Using Maximum Entropy Species Distribution Modeling For ...
Transcript of Using Maximum Entropy Species Distribution Modeling For ...
Fort Hays State UniversityFHSU Scholars Repository
Master's Theses Graduate School
Spring 2017
Using Maximum Entropy Species DistributionModeling For Long-term Conservation PlanningOf Three Federally Listed Bats In North AmericaMitchell L. MeyerFort Hays State University, [email protected]
Follow this and additional works at: https://scholars.fhsu.edu/theses
Part of the Biology Commons
This Thesis is brought to you for free and open access by the Graduate School at FHSU Scholars Repository. It has been accepted for inclusion inMaster's Theses by an authorized administrator of FHSU Scholars Repository.
Recommended CitationMeyer, Mitchell L., "Using Maximum Entropy Species Distribution Modeling For Long-term Conservation Planning Of ThreeFederally Listed Bats In North America" (2017). Master's Theses. 10.https://scholars.fhsu.edu/theses/10
USING MAXIMUM ENTROPY SPECIES DISTRIBUTION MODELING FOR LONG-
TERM CONSERVATION PLANNING OF THREE FEDERALLY LISTED BATS IN
NORTH AMERICA
being
A Thesis Presented to the Graduate Faculty
of the Fort Hays State University
in Partial Fulfillment of the Requirements for
the Degree of Masters of Science
by
Mitchell L. Meyer
B.S. University of Central Missouri
Date_____________________ Approved________________________
Major Professor
Approved________________________
Chair, Graduate Council
i
This thesis for
The Master of Science Degree
By
Mitchell L. Meyer
Has Been Approved
___________________________________
Chair, Supervisory Committee
___________________________________
Supervisory Committee
___________________________________
Supervisory Committee
__________________________________
Supervisory Committee
_________________________________
Chair, Department of Biological Sciences
ii
PREFACE
This thesis is formatted in the style of Diversity and Distributions: A Journal of
Conservation Biogeography.
Keywords: biodiversity, bats, climate change, MaxEnt, species distribution modeling
iii
ABSTRACT
We are currently in a sixth mass extinction event in which the extinction rate is
higher than it has ever been. This mass extinction event is caused by human influence on
the environment. Biodiversity is worth conserving because of its many uses to humans.
Bats are a diverse group of mammals that humans rely on for pest control services. The
gray bat, northern long-eared bat, and Indiana bat are on the Threatened and Endangered
Species List and are in need of conservation.
I built species distribution models using occurrence records, climate data, and
Maximium entropy (MaxEnt) modeling technique. I predicted the historical range and
projected the range into 2050 and 2070 in best- and worst-case future climate scenarios.
The presence of each bat was most influenced by precipitation, which influences water
availability, prey abundance, mortality, and natality. Future projected ranges became
more fragmented, shifted north from the historical range, or both; and less of the
historical range remained in the future.
Fragmentation and shifting ranges due to changing climate could have a negative
effect on each bat. I recommend conserving forested corridors especially around cave
sites used by each bat species. Forested corridors will be important for dispersal when
the range shifts or to connect fragmented areas. The models produced in this study
provide a guide for conservation management efforts for each bat. Conservation efforts
should strive to maintain or increase bat populations because of the economic and
environmental benefits they provide.
iv
ACKNOWLEDGMENTS
I give an eternal thank you to my advisor, Dr. Rob Channell, for adopting me into
your research lab. Your guidance was needed in one of the most difficult times of my
life. I learned a lot from you, educationally and about myself. Thank you to the rest of
my graduate committee, Dr. Greg Farley, Dr. Mitchell Greer, and Dr. William Stark for
being a part of this project and my graduate career. Also, thank you to the rest of the
faculty in the Department of Biological Sciences for guidance, instruction, and
encouragement.
I thank Dr. Elmer Finck and Mr. Curtis Schmidt for providing me the opportunity
to work on the Myotis septentionalis project, on which I gained much experience and
knowledge. Thank you to the Kansas Department of Wildlife Parks and Tourism and the
Department of Biological Sciences for funding during my graduate career.
I thank those who helped collect and process my data, especially Allison
Hullinger, Dr. Yass Kobayashi, Geneva McKown, Ethan Oltean, Taylor Rasmussen,
Adam Rusk, and Elizabeth Schumann. Thank you to all the people I met and befriended;
especially Elizabeth Bainbridge, Kasandra Brown, Diedre Kramer, Brendon
McCampbell, Kaitlin Moore, Samantha Pounds, and Elizabeth Tharman; I am forever
grateful for everyone’s friendship and support. Angelica Sprague and Holly Wilson were
there most; thank you for being there for me during graduate school and every aspect of
my life. Finally, thank you to my family, especially Mom, Dad, Stacey and Lori. You all
realized my potential, even when I could not, and have always pushed me to succeed.
v
TABLE OF CONTENTS
GRADUATE COMMITTEE APPROVAL ......................................................................... i
PREFACE ........................................................................................................................... ii
ABSTRACT ....................................................................................................................... iii
ACKNOWLEDGMENTS ................................................................................................. iv
TABLE OF CONTENTS .....................................................................................................v
LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ......................................................................................................... viii
INTRODUCTION ...............................................................................................................1
Climate, Biodiversity, and Importance of Bats ........................................................1
Climate and Bats ......................................................................................................3
Gray Bat ...................................................................................................................3
Northern Long-eared Bat .........................................................................................4
Indiana Bat ...............................................................................................................5
Project Purpose and Objectives ...............................................................................5
METHODS ..........................................................................................................................7
Occurrence Records .................................................................................................7
Climate Data ............................................................................................................8
Maximum Entropy Species Distribution Modeling .................................................9
RESULTS ..........................................................................................................................12
Gray Bat Results ....................................................................................................12
Northern Long-eared Bat Results ..........................................................................13
vi
Indiana Bat Results ................................................................................................14
DISCUSSION ....................................................................................................................16
Bats and Drought ...................................................................................................16
Gray Bat .................................................................................................................17
Northern Long-Eared Bat ......................................................................................18
Indiana Bat .............................................................................................................18
Conclusions ............................................................................................................19
REFERENCES ..................................................................................................................21
TABLES ............................................................................................................................33
FIGURES ...........................................................................................................................41
vii
LIST OF TABLES
1 Bioclim variables included in MaxEnt climate models (Hijmans et al., 2005). .........32
2 Permutation importance values for each bioclim variable for the MaxEnt model of
the gray bat Myotis grisescens. Higher values indicate a larger influence on the
model. .........................................................................................................................33
3 Permutation importance values for each bioclim variable for the MaxEnt model of
the northern long-eared bat (Myotis septentrionalis) Higher values indicate a larger
influence on the model. ..............................................................................................34
4 Permutation importance values for each bioclim variable for the MaxEnt model of
the Indiana bat (Myotis sodalis). Higher values indicate a larger influence on the
model. .........................................................................................................................35
5 Summary statistics for the MaxEnt models of the gray bat (Myotis grisescens),
northern long-eared bat (Myotis septentrionalis), and Indiana bat (Myotis sodalis). .36
6 Areas (km2) of the predicted historical ranges produced by the MaxEnt model of the
gray bat (Myotis grisescens), northern long-eared bat (Myotis septentrionalis), and
Indiana bat (Myotis sodalis). ......................................................................................37
