-
Identifier
-
Thesis_MES_2022Wi_LarsonE
-
Title
-
JOINT BASE LEWIS-MCCHORD AS A CASE STUDY FOR POTENTIAL WEATHER METRICS FOR MICROBAT POPULATION MONITORING
-
Date
-
March 2022
-
Creator
-
Larson, Erika
-
extracted text
-
JOINT BASE LEWIS-MCCHORD AS A CASE STUDY
FOR POTENTIAL WEATHER METRICS FOR
MICROBAT POPULATION MONITORING
by
Erika Larson
A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
March 2022
©2022 by Erika Larson. All rights reserved.
This Thesis for the Master of Environmental Studies Degree
by
Erika Larson
has been approved for
The Evergreen State College
by
________________________
Ralph Murphy, Ph. D.
Member of the Faculty
________________________
Date
ABSTRACT
Joint base Lewis-McChord as a case study for potential
weather metrics for bat population decline.
Erika Larson
Bats provide important ecosystem services. However, their populations have been declining
globally due to habitat loss, disease, and climate change. Previous studies have failed to explicitly
assess the impact of weather metrics on the number of bat sightings observed during bat surveys.
The objective of this study was to fill this knowledge gap. The ideal survey location is a place in
which habitat loss and disease are minimal. Joint Base Lewis-McChord (JBLM) meets these
criteria. It has a comprehensive forest management plan and no cases of White Nose Syndrome
have been observed. Data collected between 2018 and 2021 were provided by the Washington
Department of Fish and Wildlife (WDFW). Using these data, correlation analyses were
conducted. A negative correlation was identified between the number of bat sightings and annual
precipitation. This is consistent with previous reports that bats use more energy to fly when they
are wet. No correlations were identified between the number of bat sightings and temperature or
wind speed. This is inconsistent with previous reports about the impact of temperature and wind
speed on bat activity. These findings may have been impacted by limitations in the data due to
sample size, inconsistent record keeping, and lack of standardization. Based on this study’s
findings, future studies should utilize a combination of bat monitoring methodologies, standardize
data collection and documentation, and collaborate with other regional bat monitoring
organizations. The implementation of these recommendations would enable researchers to answer
the impacts of weather metrics on bat populations. This will be valuable in predicting the impacts
of climate change.
TABLE OF CONTENTS
LIST OF FIGURES ......................................................................................................................... v
LIST OF TABLES .......................................................................................................................... vi
ACKNOWLEDGEMENTS ........................................................................................................... vii
INTRODUCTION ................................................................................................................... 1
I.
Study Objectives .......................................................................................................................... 3
LITERATURE REVIEW ........................................................................................................ 4
II.
Contributors to Bat Population Decline ....................................................................................... 7
Habitat Loss ............................................................................................................................. 8
Disease ..................................................................................................................................... 9
Climate Change...................................................................................................................... 10
Behavioral Adaptations to Environmental Change .................................................................... 11
Utilization of Artificial Roosts............................................................................................... 14
Bat Population Monitoring ........................................................................................................ 15
Bat Surveying ............................................................................................................................ 16
III.
STUDY AREA: JOINT BASE LEWIS-MCCHORD ....................................................... 19
Site Selection Previous Studies on JBLM ................................................................................. 21
IV.
METHODS ........................................................................................................................ 24
Methods Used in Previous Studies on JBLM ............................................................................ 24
Current Study Method................................................................................................................ 26
Bat Colony Site Selection .......................................................................................................... 30
V.
ANALYSIS OF DATA ................................................................................................................. 33
Variables Used in Data Analysis ............................................................................................... 34
Research Question One .............................................................................................................. 36
Research Question Two ............................................................................................................. 38
Research Question Three ........................................................................................................... 40
VI.
DISCUSSION .................................................................................................................... 43
VII.
CONCLUSION .................................................................................................................. 52
REFERENCES .............................................................................................................................. 57
iv
LIST OF FIGURES
Figure 1. Acoustic and Mist Net Sites from 2008 Fort Lewis Survey
Figure 2. Decision Flow Diagram
Figure 3. Decision Flow Diagram JBLM Map of Main Colonies Surveyed 2018-2021
Figure 4. First Day above 40°F and Bat Sightings
Figure 5. Precipitation (inches) and Bat Sightings
Figure 6. Temperature and Bat Sightings
Figure 7. Wind (mph) and Bat Sightings
v
LIST OF TABLES
Table 1. Microbats Found in Washington State
Table 2. Roost Sites
Table 3. Total Precipitation at Roost Sites
vi
ACKNOWLEDGEMENTS
I would like to express my gratitude to my thesis reader, Dr. Ralph Murphy, who guided
me throughout this project. I would also like to thank my friends and family who
supported me.
vii
I.
INTRODUCTION
Joint-base Lewis McChord (JBLM) is a U.S. Army and Air Force Installation. It
is in the South Puget Sound region of Western Washington which has a temperate
maritime climate with long wet winters, short dry summers, and a mean annual
temperature of 51°F. According to the University of Washington, the Pacific Northwest
experienced an increase in the average daily high temperature of 1.3°F between 1895 and
2011. This change in temperature has increased the frequency of warmer extremes , with
night time temperature increases being the most noticeable (University of Washington,
2013). The trend is expected to continue. Temperatures are expected to rise
approximately 5.8°F between 2021 and 2070 (University of Washington, 2013, 2020).
Changing temperatures could cause an increase in the severity of heavy rainfall events,
increased temperature during all seasons, and both more frequent heat extremes and
fewer extreme cold events (University of Washington, 2020). Counterintuitively, it has
also been predicted that winters and summers will have increased rainfall (ED DPW,
2017; University of Washington, 2020). These changes could cause an increase in
flooding, landslides, and fires in the area.
JBLM encompasses 91,126 acres of land. It consists of 61,000 acres of forested,
woodland, and savanna areas; ideal habitats for most many species including bats (ED
DPW, 2017; Falxa, 2008a). JBLM is also home to 20,000 acres of prairie and oak
woodlands created by glacial outwash over 15,000 years ago. This habitat is home to
multiple federally endangered species, and is one of the rarest and most endangered
ecosystems in Washington with only 3% of the original prairie land remaining (Delvin,
2013; WDFW, 2021b). There are 3,850 acres of wetlands, including 1,200 acres of
1
forested wetland and riparian areas which attract insects (ED DPW, 2017). All these
geographical features and habitat types are predicted to be impacted by climate change
(ED DPW, 2017; University of Washington, 2013).
Globally, bats play important but often unappreciated roles. For example, bats
provide many ecosystem services to natural biomes and increase human well-being
(Kunz et al., 2011). These services include insect suppression, pollination, seed dispersal,
and maintaining ecosystems (Kunz et al., 2011; Law et al., 2019). Anthropogenic factors
have caused the decline of bat populations, risking the loss of the services they provide
(Kunz et al., 2011; Law et al., 2019). To understand how these factors impact bats and to
mitigate their impact, multiple studies have been conducted that focus on habitat loss,
disease, and climate change (Bat Conservation Trust, 2021; Kunz et al., 2011; Law et al.,
2019). There is no single cause for decline of bat populations, rather the cause of their
decline is multifactorial. Habitat loss, disease, and climate change are widely regarded as
having the most impact. Habitat loss typically occurs where there is an influx in
anthropogenic disturbances such as development or urbanization. The devastating bat
disease called White Nose Syndrome has spread amongst bat populations worldwide.
Finally, increased temperatures due to climate change alter food availability and
hibernation patterns. While the impact of climate change on bat populations has been a
major focus for years, comparatively little is known about the impact of weather on the
number of bat sightings during surveys. The combination of habitat loss, disease, and
climate change in most locations, makes it difficult to determine the impact of weather. In
order to isolate this impact, large sample sizes and sophisticated statistical methods could
be used. However, that approach requires significant funding. An alternative approach is
2
simply to make observations in a location where confounding factors are relatively
constant. Joint Base Lewis McChord meets this criterion and was therefore selected as
the site for this retrospective study.
JBLM is an ideal location because it has a comprehensive forest management
program that closely monitors forest health by preparing harvested timber sites for
replanting, monitoring, and removing invasive species such as Scot’s Broom, and
maintaining density harvesting to provide heterogeneity of the forests. By upholding
these practices JBLM’s forest management program has been a certified as a sustainable
forestry operation since 2002 by the Forest Stewardship Council (ED DPW, 2017). This
means that habitat loss and deforestation are not currently problems being experienced by
the installation. In addition, while White Nose Syndrome has been identified in
Washington State, it has not been observed on JBLM, despite JBLM’s diverse bat
population: 9 of the 15 species of Washington bats can be found on JBLM (Falxa, 2008a;
G. Hayes & Wiles, 2013). Consequently, habitat loss and disease are unlikely to have
major impacts on JBLM’s bat populations.
Study Objectives
Using JBLM as a case study, several research questions were asked to determine
whether the frequency of bat sightings is negatively impacted by the following weather
metrics: temperature, precipitation, and wind speed.
Question 1: Are bat sightings affected by how early in the year the temperature regularly
exceeds 40°F?
3
Briefly, the bat species on JBLM hibernate when the average low temperature is about
40°F. Thus, it was predicted that the number of bat sightings during a given observation
period would be lower if the average low temperature exceeded 40°F later in the year.
Question 2: Are bat sightings affected by total annual rainfall?
Wet bats expend more energy to fly. It was therefore predicted that fewer bat sightings
would be observed when total annual rainfall was high.
Question 3: Since day-to-day weather conditions can affect survey results, which
conditions are important to consider when conducting surveys?
Research suggests that bats are likely to be most active when temperatures are warm and
wind speeds are low. These specific predictions were evaluated in this question.
II. LITERATURE REVIEW
Bats are mammals that belong to the order Chiroptera (Bat Conservation
International, 2021; de Oliveira, 2020; Marshall, 2022). This group is the second largest
in the class Mammalia. To put this in perspective, one out of every five animals on Earth
is a member of this class. Chiroptera includes two suborders: Yinpterochiroptera and
Yangochiroptera.
Yinpterochiroptera contains bat species that are commonly referred to as megabats.
Though, megabats are not the focus of this study it is important to briefly describe this
suborder because it too provides important ecological services and has seen declining
populations. Megabats are characterized by their large size with some species weighing
up to 3.2 pounds. They do not develop echolocation because they lack a specialized
4
feature in their ears that is present in other bat species(de Oliveira, 2020; Marshall, 2022).
Their diets consist of mostly fruit and nectar. Pollen from these food sources will
frequently stick to the hairs on the bats’ bodies. The pollen from their body hairs are
transferred to other plants while feeding, resulting in pollination. Without this form of bat
pollination, known as chiropterophily, it is likely that we would not have fruits such as
bananas, avocados, or mangoes. Even alcohols, such as tequila, rely on bat pollination,
specifically of the agave plant (US Department of the Interior, 2017). Over 80 medicines
are that are made from plants rely on bats for pollination and seed dispersal (Gettler,
2013; US Department of the Interior, 2017). Some medicines are even developed using a
protein in bat saliva derived from some species of megabat which are used to help
mitigate blood clots in stroke victims (Thompson, 2021).
The second suborder of bat, and the focus of this study, is Yangochiroptera also
known as microbats. Of the 1,400 bat species, 70% of them are microbats (Bat
Conservation International, 2021). These bats are characterized by echolocation, and the
ability to hibernate (Burnett Mary Regional Group, 2021). Bats of this suborder are
smaller in size, and have large ears and a tragus used for echolocation (de Oliveira,
2020). In contrast to the foraging practices of megabats which include consumption of
fruits and nectar, microbats are predominantly insectivores (Falxa, 2008a; Frick et al.,
2012; National Science Foundation, 2012). All bats found in Washington state are
microbats, and more specifically are in the family Vespertilionidae, the most widespread
bat family. This family consists mainly of insectivores and is especially abundant in
temperate climates (Kunz et al., 2011).
