a free-ranging ectotherm. J
Therm Biol. 29(2):91–96.

79
Thierry A, Lettink M, Besson AA, Cree A. 2009. Thermal properties of artificial refuges and
their implications for retreat-site selection in lizards. Appl Herpetol. 6(4):307–326.
Tillman FE, Bakken GS, O’Keefe JM. 2021. Design modifications affect bat box
temperatures and suitability as maternity habitat. Ecol Solut Evid. 2(4):1-12.
TranBC. Why we are making a bed for 200 snakes. [accessed 2024 Mar 24].
https://www.tranbc.ca/2018/02/07/why-we-are-making-a-bed-for-200-snakes/
Trans Mountain Pipeline ULC. 2017. Lac du Bois Grasslands protected area reclamation
plan for the Trans Mountain Pipeline ULC Trans Mountain Expansion Project.
Calgary (AB): Kinder Morgan Canada Inc.
Trans Mountain. 2020. Snake species of concern overwintering den reconstruction and
monitoring plan.
Trans Mountain. 2021a. Time flies: one year of construction in the Kamloops urban area;
[accessed 2024 Feb 3].
https://www.transmountain.com/news/2021/time-flies-one-year-of-construction-inthe-kamloops-urbanarea#:~:text=What%20a%20year%20it%20has,the%20Trans%20Mountain%20Expa
nsion%20Project.
Trans Mountain. 2021b. Trans Mountain Expansion Project artificial overwintering den #1
Kamloops Area-Lac du Bois.
Trans Mountain. 2021c. Trans Mountain Expansion Project artificial overwintering den #2
Kamloops Area-Lac du Bois.
Trans Mountain. 2023a. Expansion project. [accessed 2024 Mar 24].
https://www.transmountain.com/project-overview.
Trans Mountain. 2023b. Our history. [accessed 2024 Mar 24].
https://www.transmountain.com/history.
Walker ML, Kadlec ED, Miloshewski RD, Pisani GR. 2011. Associative behavior and
affinity for anthropogenic habitats in two relocated Timber Rattlesnakes. Reptiles &
Amphibians. 18(4): 235-237.
Williams KE, Hodges KE, Bishop CA. 2015. Hibernation and oviposition sites of Great
Basin Gophersnakes (Pituophis catenifer deserticola) near their northern range limit.
J Herpetol. 49(2):207–216.
Zappalorti RT, Burger J, Burkett DW, Schneider DW, McCort MP, Golden DM. 2014.
Fidelity of northern pine snakes (Pituophis m. melanoleucus) to natural and artificial
hibernation sites in the New Jersey Pine Barrens. J. Toxicol. Environ. Health - A:
Curr. Issues. 77(21):1285–1291.

80

CHAPTER 4
CONCLUSION
SUMMARY OF THESIS
A central theme throughout this thesis is the potential scarcity of hibernacula in northern
latitudes. Stemming from this, my overarching management goal was to inform how to better
protect and properly compensate for disturbances to these critical habitat features. More
specifically, I used a community of three sympatric snake species, the Great Basin
Gophersnake, Western Yellow-bellied Racer, and Western Rattlesnake, to explore (1)
variation in hibernacula habitat use and (2) artificial hibernacula efficacy. Below are the
central objectives within each of these themes:
1. Variation in hibernacula habitat use
a. Document hibernacula and their species assemblages on our study site. (Chapter 2)
b. Investigate how the habitat features relate to the degree of communal occupation present
at hibernacula. (Chapter 2)
c. Explore how these three species vary in their selection of overwintering habitat.
(Chapter 2)
2. Artificial hibernacula efficacy
a.