7 Areas (km2) of the projected future ranges produced by the MaxEnt model of the
gray bat (Myotis grisescens), northern long-eared bat (Myotis septentrionalis), and
Indiana bat (Myotis sodalis). ......................................................................................38
8 Percent of the historical range that remained in each future climate scenario
produced by MaxEnt model of the gray bat (Myotis grisescens), northern long-eared
bat (Myotis septentrionalis), and Indiana bat (Myotis sodalis). .................................39
viii
LIST OF FIGURES
1 MaxEnt model of the gray bat (Myotis grisescens) under historical climate
conditions. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence. ...................................................................40
2 MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under
historical climate conditions. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence ...........................41
3 MaxEnt model of the Indiana bat (Myotis sodalis) under historical climate
conditions. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence. ...................................................................42
4 MaxEnt model of the gray bat (Myotis grisescens) under a best-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................43
5 MaxEnt model of the gray bat (Myotis grisescens) under a best-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................44
6 MaxEnt model of the gray bat (Myotis grisescens) under a worst-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................45
ix
7 MaxEnt model of the gray bat (Myotis grisescens) under a worst-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................46
8 MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a best-
case emission scenario for 2050. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence. ..........................47
9 MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a best-
case emission scenario for 2070. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence. ..........................48
10 MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a worst-
case emission scenario for 2050. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence. ..........................49
11 MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a worst-
case emission scenario for 2070. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence. ..........................50
12 MaxEnt model of the Indiana bat (Myotis sodalis) under a best-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................51
13 MaxEnt model of the Indiana bat (Myotis sodalis) under a best-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................52
x
14 MaxEnt model of the Indiana bat (Myotis sodalis) under a worst-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................53
15 MaxEnt model of the Indiana bat (Myotis sodalis) under a worst-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in
grey indicate a low probability of occurrence. ...........................................................54
1
INTRODUCTION
We are currently in a sixth mass extinction event in which the extinction rate is
higher than it has ever been (Barnosky et al., 2011). Traditional natural resource
management strategies were established assuming climate would remain stable (Heller
and Zavaleta, 2009); however, climatic conditions are changing due to human influence
(Huston and Marland, 2003; Thomas and Trenberth, 2003). Climate change trends
include increased mean global temperatures, prolonged and more frequent droughts, and
more frequent intense extreme precipitation events (IPCC, 2014). Climate change due to
human influence is a contributing factor to biodiversity loss (Huston and Marland, 2003)
and biodiversity will continue to decline with current climate change trends (Thomas et
al. 2004). Biodiversity will also be affected by changing species compositions through
range contractions (Hong-Wa and Arroyo, 2012), expansions, and shifts (Chen et al.,
2011; Lundy et al., 2010; Parmesan and Yohe, 2003). Fluctuating species’ ranges further
contributes to biodiversity loss by introducing new threats to ecosystems, e.g. invasive
species. (Samson and Knopf, 1993).
Biodiversity provides direct extractive uses (e.g. timber harvest), direct non-
extractive uses (e.g. recreational use), and indirect uses (e.g. pest control services;
Edwards and Abivardi, 1998). Bats benefit humans directly by providing crop security in
the form of pest control (Puig-Montserrat et al., 2015; Wagner et al., 2014; Cleveland et
al. 2006). Because climate change will continue to reduce crop yields (Challinor et al.,
2014; Lobell et al., 2011) and increase uncertainty already present in natural resource
2
strategies (Nichols et al., 2011), crop security will remain uncertain. Bats can help
reinforce crop security and because of this, they have a high economic value (Puig-
Montserrat et al., 2015; Wagner et al., 2014; Cleveland et al., 2006) estimated at $22.9
billion/year in the United States for pest control alone (Boyles et al., 2011). Without
bats, agricultural losses have been estimated at more than $3.7 billion/year (Boyles et al.,
2006). Pest control by bats also benefits the environment and public health by reducing
the need of pesticides (Cleveland et al., 2006). Because bats are taxonomically stable,
high in the food chain, and long-lived, they are effective bioindicators (Jones et al.,
2009), and therefore, useful in assessing ecosystem health and the effects of climate
change.
Bats are the second most diverse group of mammals with more than 1110 species
described (Simmons, 2005). Historically, large scale bat population declines have been
attributed to habitat destruction (Thomson, 1982). More recently, bat mortality has been
influenced by the presence of windfarms (Zimmerling and Francis, 2016), continued
habitat loss through urbanization (Russo and Ancillotto, 2015), and white-nose syndrome
(Blehert et al., 2009). White-nose syndrome is caused by a fungal pathogen
(Pseudogymnoascus destructans) that was introduced in 2006 to a cave in New York that
has been responsible for wide-spread declines in bat populations in North America
(Blehert et al., 2009). The distribution of white-nose syndrome is a result of its
transmission among colony roosting bats (Cryan et al., 2010). White-nose syndrome is
projected to spread further in North America (Ingersoll et al., 2016), resulting in
increased mortality among bats.
3
Climate influences bat activity; during increased precipitation and low
temperatures bats typically do not fly (Erikson and West, 2002). Climate change has also
been attributed to a range expansion in a European bat species (Lundy et al, 2010).
Precipitation has a strong influence on bat reproduction and survival (Adams, 2010),
natality is decreased during droughts because water and prey resources are limited
(Adams, 2010). Bat mortality is increased during droughts because of water resource
limitation (Adams and Hayes, 2008); droughts also decrease prey abundance (Jonsson et
al., 2015; Zhu et al., 2014), which could contribute to increased competition among bats.
It is probable that prolonged droughts due to climate change will have a negative effect
on bat populations. These droughts limit prey and water resources for extended times
that bat populations cannot survive.
Slow reproductive rates (Pontier et al., 1993) coupled their sensitivity to human
induced environmental stressors makes bats a prime target for conservation. Three
species of bats are in need of conservation in North America, which are listed on the
United States Threatened and Endangered Species List: the gray bat (Myotis grisescens),
the northern long-eared bat (M. septentrionalis), and the Indiana bat (M. sodalis). A
subspecies of the Townsend’s big eared bat (Corynorhinus townsendii ingens) is also
listed as endangered but is represented by few populations, not the entire species, and is
not covered in this project.
Gray Bat
The gray bat occurs in the central and southeastern portion of the United States
(Figure 1). The diet of the gray bat consists of insects and arachnids (Best et al., 1997).
4
It roosts in colonies, with aggregations of 300 to 150,000 individuals (Tuttle, 1976). The
gray bat uses caves year-round (Tuttle and Stevenson, 1977), but uses different caves
depending on the season (Hall and Wilson, 1966). Storm drains also serve as roost sites
for the gray bat (Hays and Bingman, 1964). Roosts are commonly within 1 km of major
water bodies and are rarely farther than 4 km (Tuttle, 1976). These bodies of water
provide a drinking source and foraging habitat (Tuttle, 1976). The gray bat flies under
the forest canopy (Tuttle, 1976).
On 28 April 1976, the gray bat was listed as a federally endangered species
(United States Fish and Wildlife Service, 1976). Population declines are attributed to
large scale human disturbance at roost caves (Tuttle, 1979) and pollution (Clark et al.,
1978). More recently, white-nose syndrome has infected gray bat populations (Powers et
al., 2016). Although this disease has had a detrimental effect on other bat species it does
not seem to have reduced gray bat populations (Powers et al., 2016).
Northern Long-eared Bat
The northern long-eared bat occurs in most of the central and northeastern
portions of the United States, extending into the southeastern portion of Canada (Figure
2). Northern long-eared bats require forested areas for foraging (Ford et al., 2005). Their
diet consists of insects and arachnids (Feldhammer et al., 2009). The northern long-eared
bat roosts solitarily or in colonies (Foster and Kurta, 1999) that typically do not exceed
70 individuals (Foster and Kurta, 1999; Menzel et al., 2002). In winter, the northern
long-eared bat roosts in caves (Whitaker and Rissler, 1992) and mines (Whitaker, 1992).
In the summer, this species roosts in snags and live trees in forested areas (Lacki and
5
Schwierjohann, 2001; Timpone et al., 2010). On 2 April 2015, the northern long-eared
bat was listed as a federally threatened species (United States Fish and Wildlife Service,
2015). This federal listing was due to population declines associated with the onset of
white-nose syndrome (Reynolds et al. 2016).
