In Southeast Ohio a Thesis Presented T

Home , Eastern red bat

Summer Day-Roost Selection and Thermoregulation of Eastern Red Bats (Lasiurus

borealis) in Southeast Ohio

A thesis presented to

the faculty of

the College of Arts and Sciences of Ohio University

In partial fulfillment

of the requirements for the degree

Master of Science

Maria N. Monarchino

August 2019

This thesis titled

Summer Day-Roost Selection and Thermoregulation of Eastern Red Bats (Lasiurus

borealis) in Southeast Ohio

Maria N. Monarchino

has been approved for

the Department of Biological Sciences

and the College of Arts and Sciences by

Joseph S. Johnson

Assistant Professor of Biological Sciences

Florenz Plassmann

Dean, College of Arts and Sciences

ABSTRACT

MONARCHINO, MARIA N., M.S., August 2019, Biological Sciences

Summer Day-Roost Selection and Thermoregulation of Eastern Red Bats (Lasiurus borealis) in Southeast Ohio

Director of Thesis: Joseph S. Johnson

Accurate description of habitat features that species need to survive and reproduce is central to successful wildlife conservation. Unfortunately, the process of describing these habitats is complicated and selection processes for most species are often multifaceted. In addition, human altered landscapes are often composed of a mosaic of habitat conditions that further complicate understanding habitat selection.

However, such an understanding is necessary to conserve and protecting habitat for declining species. The eastern red bat (Lasiurus borealis) is a solitary, foliage roosting bat that is widely considered to be in decline across its range. Unfortunately, due to the expanse of the eastern red bats range, there are large areas where knowledge of important day-roosting habitats is lacking. To close this information gap in southeast

Ohio, we used a multifaceted approach to examine patterns of day-roost selection by male and female eastern red bats at two study sites with different forest compositions in southeastern Ohio. We used temperature sensitive radio-telemetry to track 28 male and 25 female bats to 53 male and 74 female day-roosts during the summers of 2016–

2019. We also placed 40 temperature dataloggers in roost and random tree canopies to monitor temperature in the foliage where bats roost. Finally, we collected skin temperature data from 9 males and 10 reproductive females to better understand how 4

bats responded to thermal conditions. On the tree-, stand-, and landscape-scale, we found that day-roost selection differed between sexes and study areas. In a more structurally diverse forest, males selected larger trees located at lower, cooler elevations than females. Females used smaller diameter trees at higher warmer elevations as compared to males. However, in a less structurally diverse forest, both sexes occupied similar sized trees and elevations. The importance of elevation was mirrored in data from microclimate dataloggers, which showed that trees at higher elevations had warmer canopies than trees at lower elevations. Finally, minimum and maximum ambient temperature, but not sex, predicted heterothermic responses in roosting bats. These results highlight that heterogeneity in environmental conditions can result in different patterns of habitat selection for this species, even between sites located in a small geographic area. It also provides evidence that this species sexually segregates at both a tree and landscape scale. These results are partially explained by our skin temperature data, which provides evidence that foliage roosting species face thermal stress from cool temperatures, even during the summer, influencing where they roost. These results have a number of conservation implications. First, they illustrate the importance of maintaining forest structural diversity across an elevational gradient to provide male and female eastern red bats with roosting habitat. Our data also show that warmer microclimates are important for successful reproduction. In southeastern Ohio, this includes upland forests, but it is essential to emphasize that it is the temperature at these slope positions, and not the elevation itself, that is important to bats and should be examined on a site-specific basis. 5

ACKNOWLEDGEMENTS This thesis is a result of a massive collaboration between Ohio University,

Ohio Department of Natural Resources, and Crane Hollow Nature Preserve whom all made this possible. Through these partnerships we were able to accomplish things that would never have been possible. I am eternally grateful for all those that contributed to the success of this project.

I would like to first and foremost thank my academic advisor, Dr. Joseph

Johnson, for bestowing an enormous amount of patience, expertise, and wisdom onto me and this project over the last 2 years. He has truly shaped me as a biologist and without his unwavering support I would have never made it through graduate school. I would also like to express so much gratitude for Jennifer Norris who started this project in 2015 and trusted me to finish it for her. I feel honored to have been able to take on this challenge. I would also like to thank my committee members, Dr. Nancy

Stevens and Dr. Viorel Popescu for the support and guidance through my master’s experience.