Highlight the available microclimate within each hibernaculum and assess how well they
buffer against external environmental conditions. (Chapter 3)

b. Determine whether the internal thermal conditions of the artificial hibernacula are
comparable to those naturally experienced by snakes. (Chapter 3)
c. Monitor the adoption of the main artificial hibernaculum by the original cohort of
displaced snakes. (Chapter 3)
The principal findings of my thesis were:
•

Racers and gophersnakes appear more generalist in their selection of hibernacula,
overwintering in rodent burrows, rocky outcrops, holes under rocks and blast rock
features compared to rattlesnakes found at rocky outcrops and blast rock features.
(Chapter 2)

81
•

Little to no measured habitat features explained the numbers of gophersnakes, racers, and
rattlesnakes detected at individual hibernacula. (Chapter 2)

•

Gophersnake, racer, and rattlesnake hibernacula were situated on steeper slopes with
larger openings of the main hibernaculum entrance compared to random sites. (Chapter
2)

•

I found variation in habitat selection between these species that, in some cases, share
hibernacula. Rattlesnakes were associated with open habitat, with their hibernacula
having significantly less grass cover than random sites at the micro and macroscale. In
contrast, gophersnakes and racers were associated with further cover features at the
microscale, with gophersnakes selecting for more openings and shrub cover and racers
selecting for increased rock and shrub cover. This variation aligns with observed
behavioural differences between these species (i.e. racers and gophersnakes found away
from versus rattlesnakes conspicuously basking at the hibernacula entrance). (Chapter 2)

•

Artificial hibernacula built as mitigation can provide overwintering habitat that meets the
microclimatic conditions snakes need to survive. (Chapter 3)

•

Temperatures of overwintering Great Basin Gophersnakes within the main artificial
hibernaculum were comparable to those in natural hibernacula. In particular, the artificial
hibernacula provided thermal microsites conducive to the survival of gophersnakes at
depths ranging from 2.6 to 3.1m. (Chapter 3)

•

Despite relocation and soft-release efforts, the main artificial hibernaculum failed to show
adoption by displaced snakes two years after the disturbance. This indicates that although
these sites afforded conditions conducive to survival, successful adoption may require
more intensive monitoring and relocating of snakes to ensure the target population uses
these artificial hibernacula. (Chapter 3)

MANAGEMENT IMPLICATIONS
Habitat loss is a predominant factor threatening biodiversity globally and is one of the
primary drivers of species extinction (Hanski 2011; Banks-Leite et al. 2020; GonçalvesSouza et al. 2020). Great Basin Gophersnakes, Western Yellow-bellied Racers, and Western
Rattlesnakes are commonly threatened by habitat loss (COSEWIC, 2013, 2015a, b). The
sensitivity of animals to habitat loss depends largely on their specialization and reliance on

82
specific environmental features (Keinath et al. 2017); one such feature is snake hibernacula.
Such sites warrant identification and protection as critical habitat features for the overall
recovery and conservation of these species at risk (COSEWIC 2015b; ECCC 2019; GC
2023).
Identifying and protecting hibernacula
Within Canada, Great Basin Gophersnakes, Western Yellow-bellied Racers, and
Western Rattlesnakes all exclusively occur within the province of British Columbia
(COSEWIC, 2013, 2015a, b). Within this province, Wildlife Habitat Areas (WHA) can be
used as a conservation tool to protect 200-300 ha around communal hibernacula (MWLAP
2004; Williams et al. 2012), assuming the hibernacula are identified. I acknowledge,
however, that the characteristics I identified as separating hibernating sites from random sites
will have limited value in pre-emptive planning, i.e. proactively guiding inventory personnel
to rodent burrows used by gophersnakes and racers. Thus, “inventory” may not be
straightforward. From our research, radiotelemetry is an effective approach to identifying
hibernacula in a new area, helping us identify 15 of 28 hibernacula (54%) on our study
landscape. However, this approach is relatively invasive and requires intensive monitoring.
Thus, consideration should be given to expanding protection in areas where snakes occur
rather than relying on the protection of hibernacula that are not readily identified.
Through WHAs, British Columbia has approximately 10114 ha (average size 241 ha)
protected for rattlesnakes, 1909 ha for gophersnakes (average size 174 ha), and 100 ha for
racers (data from WHA database, last updated Feb 2024 – Ministry of Environment 2024).
From these data, it is apparent that rattlesnake hibernacula are protected to a greater extent
than gophersnake and racer hibernacula (Ministry of Environment 2024). It should be noted,
however, that only 22% of the 380 known rattlesnake hibernacula are protected (as of 2016),
indicating that even though they are protected to a greater extent, they are still not
sufficiently protected (ECCC 2019). The discrepancy between species does not come as a
surprise, as Western Rattlesnakes have garnered considerable research attention within
British Columbia (ex., Macartney et al. 1987; Lomas et al. 2015; Harvey and Larsen 2020;
Howarth et al. 2023). Further, from our research, there is evidence to suggest that rattlesnake
hibernacula are easier to identify: rattlesnakes have more specific den habitat requirements
and the animals are much more conspicuous at hibernacula than gophersnakes and racers.