5
Seventy percent of microbats consume insects and other small arthropods for
food. They can eat their body weight in insects each night, which could potentially
consist of thousands of insects (US Department of the Interior, 2017). According to the
United States Geological Survey (USGS), a single little brown bat, a species that can be
found on JBLM, can eat 4 to 8 grams of insects nightly. To put this into context, the
average mosquito weighs 2 milligrams. So in one night, the little brown bat could eat
roughly 2,000 to 4,000 mosquitoes (USGS, n.d.). The USGS goes on to report that the
Northeastern U.S. has lost over one million bats, which results in between 660 and 1320
metric tons of insects no longer being eaten by bats (USGS, n.d.). Without a healthy bat
population consistently consuming these insects, the insect populations in the U.S. could
grow, impacting the environment and the economy.
Bats are a major contributor to important ecosystem services and may be sensitive
indicators of ecosystem function and environmental threats (G. Hayes & Wiles, 2013).
First, they act as a system for maintaining the health of ecosystems by providing
pollination and seed dispersal, and secondly, they act as an effective biological pest
controller (Aubry et al., 2003; G. Hayes & Wiles, 2013; National Science Foundation,
2012).
Pollination and seed dispersal are two crucial phases in the plant reproductive
cycle. Bat pollination is performed by most species of bat worldwide and is a major
contributor to the survival of these various plant species (Mahandran et al., 2018). Bats
are effective in aiding these processes due to their increased mobility granted through
flight, their fur is able to carry and transfer large pollen loads (Fleming et al., 2009), and
their capability of defecating while midflight, allows for the movement of different
6
species of plants across the span of a landscape (Fleming et al., 2009). This defecation
process allows them to be a key source of nutrient transport from riparian areas to upland
areas (G. Hayes & Wiles, 2013). With the increase in habitat fragmentation, the ability to
pollinate over long distances has important implications for plant and ecosystem
conservation efforts (Fleming et al., 2009; Mahandran et al., 2018).
Agricultural insects and pests cost the U.S agricultural industry over $33 billion
dollars a year in crop losses. These pests also destroy 25-50% of crops across the globe
(Maslo & Kerwin, 2020). Most microbat diets consist of arthropods, with the type
depending on the species of bat and geographical location. However, they are the most
important natural predator of nocturnal insects such as mosquitoes, moths, beetles,
crickets, leafhoppers, and chinch bugs (Bat Conservation International, 2016). These
insects are some of the most serious agricultural pests. It is predicted that bat
consumption of these pests specifically saves the U.S. Agricultural Industry over $22.9
billion annually, and helps reduce the use of pesticides which can harm the environment
(Bat Conservation International, 2016; Griggs, 2015; Maslo & Kerwin, 2020).
Contributors to Bat Population Decline
The bat population is in a state of rapid decline (USGS, 2016). There are many
contributors to bat population decline which include habitat loss, disease, and climate
change. There are 15 species of bat found in the state of Washington (Table 1) (G. Hayes
& Wiles, 2013).
7
Microbats Found in Washington State
Common Name
Scientific Name
Eptesicus fuscus
1. Big Brown Bat*
Myotis californicus
2. California Myotis*
Parastrellus hesperus
3. Canyon Bat
Myotis thysanodes
4. Fringed Myotis
Lasiurus cinereus
5. Hoary Bat*
Myotis keenii
6. Keen’s Myotis*
Myotis lucifugus
7. Little Brown Myotis*
Myotis volans
8. Long-legged Myotis*
Antrozous pallidus
9. Pallid Bat
Lasionycteris noctivagans
10. Silver-haired Bat*
Euderma maculatum
11. Spotted Bat
Corynorhinus townsendii
12. Townsend’s Big-eared Bat*
Myotis evotis
13. Western Long-eared Myotis
Myotis ciliolabrum
14. Western Small-footed Myotis
Myotis yumanensis
15. Yuma Myotis*
Table 1. The species of bat identified in the Washington State Bat Conservation Plan (G.
Hayes & Wiles, 2013).
* Bat species confirmed on JBLM in the 1992 and 2008 bat surveys (Falxa, 2008a).
Habitat Loss
The most important habitat types for bats in Washington are those used in roosting
and foraging such as trees, snags, caves, mines, cliffs, buildings, and bridges.
Urbanization and human expansion have resulted in the destruction of the bats’ natural
roosting and foraging sites (Falxa, 2008a; J. P. Hayes, 2003; Loeb & O’Keefe, 2011).
While some bats have adapted to using man-made structures such as buildings, the loss of
historic roosting habitats such as old growth snags and forests, has forced many bat
species to prioritize areas suitable for foraging or for habitat, rarely both (Falxa, 2008b;
Freed & Falxa, 2010). Washington’s natural landscape used to provide habitats that were
8
in key foraging locations such as large trees or snags. However with urban development,
bats have been forced to adapt and relocate to man-made and artificial roosts such as
roofing, siding, attics, bridges, and bat boxes (Freed & Falxa, 2010). In addition to having
to relocate, many of these locations are not near ideal foraging locations such as
meadows, riparian areas, small gaps in forests, or even shrub-steppe and grassland
habitats. Although with the addition of artificial roosts to forested areas, conservationists
have been able to situate roosts near foraging resources (Baker & Lacki, 2004).
Because of deforestation and logging practices the physical distance between roosts
and foraging grounds have increased. Since a variety of bat species prefer large trees and
the loss of old growth forests has been substantial, habitat availability has been negatively
impacted (G. Hayes & Wiles, 2013; J. P. Hayes, 2003). While regulations that require the
retention of snags and buffers around riparian zones have reduced the impact of habitat
loss, their implementation is inconsistent.
Disease
There are also rapidly spreading diseases that impact bat populations, White Nose
Syndrome (WNS) for example. WNS is caused by the fungus Pseudogymnoascus
destructans. In early stages of the infection, hibernating bats infected with this fungus use
twice as much energy as uninfected bats. Eventually this prevents the bat from reentering
full hibernation. The fungus also causes the bats to awaken more readily because of warm
temperatures. In both cases, the result is that the bats deplete their fat stores, which often
leads to death (Duvergé et al., 2000; Frick et al., 2012).
9
Climate Change
Climate change affects bats in at least three different ways: 1) altering the
temperature of their roosting sites, 2) exposing them to severe weather events, and 3)
altering their foraging behavior and inadvertently exposing them to predators (Loeb &
O’Keefe, 2011). Bats increasingly live in artificial roosts such as bat boxes. Since
artificial roosts are small, rising temperatures due to climate change are a concern
because the internal temperature of an artificial roost can become very hot. According to
the University of Washington, the annual and seasonal temperatures of Washington state
are expected to increase with more frequent extreme heat events (University of
Washington, 2020). The temperatures inside artificial bat boxes can reach up to 52°C
(125.6°F) (Groc, 2021). Changes in ambient temperature also affect the availability of
prey, which in turn affects bat foraging behavior, generally leading them to forage earlier
in the day when insect activity is high. Consequently, this exposes them to increased
predation and competition with species that are active during the day (Frick et al., 2012).
Finally, as changes in the climate are expected to result in more frequent and severe
weather events, bat habitats are at a higher risk of disturbances from events such as
extreme heat which can severely impact artificial roosts or even cause the complete
destruction of roosts, e.g. by causing wild fires or severe flooding (Frick et al., 2012).
Bats in western Washington use a variety of methods to regulate their body
temperatures. For example, many species in western Washington reside in tight, warm
spaces. They can minimize thermoregulatory costs by selecting habitats with the optimal
temperatures or by colonial clustering (de Oliveira, 2020). In addition, all species of bat
10
rely on the large surface area of their wings as a form of passive thermal conductance
during flying and roosting (de Oliveira, 2020).
In the Pacific Northwest there is increasing interest in understanding how habitat
loss and climate change impact bat populations. However, to date few bat surveys have
been conducted in Washington state. Three studies were conducted on Joint Base LewisMcChord (JBLM) since the 1990s. This includes the present study. While none of the 15
species of bats that live in Washington are currently listed as endangered or threatened at
the federal or state level, two of them have been declared species of concern, specifically
Townsend’s Big-eared bat and Keen’s myotis (Table 1) (WDFW, 2021a).
Behavioral Adaptations to Environmental Change
In order to adapt to the changing climate in North America, bats have adopted
multiple behavioral strategies. These strategies include migration, hibernation, relocation,
and using artificial roosts or habitats. According to the US Geological Survey (USGS),
many species of bats are known or believed to be declining based on bat population data
collected between 1955 and 2001 (USGS, 2016). While bats adapted to selective pressure
by developing a notable array of capabilities (i.e. true flight, echolocation, and
hibernation), this process took millions of years. In the last 150 years, anthropogenic
pressures have significantly grown, resulting in drastically altered or lost habitats. The
true impact of these pressures is poorly understood because the current methods for bat
monitoring are difficult and inconsistent across organizations such as the USGS,
Department of Fish and Wildlife, Bat Conservation International, the Bat Conservation
Trust, and the Nature Conservancy.
11
According to fossil records, bats have existed for millions of years. In 2003, the
earliest well-formed bat fossil was discovered in the state of Wyoming, dating back more
than 52 million years to the Eocene period (Gunnell et al., 2008). Bat fossils are rare, due
to the bones being light weight and small (Ramel, 2020). Therefore, this discovery was
particularly exciting for researchers because it represented an evolutionary intermediate
linking modern bats to their non-flying ancestors (Gunnell et al., 2008; Gunnell &
Simmons, 2005). This new species was named Onychonyteris finneyi, and with it
researchers were able to answer many questions posed concerning how bats evolved
echolocation and flight (American Museum of Natural History, 2008). According to
Gregg Gunnell, a research scientist at the University of Michigan, the three leading
theories developed were: 1) bats developed flight before echolocation, 2) bats developed
echolocation first, and 3) bats developed both simultaneously (Gunnell et al., 2008). The
physical appearance of the fossil suggests that bats evolved flight before echolocation.
This was hypothesized because Onychonyteris lacked specific features in the skull near
the ear that are needed for echolocation in modern bats (American Museum of Natural
History, 2008; Gunnell et al., 2008). Without the use of echolocation Onychonyteris had
to rely on “visual, olfactory, or passive cues to hunt” (Gunnell et al., 2008). Since the
skull, including the orbits, was incomplete, the size of the bat’s eyes could not be
determined. Had the fossil been complete, it could have provided insight into the
evolution of bat vision. In particular, it could have helped identify similarities with
modern day bats that do not echolocate, e.g. the family Pteropodidae (Fig. 1) also known
as megabats or Old World fruit bats (Luzynski et al., 2009).
12
While seasonal survival has been extensively studied in many animal species, this
is a relatively unstudied aspect of chiroptology. The four main mechanisms that bats have
developed in order to mitigate seasonal mortalities are migration, torpor, hibernation, and
artificial structure habitation (Reusch et al., 2019). Studying these factors is challenging
because data collection methods and validated analytic tools (i.e. reference databases) are
lacking in the bat field. In contrast, bird research has benefited from long-term high
quality data and an array of validated analytic tools (Reusch et al., 2019).
To survive periods of cold weather and food and/or water shortages, bats have
developed multiple mechanisms to help them improve their chances of survivability such
as migration, torpor, and hibernation (Reusch et al., 2019).Since western Washington has
a mild maritime climate, bats in western Washington may be able to employ different
wintering strategies than bats at colder climates (Falxa, 2007). Interestingly, however,
there are still species of bats that have been identified as seasonal migrators (U.S. Fish
and Wildlife, 2020). The hoary bat (Lasiurus cinereus) is one of these species. It migrates
more than 1,000 km during its migratory season, although not much is known about
where this species of bat spends its winter months (U.S. Fish and Wildlife, 2020).