Indiana Bat
The Indiana bat occurs in the east-central and northeastern portions of the United
States and extends into the southeastern portion of Canada (Figure 3). The diet of the
Indiana bat consists of insects (Feldhammer et al., 2009). Foraging occurs under the
forest canopy (Humphrey et al., 1977; LaVal et al., 1977) and is not restricted to areas
associated with water (LaVal et al., 1977). The Indiana bat roosts in large colonies up to
100,000 individuals (United States Fish and Wildlife Service, 1999). The Indiana bat is
migratory, using cave roosts in winter and trees in forested areas the rest of the year
(Thomson, 1982). On 11 March 1967, The Indiana bat was listed as a federally
endangered species (United States Fish and Wildlife Service, 1967). Historically,
population declines have been attributed to human disturbance (Thomson, 1982); recent
population declines have been caused by white-nose syndrome (Thogmartin et al., 2012).
Project Purpose and Objectives
The purpose of this project was to determine the historical and projected future
distributions in response to climate for two federally endangered (gray bat and Indiana
bat) and one federally threatened (northern long-eared bat) species in North America.
The objectives of this project were to develop species distribution models using climate
variables, identify which variable had most influenced the distribution of the species, and
6
make recommendations for management. I hope to contribute to the long-term
management goals regarding these species.
7
METHODS
I built species distribution models (SDM) for three Myotis bat species in North
America: the gray bat (M. grisescens) the northern long-eared bat (M. septentrionalis),
and Indiana bat (M. sodalis) by using historical occurrence records and bioclimatic
(bioclim) variables. I predicted the historical range and projected the ranges into the
years 2050 and 2070 under a best and worst-case emission scenario for climate change. I
also quantified the percent of the historical range that remained in each future projected
range for each future climate scenario.
Occurrence Records
I obtained occurrence records from Global Biodiversity Information Facility
(GBIF, 2012) (received from www.gbif.org in September 2016) and VertNet (received
from www.vertnet.org in September 2016). I included records that had latitude and
longitude coordinates and those with adequate locality information to georeference using
GEOLocate Web Application (Rios and Bart, 2010). Additional occurrence records for
M. septentrionalis were obtained from an IACUC-approved field study (protocol number
15-0002) in 2015 and 2016 by Fort Hays State University. Duplicate records, records
with the same latitude and longitude coordinates, were removed. I vetted occurrences by
reviewing spatial relationships in ArcGIS version 10.3.1 and removed outlying
occurrences, i.e. any records that occurred in the oceans or great lakes.
8
Climate Data
I retrieved 19 bioclim variables (Table 1) for historical, best-case, and worst-case
future climate scenarios for 2050 and 2070 from WorldClim (Hijmans et al., 2005).
Bioclim variables are derived from global temperature and precipitation data (Hijmans et
al., 2005). Historical bioclim variables are averaged from 1960 to 1990 (Hijmans et al.,
2005). Future bioclim variables for 2050 are averaged from the projected values for 2041
to 2060, while 2070 are averaged from the projected values for 2061 to 2080 (Hijmans et
al., 2005). Future bioclim variables were derived from projected greenhouse gas
emissions.
Representative concentration pathways (RCP) differ in amounts of greenhouse
gas emissions over time (Hijmans et al., 2005). I used best- and worst-case future climate
scenarios represented by RCPs of 2.6 W/m2 and 8.5 W/m2, respectively. The RCP of 2.6
W/m2 is characterized by low greenhouse gas emissions over time that are expected to
rise and slightly decline (van Vuuren et al., 2011). The RCP of 8.5 W/m2 is characterized
by high greenhouse gas emissions over time caused by high human population growth
(van Vuuren et al., 2011). In the worst case climate scenario, greenhouse gas emissions
experience a sharp increase over time and do not decline (van Vuuren et al., 2011).
The climate data was a raster: geographically projected grid cells that were
homogenous for a physical measurement. The spatial resolution was 2.5 arcminutes
(approximately 5 km2). I downloaded the future climate scenarios from the Beijing
Climate Center Climate System Model (BCC-CSM1) because it includes both best- and
9
worst-case climate scenarios, and its performance is similar to other climate models
(Tongwen et al., 2014).
Maximum Entropy Species Distribution Modeling
I used climate envelopes, the associations of an organisms current geographic
distribution with current climate (Thomas et al., 2004), to predict species ranges
(Hijmans and Graham, 2006). Species distribution models predict occurrence to un-
sampled sites or into past or future climate scenarios (Elith and Leathwick, 2009).
Maximim entropy (MaxEnt) is an algorithmic technique (Phillips et al., 2006; Phillips et
al., 2004) used for species distribution modeling with many other applications, such as
studying climate change effects, invasive species management, and mapping disease
spread and risk (Miller, 2010; Phillips and Dudík, 2007) and has high predictive accuracy
(Elith et al., 2006). MaxEnt uses presence-only modelling (Phillips et al., 2006), this is
advantageous because absence data are rarely available and when available could
represent the organism was present but undetected (Graham et al., 2004).
Occurrence records are often biased because sampling efforts are most likely to be
in areas with easier access, e.g. closer to roads (Phillips et al., 2009). Sampling effort
bias causes inaccurate predictions (Kadmon et al., 2004). MaxEnt uses background
points with similar bias as the occurrence data to increase model performance (Phillips et
al., 2009). Many bioclim variables are highly correlated; MaxEnt is an appropriate
method because it is not sensitive to multicollinearity (Evangelista et al., 2011). I used
MaxEnt version 3.3.3k to build the species distribution models.
10
I cross-validated each model by randomly dividing the occurrence data into two
parts: training data (90%) and test data (10%). I used the training data to build the model
and the test data to test the model (Hijmans, 2012). Because the species is known to
occur at the locations in the test data, I could compare the models’ predictions (derived
from the training data) to the observed occurrences from the test data. If the test data
occurrences were in places predicted by the model, the model had high performance; if
the test data occurrences were in areas the model predicted them not to be, the model had
low performance. To evaluate model performance, I used the area under the curve
(AUC) value of the receiver operating characteristic (ROC) curve produced from the
MaxEnt program. AUC values range from 0 to 1 (Phillips et al., 2006): values >0.9 are
considered excellent; 0.7-0.9 are considered good; and <0.7 are considered uninformative
(Swets, 1988).
Each model predicted to new environmental conditions that were not encountered
during model training. Values become inflated when they are predicted outside of what
an organism has been observed (Phillips et al., 2006). To eliminate the inflation of
predicted values, I used a clamping procedure to set an upper and lower limit of suitable
conditions during model training (Phillips et al., 2006).
To assess the importance of each bioclim variable in the MaxEnt model, I used a
jackknife procedure and the permutation importance values produced by the MaxEnt
program. The jackknife procedure omits each variable, constructs a model without that
variable, then constructs a model using only the omitted variable (Baldwin, 2009). The
models created in the jackknife procedure are then compared to the model that includes
11
every variable. MaxEnt returns permutation importance values, for each variable, that
estimate the importance of each variable on the final model (Songer et al., 2012).
MaxEnt returns a geographical representation of the species ecological niche
made of grid cells based on input variables. Each grid cell has a probability of
occurrence for the species. For each distribution, I used a threshold to establish the
minimum probability that indicated presence for each grid cell (Urbani et al., 2015).
Using a fixed cumulative value 10 logistic threshold, I converted the grid cells to binary
values that represented presence (1) or absence (0) (Urbani et al., 2015). Using the fixed
cumulative value 10 logistic threshold assumes 10% of the occurrence records were
misidentified or incorrectly georeferenced (Raes et al., 2009). This fixed cumulative
value is frequently used in species distribution modeling (Bosso et al., 2013); it is
conservative and recommended for studies with large datasets in which data are collected
without standardized methods over long time spans (Rebelo and Jones, 2010). Errors
associated with GPS devices, the GEOLocate web application, or incorrectly collected
data should not have a significant negative effect on the model. Because the climate data
had high spatial autocorrelation, climate did not vary drastically across large geographic
space at this grain and would not contribute to error in the MaxEnt model.