83
This discrepancy of protection paired with arguably harder-to-identify hibernacula (i.e.
rodent burrows) indicates a need for further inventory of gophersnake and racer hibernacula
in particular.
A multispecies approach has been recommended across management and recovery
strategies (SIRART 2008; ECCC 2019). Some even emphasize this approach because
gophersnakes and rattlesnakes use the same hibernacula (ECCC 2019). However, within this
study, I showcase variation in habitat selection between these species that do not universally
share hibernacula. Given this, I advised in Chapter 2 against using the selection of wellstudied and inventoried species (e.g., Western Rattlesnakes) to extrapolate to other species
with less well-known denning habits. Overall, a multispecies approach will only be practical
if one first fully understands the unique variation between the species.
Building artificial hibernacula
As critical habitat features, the loss of hibernacula demands compensation. Within
Chapter 3, I demonstrated that artificial hibernacula built as mitigation measures can provide
microclimatic conditions conducive to the survival of snakes overwintering at a northern
range limit. Managers wanting to construct such habitat should give particular attention to:
•

Utilizing depth and distance to hibernacula entrance to provide a gradient of
microclimatic conditions to facilitate hibernation.

•

Ensuring the structure is sufficiently buffered from external conditions and is not
dependent on snow insulation.

•

Tailoring the construction approach to the needs of the focal species (i.e., temperature
and humidity).
Currently, the “Best Management Practices for Amphibian and Reptile Salvages in

British Columbia” offer artificial hibernacula as a potential site for the relocation of salvaged
reptiles (MFLNRO 2016). Although, artificial hibernacula can be conducive to survival, I
found that salvaged snakes did not readily adopt the main hibernacula even with soft-release
and relocation efforts. Based on these results, I believe using artificial hibernacula to provide
immediate overwintering habitat for salvaged snakes may be ineffective. However, there do
not appear to be alternative mitigation options in times of disturbance. Thus, I want to
emphasize the importance of adequately surveying and avoiding disturbance of critical

84
habitat beforehand (Whiting and Booth 2012). Further research still is needed to understand
the effectiveness of the use of artificial hibernacula as mitigation measures in the long term
(see Adoption and long-term viability of artificial hibernacula below).
Incidental artificial hibernacula
Incidental artificial hibernacula may be a significant source of hibernacula at northern
extremes where natural habitat might be limited (Costanzo 1986). Within our study site, there
are numerous examples of ‘accidentally’ created artificial hibernacula (i.e. overtop of old
pipeline construction, in blast rock on the sides of roads, or in other previously disturbed
habitats). Excluding the newly installed artificial hibernacula in this study, 9 of the 28 (32%)
known hibernacula on the landscape were found in anthropogenically disturbed habitats.
Further, the three previously unknown hibernacula excavated by pipeline construction were
found near the old pipeline installed in 1951. These observations of incidental anthropogenic
hibernacula are not unique to this study; hibernacula have previously been found on humancreated rocky slopes (Williams et al. 2015), in rockpiles beside buildings (Walker et al.
2011), and in other anthropogenic features (Brown et al. 1974; Costanzo 1986; Burger et al.
1988; Edkins et al. 2018). Although I could not find support in the literature, perhaps a
preference for these sites comes from the interstitial spaces created by dumping rocks or
garbage. Further, the preference for sites along the pipeline may be related to the presence of
heat from the pipeline during the transportation of crude oil (Dong et al. 2019). For example,
it has been found that natural gas pipelines can alter the thermal regime of soil due to the
external heating of the pipeline (Naeth et al. 1993). During winter, these natural gas pipelines
were found to affect the depth of the frost line in the adjacent soil (Naeth et al. 1993). Not
only does this indicate the willing usage of anthropogenic structures (Eye 2022), but it also
underscores the responsibility of managers to guarantee the absence of snakes at these sites
prior to further disturbance.
Management of snakes in Lac du Bois
It is essential to consider direct management implications for this peripheral
population of snakes as the Lac du Bois Grassland Protected Area lies within the coincidental
northern limits of the Great Basin Gophersnake, Western Yellow-bellied Racer, and Western
Rattlesnake (Matsuda et al. 2006). Through this research, I recommend the following:

85
•

Providing protection through WHAs to all identified hibernacula found within this study.