It is suggested that some bats that live in western Washington remain active
regardless of the season (U.S. Fish and Wildlife, 2020). The California myotis (Myotis
californicus) and silver-haired bat (Lasionycteris noctivagans) are two species in the
Pacific Northwest that have been observed outside of roosts throughout the winter
months (U.S. Fish and Wildlife, 2020). It is believed that they can avoid going into
hibernation by going into a state known as torpor which is a shortened state of
13
hibernation that is triggered during inclement weather in order to conserve and slow the
expenditure of energy. This enables them to awaken from torpor multiple times
throughout the winter when the weather is warm enough.
Hibernation is one of the most crucial mechanisms bats have evolved to survive
cold winters. Hibernation is a state in which body temperature, metabolic rate, heart rate,
and respiration are reduced to prevent the depletion of fat reserves (Briggs, 2021; Link,
2004; U.S. National Park Service, 2020). When ambient temperatures, drop below 40° F
many bats will hibernate. Microbats’ hibernation sites (also called roosting sites or their
hibernaculum) include large trees, caves, mine shafts, tunnels, old wells, artificial roosts,
and other man-made structures (Link, 2004). To survive winter, the hibernaculum needs
to be quiet and cool, with a consistent temperature and humidity. A hibernaculum
protects bats from predation, light, noise, and other types of disturbances (Link, 2004;
WDFW, 2020). These roosts allow them to safely hibernate when food sources become
scarce (Briggs, 2021). Unfortunately, because of the lack of standardization in
chiroptology, it has been suggested that the term torpor and hibernation are used
interchangeably in some of the literature (Lactis, 2020; U.S. National Park Service, 2020;
Weller et al., 2016; Wildlife Online, n.d.).
Utilization of Artificial Roosts
All species of female bats birth one large pup during each pregnancy. While bats
only have one breeding season, they can have multiple litters during that one season (Bat
World Sanctuary, 2013). These pups are born hairless and have limited thermoregulatory
abilities (de Oliveira, 2020). Similar to large mammals, bats rely on their mother’s milk
for nutrition until weaning, which can take anywhere from six weeks to four months
14
depending on when their wings become fully developed (Bat World Sanctuary, 2013; de
Oliveira, 2020). Non-nutritional care provided by the mother consists of protection,
sensory stimulation, thermal influence, and transport. Females must frequently leave their
offspring while foraging which is why habitat loss is impacting the bat population so
heavily. Instead of leaving the pup in the colony, some females take their young to their
day roosts. This is not only to keep their pup close to them for safety, but their colony is
more vulnerable to disturbance and can be destroyed by a single destructive act such as a
flood, severe rain event, or fire (Ammerman et al., 2012).
Once weaned, juvenile bats venture out of the colony to begin foraging for
themselves (Ammerman et al., 2012). When a bat reaches adulthood, it can be
challenging to determine its age. The primary method of identifying adults is by
morphological traits, including their long bones, body mass, pelage coloring (facial and
body markings), tooth wear, and the size of the pulp cavity. The pulp cavity is an internal
opening of a vertebrate tooth that contains connective tissues, blood vessels, and nerves
(Brunet-Rossini & Wilkinson, 2009).
Bat Population Monitoring
Microbats are difficult to study. They tend to be well hidden in trees or caves and
exit their roosts at night to hunt for insects (Loeb et al., 2015). Consequently, various
survey methods have been developed which include: visual or direct counts, nets, and
acoustic surveys. These methods are commonly used for observing and documenting bat
population data (National Park Service, 2018). According to the United States
Department of Agriculture, there is no public or private program that is conducting a
standardized monitoring of bat species in North America (Loeb et al., 2015).
15
In 1994, the USGS, began the Bat Population Data Project. The project identified
significant knowledge gaps concerning bat populations in the United States (USGS,
2016). This led to a multi-phase and comprehensive effort to compile bat population data
across organizations in the U.S. and its territories. The results were published in 2003.
An article published in 2016 suggested that the database was being updated, however,
further information regarding this specific project was not found (USGS, 2016). USGS
has been leading, managing, and coordinating the North American Bat Monitoring
Program (NABat). According to the NABat management Plan, the Bat Population Data
Project has become a repository of bat population data called the Bat Population
Database. The plan for this database is for NAbat partners to upload their data with sitespecific information such location, grid cell number, date, times, environmental
conditions, variables related to bat detectors, species identification information, years of
experience identifying species or counting bats, and the metadata associated with acoustic
surveys (Loeb et al., 2015).
Currently, this program is monitoring 47 species of North American bats, which is
a huge step toward data standardization. However, NAbat does not provide any
monitoring equipment or software to any agencies or organizations, which likely limits
the rate of data deposition (Loeb et al., 2015).
Bat Surveying
Without acoustic sampling or bat detection equipment, visual bat sightings are
one of the main indicators of whether bats are present in an area (Predator Free New
Zealand Trust, n.d.). There are other indicators of bat activity such as the presence of
guano or changes to bat roosts or hibernacula. However, the latter indicators do not reveal
16
how recently bat activity occurred. While bat sightings tend to underestimate true bat
counts by roughly 4%, visual counts are still the most cost effective, easiest method for
data collection, and does not require time for set up, unless infrared monitors or cameras
are being utilized (Sedgeley, 2012).
There are a range of reasons why bats are difficult to monitor and survey. These
range from their small size (as small as 2-2.6 grams), their nocturnal and elusive habits,
and the lack of standardized monitoring programs and procedures (Loeb et al., 2015;
United States Geologic Survey, 2016). While bats consist of one fifth of the world’s
mammalian biodiversity with over 1400 species (Frick et al., 2019), there is a lot about
bats that is unknown, which makes attempts to prioritize and plan conservation efforts
difficult and demanding (Frick et al., 2019; Loeb et al., 2015). In a report published by
the U.S, Geological Survey in 2003, it was stated that “scientific validity of the past and
current efforts have not been critically examined and there have been no efforts to
synthesize and summarize these efforts (O’Shea & Brogan, 2003)”. However there are
multiple state and government agencies that are collaborating on bat conservation,
including the National Fish and Wildlife Foundation, Bat Conservation International, the
U.S. Forest Service, the U.S. Bureau of Land Management, and the U.S. Geological
Survey (O’Shea & Brogan, 2003).
Each organization uses their own criteria or key words to categorize survey types.
After reviewing multiple organizations’ methods, a pattern was identified, allowing for
the most frequently referred to surveys to be classified based on general categories. The
methods used by the following organizations and others not listed were reviewed:
NABat, Bat Conservation International, USGS, the National Park Service, Bat Survey
17
Ireland. Many terms were used to describe visual surveys, including‘dusk’, ‘dawn’,
‘emergence’, ‘reentry’, ‘activity’, ‘return to roost’, ‘sunset’, ‘roost’, and ‘hibernation’.
Because so many terms were used to describe the same approach, a classification scheme
was devised to differentiate the main approaches. Three categories were identified:: 1)
visual or count, 2) capture and release, and 3) acoustic. There were only two
organizational surveying guides identified during the research process. This included the
“Bat Survey Guidelines for Professional Ecologists” from Bat Conservation International
(Collins et al., 2016) and “A plan for North American Bat Monitoring Program” from the
Department of Agriculture (Loeb et al., 2015). These are the only two publications that
described the specific criteria that determine when particular survey types are required.
Visual or Count
The most widely used survey is the visual or count survey. This type of survey
serves two purposes: 1) to identify the presence of bat and 2) to quantify the level of
activity (units being the number of bats). This method is the most widely used because it
is cost effective, and no equipment is needed. However, compared to the other three
method types, it is not the most efficient. Types of field surveys include visual, colony
emergence or re-entry counts, roost searches, and identification of activity.
Capture and Release Methods
Capture and release methods are conducted using traps or nets set up by
researchers. This approach uses mist nets or harp-traps. Nets are the most common form
of capture and release method conducted on bat populations and provide life history
information. For example, a captured bat can be used to measure morphology, acquire
samples (blood, tissue, parasites), and assess physiological status such as species, gender,
18
or breeding status (Collins et al., 2016; Darras et al., 2021). Mist nets are ultrafine
monofilament nylon nets that are 12m in length and 3m in height. Setting up mist nets
takes place 3 hours prior to dusk and require personnel to monitor them consistently until
the end of the netting session (Brunet-Rossini & Wilkinson, 2009). Harp-traps help to
capture bats without exposing them to the entanglement that occurs in the netting of mist
nets. It uses vertical nylon strings, similar to those in the harp instrument, and funnels the
bats during flight into a canvas collecting bag at the bottom (NHBS, 2019). Both methods
can cause injury or death to bats. For the netting mechanisms to function, bats are
required to fly directly into them. This can cause a bias in the species that are captured.
For example, species of bat that fly at higher altitudes will be underrepresented when
using these methods. On the other hand, capture and release methods provide life history
information that neither visual count or acoustic methods can provide.
Acoustic Methods
All bats in the United States use echolocation (Falxa, 2008b). Thus, acoustic sampling is
one of the best methods for bat monitoring since it is the least obtrusive and most
efficient way to speciate and record bat calls (United States Geologic Survey, 2016).
Most bat species can be distinguished based on the varying frequencies of their calls.
This allows for the detection of a larger number of species of bats in a short amount of
time. However, this approach entails a reliance on sensitive, battery-powered devices.
III. STUDY AREA: JOINT BASE LEWIS-MCCHORD
Joint Base Lewis-McChord (JBLM) is an ideal location for identifying how weather
metrics impact bats survey results. Specifically, many of the factors that might confound
analyses in other location are controlled for on JBLM. Two key variables are factors that
19
have been implicated in the decline of bat population sizes elsewhere: habitat loss and
disease, namely White Nose Syndrome. Since 1) White Nose Syndrome has not been
observed on JBLM and 2) JBLM has an intensive forest management plan that puts strict
limitations and controls over the habitats on JBLM, these factors cannot confound the
analysis. This allows for the unconfounded determination of the relation between weather
metrics such as temperature, precipitation, and wind on the number of bat sightings.
JBLM has eight different habitat types which helps to provide a variety of locations
for bats to colonize and forage. These eight habitat types include: 1) marsh (wetland
without significant open water), 2) forest edge (clearing, non-native vegetation), 3)
riparian forest (stream in the forest), 4) dry forest corridor (road, pipeline), 5) savanna
with sparse trees (near a marsh, pond, or stream), 6) savanna (grass or shrub land with
sparse oak or pine stands), 7) large open water (lake, pond, or river), 8) open field (nonnative or no vegetation). According to the Washington State Bat Conservation Plan, all
these habitat types are ideal for bat foraging and habitation (ED DPW, 2017).
Not only does JBLM have a variety of habitat types ideal for bat habitation, but it
also consists of huge swaths of land reserved for restoration and the conservation of
threatened and endangered species. JBLM has a comprehensive forest management
program and has been certified since 2002 by the Forest Stewardship Counsel as a
sustainable forestry operation. With JBLM’s forest management plan and conservation
practices (ED DPW, 2017), one of the major contributors to bat population decline,
habitat loss, is not a current threat. Because literature on bat research is limited, the
inclusion of rare habitats such as lowland forests and prairies make JBLM one of
relatively few contributors of data for these types of habitats (Falxa, 2008a).
20
JBLM consists of 91,126 acres of land, and is home to nine of the 15 species of bat
found in Washington state make JBLM their home which includes the two Washington
State Candidate species, Keen’s Myotis (Myotis keenii) and Townsend’s Big Eared Bat
(Corynorhinus townsendii) (G. Hayes & Wiles, 2013). These bats have been observed in
two of the three major bat surveys that have been previously conducted on JBLM. The
first survey was conducted in 1992 by the USDA Forest Service, and the second survey
was conducted in 2008 by Cascadia Research (Falxa, 2008a). JBLM is important for the
monitoring of these species of bats. Keen’s myotis has one of the smallest ranges in
North America and is assumed to be rare. Townsend’s Big-eared bat is a species at risk
especially in areas with pesticide spraying activities that can be seen in many forests and
agricultural areas. Since JBLM has a comprehensive forest management plan, it is likely
that the base would be an ideal location to monitor this population. While these bats do
prefer caves and mines, they do use large snags as habitats (J. P. Hayes, 2003). Part of
JBLM’s forest management practice is to try to create and maintain snags for habitat and
species conservation (ED DPW, 2017).