12
RESULTS
Gray Bat
I obtained 330 unique occurrence records for the gray bat. The MaxEnt model
performance was high (AUC of 0.915), indicating this model can be used for prediction.
The most influential variable for the gray bat model was precipitation of driest month.
This variable had the highest permutation importance value (26.5). In the jackknife
procedure, the model performed worse without this variable than it did without any other
variable.
The historical range predicted by the MaxEnt model for the gray bat (Figure 1)
had an area of 823,020 km2, and was located in central and southeastern portion of the
United States. In each future climate scenario, there was a range expansion predicted to
the western portion of the United States from Mexico, through Canada and into Alaska.
In a best-case climate scenario in 2050 (Figure 4) and 2070 (Figure 5) the areas were
3213100 km2 and 3097150 km2, respectively. The eastern portion of the range did not
become more fragmented but it shifted north relative to the historical range. The western
portion of the range was larger and less fragmented than the eastern portion in 2050 and
2070. In a worst-case climate scenario for 2050 (Figure 6) and 2070 (Figure 7) the areas
were 2631130 km2 and 2498100 km2, respectively. The eastern portion of the range
became highly fragmented and shifted north relative to the historical range. The western
portion of the range remained larger and less fragmented than the eastern portion of the
range.
13
When projected into the future, the historical range decreased with each future
climate scenario. In a best-case climate scenario for 2050, 42% of the historical range
remained, while for 2070, 48% of the historical range remained. In a worst-case
emission scenario for 2050, 23% of the historical range remained; while for 2070, 17% of
the historical range remained.
Northern Long-eared Bat
I obtained 500 unique occurrence records for the northern long-eared bat. The
MaxEnt model had moderately high performance (AUC of 0.886), indicating this model
can be used for prediction. The most influential variable for the northern long-eared bat
model on the model was precipitation of warmest quarter. This variable had the highest
permutation importance value (20.6) and in the jackknife procedure, this variable
performed well when it was the only variable included.
The historical range produced by the MaxEnt model for the northern long-eared
bat (Figure 2) had an area of 2414430 km2. The historical range was located in the
central and northeastern portion of the United States, into the southeastern portion of
Canada, with small fragments in western Canada and Alaska. In a best-case emission
scenario for 2050 (Figure 8) and 2070 (Figure 9) range areas were 2110910 km2 and
2517100 km2, respectively. For both years, the range shifted north and became
increasingly fragmented relative to the historical range while large areas became
climatically suitable in the western part of North America. In a worst-case emission
scenario for 2050 (Figure 10) and 2070 (Figure 11) the areas were 2415000 km2 and
14
2384780 km2, respectively. The ranges became more fragmented, shifted north, and
larger areas existed in the western part of North America relative to the historical range.
When projected into the future, the historical range decreased with each future
climate scenario. In a best-case climate scenario for 2050, 67% of the historical range
remained; while for 2070, 76% of the historical range remained. In a worst-case climate
scenario for 2050, 57% of the historical range remained; while for 2070, 40% of the
historical range remained.
Indiana Bat
I obtained 383 unique historical occurrence records for the Indiana bat. The
MaxEnt model had moderately high performance (AUC of 0.845), indicating this model
can be used for prediction. The most influential variable for the Indiana bat model was
precipitation seasonality. This variable had the highest permutation importance value
(29).
The historical range produced by the MaxEnt model for the Indiana bat (Figure 3)
had an area of 1562840 km2. The historical range was located in the central and
northeastern portion of the United States into the southeastern portion of Canada. In each
future climate scenario, there is range expansion predicted in the western portion of the
North America, from Mexico through Canada and into Alaska. This range expansion
consists of small fragmented areas. In a best-case climate scenario for 2050 (Figure 12)
and 2070 (Figure 13) the areas were 1888430 km2 and 2258400 km2, respectively. In
2050 and 2070, the areas in the west did not differ drastically. The range in the eastern
part of the United States was projected to become highly fragmented and shifted north for
15
2050 and 2070, relative to the historical range. In a worst-case climate scenario for 2050
(Figure 14) and 2070 (Figure 15) the areas for each range were 2263230 km2 and
2352950 km2, respectively; both ranges shifted north and became more fragmented. The
western portion of the range increased in size in the western United States from 2050 to
2070.
When projected into the future, the historical range decreased with each future
climate scenario. In a best-case climate scenario for 2050, 54% of the historical range
remained; while for 2070, 76% of the historical range remained. In a worst-case climate
scenario for 2050, 23% of the historical range remained; for 2070, 17% of the historical
range remained.
16
DISCUSSION
Bats and Drought
Precipitation had the most influence on predicting presence of each bat species.
During times of decreased precipitation, bats have low survival and low reproductive
output (Amorim et al., 2015; O’Shea et al., 2010). Prolonged droughts due to climate
change might limit resources important for survival. Insect abundance decreases with
decreases in precipitation (Zhu et al., 2014; Janzen and Schoener, 1968); prolonged
droughts could influence species occurrence of bats because it negatively impacts prey
abundance. Decreased precipitation could also limit water sources for drinking by bats.
Limited water sources might increase mortality because, during lactation, bats need to
increase water intake for milk production (Adams and Hayes, 2008).
Precipitation of driest month was the most influential variable for determining
presence of the gray bat. Because gray bats are restricted in distribution by roost distance
to major water bodies (Tuttle, 1976), prolonged droughts due to climate change might
reduce the presence of available water and further limit where the gray bat can occur.
Precipitation of warmest quarter was the most influential variable for determining
presence of the northern long-eared bat. Prolonged droughts from climate change might
negatively affect gestation, parturition, and lactation as these stages of reproduction occur
in warmer months (Caceres and Barclay, 2000). Precipitation seasonality had the most
influence for the presence of the Indiana bat. Prolonged droughts might restrict drinking
sources important during winter when Indiana bats arouse to drink (Boyles et al., 2006).
17
Gray Bat
Each projected range of the gray bat was larger than the historical range predicted
by the MaxEnt model. This suggests that there could be a larger area for the gray bat to
inhabit in the future. However, in the models for each future climate scenario there was a
decrease in the total area from 2050 to 2070. Also, less of the historical range remains in
the future climate scenarios. The gray bat might be negatively affected by climate
change because of the reduction in size of the historical range over time, increase in
fragmentation, and shifts in the climatic envelope.
The gray bat will need to disperse to new areas for continued survival. As
suitable areas shift north, forested corridors will be necessary to allow for dispersal,
because the gray bat flies under the forest canopy (Tuttle, 1976). Suitable habitat
corridors would be necessary because energetic demands of long distance flight increase
the chance of mortality in the gray bat (Tuttle and Stevenson, 1977). Management should
focus on creating and maintaining forested corridors that make connections among caves
and large water bodies, with an emphasis on conserving already existing corridors used
by the gray bat. It is unlikely that the gray bat will disperse to the projected range in the
western United States in each future climate scenario, unless habitat corridors can be
maintained that lead from east to west that include suitable habitat. It is also unlikely that
the forests in the western range expansions are suitable because they are coniferous and
the gray bat inhabits deciduous forests in the eastern United States. However, the
expanded range in the east could be used as introduction habitat.
18
Northern Long-eared Bat
The areas of the projected ranges by the MaxEnt model for the northern long-
eared bat were similar to the predicted historical range. Every projected range for the
northern long-eared bat shifts north and is more fragmented than the historical range,
with more areas projected in the northwest region of North America. This shifting and
fragmented range indicates the northern long-eared bat might be negatively affected by
climate change.