•

Continue monitoring the identified hibernacula to understand population dynamics in the
long term.

•

Expand inventory and protection of hibernacula within the park and periphery. It is
crucial to consider the periphery landscape as some snakes may use the park during the
active season and overwinter in critical habitat outside the park.

•

Continue monitoring the artificial hibernacula to understand the uptake in the long term.

•

Lac du Bois is impacted by many anthropogenetic activities (e.g., construction, hiking,
biking, horseback riding, off-roading, cattle grazing, etc.) (BC Parks 2017). Thus,
restricting activities around hibernacula (e.g., those near trails) may be beneficial to
reduce potential human-wildlife conflict and the overall impact on snakes.

•

Rodent burrow hibernacula have been found to collapse on our site. These events were
suspected to be the result of cattle grazing. Similar occurrences of rodent burrow collapse
by ungulates have been described elsewhere (Lovegrove and Painting 1987; Weihs and
Shroder 2011). Although the overall impact of this event is unclear, future research
should be conducted, and consideration should be given to the scope of future grazing.

•

If development takes place within or around Lac du Bois, surveying efforts for
hibernacula should occur around egress or ingress windows. Behavioural variation
between the species should be considered during surveys (i.e. racers and gophersnakes
found away from versus rattlesnakes conspicuously basking at the hibernacula
entrance).

FUTURE RESEARCH CONSIDERATIONS
Continued identification of hibernacula
I have previously suggested that an increased inventory of hibernacula is needed,
particularly for gophersnakes and racers, but that is relatively more difficult than for
rattlesnakes. With increased inventory, I recommend collecting data on the habitat features at
more overwintering sites. Future research could use this data to further elucidate the habitat
selection of gophersnakes and racers at hibernacula. In doing so, I recommend focusing on
certain types of hibernacula, such as earthen sites, to provide further context in their
identification and subsequent protection.

86
Availability of hibernacula and communal denning
As previously mentioned, a central theme throughout the thesis is the limited
availability of hibernacula at northern latitudes. One hypothesis in the literature theorizes that
the limited availability of adequate overwintering sites results in increased instances of
communal denning (Gregory 1984). This phenomenon is quoted as being “undoubtedly the
main cause of communal overwintering” (Gregory 1984) and although many studies
referenced this theory (Prior and Weatherhead 1996; Harvey and Weatherhead 2006;
Gardiner and Sonmor 2011; Gardiner et al. 2013; Bruckerhoff et al. 2021), few have tried to
assess the availability of hibernacula (Prior and Weatherhead 1996; Harvey and Weatherhead
2006; Williams et al. 2014; Williams et al. 2015). Prior and Weatherhead (1996) found the
results were equivocal to whether communal hibernacula of Black Rat snakes in Ontario
were limited after comparing their habitat features to the broader landscape. Increased rates
of overwinter mortality for Massasauga Rattlesnakes (Sistrurus catenatus catenatus) in
Ontario suggest a potential limitation in hibernacula that meet the requirements for survival
(Harvey and Weatherhead 2006). William et al. (2014,2015) found that the solitary use of
hibernacula, decreased fidelity, and high survival overwinter suggest that Great Basin
Gophersnakes in the Okanagan have increased availability of hibernation sites. Further,
within this study (and a concurrent study -Ragsdale MSc in prep), I observed snakes finding
alternative hibernacula after they lost their original overwintering sites. Such plastic
behaviour indicates a greater availability of hibernacula for gophersnakes and racers than
first thought. To our knowledge, however, no research has concretely quantified the
availability of hibernacula. Ideally, future research should assess the availability of
hibernacula on a given landscape and then connect this to the occurrence of communal
denning. Not only will this help to highlight the potential scarcity of these sites, but it will
also help provide a theoretical understanding of the communal overwintering behaviour in
northern snakes.
Adoption and long-term viability of artificial hibernacula
Following our research on the adoption of artificial hibernacula built as mitigation
measures, many pertinent research questions have arisen. Although I indicate that the
immediate adoption by salvaged snakes is poor, it remains unclear how long it takes for a
hibernaculum (artificial or not) to be adopted by snakes. A long-term study on artificial