Site Selection Previous Studies on JBLM
According to the 2008 bat study conducted on Fort Lewis, the observation locations
were chosen in order to represent the eight habitat types that could be found on present
day JBLM. By including these eight habitat types, it allowed for the inclusion of nondeveloped, accessible land within the constraints of troop training schedules, since JBLM
is a military installation, and ensuring the safety of survey personnel is critical. While
some high security areas (such as artillery impact zones) were off limits to the 2008
study, the ability to sample the perimeters of these areas or sampling areas with a similar
21
habitat type would allow for the prediction of species variation based on the activity
levels among different bat species in the various habitat types (Falxa, 2008a).
The survey conducted on Fort Lewis in 2008, monitored a total of 82 sites which
included 67 acoustic sites and 15 net capture sampling sites which are depicted in Figure
1. The sites selected for net capture were identified as productive sites for bat sightings in
the 1992 study, conducted by the USDA Forest Service. The 1992 report was
unpublished and thus unavailable at the time of this study’s research. However, the 2008
study did have access to the 1992 report and were able to summarize the data and the
study’s findings. Thus allowing, the 2008 study to use the population data acquired
during the 1992 survey to determine where to set up sampling equipment, and eventually
allowing them to resample the sites in order to run a comparative analysis.
22
Figure 1. A map of the sampling locations used in the 2008 survey on Fort Lewis.
Both acoustic (red circles) and mist net sites (red squares) are included (Falxa, 2008a).
23
IV. METHODS
Methods Used in Previous Studies on JBLM
There have been three bat survey studies conducted on JBLM over the past three
decades, the first was conducted in 1992, the second in 2008, and the third spanning 2018
to 2021. While these surveys all had similar objectives, the ability to compare their
research is limited due to the inconsistency in survey monitoring methods between the
three studies. The 1992 and 2018-2021 studies were focused on gauging the abundance of
bat populations on JBLM. In contrast, the 2008 study was more of a species inventory.
The methodologies of each study differed in important ways – as detailed below.
The 1992 survey conducted by the USDA Forest Service, incorporated both netting
and acoustic recording. However, the report was unpublished, and after significant effort,
a copy of the report could not be found. Nevertheless, some of the information from that
survey are included in the 2008 Fort Lewis Bat Survey conducted by Cascadia Research
and the Nature Conservancy (Falxa, 2008b). The acoustic sampling technology used in
the 1992 study did not have the capability of determining the species of bat based on the
echo or frequency emitted by the bat. This technology was used merely to identify
whether bats were present. Notably, however, it could determine differences in
frequencies, allowing for some very limited degree of speciation. There was little
information provided about the mist netting practices used in the 1992 survey other than
they preferred to use harp traps instead of mist-nets. It was also reported that Big brown
bats evade harp traps more than mist net traps which may indicate that there could be
some species-specific bias in their findings (Falxa, 2008a).
24
Similar to the 1992 survey, a mixture of both netting and acoustic sampling were used
in 2008. While both used acoustic identification methods, the technology used in 1992
was more simplistic, using heterodyne style tunable units which fail to collect species
level information. In contrast, the 2008 survey used time expansion ultrasonic detectors
which can be left unattended and when combined with the Sonobat software and a
database of reference calls can obtain detailed species identifications (Falxa, 2008a).
Acoustic monitoring could potentially identify specific species of bat; however, some
species have similar echolocating frequencies which may not provide an accurate
identification of the species (Falxa, 2008a).
However, even with this improved technology, there are still some bat species that are
difficult to differentiate, e.g. Little Brown Myotis and Long-legged Myotis. Positive
identification can be determined using the catch and release method which is the only
method that allows for speciation based on physical characteristics. Net capture methods
enable surveyors to get a positive identification on most of the species being surveyed. In
addition, they can provide useful life history data, which cannot be obtained by visual
emergence counts or acoustic recordings. While nets were used in the 1992 survey, the
one conducted in 2008 used mist-nets. These are different than harp traps because they
are situated lower to the ground. Similar to harp traps, mist-nets can induce some speciesspecific biases (Falxa, 2008a). More specifically, mist-nets underrepresent the abundance
of bats that tend to fly at higher altitudes.
Acoustic sampling methods are the most technologically sophisticated surveying
method. Though these methods deliver significant value, their implementation is more
complicated. Specifically, acoustic sampling requires a battery to be charged, potentially
25
limiting the survey duration. On the other hand, the surveyor does not need to be present
reducing the risk that the surveyor will influence bat behavior. In addition, acoustic
methods enable speciation. However, they can have difficulties in differentiating bats that
are part of the same genus. In addition, acoustic methods do not reveal bat life history
data in the way that live capture or netting can provide. However, the latter methods can
introduce species-specific biases and could potentially injure the bats that are captured
using these methods (Falxa, 2008a).
Current Study Method
In this retrospective study, data collected and provided by the Washington State
Department of Fish and Wildlife (WDFW) were analyzed. The data includes four years
of bat emergence survey data collected in the summers of 2018 through 2021 across
multiple roosting sites on JBLM by volunteers chosen from the Department of Fish and
Wildlife Internship located on JBLM. The data were collected using the guidelines
provided in the Colony Assessment Protocol (CAP) generated and provided by the
Washington Department of Fish and Wildlife. The data was documented using field data
sheets that the survey volunteers were provided during the time of the survey. These
forms were later transcribed to an excel spreadsheet for this study. In contrast to the 1992
and 2008 surveys, the current survey conducted from 2018 to 2021 used only visual
emergence to survey the bat populations instead of acoustic or catch and release methods.
The Colony Assessment Protocol states that a survey team must be limited to
three personnel. This is intended to limit the amount of noise disturbance added to the
environment while conducting the survey (Washington Department of Fish and Wildlife,
2018). Communication between surveyors should be made using low soft tones when
26
speaking. In addition, the surveyors must avoid wearing clothing that could make noise
(e.g., synthetic material like nylon jackets or pants can make ‘swishing’ noises when
moved). Headlamps and flashlights are prohibited from being shone directly onto the bats
for more than a few seconds since bats are sensitive to changing light levels (Washington
Department of Fish and Wildlife, 2018). McMaster University’s Bat Lab states that light
is the primary cue that bats use to sense nighttime, indicating it is time to become active
(Conger, 2020). This is important because bats are nocturnal and use the cue of the sun
setting as an indication that it is safe to emerge from their roosts (Rehm, 2018).
Nighttime helps bats to elude daytime predators such as weasels, snakes, skunks, foxes,
and humans. Dusk also increases food availability for bats because birds become scarce
at night, thus, reducing the competition for the same food resources (Rehm, 2018).
In order to avoid influencing the emergence data, the surveyors must be in their
positions by the roosts’ exits at least 30 minutes prior to sunset. It is important to conduct
a pre-survey site analysis to identify where the primary exits are located (Washington
Department of Fish and Wildlife, 2018). It is a best practice to space the surveyors so that
all sides and angles of the structure can be observed at the same time. The detailed flow
chart, Figure 2 outlines the decision-making process for overcoming challenges unique to
the roost site (Fig. 2). The CAP, requires that the surveyors use a double-surveyor effort
which means that two surveyors will be counting from the same exit (Washington
Department of Fish and Wildlife, 2018). The surveyors will sit or stand at the colony site
and collect the data for 1-2 hours. The Flow Diagram in Figure 2 was designed to be a
resource for helping surveyors trouble shoot issues that could arise while conducting the
survey. For example, if a roost had three exists being used by the bats at a specific roost,
27
and only two surveyors were available, the flow diagram can be referenced for guidance
on how to proceed. According to Figure 3, the surveyors would determine which exits the
majority of the bats were using. The surveyors would then monitor only those two exists
instead of trying to monitor all three. If this was not possible, there is also an option to
temporarily block one of the entrances forcing the bats to exit through an easier exit to
monitor. There was no indication of this occurring in any of the emergence count surveys
conducted. The flow diagram also accounts for special circumstances when an internal
count of the roost may be necessary such as the inability to clearly see bats emerging
from exits. The blocking of one or more exits is preferred since an internal count is very
invasive. It consists of either opening up a bat roost or attempting to look inside to see if
there are any indications of active bats such as droppings (guano), feeding remains, the
absence of cobwebs, an identifiable odor, and scratching or markings around the roost
exits (Orme, 2020). This method is very intrusive and can severely disrupt the bats.
However, this type of survey was not reported in the 2018 to 2021 data set.
28
Figure 2. Decision Flow Diagram. The decision flow diagram to determine the best form of
count survey to obtain a bat population sample. This figure is from the Bat Colony Protocol used
by researchers on JBLM to conduct emergence bat surveys (Washington Department of Fish and
Wildlife, 2018).
29
Bat Colony Site Selection
The bat roost sampling locations for 2018 to 2021 were chosen by using the sites
selected during the survey conducted in 2008. Due to limited number of staff available to
conduct the emergence surveys, only a portion of the previous survey’s roost locations
could be studied. Moreover, only the roosts with the most activity were chosen for
preliminary inspections to evaluate the roost’s activity level. An active roost was
identified by having either 1) visible evidence of bats emerging and/or returning to the
roost or 2) guano near the roost entrances. While these inspections took place, no
documentation or results for these evaluations were available. Thus, it was largely
unknown why specific roosts were chosen for a given survey year.
Table 2 shows that the number of roosts being surveyed varied by year. In
addition, there was no information to identify why roosts were selected for observation
each year. There were only three roosts that were observed in all four years. These roosts
are illustrated in Figure 3. The 2018 survey monitored 7 roost sites. However, four of
these sites (OP2, OP3, Spanaway, and TA15 Triangle) did not have associated
information: the type of habitat (i.e., bridge, wetland, snag) or colony type (i.e., natural or
artificial roost). The 2019 survey monitored only three sites, which severely restricted the
options for comparing the roosts across the four years. There was no information to
indicated why only three sites were monitored. Finally, 7 sites were monitored in 2020,
and 8 in 2021.
30
Figure 3. JBLM Map of Main Colonies Surveyed 2018-2021. The three bat roosts that
were observed during all four years of bat emergence surveys. The red circles identify
the roost or colony site. Image Credit: Erika Larson, 2021.
31
Survey
Year
2018
Site/Roost Name
Frequency
1. Chamber’s Bridge
2
2. Military Museum
2
3. OP2*
1
4. OP3*
1
5. Spanaway*
2
6. TA15 Triangle*
2
7. Watkin’s Boxes
2
2019
1. Chamber’s Bridge
2
2. Military Museum
2
3. Watkin’s Boxes
2
2020
1. Batman House
2
2. Chamber’s Boxes
2
3. Chamber’s Bridge
2
4. Holden Woods RB
1
5. Military Museum
2
6. Tank Bridge
2
7. Watkin’s Boxes
2
2021
1. 8th Ave RR Bridge
2
2. Batman House
2
3. Chamber’s Boxes
2
4. Chamber’s Bridge
2
5. Fish Hatchery
1
6. Military Museum
2
7. Tank Bridge
2
8. Watkin’s Boxes
2
Table 2. Roost Sites. The roost names that were observed
during each survey year and how many times the surveys
were conducted on the specific sites.
* - no description or other information was provided
about the site other than the site/roost name
The Colony Assessment Protocol requires that selected roosts be observed one or
more times per season. Table 2 demonstrates what roosts sites were being monitored each
year. Most of the roosts were surveyed twice during the season. The ideal timeframe for
observing a bat population is over the summer months. This is the time when bat
populations emerge from their winter roosts and begin to forage and start the mating
process. Each species has their own time windows for peak activity. For example, myotis
32
and big brown bats (both observed on JBLM) tend to emerge and become active between
the dates of June 15 to July 15, while the Townsend’s Big-eared bat (also observed on
JBLM) is primarily active between July 1 to July 22 (Washington Department of Fish and
Wildlife, 2018). Given these time frames, for all nine species of bat on JBLM, the best
time frame to conduct bat surveys is between the beginning of June and the middle of
August (Washington Department of Fish and Wildlife, 2018).