Corridors will be necessary for the northern long-eared bat to move among
suitable areas because future projected ranges were highly fragmented. Northern long-
eared bats use forested corridors in fragmented landscapes (Yates and Muzika, 2006). I
recommend the management of forested corridors from the southern portion to the
northern portion of the projected range to coincide with a north shifting range over time.
I also recommend conserving forested areas currently occupied by the northern long-
eared bat.
Indiana Bat
Each projected range became more fragmented than the historical range.
However, each projected range was larger than the historical range and over half of the
historical range remained in each future climate scenario. This could mean a large for the
Indiana bat to inhabit in the future. Despite the large projected ranges, some populations
will be impacted by the fragmentation and shifting of the projected range and, thus, the
Indiana bat might be negatively impacted by climate change.
19
The expanded range in the western United States projected in each future-climate
scenario was highly fragmented and disjunct from the eastern portion. Dispersal to the
expanded western range is unlikely because of the disjunction of the eastern and western
ranges. It is also unlikely that the forests in the western range expansions are suitable
because they are coniferous and the Indiana bat inhabits deciduous forests in the eastern
United States. Because the range is projected to shift north, I recommend the
management of suitable habitat in the northern portion of the Indiana Bat range. Suitable
forested habitat should be conserved to provide for corridors to allow for movement and
foraging sites. Forested areas associated with cave sites should be conserved to provide
appropriate habitat for the Indiana bat.
Conclusions
Researchers have claimed that changing climate influences bat behavior (Erikson
and West, 2002), natality (Adams, 2010), survival (Frick et al., 2010) and distribution
(Lundy et al., 2010). Overall, for each species covered in this project, the future
projected ranges become more fragmented and shift north from the historical ranges.
Each species must disperse to new areas for continued survival. Understanding where
suitable climactic conditions occur currently and in the future, will help to identify places
suitable for habitat management. Because of climatological influence on bats
populations, MaxEnt SDM’s will provide useful in management for each of the species
modeled in this study. The models produced in this study provide a guide for
conservation management efforts for each bat. Conservation efforts should strive to
20
maintain or increase bat populations because of the economic and environmental benefits
they provide.
21
REFERENCES
Adams, R.A. 2010. Bat reproduction declines when conditions mimic climate change
projections for western North America. Ecology, 91, 2437-2445.
Adams, R.A. & Hayes, M.A. 2008. Water availability and successful lactation by bats as
related to climate change in arid regions of western North America. Journal of
Animal Ecology. 77, 1115-1121.
Amorim, F., Mata, V.A., Beja, P., Rebelo, H. 2015. Effects of a drought episode on the
reproductive success of European free-tailed bats (Tadarida teniotis).
Mammalian Biology, 80, 228-236.
Baldwin, R.A. 2009. Use of maximum entropy modeling in wildlife research. Entropy,
11, 854-866.
Barnosky, A.D., Matzke, N., Tomiya, S., Wogan, G.O.U., Swartz, B., Quental, T.B.,
Marshall, C., McGuire, J.L., Lindsey, E.L., Maguire, K.C., Mersey, B., & Ferrer,
E.A. 2011. Has the Earth’s sixth mass extinction already arrived? Nature, 471,
51-57.
Best, T.L., Milam, B.A., Haas, T.D., Cvilikas, W.S., & Saidak, L.R. 1997. Variation in
diet of the Gray Bat (Myotis grisescens). Journal of Mammalogy, 78, 569-583.
Blehert, D.S., Hicks, A.C., Behr, M., Meteyer, C.U., Berlowski-Zier, B.M., Buckles,
E.L., Coleman, J.T.H., Darling, S.R., Gargas, A., Niver, R., Okoniewski, J.C.,
Rudd, R.J., Stone, W.B. 2009. Bat white-nose syndrome: an emerging fungal
pathogen? Science, 323, 227.
22
Bosso, L., Rebelo, H., Garonna, A.P., and Russo, D. 2013. Modelling geographic
distribution and detecting conservation gaps in Italy for the threatened beetle
Rosila alpina. Journal for Nature Conservation, 21,72-80.
Boyles, J.G., Cryan, P.M., McCracken, G.F., & Kunz, T.H. 2011. Economic importance
of bats in agriculture. Science, 332, 41-42.
Boyles, J.G., Dunbar, M.B., & Whitaker, J.O., Jr. 2006. Acitivty following arousal in
winter in North American vespertilionid bats. Mammal Review, 36, 267-280.
Caceres, M.C., & Barclay, R.M.R. 2000. Myotis septentrionalis. Mammalian Species,
634, 1-4.
Challinor, A.J., Watson, J., Lobell, D.B., Howden, S.M., Smith, D.R., & Chhetri, N.
2014. A meta-analysis of crop yield under climate change and adaptaion. Nature
Climate Change, 4, 287-291.
Chen, I.C., Hill, J.K., Ohlemüller, R., Roy, D.B., & Thomas, C.D. 2011. Rapid ranges
shifts of species associated with high levels of climate warming. Science, 333,
1024-1026.
Clark, D.R., Jr., LaVal, R.K., & Swineford, D.M. 1978. Dieldrin-induced mortality in an
endangered species, the Gray Bat (Myotis grisescens). Science, 199, 1357-1359.
Cleveland, C.J., Betke, M., Federico, P., Frank, J.D., Hallam, T.G., Horn, J., López, J.D.
Jr., McCracken, G.F., Medellín R.A., Moreno-Valdez, A., Sansone, C.G.,
Westbrook, J.K., & Kunz, T.H. 2006. Economic value of the pest control service
provided by Brazilian free-tailed bats in south-central Texas. Frontiers in
Ecology and the Environment, 4, 238-243.
23
Cryan, P.M., Meteyer, C.U., Boyles, J.G., & Blehert, D.S. 2010. Wing pathology of
white-nose syndrome in bats suggests life-threatening disruption of physiology.
BMC Biology, 8, 135.
Edwards, P.J. and Abivardi, C. 1998. The value of biodiversity: where ecology and
economy blend. Biological Conservation, 83, 239-246.
Elith, J., Graham, C.H., Anderson, R.P., Dudík, M., Ferrier, S., Guisan, A., Hijmans, R.J.,
Huettmann, F., Leathwick, J.R., Lehmann, A., Li, J., Lohmann, L.G., Loiselle,
B.A., Manion, G., Moritz, C., Nakamura, M., Nakazawa, Y., Overton, J.,
Peterson, A.T., Phillips, S.J., Richardson, K., Scachetti-Pereira, R., Schapire,
R.E., Soberón, J., Williams, S., Wisz, M.S., & Zimmermann, N.E. 2006. Novel
methods improve prediction of species’ distributions from occurrence data.
Ecography, 29, 129-151.
Elith, J. and Leathwick, J.R. 2009. Species distribution models: ecological explanation
and prediction across space and time. Annual Review of Ecology Evolution and
Systmatics, 40, 677-697.
Erikson, J.L. & West, S.D. 2002. The influence of regional climate and nightly weather
conditions on activity patterns of insectivorous bats. Acta Chiropterologica, 4,
17-24.
Evangelista, P.H., Kumar, S., Stohlgren, T.J., & Young, N.E. 2011. Assessing forest
vulnerability and the potential distribution of pine beetles under current and future
climate scenarios in the Interior West of the US. Forest Ecology and
Management, 262, 307-316.
24
Feldhammer, G.A., Carter, T.C., & Whitaker, J.O., Jr. 2009. Prey consumed by eight
species of insectivorous bats from southern Illinois. American Midland
Naturalist, 162, 43-51.
Ford, W.M., Menzel, M.A., Rodrigue, J.L., Menzel, J.M., & Joshua, J.B. 2005. Relating
bat species presence to simple habitat measures in a central Appalachian forest.