87
hibernacula will elucidate this and possible mechanisms at work (i.e. scent trails, visual cues,
conspecific attraction, etc.). Further, an experimental approach may be valuable to determine
if seeding hibernacula with snakes is effective to aid in adoption. Additionally, despite the
recent and increasing use of artificial hibernacula, it is unclear whether they have a lifetime
or are at risk of environmental disturbances (i.e. flooding). Such research will shed light on
the overall viability of such structures as a continued mitigation or conservation strategy.

CONCLUSION
Life in colder climates requires specialized adaptations to prevail during winter. Here,
snakes must find an adequate hibernaculum to survive. Due to the biological importance and
potential scarcity of hibernacula in northern climes, proper identification and protection are
critical for conserving northern snakes. Based on this, I sought to understand how to identify
and mitigate future disturbances of these critical habitat features. Using a community of three
at-risk sympatric snake species at their northern range limits, I found that the patterns of
hibernacula use in communities of snakes are more complex than initially understood. In
particular, I discovered distinct interspecific variations in behaviour and hibernacula
selection between three species that, in some instances, share the same hibernacula. Such
results indicate the need to rethink how we identify and protect hibernacula for even closely
related species, such as a community of snakes. Based on this variation, utilizing speciesspecific management approaches may be more effective than relying on the habitat selection
and protection garnered by well-studied species. In cases where hibernacula are irrevocably
destroyed, we showcase that artificial hibernacula can be built conducive to the survival of
snakes but have a low immediate adoption from a displaced population. Thus, rather than just
relying on artificial hibernacula, I recommend putting in every effort to avoid disturbances of
these critical habitat features. Overall, I recommend larger protection at the landscape level
to reduce the occurrence of hibernacula loss and compensate for the lack of protection around
hard-to-identify hibernacula. If destruction is unavoidable, we need to use it as a transparent
learning opportunity with ample discussion and attention paid to research design of artificial
structures. Such an approach will help to answer the many questions that have arisen
regarding the overwintering site selection of northern snakes and the long-term use of
artificial hibernacula. Moving forward, the results of this study provide critical information to

88
inform the management strategies of these at-risk species and will influence the
identification, protection, and response to the disturbance of hibernacula.

LITERATURE CITED
Banks-Leite C, Ewers RM, Folkard-Tapp H, Fraser A. 2020. Countering the effects of habitat
loss, fragmentation, and degradation through habitat restoration. One Earth. 3(6):672–
676.
BC Parks. 2017. Lac du Bois Grasslands Protected Area Management Plan. Kamloops.
Brown WS, Parker WS, Elder JA. 1974. Thermal and spatial relationships of two species
of Colubrid snakes during hibernation. Herpetol. 30(1):32–38.
Bruckerhoff LA, Kamees LK, Holycross AT, Painter CW. 2021. Patterns of survival of a
communally overwintering Rattlesnake using an artificial hibernaculum. Ichthyol
Herpetol. 109(1):64–74.
Burger J, Zappalorti RT, Gochfeld M, Boarman WI, Caffrey M, Doig V, Garber SD, Lauro
B, Mikovsky M, Safina C, et al. 1988. Hibernacula and summer den sites of Pine
Snakes (Pituophis melanoleucus) in the New Jersey Pine Barrens. J Herpetol.
22(4):425–433.
[ECCC] Environment and Climate Change Canada. 2019. Recovery strategy for the Western
Rattlesnake (Crotalus oreganus), the Great Basin Gophersnake (Pituophis catenifer
deserticola) and the Desert Nightsnake (Hypsiglena chlorophaea) in Canada.
COSEWIC. 2013. COSEWIC assessment and status report on the Great Basin Gopersnake
Pituophis catenifer deserticola in Canada. Ottawa.
COSEWIC. 2015a. COSEWIC assessment and status report on the Western Rattlesnake
Crotalus oreganus in Canada. Ottawa.
COSEWIC. 2015b. COSEWIC assessment and status report on the Eastern Yellow-bellied
Racer Coluber constrictor flaviventris and the Western Yellow-bellied Racer Coluber
constrictor mormon in Canada. Ottawa.
Costanzo JP. 1986. Influences of hibernaculum microenvironment on the winter life history
of the garter snake (Thamnophis sirtalis). OJS. 86(5):199–204.
Dong H, Zhao J, Zhao W, Si M, Liu J. 2019. Study on the thermal characteristics of crude oil
pipeline during its consecutive process from shutdown to restart. Case Stud Therm
Eng. 14: 1-20.
Edkins TL, Somers CM, Vanderwel MC, Sadar MJ, Poulin RG. 2018. Variable habitat
selection and movement patterns among Bullsnake (Pituophis catenifer sayi)
populations in Saskatchewan. Can Field-Nat. 132(2):126–139.
Eye DM. 2022. Reproductive ecology of female Western Rattlesnakes (Crotalus Oreganus)
in Canada [thesis]. [Kamloops (BC)]: Thompson Rivers University.