V. ANALYSIS OF DATA
Colony site, observation date, and emergence counts were documented by the
Department of Fish and Wildlife volunteers conducting the survey. These data were
digitally transcribed in a Microsoft Excel workbook. A series of correlation analyses
were conducted to identify associations between variables of interest. The following
subsections describe the analyses and results for each research question. For all analyses,
the same analytic strategy was used: the data were plotted to visualize trends and then
Spearman’s correlation was calculated.
Roost sites were treated as independent observations. Treating the observations as
independent makes sense because temperate bats species are expected to return to the
same roosts over the 4-year timescale during which they were observed (Pettit &
O’Keefe, 2017). The data for each research question were plotted to observe trends. No
linear relationships were observed. Moreover, there is no reason to suspect that the
variables of interest should be linearly correlated. Thus, it would be inappropriate to use a
statistical measure of correlation that assumes linearity, i.e. Pearson’s correlation. Rather,
a non-parametric method, i.e. Spearman’s correlation, would be more appropriate.
33
Spearman’s correlation ranges from -1 to +1. Numbers close to +1 indicate a positive
association between the two variables, while numbers close to -1 indicate a negative
association (Andri et al., 2021).
All data analyses were performed using R Version 4.1.2 (R Core Team, 2019).
Spearman’s correlation and associated P values and confidence intervals were calculated
using the SpearmanRho function in the DescTools R package (version 0.99.44)(Andri et
al., 2021).
Scatterplots were generated using the package ggplot2 (Wickham, 2016). Finally,
power analysis was performed using the package Webpower (Zhang & Mai, Yujiao,
2021).
Variables Used in Data Analysis
Research Question One: Are bat sightings affected by how early in the year the
temperature regularly exceeds 40°F?
Microbats, especially the bats found on JBLM, hibernate when the low temperature is
around 40°F (Wildlife Online, n.d.). There are contrasting reports in the literature as to
the exact temperature at which bats begin to awaken from torpor or hibernation. Reported
values ranged from approximately 35°F to 66°F and could be attributed to variations in
the bat species (Anufriev et al., 2003; Frick et al., 2012; Wildlife Online, n.d.). Since
microbats prefer to hibernate at 40°F, a temperature ≥ 40°F was chosen to identify when
bats are likely begin to emerge from their hibernaculum or winter colonies (Frick et al.,
2012). In this study, this variable is referred to as The First Day Above 40°F
(FDA40). The FDA40 specifically identifies the first day in the calendar year on which
the seven-day average low temperature exceeded 40 degrees Fahrenheit. The FDA40 was
34
calculated using Microsoft Excel by taking daily temperatures pulled from the Weather
Underground website for the months January through May of the respective years (The
Weather Company, 2021). This time frame was selected because January marked the
beginning of the calendar year and the average low temperature in the Pacific Northwest
was above 40 degrees in the Month of May for all observation years. Thus, any data
beyond May was not required.
Research Question Two: Are bat sightings affected by total annual rainfall?
Addressing the relationship between the number of bat sightings and total annual
rainfall required calculating the total annual rainfall for each of the observation years.
Total annual rainfall was chosen because it is a metric, albeit not a perfect one, that relates
to both direct effects on bats and indirect effects via the impact of rainfall on the
availability of food sources. Moreover, precipitation data were inconsistently collected at
the time of the surveys. Thus, daily precipitation data for individual sites were unreliable.
Consequently, annual precipitation data were downloaded from Weather Underground
(The Weather Company, 2021) and imported into Excel for analysis. The total
precipitation between January 1 and December 31 of each year were calculated by
summing the monthly precipitation in each respective survey year. These values are
reproduced in Table 3 below.
Year
2018
2018
2018
2019
2019
2019
2020
Precipitation
29.68
29.68
29.68
25.7
25.7
25.7
34.61
Roost/Site
Chamber’s Bridge
Military Museum
Watkin’s Boxes
Chamber’s Bridge
Military Museum
Watkin’s Boxes
Chamber’s Bridge
Sightings
31
26.5
14
28.5
31.5
24.5
30
35
2020
2020
2021
2021
2021
34.61
34.61
37.71
37.71
37.71
Military Museum
Watkin’s Boxes
Chamber’s Bridge
Military Museum
Watkin’s Boxes
27
22
16
0
7.5
Table 3. Total Precipitation at Roost Sites. It shows the annual precipitation for each
roost location and the corresponding bat sightings for each roost in each survey year.
Research Question Three: Since day-to-day weather conditions can affect survey
results, which conditions are important to consider when conducting surveys?
Weather metrics such as temperature and windspeed were collected by the surveyors
at the time of observation. These data were documented at the beginning of the
observation period in the survey field observation form which was later digitally
transcribed. While the other research questions used data that spanned over all four
survey years, temperature and windspeed at the time of observation were only recorded
for survey years 2020 and 2021. The reason for the inconsistency in this aspect of the
surveying methodology is unknown. As a consequence, the data used to address research
question three includes only the years 2020 and 2021.
However, despite the data only being collected over the span of two years, survey
year 2020 observed 7 roost sites and 2021 observed 8 roost sites as depicted in Table 2.
The analyses in this section are organized according to the two types of weather
conditions recorded at the time of observation: temperature and wind speed. Precipitation
would be another meaningful metric that would be taken into consideration, however, it
was not collected at the time of the surveys.
Research Question One
Are bat sightings affected by how early in the year the temperature regularly
exceeds 40°F?
36
To determine whether there is a correlation between the variable First Day Above
40 Degrees (FDA40) and the number of bat sightings (Fig. 4), Spearman’s correlation
was calculated. All available data without missing values were used for this analysis: the
years 2018 to 20201 and three roosts (n=12). Spearman’s correlation was 0.02 with a
95% Confidence Interval (CI) of -0.56 to 0.59. The associated p-value was 0.95,
indicating that the FDA40 and bat sightings are not correlated. Figure 4 shows the three
different roosts represented using a different color for each roost: red for Chamber’s
Bridge, green for the Military Museum, and blue for Watkin’s Bat Boxes. Each survey
year is represented using a different symbol shown as a circle for 2018, a triangle for
2019, a square for 2020, and a plus sign for 2021. No statistically significant correlation
between FDA40 and the number of annual bat sightings was identified. This could be
interpreted in several ways: 1) the variables are not correlated, 2) the sample size is too
low n=12 (statistically underpowered) to detect the true effect size, or 3) the variables
FDA40 and bat sightings may not capture salient environmental factors or bat population
size. In other words, FDA40 may not be a good indicator of when bats emerge from
hibernation or torpor. Although this study has a small sample size (n=12), this is quite
common for bat studies and points to the need for a coordinated, large-scale, and
comprehensive monitoring effort.
37
First Day above 40°F (FDA40) and Bat Sightings
n=12
Figure 4. First Day above 40°F and Bat Sightings. Relationship between the FDA40 and
number of bat sightings. The FDA40 indicates the first calendar day at which the 7-day
average low temperature exceeded 40°F. Colors identify roosts and shapes identify
observation year. The FDA40 and number of bat sightings were uncorrelated (p=0.95).
Spearman’s correlation was 0.02 with a 95% CI of -0.56 to 0.59. n=12. Image Credit: Erika
Larson, 2021.
Research Question Two
Are bat sightings affected by total annual rainfall?
To determine if there was a correlation between precipitation and annual bat
sightings (Fig. 5), Spearman’s correlation was calculated. All available data were
included in this analysis, including the years 2018 through 2021 and three roosts (n=12).
Spearman’s Correlation was -0.63 with a 95% CI of -0.88 to -0.08. The associated pvalue was 0.029, indicating that total annual rainfall and bat sightings are negatively
correlated, which is significant at the alpha=0.05 level. However, it should be noted that
38
the confidence interval is very wide, highlighting that significant uncertainty remains as
to the true size of this effect. Nevertheless, a statistically significant correlation of -0.63
suggests that total annual rainfall has a large impact on bat emergence behavior.
Figure 5 shows the three different roosts represented using a different color for
each roost: red for Chamber’s Bridge, green for the Military Museum, and blue for
Watkin’s Bat Boxes. Each survey year is represented using a different symbol shown as a
circle for 2018, a triangle for 2019, a square for 2020, and a plus sign for 2021.
Annual Precipitation (inches) and Bat Sightings per Year
n=12
Figure 5. Annual Precipitation (inches) and Bat Sightings. Relationship between the
annual precipitation (inches) and number of bat sightings. Colors identify roosts and
shapes identify observation year. Total annual rainfall and the number of bat sightings
were negatively correlated (p=0.029). Spearman’s Correlation was -0.63 with a 95% CI
of -0.88 to -0.08. n=12. Image Credit: Erika Larson, 2021
39
Research Question Three
Since day-to-day weather conditions can affect survey results, which conditions are
important to consider when conducting surveys?
Temperature and windspeed recorded during bat surveys were analyzed to
determine if they are associated with the number of bat sightings. While research
question one addressed temperature, it did not consider temperature at the time of
observation.
To determine if there was a correlation between temperature at the time of
observation and bat sightings (Fig. 6), Spearman’s correlation was calculated. Data from
the years 2020 and 2021 were included which were 8 roosts 7 from 2021 due to
inconsistencies in data collection. All the roosts in the 2020 and 2021 data were surveyed
twice per year, except for Holden Woods RB from 2020 and Fish Hatchery from 2021
which were only surveyed once during those respective years (n=28). Spearman’s
Correlation was 0.01 with a 95% CI of -0.37 to 0.39. The associated p-value was 0.97
indicating, that temperature and the number of bat sightings are not correlated.
40
Temperature and Bat Sightings for 2020 and 2021
n=28
Figure 6. Temperature and Bat Sightings. Relationship between the temperature and
number of bat sightings. Roosts are indicated by color and survey year is indicated by
point shape (circle=2020 and triangle=2021). These data identified no correlation
(p=0.97) between the temperature documented at the time of the survey and the
number of bat sightings. Spearman’s correlation was 0.01 with a 95% CI of -0.37 to
0.39. n=28. Image Credit: Erika Larson
To determine if there was a correlation between wind speed at the time of
observation and the number of bat sightings (Fig. 7), Spearman’s correlation was
calculated. The same data and sample size (n=28) were used in this analysis as in the
temperature analysis described above (Figure 6). Spearman’s Correlation was 0.11 with a
95% CI of -0.28 to 0.47. The associated p-value was 0.60 indicating that wind speed and
the number of bat sightings are not correlated.
41
Wind (mph) and Bat Sightings
n=28
Figure 7. Wind (mph) and Bat Sightings. Relationship between the windspeed and
number of bat sightings. Wind speed was recorded at the time of the survey. Roosts are
indicated by color and survey year is indicated by point shape (circle=2020 and
triangle=2021). Wind speed and the number of bat sightings were uncorrelated (p=0.60)
with a Spearman’s correlation of 0.11 (95% CI -0.28, -0.47). Image Credit: Erika Larson,
2021
42
VI.
DISCUSSION
The objective of this study was to determine which weather metrics impact the
number of bat sightings during surveys. Prior research suggested that there was an
interplay of three different contributors to bat population decline: disease, habitat loss, and
climate change. However, there were few studies that took into consideration weather
metrics such as temperature, precipitation, and wind. These are important factors to
evaluate since weather is predicted to change in the future due to climatic changes and
global warming. This study attempted to fill that knowledge gap.
Question 1: Are bat sightings affected by how early in the year the temperature
regularly exceeds 40°F?