This project wouldn’t have been possible without the enormous amount of field help that I had through Ohio University, the Division of Wildlife, and Crane

Hollow Nature Preserve. An amazing group of technicians have worked tirelessly on this project over the years including, Marnie “The Bad One” Behan, Erica Ebert, Jeff

Geiger, Brent Wittich, Izzy Ruta, Holden Slagel, and Tyler Black. Crane Hollow has provided amazing help with their interns including, Ben Rechel, Brad Cordle, Adrian

Shields, Lucy Williams, and Katie Kelly. I have to personally thank Marnie Behan for saving the day so many times and providing an endless amount of sass. In addition to 6

technicians, I’ve had numerous wildlife professionals, colleagues, and friends provide support for this project. In particular, I have to thank Katrina Schultes and Bridget

Brown for providing endless wisdom, support, and field help since day one.

Additionally, I would also like to thank Missy Meierhofer and Sam Leivers for coming to the rescue and helping with statistical analyses.

I also would like to express the deepest gratitude for my boyfriend and greatest supporter Daniel Chappell. He believes in me even when I can’t believe in myself and this thesis is dedicated to him. I am also so incredibly grateful for my loving parents and two sisters, Anna and Mandy, for always supporting my dreams and passions.

Thank you to the incredible group of graduate students at Ohio University for their ongoing support. In particular, I would like to thank the bat lab graduate students,

Aüstin Waag, Mattea “Tater” Lewis, and Eli Lee, for the endless comradery and laughs.

This project would not have been possible without the generous funding provided by the Ohio Department of Natural Resources. A very big thank you to Kate

Parsons, Melissa Moser, Erin Hazelton, and Ryan Harris, for all the support. I would also like to thank Ohio University Department of Biological Sciences for the additional generous financial support. In addition, I would also like to express the deepest gratitude to all those at Crane Hollow Nature Preserve for providing housing, personal support, training, equipment, and trusting me with your bats all these years. I would like to thank the Crane Hollow Board of Directors, Heather Stehle, and Joseph

Moosbrugger for allowing this research in the strikingly beautiful place that is Crane 7

Hollow Nature Preserve with me. A very special shout out goes Joseph Moosbrugger and Lauren Metcalf for the botany lessons, mentorship, field support, and cherished friendship. Working at Crane Hollow Nature Preserve will forever be one of the greatest privileges of my life.

Lastly, I have to express so much gratitude for all the eastern red bats that were a part of this study, as without them this project would have never been successful. I am eternally grateful for their fuzzy butts and roosting knowledge. All animal capture and handling procedures were approved by the Institutional Animal Care and Use

Committee of Ohio University (17-H-008).

TABLE OF CONTENTS Page

Abstract ...... 3 List of Tables ...... 10 List of Figures ...... 11 Chapter 1: Introduction ...... 12 Chapter 2: Summer Day-Roost Selection by Eastern Red Bats (Lasiurus borealis) Varied Between Areas with Different Land Use Histories ...... 15 Abstract ...... 15 Introduction ...... 16 Materials and Methods ...... 20 Capture and Radio-telemetry ...... 22 Day-roosting Habitat ...... 22 Data Analysis ...... 24 Results ...... 25 Capture and Radio-telemetry ...... 25 Day-roosting Habitat ...... 25 Discussion ...... 28 Tables and Figures ...... 35 Appendix ...... 42 Chapter 3: Roost Elevation and Daily Temperature, but Not Sex, Predict Patterns of Summer Heterothermy in Eastern Red Bats (Lasiurus borealis) in Southern Ohio ..... 43 Abstract ...... 43 Introduction ...... 44 Methods...... 46 Study Area ...... 46 Capture and Temperature-sensitive Radio-telemetry ...... 47 Roost Microclimates ...... 48 Data Analyses ...... 49 Results ...... 51 Capture and Temperature Sensitive Radio Telemetry ...... 51 Discussion ...... 53 9

Tables and Figures ...... 58 Literature Cited ...... 65

LIST OF TABLES Page

Table 2.1: Roost and random tree species at Crane Hollow and Sunday Creek ...... 35

Table 2.2: Logistic regression model rankings for Crane Hollow ...... 36

Table 2.3: Logistic regression model rankings for Sunday Creek ...... 36

Appendix 2.1: Biologically informed a priori models ...... 42

Table 3.1: Linear mixed effect model rankings for average skin temperature ...... 58

Table 3.2: Linear mixed effect model rankings for minimum skin temperature ...... 58