Biological Conservation, 126, 528-539.
Foster, R.W. & Kurta, A. 1999. Roosting ecology of the Northern Bat (Myotis
septentrionalis) and comparisons with the endangered Indiana Bat (Myotis
sodalis). Journal of Mammalogy, 80, 659-672.
Frick, W.F., Reynolds, D.S., & Kunz, T.H. 2010. Influence of climate and reproductive
timing on demography of little brown myotis Myotis lucifugus. Journal of Animal
Ecology, 79, 128-136.
GBIF. 2012. Recommended practices for citation of the data published through the
GBIF Network. Version 1.0 (Authored by Vishwas Chavan), Copenhagen: Global
Biodiversity Information Facility. Pp.12, ISBN: 87-92020-36-4. Accessible
at http://links.gbif.org/gbif_best_practice_data_citation_en_v1.
Graham, C.H., Ferrier, S., Huettman, F., Moritz, C., & Peterson, A.T. 2004. New
developments in museum-based informatics and applications in biodiversity
analysis. Trends in Ecology & Evolution, 19, 497-503.
Hall, J.S. & Wilson, N. 1966. Seasonal populations and movements of the gray bat in
the Kentucky area. The American Midland Naturalist, 75, 317-324.
25
Hays, H.A. & Bingman, D.C. 1964. A colony of gray bats in southeastern Kansas.
Journal of Mammalogy, 45: 150.
Heller, N.E. & Zavaleta, E.S. 2009. Biodiversity management in the face of climate
change: a review of 22 years of recommendations. Biological Conservation, 142,
14-32.
Hijmans, R.J. 2012. Cross-validation of species distribution models: removing spatial
sorting bias and calibration with a null model. Ecology, 93, 679-688.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., & Jarvis, A. 2005. Very high
resolution interpolated climate surfaces for global land areas. International
Journal of Climate, 25, 1965-1978.
Hijmans, R.J. & Graham, C.H. 2006. The ability of climate envelope models to predict
the effect of climate change on species distributions. Global Change Biology, 12,
2272-2281.
Hong-Wa, C. & Arroyo, T.P.F. 2012. Climate-induced rage contraction in the Malagasy
endemic plant Mediusella and Xerochlamys (Sarcolaenaceae). Plant Ecology and
Evolution, 145, 302-312.
Humphrey, S.R., Richter, A.R., & Cope, J.B. 1977. Summer habitat and ecology of the
endangered Indiana Bat, Myotis sodalis. Journal of Mammalogy, 58, 334-346.
Huston, M.A & Marland, G. 2003. Carbon management and biodiversity. Journal of
Environmental Management, 67, 77-86.
26
Ingersoll, T.E., Sewall, B.J., & Amelon, S.K. 2016. Effects of white-nose syndrome on
regional population patterns of 3 hibernating bat species. Conservation Biology,
30, 1048-1059.
IPCC. 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups
I, II, and III to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change [Core Wrtiting team, R.K. Pachauri and L.A. Meyer (eds.)].
IPCC, Geneva, Switzerland, 151 pp.
Janzen, D.H., & Schoener, T.W. 1968. Differences in insect abundance and diversity
between wetter and drier sites during a tropical dry season. Ecology, 49, 96-110.
Jones, G., Jacobs, D.S., Kunz, T.H., Willig, M.R., & Racey, P.A. 2009. Carpe noctem:
the importance of bats as bioindicators. Endangered Species Research, 8, 93-115.
Jonsson, M., Hedström, P., Stenroth, K., Hotchkiss, E.R., Vasconcelos, F.R., Karlsson, J.,
& Byström, P. 2015. Climate change modifies the size structure of assemblages
of emerging aquatic insects. Freshwater Biology, 60, 78-88.
Kadmon, R., Farber, O., & Danin, A. 2004. Effect of roadside bias on the accuracy of
predictive maps produced by bioclimatic models. Ecological Applications, 14,
401-413.
Lacki, M.J. & Schwierjohann, J.H. 2001. Day-roost characteristics of Northern Bats in
mixed mesophytic forest. The Journal of Wildlife Management, 65, 482-488.
LaVal, R.K., Clawson, R.L., LaVal, M.L., & Caire, W. 1977. Foraging behavior and
nocturnal activity patterns of Missouri bats, with emphasis on the endangered
27
species Myotis grisescens and Myotis sodalis. Journal of Mammalogy, 58, 592-
599.
Lobell, D.B., Schlenker, W., & Costa-Roberts, J. 2011. Climate trends and global crop
production since 1980. Science, 333, 616-620.
Lundy, M., Montgomery, I., & Russ, J. 2010. Climate change-linked range expansion of
Nathusius’ Pipistrelle Bat, Pipistrellus nathusii (Keyserling & Blasius, 1839).
Journal of Biogeography, 37, 2232-2242.
Menzel, M.A., Owen, S.F., Ford, W.M., Edwards, J.W., Wood, P.B., Chapman, B.R., &
Miller, K.V. 2002. Roost tree selection by northern long-eared bat (Myotis
septentrionalis) maternity colonies in an industrial forest of the central
Appalachian Mountains. Forest Ecology and Management, 155, 107-114.
Miller, J. 2010. Species distribution modeling. Geography Compass, 490-509.
Nichols, J.D., Koneff, M.D., Heglund, P.J., Knutson, M.G., Seamans, M.E., Lyons, J.E.,
Morton, J.M., Jones, M.T., Boomer, G.S., & Williams, B.K. 2011. Climate
change, uncertainty, and natural resource management. The Journal of Wildlife
Management, 75, 6-18.
O’Shea, T.J., Ellison, L.E., Neubaum, D.J., Neubaum, M.A., Reynolds, C.A., & Bowen,
R.A. 2010. Recruitment in a Colorado population of big brown bats: breeding
probabilities, litter size, and first year survival. Journal of Mammalogy, 91, 418-
428.
Parmesan, C. & Yohe, G. 2003. A globally coherent fingerprint of climate change
impacts across natural systems. Nature, 421, 37-42.
28
Phillips, S.J., Anderson, R.P., & Schapire, R.E. 2006. Maximum entropy modeling of
species geographic distributions. Ecological Modelling, 190, 231-259.
Phillips, S.J., Dudík, M., & Schapire, R.E. 2004. A maximum entropy approach to
species distribution modeling. Proceedings of the Twenty-First International
Conference on Machine Learning, 655-662.
Phillips, S.J., & Dudík, M. 2007. Modeling of species distributions with MaxEnt: new
extensions and a comprehensive evaluation. Ecography, 31, 161-175.
Phillips, S.J., Dudík, M., Elith, J., Graham, C.H., Lehmann, A., Leathwick, J., & Ferrier,
S. 2009. Sample selection bias and presence-only distribution models:
implications for background and pseudo-absence data. Ecological Applications,
19, 181-197.
Pontier, D., Gaillard, J.M., & Allainé, D. 1993. Maternal investment per offspring and
demographic tactics in placental mammals. Oikos, 66, 424-430.
Powers, K.E., Reynolds, R.J., Orndoff, W., Hyzy, B.A., Hobson, C.S., & Ford, W.M.
2016. Monitoring the status of Gray Bats (Myotis grisescens) in Virginia, 2009-
2014, and potential impacts of white-nose syndrome. Southeastern Naturalist, 15,
127-137.
Puig-Montserrat, X., Torre, I., López-Baucells, A., Guerrieri, E., Monti, M.M., Ràfols-
García, R., Ferrer, X., Gisbert, D., & Flaquer, C. 2015. Pest control service
provided by bats in Mediterranean rice paddies: linking agroecosystems structure
to ecological functions. Mammalian Biology, 80, 237-245.