It was predicted that bat sightings would be affected by how early in the year the
temperature exceeded 40 degrees. Specifically, it was predicted that there would be more
annual bat sightings for the years that saw a later day of emergence from the bat winter
roosting sites or hibernaculum. To evaluate this prediction, the variable First Day Above
40 Degrees, or FDA40 was created to quantitatively define the time at which the
temperatures of JBLM became amenable to bat emergence. Note that there is debate as
to what specific temperature defines this criterion. Regardless, the present analysis failed
to identify a correlation between the FDA40 and number of bat sightings at the respective
roosts (p=0.95). Spearman’s correlation was 0.02 with a 95% Confidence Interval (CI) of
-0.56 and 0.59.
These statistical results agree with visual interpretation of the data. As illustrated
in Figure 4, the variables FDA40 and the number of bat sightings are not visibly
43
correlated. This result is surprising since the lack of correlation suggests that the changes
in the temperature from the winter months to the spring does not impact bat emergence.
So, either this is not the correct metric to be using to predict the occurrence of annual bat
sightings, or there could be other factors in play. Reports in the literature indicate that if a
bat emerges too early there could be limitations on the food that is available for foraging
(Frick et al., 2012). On the other hand, bats survive the winter months by using their fat
stores while hibernating, meaning that long hibernation times could also put them at risk.
The short duration of this study (four years) could also be creating limitations necessary
to detect climatic changes (Frick et al., 2012). According to the National Oceanic and
Atmospheric Administration (NOAA), climate is the average of weather patterns in a
location of 30 years or more (NOAA, 2016). Climate is slow to change over time, thus, a
longer duration of study would be required to begin seeing the impacts of these changes
on the hibernation times. So, extending the number of years included in this survey may
be necessary to generate results that support the use of FDA40 to predict the number of
bat sightings.
According to the U.S. Fish and Wildlife Service, bats that live in coastal and
Puget sound lowlands could be active all year long (WDFW, 2020). Two species have
been observed foraging in the winter months (Falxa, 2007; U.S. Fish and Wildlife, 2020).
So, there is the possibility that the FDA40 is the wrong metric to be using. Pacific
Northwest bats could be going into mild states of torpor instead of a full hibernation. For
example, species such as the silver-haired bats (Lasionycteris noctivagans) and the
California myotis (Myotis californicusI), both species which are located on JBLM (Table
2), go into a state of torpor during severe weather, and become active during good
44
weather (WDFW, 2020). This suggests that it is possible that Pacific Northwest bats have
acquired a type of adaptive emergence behavior (Frick et al., 2012).
It has also been reported that early emergence can lead to either: 1) an increased
risk of predation and competition with other diurnal insectivores or 2) the forfeiture of
optimum foraging opportunities during peak prey availability (Frick et al., 2012). If bats
are capable of adaptive emergence behavior then it is possible that some of the species on
JBLM migrate during the winter. In the 2008 survey, the hoary bat (Lasiurus cinereus)
was identified on JBLM. This species of bat survives the winter by migrating more than
1,000 km (WDFW, 2020). However, the same species of bat has also been known to
hibernate in the winter (Weller et al., 2016). The hibernation, torpor, and migratory
strategies of bats in the Pacific Northwest are not well understood. This points to the need
for more comprehensive surveys that include speciation.
Lastly, it is also possible that the short duration of the current survey (four years)
could have caused this study to fail to detect associations between variables that are
logically connected. For example, it is possible that the number of bat sightings could be
related to events from the previous year such as the impact of temperature on food
supply, lower annual summer or fall temperature. These are things that could potentially
affect bat population size, thus, affecting the number of bat sightings in the subsequent
year.
Question 2: Are bat sightings affected by total annual rainfall?
It was predicted that total annual rainfall would negatively impact the number of
bat sightings. A negative correlation was identified between precipitation and the number
45
of bat sightings (p=0.029). Spearman’s correlation was -0.63 with a 95% CI of -0.88 to 0.08. This supports the prediction the number of bats sightings will decrease as total
annual rainfall increases.
According to a study conducted by Voight et al., when a bat becomes wet, its
pelage and wings moisten, thus increasing their metabolic requirements for flight. It is
unknown whether this is caused by the rainfall lowering their body temperatures and
increasing their metabolisms or if it makes them less aerodynamic. For example, it has
been proposed that rain induces clumping of their fur which could hinder flight and
increase metabolic rates (Braconnier, 2011). While this study was conducted in Costa
Rica, the bats are in the same suborder as the bats found on JBLM. Leaf Nosed Bats, such
as those studied by Braconnier et al., have similar physical characteristics and habitats as
Vesper bats seen on JBLM, which suggests that their fur and skin could also share similar
characteristics (Birkett et al., 2014; Braconnier, 2011). Voight et al. reported that aberrant
flight during rain is due to moisture on the bat rather than disorientation due to the
raindrops. Specifically, wet fur reportedly causes bats to have higher metabolic rates
(Voight et al., 2011). These reports support the present study’s findings.
The Colony Assessment Protocol that provides the surveying guidelines instructs
that the surveyors are not to continue with the emergence counts if there is heavy
precipitation or fog present (Washington Department of Fish and Wildlife, 2018). There
is no information regarding why these guidelines were included. The guidelines did not
provide a measurement guide to indicate how much precipitation is considered heavy or
how much fog needs to be present to discontinue the survey. The duration of the bat
surveys could also be a factor since the surveyors were only conducting emergence
46
counts for 1-2 hours starting at dusk. However, there is no documentation or information
provided to account for any survey cancellations or reschedules due to inclement weather
events.
Question 3: Since day-to-day weather conditions can affect survey results, which
conditions are important to consider when conducting surveys?
The goal of question three was to define the relationship between the number of
bat sightings and specific weather metrics recorded at the time of observation:
temperature and wind speed. The only two weather metrics documented during the bat
surveys conducted during this study were temperature and windspeed, both of which
were recorded at the beginning of the surveying session. A major limitation in evaluating
these metrics is that they were only recorded for survey years of 2020 and 2021. There
was no information provided as to why this change in protocol occurred or why these
readings might have been left out for years 2018 and 2019. Although question three
includes fewer years of data, it includes eight different roosts in 2020 and seven in 2021.
Thus, the resulting sample size was 28.
Metric: Temperature
It was predicted that a higher temperature would increase the amount of bat
sightings. Spearman’s correlation was 0.01 with a 95% CI of -0.37 to 0.39 and a p-value
of 0.97. Interestingly, the temperatures at the beginning of the observation periods were
not correlated with the number of bat sightings. This suggests that an increase in
temperature does not affect the number of bat sightings on JBLM. This is counter
intuitive since bats have been reported to be more active in warm weather (Minnesota
47
Wild Animal Management, 2020). Many studies suggest that at moderate temperatures,
activity levels are maximized (Anufriev et al., 2003; Dahl et al., n.d.; Frick et al., 2012).
If this is the case, then Pearson’s and Spearman’s correlations would not be good metrics
because they detect only strictly increasing or strictly decreasing relationships, and not
those that increase then subsequently decrease. The latter relationship was not observed
in this study, as illustrated in Fig. 6. Therefore, this observation supports the selection of
statistical test used in this study. On the other hand, it is possible that this relationship
exists but was obfuscated by the high variability in the number of bat sightings across the
narrow temperature range - spanning from 57°F to 88°F. Since this is a typical
temperature range for Western Washington, a substantially enlarged sample size may be
required to detect such a relationship. While Hoeffding’s D statistic could detect this kind
of relationship, it 1) would be statistically underpowered and 2) is not easily interpretable
ranging from -0.5 to 1 depending on whether there are ties in the ranked data (Harrell Jr.
& Dupont, 2021; Hoeffding, 1948).
To gain a comprehensive understanding of the relationship between temperature
and bat activity, observations would be required across a broader range of temperatures.
This is important because models produced by the University of Washington predict that
there will be temperature increases in the Pacific Northwest (University of Washington,
2020). While the present study failed to detect a correlation, other studies have identified
correlations between bat sightings and temperature (Duvergé et al., 2000; Pettit &
O’Keefe, 2017). These changes in temperature affect each bat species differently (Groc,
2021). The differential effects can be due to bat size, metabolism, behavior, and roosting
location. For example, bats that roost in artificial roosts such as bat boxes are at risk for
48
rapid heating, potentially leading to overheating and dehydration (Groc, 2021). As
climate change progresses, these artificial roosts may no longer be safe. According to
Cori Lausen from Wildlife Conservation Society of Canada, “Bats walk a tightrope
because a few degrees can make a difference to whether they live or die” (Groc, 2021).
Due to habitat loss, artificial roosts are becoming more common. If heat waves like the
one experienced in Western Washington in June 2021 become more common, then bat
mortality is likely to increase. Since bats have a slow reproductive rate it is difficult for
their population to recover from these heat events (Groc, 2021).
To mitigate the impact of heat on bats roosting in artificial roosts, researchers
have begun investigating different bat box designs and the impact of their installation
locations. They have found that some bat boxes can reach a temperature of 52°C (Groc,
2021). Researchers have found that different modifications to these artificial roosts such
as adding a chimney, water chamber, or painting the boxes a lighter color could be
helpful in stabilizing the temperature inside these roosts (Groc, 2021). Interestingly, bats
seem to choose their roosting site based on its location rather than these other factors. All
of the above observations point to a substantial lack of information concerning: 1)
Species specific responses to increased temperatures, 2) how bats select their roosting
sites, and 3) how the impact of severe weather events involving bats can be mitigated.
Metric: Wind Speed
It was predicted that an increase in wind speeds would decrease the number of bat
sightings. However, Spearman’s correlation was 0.11 with a 95% CI of -0.28 to 0.47 and
a p-value of 0.60. Since the data were only documented for two years it was difficult to
49
determine whether there was any significant trends over time. Wind speeds at the
beginning of the observation periods during 2020 and 2021 were not correlated with the
number of bat sightings. This suggests that wind speed does not affect bat activity. This is
surprising, and contrasts with reports that bat activity is associated with wind speed (Bach
et al., 2011). Moreover, there were no obvious non-monotonic relationships between the
variables (e.g. increasing then decreasing) (Fig. 7). Thus, failure to detect a correlation
cannot be attributed to improper use of statistical methods. In other words, the use of
Spearman’s correlation was justified for this analysis.
Previous studies have indicated that wind speed does affect the number of bat
sightings (Bach et al., 2011; Pettit & O’Keefe, 2017; WINDExchange, n.d.). According
to German Nature Conservation Organizations and Authorities, different species of bats
show different tolerances to windspeed. This is at least in part due to bat size (Bach et al.,
2011). Since bat species in the Pacific Northwest are small, they could be more
susceptible to an increase in wind speeds than larger bats would be. The literature in this
regard is limited with most studies focusing on the interactions between bats and wind
turbines, not the wind itself. The lack of data in this regard is especially concerning since
wind speeds can be highly variable. Interestingly, in the present study, there appears to be
a substantial difference in wind speed readings between years, with 2020 having higher
average wind speeds (8.5 mph) compared with 20201 (4.6 mph). The average number of
bat sightings per roost in 2020 and 2021 were 18.08 and 3.73, respectively. These
surprising results could potentially be an artifact of the method of data collection.
However, that is difficult to discern, as there was no indication as to how the wind speed
readings were collected by the volunteers conducting the surveys.
50
While these results do not support the hypothesis that these weather metrics are
associated with the number of bat sightings on JBLM, they do highlight that more
research needs to be conducted to include a larger sample size, longer duration of
surveying practices, standardization of surveying practices, and consistent training of
personnel. All of these could have had impacts on the underlying data and thus the
conclusions for the research questions being evaluated during this study.
51
VII.
CONCLUSION
The objective of this study was to determine which weather metrics impact the
number of bat sightings. JBLM was used as a case study because the confounding factors
that cause bat population decline elsewhere are minimal or absent. This strategy allowed
for three key weather metrics (temperature, precipitation, and wind) to be evaluated using
data collected between 2018 and 2021. These data were collected by surveyors with the
Department of Fish and Wildlife, publicly available sources. Using these data, three
questions were asked:
Question 1: Are bat sightings affected by how early in the year the temperature regularly
exceeds 40°F?