Table 3.3: Linear mixed effect model rankings for canopy microclimate ...... 59

LIST OF FIGURES Page

Figure 2.1: Capture rates of eastern red bats at Crane Hollow and Sunday Creek ...... 37

Figure 2.2: Box and whisker plots of tree elevation ...... 38

Figure 2.3: Box and whisker plots of tree diameter ...... 39

Figure 2.4: Box and whisker plots of tree distance to water ...... 40

Figure 2.5: Average daytime temperatures at three different elevations ...... 41

Figure 3.1: Box and whisker plot of average skin temperatures...... 60

Figure 3.2: Box and whisker plot of minimum skin temperatures ...... 61

Figure 3.3: Example skin temperature readings from male and female red bat ...... 62

Figure 3.4: Temperature vs. Elevation of microclimate dataloggers in trees ...... 63

Figure 3.5: Example skin temperature reading from pregnant female red bat ...... 64

CHAPTER 1: INTRODUCTION

The eastern red bat (Lasiurus borealis) is a solitary, foliage roosting bat species that ranges across the entirety of the eastern United States, as well as into

Canada and Mexico (Shump and Shump 1982). This species is thought to be in decline due to several factors, including fatal collisions with wind turbines and loss of forested habitat (Winhold et al. 2008, Frick et al. 2017). Mortalities at wind energy facilities have emerged as a significant threat to the eastern red bat in last two decades, particularly along fall and spring migration routes (Arnet et al. 2008, Arnett and

Baerwald, 2013). However, habitat loss has long been a concern for this species due to its reliance on forested habitat throughout the year (Carter et al. 2003, Kunz et al.

2003). Forested habitats are essential in that they provide day-roosting habitat, which in turn provides protection from predators, shelter from the elements, and a place to raise young (Barclay and Kurta 2007, Kunz and Lumsden 2003). Without suitable forested habitats, eastern red bats cannot reproduce successfully, exacerbating their decline. Nevertheless, concerns over the large-scale mortalities at wind energy facilities often overshadow concerns over the availability of suitable habitat.

Further troubling for the eastern red bat is that although the decline of the eastern red bat has come to light in recent years, it is frequently overshadowed by other threats facing bats in the United States. For example, white-nose syndrome

(WNS), a fungal pathogen caused by the cold-loving fungus Pseudogymnoascus destructans, is a deadly disease that has decimated populations of cavernicolous bats in the eastern United States (Frick et al. 2006). WNS and the decline of most cave- 13

hibernating species has been the focus of considerable bat research since 2006, but not all species are susceptible. Due to the unique behavior of hibernating under leaf litter, the eastern red bat does not appear to be vulnerable to the disease (Dunbar and Tomasi

2006, Pettit and O’Keefe 2017). Nevertheless, the eastern red bat is still in decline, and merits conservation.

This project was initiated in 2015 by the Ohio Division of Natural Resources as a broadscale initiative to better understand and document bat species in Ohio. After preliminary mist-net surveys were conducted in the southeastern region of Ohio from

2015–2016, the project was narrowed in scope when reproductive female eastern red bats were documented in some areas of the state but not in others. Due to the expansive rage of the eastern red bat, there are large stretches of the range that do not have local habitat selection data, making the trends in bat captures difficult to interpret. This represented an important opportunity to better understand important habitat features for this species, which can then be used to help generate management recommendation for all bat species.

The goal of my Master’s research at Ohio University was to continue this work started by the Ohio Division of Natural Resources in order to understand important habitat features for eastern red bats and further investigate trends that suggested that sexes of this species were occupying different habitats. To achieve this goal, I examined patterns of day-roost selection by male and female eastern red bats at two study sites in southeastern Ohio. By conducting mist-net surveys and using temperature-sensitive radio-telemetry, we were able to not only acquire habitat data 14

for this species, but we were able to start to understand the underlying reasons behind the observed patterns of selection. The results of my study yielded interesting trends that are seen in other parts of the eastern bat range, but it also revealed other aspects of this species’ habitat selection and thermoregulation that I was not expecting. I hope these data can be used by land stewards and managers in protecting and enhancing habitat for the eastern red bat. To that end, Chapter 2, “Summer Day-Roost Selection by Eastern Red Bats (Lasiurus Borealis) Varies Between Areas with Different Land

Use Histories”, is written as a stand-alone publication detailing the differences in day- roost selection in areas that have distinctly different forest structure and composition.