29
Raes, N., Roos, M.C., Slik, J.W.F., van Loon, E.E., and ter Steege, H. 2009. Botanical
richness and endemicity patterns of Borneo derived from species distribution
models. Ecography, 32, 180-192.
Rebelo, H. & Jones, G. 2010. Ground validation of presence-only modelling with rare
species: a case study on barbastelles Barbastella barbastellus (Chiroptera:
Vespertionionidae). Journal of Applied Ecology, 47, 410-420.
Reynolds, R.J., Powers, K.E., Orndorff, W. Ford, W.M., & Hobson, C.S. 2016. Changes
in rates of capture and demographics of Myotis septentrionalis (Northern Long-
eared Bat) in western Virginia before and after onset of white-nose syndrome.
Northeastern Naturalist, 23, 195-204.
Rios, N.E. & Bart, H.L. 2010. GEOLocate (Version 3.22) [Computer Software]. Belle
Chasse, LA: Tulane University Museum of Natural History.
Russo, D. & Ancillotto, L. 2015. Sensitivity of bats to urbanization: a review.
Mammalian Biology, 80, 205-212.
Samson, F.B. & Knopf, F.L. 1993. Managing biological diversity. Wildlife Society
Bulletin, 21, 509-514.
Simmons, N.B. 2005. Order Chiroptera. Pp. 312-529, In: Mammal Species of the
World: A Taxonomic and Geographic Reference (D.E. Wilson and D.M. Reeder,
eds.). Smithsonain Institution Press, Washington, D.C.
Songer, M., Delion, M., Biggs, A. & Huang, Q. 2012. Modeling impacts of climate
change on giant panda habitat. International Journal of Ecology, 2012, 1-12.
30
Swets, J. 1988. Measuring the accuracy of diagnostic systems. Science, 240, 1285-
1293.
Thogmartin, W.E., King, R.A., McKann, P.C., Szymanski, J.A., & Pruitt, L. 2012.
Population-level impact of white-nose syndrome on the endangered Indiana Bat.
Journal of Mammalogy, 93, 1086-1098.
Thomas, C.D., Cameron, A., Green, R.E., Bakkenes, M., Beaumont, L.J., Collinham,
Y.C., Erasmus, B.F.N., de Siqueira, M.F., Grainger, A., Hannah, L., Hughes, L.,
Huntley, B., van Jaarsveld, A.S., Midgley, G.F., Miles, L., Ortega-Huerta, M.A.,
Peterson, A.T., Phillips, O.L., & Williams, S.E. 2004. Extinction risk from
climate change. Nature, 427, 145-148.
Thomas, T.R. & Trenberth, K.E. 2003. Modern global climate change. Science, 302,
1719-1723.
Thomson, C.E. 1982. Myotis sodalis. Mammalian Species, 163, 1-5.
Timpone, J.C., Boyles, J.G., Murray, K.L., Aubrey, D.P., & Robbins, L.W. 2010.
Overlap in roosting habits of Indiana Bats (Myotis sodalis) and Northern Bats
(Myotis septentrionalis). American Midland Naturalist, 163, 115-123.
Tongwen, W., Lianchun, S., Weiping, L., Zaizhi, W., Hua, Z., Xiaoge, X., Yanwu, Z., Li,
Z., Jianglong, L., Fanghua, W., Yiming, L., Fang, Z., Xueli, S., Min, C., Jie, Z.,
Yongjie, F., Fang, W., Yixiong, L., Xiangwen, L., Min, W., Qianxia, L., Wenyan,
Z., Min, D., Qigeng, Z., Jinjun, J., Li, L., & Mingyu, Z. 2014. An overview of
BCC climate system model development and application for climate change
studies. Journal of Meteorological Research, 28, 34-56.
31
Tuttle, M.D. 1976. Population ecology of the gray bat (Myotis grisescens): factors
influencing growth and survival of newly volant young. Ecology, 57, 587-595.
Tuttle, M.D. 1979. Status, causes of decline, and management of endangered gray bats.
The Journal of Wildlife Management, 43, 1-17.
Tuttle, M.D., & Stevenson D.E. 1977. An analysis of migration as a mortality factor in
the gray bat based on public recoveries of banded bats. The American Midland
Naturalist, 91, 235-240.
Urbani, F., D’Alessandro, P., Frasca, R., & Biondi, M. 2015. Maximum entropy
modeling of geographic distributions of the flea beetle species endemic in Italy
(Coleoptera: Galerucinae: Alticini). Zoologischer Anzeiger, 258, 99-109.
United States Fish and Wildlife Service. 1967. Endangered Species. Federal Register,
32, 4001.
United States Fish and Wildlife Service. 1976. Determination that two species of
butterflies are threatened species and two species of mammals are endangered
species. Federal Register, 41, 17736-17740.
United States Fish and Wildlife Service. 2015. Endangered and threatened wildlife and
plants; threatened species status for the Northern Long-eared Bat with 4(d) rule.
Federal Register, 80, 17974-18033.
United States Fish and Wildlife Service. 1999. Agency draft. Indiana bat (Myotis
sodalis) revised recovery plan. United States Fish and Wildlife Service, Fort
Snelling, Minnesota.
32
van Vuuren, D.P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K.,
Hurtt, G.C., Kram, T., Krey, V., Lamarque, J., Masui, T., Meinshausen, M.,
Nakicenovic, N., Smith, S.J., & Rose, S.K. 2011. The representative
concentration pathways: an overview. Climatic Change, 109, 5-31.
Wagner, T.C., Darras, K., Bumrungsri, S., Tscharntke, T., Klein, A. 2014. Bat pest
control contributes to food security in Thailand. Biological Conservation, 171,
220-223.
Whitaker, J.O. Jr. 1992. Winter activity of bats at a mine entrance in Vermillion County,
Indiana. The American Midland Naturalist, 127, 52-59.
Whitaker, J.O. Jr. & Rissler, L.J. 1992. Seasonal Activity of bats at Copperhead Cave.
Proceedings of the Indiana Academy of Science, 101, 127-135.
Yates, M.D. & Muzika, R.M. 2006. Effect of forest structure and fragmentation on site
occupancy of bat species in Missouri Ozark forests. The Journal of Wildlife
Management, 70, 1238-1248.
Zimmerling, R.J. & Francis, C.M. 2016. Bat mortality due to wind turbines. The
Journal of Wildlife Management, 80, 1360-1369.
Zhu, H., Wang, D., Wang, L., Fang, J., Sun, W., & Ren, B. 2014. Effects of altered
precipitation on insect community composition and structure in a meadow steppe.
Ecological Entomology, 39, 453-4
33
TABLES
Table 1. Bioclim variables included in MaxEnt climate models (Hijmans et al., 2005).
1 Annual Mean Temperature
2 Mean Diurnal Range (Mean of Montly (Max Temperature - Min Temperature))
3 Isothermality (Mean Diurnal Range/Temperature Annual Range) (* 100)
4 Temperature Seasonality (Standard Deviation * 100)
5 Max Temperature of the Warmest Month
6 Min Temperature of the Coldest Month
7 Temperature Annual Range (Max Temperature of Warmest Month - Min
Temperature of Coldest Month)
8 Mean Temperature of Wettest Quarter
9 Mean Temperature of Driest Quarter
10 Mean Temperature of Warmest Quarter
11 Mean Temperature of Coldest Quarter
12 Annual Precipitation
13 Precipitation of Wettest Month
14 Precipitation of Driest Month
15 Precipitation Seasonality (Coefficient of Variation)
16 Precipitation of Wettest Quarter
17 Precipitation of Driest Quarter
18 Precipitation of Warmest Quarter
19 Precipitation of Coldest Quarter
34
Table 2. Permutation importance values for each bioclim variable for the MaxEnt model
of the gray bat Myotis grisescens. Higher values indicate a larger influence on the model.