89
Gardiner LE, Somers CM, Martino JA, Parker DL, Poulin RG. 2013. Balancing the
dumbbell: summer habitats need protection in addition to winter dens for northern
snake communities. J Wildl Manag. 77(5):975–982.
Gardiner LE, Sonmor KW. 2011. Reptiles major slump event at Grasslands National Park
Snake Pit in southwestern Saskatchewan. Blue Jay. 69(3):120–124.
Gonçalves-Souza D, Verburg PH, Dobrovolski R. 2020. Habitat loss, extinction
predictability and conservation efforts in the terrestrial ecoregions. Biol Conserv. 246:
1-8.
[GC] Government of Canada. 2023. Critical habitat for species at risk national dataset –
Canada; [accessed 2024 Mar 2]. https://datadonnees.az.ec.gc.ca/data/species/protectrestore/critical-habitat-species-at-risk-canada/
Gregory PT. 1984. Communal denning in snakes. In: Seigel R, Hunt L, Knight J, Malaret L,
Zuschlag N, editors. Vertebrate Ecology and Systematic A Tribute to Henry S. Fitch.
Lawrence: University of Kansas Publications. p. 57-76.
Hanski I. 2011. Habitat loss, the dynamics of biodiversity, and a perspective on
conservation. Ambio. 40(3):248–255.
Harvey DS, Weatherhead PJ. 2006. Hibernation site selection by Eastern Massasauga
Rattlesnakes (Sistrurus catenatus catenatus) near their northern range limit. J
Herpetol. 40(1):66–73.
Harvey JA, Larsen KW. 2020. Rattlesnake migrations and the implications of thermal
landscapes. Mov Ecol. 8(1).
Howarth, Bishop CA, Larsen KW. 2023. Western Rattlesnake (Crotalus oreganus) spring
migration in British Columbia: a comparative study of juveniles and adults. Can J
Zool. 101(7): 530-540.
Keinath DA, Doak DF, Hodges KE, Prugh LR, Fagan W, Sekercioglu CH, Buchart SHM,
Kauffman M. 2017. A global analysis of traits predicting species sensitivity to habitat
fragmentation. Global Ecology and Biogeography. 26(1):115–127.
Lomas E, Larsen KW, Bishop CA. 2015. Persistence of Northern Pacific Rattlesnakes
masks the impact of human disturbance on weight and body condition. Anim
Conserv. 18(6):548–556.
Lovegrove B, Painting S. 1987. Variations in the foraging behaviour and burrow structures of
the Damara Molerat Cryptomys damarensis in the Kalahari Gemsbok National Park.
KOEDOW. 30:149–163.
Macartney JM, Larsen KW, Gregory PT. 1987. Body temperatures and movements
of hibernating snakes (Crotalus and Thamnophis) and thermal gradients of natural
hibernacula. Can J Zool. 67:108–114.
Matsuda BM, Green DM, Gregory PT. 2006. Amphibians and Reptiles of British Columbia.
Victoria: Royal BC Museum.
Ministry of Environment. 2024. Approved wildlife habitat areas.
https://www.env.gov.bc.ca/wld/frpa/iwms/wha.html.