Question 2: Are bat sightings affected by total annual rainfall?
Question 3: Since day-to-day weather conditions can affect survey results, which
conditions are important to consider when conducting surveys?
A statistically significant association was detected for question two. No associations
were detected for questions one and three. This points to the conclusion that bat sightings
during the summer are less frequent in years when the total annual precipitation is high.
However, these findings, both positive and negative, may have been impacted by the
study’s limitations: 1) metrics that inaccurately reflect the variable of interest, 2) small
sample size, and 3) lack of standardization and training.
Weather data are widely available. However, the data are in a format that cannot
readily be used to assess the impact of weather on the number of bat sightings. In
addition, seasonal and daily trends complicate analysis. To relate the number of bat
52
sightings to weather metrics, weather data must be aggregated into simple and
interpretable values. In this study, an attempt was made to devise variables that are
meaningful in the context of bat biology and robust to variation. For example, the
variable First Day Above 40°F (FDA40) was used in this study. While it is defined as the
first day in the calendar year when the seven-day average of daily low temperatures
exceeds 40°F, it can be interpreted as the first day of the year on which daily lows
regularly exceed 40°F. Though this temperature was selected because bats tend to emerge
from hibernation or torpor at this temperature, bat species may be differentially affected
by temperature. Therefore, FDA40 may be a good metric for some species but a poor
metric for others. Weather metrics may also have impacted the data by introducing an
unexpected bias. Specifically, the guidelines outlined in the Colony Assessment Protocol
(CAP) concerning precipitation and wind are likely to introduce a bias. The CAP
guidelines state that surveyors should stop an emergence survey if the following weather
conditions are met: heavy precipitation or high wind speed (Washington Department of
Fish and Wildlife, 2018). Since the guidelines do not specify exact thresholds for these
conditions, the subjectivity of the surveyors is introduced.
To improve the quality of survey data, a protocol with more specific guidelines is
needed. It should incorporate specific thresholds for rain and windspeed that identify
when a survey should be terminated. These thresholds are needed for multiple reasons
including the impact of these metrics on bat activity and the visibility of the bats to the
surveyors. To overcome the latter issue, visual aids different could be used (e.g., infrared
or cameras). Alternatively, a combination of methods could be employed. Mist-netting
53
and acoustic sampling have proven to be the most effective for bat monitoring and should
be considered for use in future studies on JBLM.
The second major limitation was sample size. The sample sizes used in this study
were low. Thus, it was not surprising that this study failed to detect associations between
variables that are logically connected (e.g. number of bat sightings and wind speed). Even
though, there were four years of data collected, the fact that the survey conducted in 2019
only observed three roosts severely limited the data that could be analyzed. Consistency
of site monitoring across survey years would improve the identification of factors that are
associated with the number of bat sightings.
In addition to inconsistent observation of roost sites, some variables were missing
in 2018 and 2019 data. Temperature and wind speed were only collected in 2020 and
2021. The inconsistent documentation of these metrics further limited the sample size,
resulting in only two years of usable data. Two years is insufficient to identify a trend. It
is recommended that these metrics be documented consistently at the start of each
observation, period. More complete and accurate documentation of weather metrics
expanding beyond just temperature, precipitation, and wind are also recommended (e.g.
humidity, fog, pressure). This strategy would facilitate a conversation about the types of
conditions that affect bat sightings. Importantly, it could facilitate the prediction of the
impacts of climate change on bat population health.
The third major limitation is two-fold, a lack of standardization and a lack of
training for surveyors. Standardization efforts should focus on bat speciation, counting,
and recording weather conditions. Historically, three surveying strategies have been used:
visual emergence counts, netting, and acoustic sampling. However, to incorporate these
54
methods, more training needs to be provided to the volunteers conducting the bat surveys
on JBLM. Ideally, this would involve a collaborative effort amongst bat monitoring
organizations in the region such as the Washington Department of Fish and Wildlife, Bats
Northwest, Western Bat Working Group, Cascadia Research, and the Nature
Conservancy. Standardized training would go hand-in-hand with standardization of
regional surveying methodologies. Sharing the expertise of representatives from these
organizations could reduce biases resulting from the methods employed or lack of
surveyor experience. Currently, JBLM only conducts emergence count surveys. While
emergence count surveys are cost effective, they generate low quality data. Specifically,
they produce unreliable data when the surveyors are inadequately trained. To circumvent
this problem, it is recommended that future surveys incorporate acoustic sampling
technologies and live capture tactics such as mist-nets. This will produce a more accurate
representation of JBLM’s bat population. While this would require more comprehensive
training for surveyors, it would dramatically improve the resulting data.
In summary, this study identified a negative correlation between the number of
bat sightings and total annual rainfall. The study’s conclusions may have been impacted
by 1) metrics that inaccurately reflect the variable of interest, 2) small sample size, and 3)
lack of standardization and training. These findings and limitations lead to the following
recommendations. 1) A combination of all three surveying methods should be utilized
(emergence count surveys, live-capture, and acoustic sampling). This may require a longterm funding source since the upfront costs would include equipment and training. 2)
Surveys should be conducted over a longer timeframe and at consistent roosting sites.
This will enable the detection of the impacts of climatic changes. 3) Subject matter
55
experts should carefully consider and devise weather metrics that are interpretable and
meaningful in the context of bat biology. 4) The incorporation of data sharing and
regional training tactics would create transparency among organizations in the area and
increase awareness and communication. Overall, the implementation of these
recommendations would significantly improve future research and conservation efforts
by enabling important questions to be asked and ensuring that data are both high quality
and comprehensive, incorporating all factors of interest.
56
REFERENCES
American Museum of Natural History. (2008, February). Researchers Find Bats
Evolved Ability To Fly Before Echolocation | AMNH. American Museum of
Natural History. https://www.amnh.org/research/sciencenews/2008/researchers-find-bats-evolved-ability-to-fly-before-echolocation
Ammerman, L. K., Hice, C. L., & Schmidley, D. J. (2012). Bat-Watching Sites of
Texas. Texas A&M University Press, November.
https://tpwd.texas.gov/publications/pwdpubs/media/pwd_bk_w7000_1411.
pdf
Anufriev, A. I., Solomonova, T. N., Arkhipov, G. G., Turpanov, A. A., &
Solomonov, N. G. (2003). Bat (Vespertilionidae) Hibernation in the
Northeasternmost Part of Their Geographic Range. Doklady Biological
Sciences, 392(1), 413–415. https://doi.org/10.1023/A:1026175805053
Aubry, K. B., Hayes, J. P., Biswell, B. L., & Marcot, B. G. (2003). The ecological
role of tree-dwelling mammals in western coniferous forests. In C. J. Zabel
& R. G. Anthony (Eds.), Mammal Community Dynamics (1st ed., pp. 405–
443). Cambridge University Press.
https://doi.org/10.1017/CBO9780511615757.013
Bach, P., Niermann, I., & Bach, L. (2011). Impact of wind speed on the activity of
bats-at the coast and inland. 1.
Baker, M. D., & Lacki, M. J. (2004). Forest Bat Communities in the east Cascade
Range, Washington. Northwest Science, 78(3), 234–241.
57
Bat Conservation International. (2016). Calculate the Value of Bats.
https://www.fs.usda.gov/Internet/FSE_DOCUMENTS/fseprd476773.pdf
Bat Conservation International. (2021). Bats 101. Bat Conservation International.
https://www.batcon.org/about-bats/bats-101/
Bat Conservation Trust. (2021, February). Drivers of bat declines—News—Bat
Conservation Trust. https://www.bats.org.uk/news/2021/02/drivers-of-batdeclines
Bat World Sanctuary. (2013). Bat Reproduction—Bat Facts and Information.
https://www.batworlds.com/bat-reproduction/
Birkett, K. M., Weidman, K. S., & Woo, Y. (2014). Vespertilionidae (evening bats
and vesper bats). Animal Diversity Web.
https://animaldiversity.org/accounts/Vespertilionidae/
Braconnier, D. (2011, May). Why does rain keep bats grounded?
https://phys.org/news/2011-05-grounded.html
Briggs, P. (2021). Where do bats go in the winter? Bat Conservation Trust.
https://www.discoverwildlife.com/animal-facts/mammals/where-do-batsgo-in-the-winter/
Brunet-Rossini, A. K., & Wilkinson, G. (2009). Ecological and Behavioral
Methods for the Study of Bats. John’s Hopkins University Press.
https://www.researchgate.net/publication/260244816_Methods_for_age_e
stimation_and_the_study_of_senescence_in_bats
Burnett Mary Regional Group. (2021). Micro-bats. All About Bats.
https://www.allaboutbats.org.au/micro-bats/
58
Collins, J. (BCT), Bat Conservation Trust, & Chartered Institute of Ecology and
Environmental Management. (2016). Bat survey guidelines for
professional ecologists: Good practice guidelines.
Dahl, F., Davis, J., & Buckingham, D. (n.d.). Bat Conservation on Joint Base
Lewis-McChord and Weather Effects on Emergence Times. 1.
Darras, K. F. A., Yusti, E., Huang, J. C.-C., Zemp, D.-C., Kartono, A. P., &
Wanger, T. C. (2021). Bat point counts: A novel bat sampling method
shines light on flying bat communities [Preprint]. Preprints.
https://doi.org/10.22541/au.163039378.81774988/v1
de Oliveira, F. V. (2020). Microchiroptera Life History. In J. Vonk & T.
Shackelford (Eds.), Encyclopedia of Animal Cognition and Behavior (pp.
1–15). Springer International Publishing. https://doi.org/10.1007/978-3319-47829-6_1158-1
Delvin, E. (2013). Restoring Abandoned Agricultural Lands in Puget Lowland
Prairies: A New Approach [Thesis].
https://digital.lib.washington.edu:443/researchworks/handle/1773/23703
Duvergé, P. L., Jones, G., Rydell, J., & Ransome, R. D. (2000). Functional
significance of emergence timing in bats. Ecography, 23(1), 32–40.
https://doi.org/10.1111/j.1600-0587.2000.tb00258.x
ED DPW, F. B. (2017, October). JBLM Forest Management Plan 2017. Army.Mil.
https://home.army.mil/lewismcchord/application/files/7515/5371/9197/JBLMForestManagementPlan.p
df
59
Falxa, G. (2007). WINTER FORAGING OF SILVER-HAIRED AND CALIFORNIA
MYOTIS BATS IN WESTERN WASHINGTON. Northwestern Naturalist,
88(2), 98–100. https://doi.org/10.1898/10511733(2007)88[98:WFOSAC]2.0.CO;2
Falxa, G. (2008a, October 31). 2008 Fort Lewis Bat Survey. Cascadia Research.
https://www.cascadiaresearch.org/files/Projects/Archived_projects/Bats/Fo
rt_Lewis_2008_Bat_Survey_Final_sm.pdf
Falxa, G. (2008b). Fort Lewis 2008 Bat Survey -Final. The Nature Conservancy.
https://www.cascadiaresearch.org/files/Projects/Archived_projects/Bats/Fo
rt_Lewis_2008_Bat_Survey_Final_sm.pdf
Fleming, T. H., Geiselman, C., & Kress, W. J. (2009). The evolution of bat
pollination: A phylogenetic perspective. Annals of Botany, 104(6), 1017–
1043. https://doi.org/10.1093/aob/mcp197
Freed, S., & Falxa, G. (2010). Bat Box Preference Study on Fort Lewis,
Washington. The Nature Conservancy.
https://www.cascadiaresearch.org/files/Projects/Archived_projects/Bats/Ba
tBoxPreference_screen-view.pdf
Frick, W. F., Kingston, T., & Flanders, J. (2019). A review of the major threats
and challenges to global bat conservation. Annals of the New York
Academy of Sciences, 1469(1), 5–25. https://doi.org/10.1111/nyas.14045
Frick, W. F., Stepanian, P. M., Kelly, J. F., Howard, K. W., Kuster, C. M., Kunz, T.