Variable Permutation Importance
Precipitation of Driest Month 26.5
Precipitation Seasonality (Coefficient of Variation) 16.7
Mean Diurnal Range (Mean of Monthly (Max Temperature
- Min Temperature))
10.7
Max Temperature of Warmest Month 7.2
Annual Precipitation 5.1
Mean Temperature of Driest Quarter 4.3
Annual Mean Temperature 3.9
Isothermality (Mean Dirunal Range/Temperature Annual
Range)(*100)
3.6
Mean Temperature of Coldest Quarter 3.4
Precipitation of Wettest Quarter 3
Precipitation of Warmest Quarter 3
Temperature Seasonality (Standard Deviation * 100) 2.5
Mean Temperature of Wettest Quarter 1.8
Mean Temperature of Warmest Quarter 1.6
Precipitation of Driest Quarter 1.6
Min Temperature of Coldest Month 1.4
Temperature Annual Range (Max Temperature of Warmest
Month - Min Temperature of Coldest Month)
1.4
Precipitation of Coldest Quarter 1.3
Precipitation of Wettest Month 1
35
Table 3. Permutation importance values for each bioclim variable for the MaxEnt model
of the northern long-eared bat (Myotis septentrionalis). Higher values indicate a larger
influence on the model.
Variable Permutation Importance
Precipitation of Warmest Quarter 20.6
Mean Temperature of Warmest Quarter 12
Precipitation of Wettest Quarter 9.8
Temperature Seasonality (Standard Deviation * 100) 9.8
Isothermality (Mean Dirunal Range/Temperature Annual
Range)(*100)
8.5
Annual Precipitation 8
Max Temperature of Warmest Month 7.3
Annual Mean Temperature 4.5
Mean Temperature of Wettest Quarter 3.8
Mean Diurnal Range (Mean of Monthly (Max Temperature
- Min Temperature))
3.6
Precipitation Seasonality (Coefficient of Variation) 3.1
Temperature Annual Range (Max Temperature of Warmest
Month - Min Temperature of Coldest Month)
2.3
Mean Temperature of Driest Quarter 2
Min Temperature of Coldest Month 1.7
Mean Temperature of Coldest Quarter 1.5
Precipitation of Wettest Month 0.7
Precipitation of Driest Month 0.4
Precipitation of Coldest Quarter 0.4
Precipitation of Driest Quarter 0
36
Table 4. Permutation importance values for each bioclim variable for the MaxEnt model
of the Indiana bat (Myotis sodalis). Higher values indicate a larger influence on the
model.
Variable Permutation Importance
Precipitation Seasonality (Coefficient of Variation) 29
Annual Precipitation 18.3
Precipitation of Driest Month 12.9
Temperature Annual Range (Max Temperature of Warmest
Month - Min Temperature of Coldest Month)
7.9
Annual Mean Temperature 5.3
Mean Temperature of Coldest Quarter 5.1
Temperature Seasonality (Standard Deviation * 100) 4.1
Mean Temperature of Wettest Quarter 3.8
Isothermality (Mean Dirunal Range/Temperature Annual
Range)(*100)
3.4
Precipitation of Driest Quarter 2.9
Precipitation of Warmest Quarter 1.7
Mean Diurnal Range (Mean of Monthly (Max Temperature
- Min Temperature))
1.7
Mean Temperature of Driest Quarter 1.7
Max Temperature of Warmest Month 1.1
Precipitation of Wettest Month 0.7
Precipitation of Wettest Quarter 0.2
Precipitation of Coldest Quarter 0.2
Min Temperature of Coldest Month 0.1
Mean Temperature of Warmest Quarter 0
37
Table 5. Summary statistics for the MaxEnt models of the gray bat (Myotis grisescens),
northern long-eared bat (Myotis septentrionalis), and Indiana bat (Myotis sodalis).
Species Scenario Year Test
AUC
StDev
AUC
Training
Points
Test
Points
Myotis
grisescens
Best-
Case
2050 0.915 0.0218 153 17
2070 0.915 0.0218 153 17
Worst-
Case
2050 0.915 0.0218 153 17
2070 0.915 0.0218 153 17
Myotis
septentrionalis
Best
Case
2050 0.886 0.0255 304 33
2070 0.886 0.0255 304 33
Worst-
Case
2050 0.886 0.0255 304 33
2070 0.886 0.0255 304 33
Myotis sodalis
Best-
Case
2050 0.845 0.0282 193 21
2070 0.845 0.0282 193 21
Worst-
Case
2050 0.845 0.0282 193 21
2070 0.845 0.0282 193 21
38
Table 6. Areas (km2) of the predicted historical ranges produced by the MaxEnt model of
the gray bat (Myotis grisescens), northern long-eared bat (Myotis septentrionalis), and
Indiana bat (Myotis sodalis).
Species Area of Historical Range (km2)
Myotis grisescens 823020
Myotis septentrionalis 2414430
Myotis sodalis 1562840
39
Table 7. Areas (km2) of the projected future ranges produced by the MaxEnt model of
the gray bat (Myotis grisescens), northern long-eared bat (Myotis septentrionalis), and
Indiana bat (Myotis sodalis).
Species Scenario Year Area (km2)
Myotis grisescens
Best-Case 2050 3213100
2070 3097150
Worst-Case 2050 2631130
2070 2498100
Myotis septentrionalis
Best Case 2050 2110910
2070 2517100
Worst-Case 2050 2415000
2070 2384780
Myotis sodalis
Best-Case 2050 1888430
2070 2263230
Worst-Case 2050 2258400
2070 2352950
40
Table 8. Percent of the historical range that remained in each future climate scenario
produced by MaxEnt model of the gray bat (Myotis grisescens), northern long-eared bat
(Myotis septentrionalis), and Indiana bat (Myotis sodalis).
Species Scenario Year Remnant historical range (%)
Myotis grisescens
Best-Case 2050 42
2070 48
Worst-Case 2050 23
2070 17
Myotis septentrionalis
Best Case 2050 67
2070 76
Worst-Case 2050 57
2070 40
Myotis sodalis
Best-Case 2050 54
2070 76
Worst-Case 2050 64
2070 55
41
FIGURES
Figure 1. MaxEnt model of the gray bat (Myotis grisescens) under historical climate
conditions. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
42
Figure 2. MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under
historical climate conditions. Areas in black indicate a high probability of occurrence,
areas in grey indicate a low probability of occurrence.
43
Figure 3. MaxEnt model of the Indiana bat (Myotis sodalis) under historical climate
conditions. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
44
Figure 4. MaxEnt model of the gray bat (Myotis grisescens) under a best-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
45
Figure 5. MaxEnt model of the gray bat (Myotis grisescens) under a best-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
46
Figure 6. MaxEnt model of the gray bat (Myotis grisescens) under a worst-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
47
Figure 7. MaxEnt model of the gray bat (Myotis grisescens) under a worst-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
48
Figure 8. MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a
best-case emission scenario for 2050. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence.
49
Figure 9. MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a
best-case emission scenario for 2070. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence.
50
Figure 10. MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a
worst-case emission scenario for 2050. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence.
51
Figure 11. MaxEnt model of the northern long-eared bat (Myotis septentrionalis) under a
worst-case emission scenario for 2070. Areas in black indicate a high probability of
occurrence, areas in grey indicate a low probability of occurrence.
52
Figure 12. MaxEnt model of the Indiana bat (Myotis sodalis) under a best-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
53
Figure 13. MaxEnt model of the Indiana bat (Myotis sodalis) under a best-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
54
Figure 14. MaxEnt model of the Indiana bat (Myotis sodalis) under a worst-case emission
scenario for 2050. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.
55
Figure 15. MaxEnt model of the Indiana bat (Myotis sodalis) under a worst-case emission
scenario for 2070. Areas in black indicate a high probability of occurrence, areas in grey
indicate a low probability of occurrence.