90
[MFLNRO] Ministry of Forest Lands and Natural Resource Operations. 2016. Best
management practice for amphibian and reptile salvages in British Columbia.
[MWLAP] Ministry of Water Land and Air Protection. 2004. Identified wildlife
management strategy accounts and measures for managing identified wildlife
southern interior forest region.
Naeth M, Chan Yk DA, McGILL W, Bailey A. 1993. Soil temperature regime in mixed
prairie rangeland after pipeline construction and operation. Can Agric Eng. 35(2):89–
95.
Prior KA, Weatherhead PJ. 1996. Habitat features of black rat snake hibernacula in
Ontario. J Herpetol. 30(2):211–218.
[SIRART] Southern Interior Reptile and Amphibian Recovery Team. Recovery strategy
series rcovery strategy for the Gopher Snake, deserticola Subspecies (Pituophis
catenifer deserticola) in British Columbia.
Walker ML, Kadlec ED, Miloshewski RD, Pisani GR. 2011. Associative behavior and
affinity for anthropogenic habitats in two relocated Timber Rattlesnakes. Reptiles &
Amphibians. 18(4): 235-237.
Weihs BJ, Shroder JF. 2011. Mega-terracettes and related ungulate activities in Loess Hills,
Iowa, USA. ZfG. 55(1):45–61.
Whiting C, Booth HJ. 2012. Adder Vipera berus hibernacula construction as part of a
mitigation scheme for flood defence works, Horsey Corner, Norfolk, England. Conser
Evid. 9: 9-16.
Williams KE, Hodges KE, Bishop CA. 2012. Small reserves around hibernation sites may
not adequately protect mobile snakes: The example of great basin gophersnakes
(Pituophis catenifer deserticola) in British Columbia. Can J Zool. 90(3):304-312.
Williams KE, Hodges KE, Bishop CA. 2014. Phenology and demography of great basin
gophersnakes (Pituophis catenifer deserticola) at the Northern edge of their range.
Herpetol Conserv Biol. 9(2):246–256.
Williams KE, Hodges KE, Bishop CA. 2015. Hibernation and oviposition sites of Great
Basin Gophersnakes (Pituophis catenifer deserticola) near their northern range limit.
J Herpetol. 49(2):207-216.

91

APPENDIX A
DIRECTION OF RELATIONSHIP OF HABITAT VARIABLES CORRESPONDING TO THE
POISSON REGRESSION

Direction of relationship of habitat variables associated with top models with ⃤ AICc value
<2.0 (Burnham and Anderson 2002) corresponding to the Poisson regression results for Great
Basin Gophersnake, Western Yellow-bellied Racer, and Western Rattlesnake models at the
microscale (top row) and macroscale (bottom row) (Table 2.2). Points correspond to the
estimates of the coefficients (log scale with 95% confidence intervals (CI)) of the top models.
Positive values describe habitat features associated with increased snakes, and negative
values describe variables not associated with them. If the CI crosses the dashed line, the
coefficient is not significant. Full names and detailed descriptions of explanatory variables
are found in Table 2.1.

92

APPENDIX B
DIRECTION OF RELATIONSHIP OF HABITAT VARIABLES CORRESPONDING
TO THE CONDITIONAL LOGISTIC REGRESSION

Direction of relationship of habitat variables associated with top models with ⃤ AICc value
<2.0 (Burnham and Anderson 2002) corresponding to the conditional logistic regression
results for Great Basin Gophersnake, Western Yellow-bellied Racer, and Western
Rattlesnake models at the microscale (top row) and macroscale (bottom row) (Table 2.3).
Points correspond to the estimates of the coefficients (log scale with 95% confidence
intervals (CI)) of the top models. The rattlesnake model did not converge at the microscale.
Positive values describe habitat features selected for, and negative values describe variables
selected against. If the CI crosses the dashed line, the coefficient is not significant. Full
names and detailed descriptions of explanatory variables are found in Table 2.1.