H., & Chilson, P. B. (2012). Climate and Weather Impact Timing of
60
Emergence of Bats. PLOS ONE, 7(8), e42737.
https://doi.org/10.1371/journal.pone.0042737
Gettler, L. (2013, January 25). Understanding bat evolution could lead to new
treatments for viruses and aging. New Atlas. https://newatlas.com/batscsiro-cancer-immmune-systems/25953/
Griggs, M. B. (2015, September 15). Bats Are Worth $1 Billion To Corn Industry.
Popular Science. https://www.popsci.com/bats-are-worth-1-billion-to-cornindustry/
Groc, I. (2021, February 24). Hot houses: The race to save bats from overheating
as temperatures rise. The Guardian.
https://www.theguardian.com/environment/2021/feb/24/race-to-save-batsflying-foxes-from-overheating-as-temperatures-rise-aoe
Gunnell, G., & Simmons, N. (2005). Fossil Evidence and the Origin of Bats.
Journal of Mammalian Evolution, 12. https://doi.org/10.1007/s10914-0056945-2
Gunnell, G., Simmons, N., & Seymour, K. L. (2008, February). Bats Flew First,
Developed Echolocation Later, Fossilized Missing Link Shows.
ScienceDaily.
https://www.sciencedaily.com/releases/2008/02/080213121444.htm
Harrell Jr., F. E., & Dupont, C. (2021). Hmisc: Harrell Miscellaneous (4.6-0)
[Computer software]. https://CRAN.R-project.org/package=Hmisc
Hayes, G., & Wiles, G. J. (2013). Washington State Bat Conservation Plan. 150.
61
Hayes, J. P. (2003). Habitat ecology and conservation of bats in western
coniferous forests. In C. J. Zabel & R. G. Anthony (Eds.), Mammal
Community Dynamics: Management and Conservation in the Coniferous
Forests of Western North America (pp. 81–119). Cambridge University
Press. https://doi.org/10.1017/CBO9780511615757.005
Hoeffding, W. (1948). A non-parametric test of independence. Ann Math Stat, 19,
546–557.
Kunz, T. H., Braun de Torrez, E., Bauer, D., & Lobova, T. (2011). Bats
Ecosystem Services Provided by Bats.
https://www.documentcloud.org/documents/6888737-BatsEcosystem.html
Lactis, E. (2020, October 4). Before our Northwest bats go into hibernation, a
little sympathy. No, they don’t carry the coronavirus. They just eat bugs.
The Seattle Times. https://www.seattletimes.com/seattle-news/before-ournorthwest-bats-go-into-hibernation-a-little-sympathy-no-they-dont-carrythe-coronavirus-they-just-eat-bugs/
Law, A. S., Khushali Shah, McAndrews, E., & Stumpf, J. (2019). Impacts of
anthropogenic disturbance and insect abundance on Sonoran Desert bat
activity [Pdf]. https://doi.org/10.21973/N3K378
Link, R. (2004). Living with Wildlife—Bats. Washington Department of Fish &
Wildlife.
https://wdfw.wa.gov/sites/default/files/publications/00605/wdfw00605.pdf
62
Loeb, S. C., & O’Keefe, J. M. (2011). Bats and Gaps: The Role of Early
Successional Patches in the Roosting and Foraging Ecology of Bats.
https://www.academia.edu/11468651/Bats_and_Gaps_The_Role_of_Earl
y_Successional_Patches_in_the_Roosting_and_Foraging_Ecology_of_Ba
ts
Loeb, S. C., Rodhouse, T. J., Ellison, L. E., Lausen, C. L., Reichard, J. D., Irvine,
K. M., Ingersoll, T. E., Coleman, J. T. H., Thogmartin, W. E., Sauer, J. R.,
Francis, C. M., Bayless, M. L., Stanley, T. R., & Johnson, D. H. (2015). A
plan for the North American Bat Monitoring Program (NABat) (SRS-GTR208; p. SRS-GTR-208). U.S. Department of Agriculture, Forest Service,
Southern Research Station. https://doi.org/10.2737/SRS-GTR-208
Luzynski, K. C., Sluzas, E. M., & Wallen, M. M. (2009). Pteropodidae (Old World
fruit bats). Animal Diversity Web.
https://animaldiversity.org/accounts/Pteropodidae/
Mahandran, V., Murugan, C. M., Marimuthu, G., & Nathan, P. T. (2018). Seed
dispersal of a tropical deciduous Mahua tree, Madhuca latifolia
(Sapotaceae) exhibiting bat-fruit syndrome by pteropodid bats. Global
Ecology and Conservation, 14, e00396.
https://doi.org/10.1016/j.gecco.2018.e00396
Marshall, M. (2022, January 26). A tiny change to an inner ear bone led bats to
evolve into two groups. New Scientist.
https://www.newscientist.com/article/2305724-a-tiny-change-to-an-innerear-bone-led-bats-to-evolve-into-two-groups/
63
Maslo, B., & Kerwin, K. (2020, December). Ecological and Economic Importance
of Bats in Integrated Pest Management. NJAES Rutgers.
https://njaes.rutgers.edu/fs1270/
Minnesota Wild Animal Management. (2020, March). Bats Get Active in Warm
Weather | Minnesota Wild Animal Management. Minnesota Wild Animal
Management, Inc.
https://www.minnesotawildanimalmanagement.com/bats-get-active-inwarm-weather/
National Park Service. (2018). Studying Bats—Bats (U.S. National Park Service).
https://www.nps.gov/subjects/bats/studying-bats.htm
National Science Foundation. (2012, October 31). The Night Life: Why We Need
Bats All the Time--Not Just on Halloween.
https://www.nsf.gov/discoveries/disc_summ.jsp?cntn_id=125883
NHBS. (2019, June 26). Natural History Book Service: Harp Trap. Natural History
Book Service. https://www.nhbs.com/blog/the-nhbs-harp-trap
NOAA. (2016, March 9). What’s the difference between climate and weather? |
National Oceanic and Atmospheric Administration. National Oceanic and
Atmospheric Administration. https://www.noaa.gov/explainers/what-sdifference-between-climate-and-weather
Orme, J. (2020, June 3). Bat Surveys and Solutions: The Ultimate Guide.
Homebuilding & Renovating. https://www.homebuilding.co.uk/advice/batsurveys
64
O’Shea, T. J., & Brogan, M. A. (2003). Monitoring trends in bat populations of the
United States and territories: Problems and prospects (p. 287). U.S.
Geological Survey.
Pettit, J. L., & O’Keefe, J. M. (2017). Day of year, temperature, wind, and
precipitation predict timing of bat migration. Journal of Mammalogy, 98(5),
1236–1248. https://doi.org/10.1093/jmammal/gyx054
Predator Free New Zealand Trust. (n.d.). How to monitor native bats. Predator
Free NZ Trust. Retrieved December 6, 2021, from
https://predatorfreenz.org/toolkits/is-your-predator-control-working/how-tomonitor-native-bats/
Ramel, G. (2020, April 2). Bat Anatomy 101: The Various Bones of The Wing &
Skeleton. Earth Life. https://www.earthlife.net/mammals/bat-anatomy.html
Rehm, J. (2018, November 28). Why some bats hunt during the day. National
Geographic. https://www.nationalgeographic.com/animals/article/daytimebats-help-explain-nocturnal-evolution
Reusch, C., Gampe, J., Scheuerlein, A., Meier, F., Grosche, L., & Kerth, G.
(2019). Differences in seasonal survival suggest species‐specific reactions
to climate change in two sympatric bat species. Ecology and Evolution,
9(14), 7957–7965. https://doi.org/10.1002/ece3.5292
Sedgeley, J. (2012). Bats: Exit counts at bat roosts-simple visual counts [Bat
Inventory]. Department of Conservation Te Papa Atawhai.
The Weather Company. (2021). SeaTac, WA Weather History | Weather
Underground.
65
https://www.wunderground.com/history/monthly/us/wa/seatac/KSEA/date/
2021-5
Thompson, B. (2021, October 28). Helpful, not haunting: Why bats are vital to our
ecosystem. Wcnc.Com. https://www.wcnc.com/article/news/local/connectthe-dots/bats-vital-to-ecosystem-how-they-help-scientists-mademedicine/275-44b344a6-18f7-414e-a577-031cc0e878b4
United States Geologic Survey. (2016). North American Bat Monitoring Program
(NABat) | U.S. Geological Survey. USGS.
https://www.usgs.gov/centers/fort-collins-science-center/science/northamerican-bat-monitoring-program-nabat
University of Washington. (2013). Climate Change. Climate Impacts Group.
https://cig.uw.edu/learn/climate-change/
University of Washington. (2020, December). How is the Pacific Northwest
Climate Projected to Change? Climate Impacts Group.
https://cig.uw.edu/wpcontent/uploads/sites/2/2020/12/snoveretalsok2013sec5.pdf
US Department of the Interior. (2017, October 24). 13 Awesome Facts About
Bats. U.S. Department of the Interior. https://www.doi.gov/blog/13-factsabout-bats
U.S. Fish and Wildlife. (2020). What Happens to Bats in the Pacific Northwest
Winter? USFWS - Medium. https://usfws.medium.com/what-happens-tobats-in-the-pacific-northwest-winter-5c47b4ed2827
66
U.S. National Park Service. (2020, June 12). Hibernate or Migrate—Bats.
National Park Service. https://www.nps.gov/subjects/bats/hibernate-ormigrate.htm
USGS. (n.d.). What do bats eat? | U.S. Geological Survey. United States
Geological Service. Retrieved February 2, 2022, from
https://www.usgs.gov/faqs/what-do-bats-eat
USGS. (2016). Bat Population Data Project | U.S. Geological Survey.
https://www.usgs.gov/centers/fort-collins-science-center/science/batpopulation-data-project
Voight, C. C., schneeberger, K., Voight-Heucke, S. L., & Lewanzik, D. (2011,
May). Rain increases the energy cost of bat flight | Biology Letters.
Https://Royalsocietypublishing.Org/Doi/10.1098/Rsbl.2011.0313.
https://royalsocietypublishing.org/doi/10.1098/rsbl.2011.0313
Washington Department of Fish and Wildlife. (2018). Bat Colony Emergence
Count Protocol 2018 Washington Department of Fish and Wildlife.
WDFW. (2020, November 12). What Happens to Bats in the Pacific Northwest
Winter? Medium. https://usfws.medium.com/what-happens-to-bats-in-thepacific-northwest-winter-5c47b4ed2827
WDFW. (2021a). State Listed Candidate Species. Washington State Department
of Wildlife. https://wdfw.wa.gov/sites/default/files/202110/statelistedcandidatespecies_10132021.pdf
WDFW. (2021b). Westside prairie. Washington Department of Fish & Wildlife.
https://wdfw.wa.gov/species-habitats/ecosystems/westside-prairie
67
Weller, T. J., Castle, K. T., Liechti, F., Hein, C. D., Schirmacher, M. R., & Cryan,
P. M. (2016). First Direct Evidence of Long-distance Seasonal Movements
and Hibernation in a Migratory Bat. Scientific Reports, 6(1), 34585.
https://doi.org/10.1038/srep34585
Wickham, H. (2016). ggplot2: Elegant Graphics for Data Analysis. SpringerVerlag.
Wildlife Online. (n.d.). Bats—Hibernation & Torpor | Wildlife Online. Wildlife
Online. Retrieved February 12, 2022, from
https://www.wildlifeonline.me.uk/animals/article/bats-hibernation-torpor
WINDExchange. (n.d.). WINDExchange: Wind Energy in Washington. Retrieved
November 7, 2021, from https://windexchange.energy.gov/states/wa
Zhang, Z., & Mai, Yujiao. (2021). Webpower: Basic and Advanced Statistical
Power Analysis (0.6) [R]. https://CRAN.R-project.org/package=WebPower
68