93

APPENDIX C
UNIVARIATE ANALYSIS RESULTS OF CANDIDATE VARIABLES FOR CONDITIONAL
LOGISTIC REGRESSION
Univariate analysis results of candidate variables for conditional logistic regression analysis
at the microscale for Great Basin Gophersnakes, Western Yellow-Bellied Racers, and
Western Rattlesnakes. Candidate variables were selected following univariate analysis
(Paired t-test or Wilcoxon signed-rank tests, depending on normality results, between used
and random sites) with P<0.25 and Spearman rank correlation tests with correlation
coefficients <0.7. When the correlation coefficient was greater than or equal to 0.7, the
variable with the least variation, as determined from the univariate analysis, was dropped,
and the other was selected for further analysis (Dormann et al. 2013). Significance of
univariate analysis at P<0.05 is denoted by a *.
Group
Gophersnake

Racer

Rattlesnake

Variable
AREA
SLOPE
G1
S1
OPEN
AREA
SLOPE
OPEN
R1
S1
AVGRV
ASP
AREA
SLOPE
G1
S1
D1

Test
Wilcoxon
Wilcoxon
Wilcoxon
Wilcoxon
Wilcoxon
Wilcoxon
Paired t-test
Wilcoxon
Wilcoxon
Wilcoxon
Wilcoxon
Paired t-test
Wilcoxon
Paired t-test
Wilcoxon
Wilcoxon
Wilcoxon

P
<0.0001*
0.0001*
0.0002*
0.0056*
0.0011*
0.0001*
0.0002*
0.0084*
0.0007*
0.0133*
0.0376*
0.1602
0.0001*
0.0023*
0.0007*
0.0677
0.0603

94

APPENDIX D
COMPARISON OF WEEKLY TEMPERATURE OSCILLATION AT TIMES WITH SNOW
AND WITHOUT SNOW
The comparison of weekly temperature oscillation at times with snow and without snow from
November 1, 2021, to March 31, 2022. ASH1 refers to artificial snake hibernacula 1, and
ASH2 refers to artificial snake hibernacula 2. Position refers to the distance of loggers to the
hibernaculum mouth, where the back is the furthest and the front is the closest (Figure 3.1).
Comparisons were run individually for each position. ∗Significant at p < 0.05. Average
weekly temperature oscillations for periods with snow and without snow ± the standard
deviations are displayed.
Hibernacula
ASH1

ASH2

Position
Back
Middle
Front
Back
Middle
Front

With Snow (°C)
1.0±0.10
1.3±0.2
0.9±0.7
1.1±1.0
0.90±0.5
1.7±0.6

No Snow (°C)
0.9±0.2
1.1±0.2
0.8±-0.3
0.9±-0.2
1.0±0.4
1.7±00.8

Test
t-test
t-test
Wilcox
Wilcox
Wilcox
t-test

p
0.1103
0.0893
0.6982
0.341
0.8046
0.9321

95

APPENDIX E
POST HOC EXACT TEST RESULTS FOR THE COMPARISON OF AVERAGE DAILY
TEMPERATURE BETWEEN DIFFERENT POSITIONS IN ARTIFICIAL SNAKE
HIBERNACULUM 1 AND 2 AND THE AVERAGE, MAXIMUM, AND MINIMUM SNAKE
TEMPERATURE
The post hoc exact test results for the comparison of average daily temperature between
different positions in artificial snake hibernaculum 1 and 2 and the average, maximum, and
minimum snake temperature (n=4) between November 1, 2022, and March 31, 2023 (Figure
3.6). ASH1 refers to artificial snake hibernaculum 1, and ASH2 refers to artificial snake
hibernaculum 2. Front, middle, and back refer to the distance of loggers to the hibernaculum
mouth, where the back is the furthest and the front is the closest (Figure 3.1). The values
correspond to the p values from the test results. ∗Significant at p < 0.05. Non significance is
bolded. Average is the overall average daily temperature ± the standard deviation.
ID
ASH1Front
ASH1Middle
ASH2Back
ASH2Front
ASH2Middle
Average Snake
Maximum Snake
Minimum Snake

Average (°C)
8.020±2.833
9.931±2.879
9.039±2.873
6.628±2.589
7.945±2.406
7.778±2.697
8.728±3.140
6.758±2.192

Average Snake Maximum Snake
0.3674
<0.0001*
<0.0001*
<0.0001*
<0.0001*
0.5414
<0.0001*
<0.0001*
<0.0001*
<0.0001*
<0.0001*
<0.0001*
<0.0001*

Minimum Snake
<0.0001*
<0.0001*
<0.0001*
1.0000
<0.0001*
-