Chapter 3, “Roost Elevation and Daily Temperature, but Not Sex, Predict Summer

Heterothermy in Eastern Red Bats (Lasiurus Borealis) in Southeastern Ohio”, is also a stand-alone publication that details the first study to investigate thermoregulation in eastern red bats by recording skin temperatures.

Bats are the among the most diverse orders of mammals on the planet and are essential to healthy, functioning ecosystems worldwide (Voit and Kingston 2016).

However, a troubling number of bat species are in decline, adding to the growing list of taxa that are suffering populations losses around the world. Although this study is limited in scope and cannot address population losses of all bats, I hope that the information presented in the next chapters can provide information to help curb the decline of the eastern red bat before populations decline to alarming levels.

CHAPTER 2: SUMMER DAY-ROOST SELECTION BY EASTERN RED BATS (LASIURUS BOREALIS) VARIED BETWEEN AREAS WITH DIFFERENT LAND USE HISTORIES

Abstract

The eastern red bat (Lasiurus borealis) is widely considered to be in decline, inspiring interest in identifying important habitats so that they may be protected.

Unfortunately, knowledge of important day-roosting habitats is lacking for much of the species’ range. We examined patterns of day-roost selection by male and female eastern red bats at two study sites in southeastern Ohio to help fill this information gap. We radio-tagged 28 male and 25 female bats during the summers of 2016–2019 and located 53 male and 74 female day-roosts. Day-roost selection differed between sexes and study areas. In a mostly even-aged forest with significant historical disturbance, we found males and females roosting in trees located at higher elevations, and farther from water sources, than available trees, with no clear selection based on tree or stand characteristics. However, in a forest with less historical disturbance and more structural diversity, we found sexes differed in how they selected from available habitats. Specifically, males selected trees with larger diameters located at lower, cooler elevations than females. Females selected roosts primarily at higher, warmer elevations. These data show that heterogeneity in environmental conditions can lead to different patterns in habitat selection, even between sites located in a small geographic area. They also show that eastern red bats sexually segregate on the local landscape, provided the presence of diverse forest conditions, but may not do so in the absence of such diversity. Similar to management recommendations for other bat species, we 16

recommend managing forests using uneven-aged techniques, maintaining structural diversity across an elevational gradient to provide male and female eastern red bats with suitable day-roosting habitat in Southeast Ohio.

Introduction

Successful wildlife conservation relies upon accurate descriptions of the habitat features and resources that species need to survive and reproduce (Mayor et al.

2009, Morrison et al. 2012). However, such descriptions may be difficult to formulate for species with large geographic ranges, as they may occupy a wide range of environmental conditions and ecological communities (Ruhl et al. 2018). Further hindering identification of important resources, some wide-ranging species also occupy human-altered landscapes composed of a mosaic of habitat patches, each with a distinct vegetative composition and structure, influencing habitat use and quality

(Betts et al. 2014, Duchamp et al. 2008). Although this can be a challenge for conservation, these patches represent opportunities to study how species utilize or fail to utilize different habitats within the same region, illuminating how species respond to a range of conditions and potentially informing wildlife conservation (Russo et al.

2010).

Protecting habitat for wildlife remains a priority for conservation as the number of factors threating biodiversity continue to increase (Rawat et al. 2015, Voigt and Kingston 2016). In North America, the eastern red bat (Lasiurus borealis) is believed to be in decline from collisions with commercial wind turbines and habitat loss (Winhold et al. 2008, Frick et al. 2017). Several species suffer from fatal 17

collisions at commercial wind energy facilities, but the eastern red bat is the second most commonly killed (Arnet et al. 2008, Arnett and Baerwald, 2013). Exact mortality rates are unclear and vary by year, but it was estimated that 143,023–287,403 eastern red bats were killed at wind facilities in the United States from 2000 to 2011 (Arnett and Baerwald, 2013). Mortalities are expected to continue as this wind energy technology becomes more popular and efficient, leading to concerns over population viability. While large-scale mortalities at wind energy facilities are a relatively new phenomenon, it has long been assumed that eastern red bats are negatively affected by loss of forest habitat across their range (Carter et al. 2003, Kunz et al. 2003). As a species that completely relies on forests, red bats are particularly vulnerable to changes in forest communities, and forest management conducted with this species in mind may help slow its decline (Carter and Menzel 2007, Hayes and Loeb 2007).

However, studies of eastern red bat habitat selection are sparsely distributed across the species range, which spans across the entirety of the eastern United States into Canada and Mexico (Shump and Shump 1982).

Forests are central to the ecology of many bat species, with 27 of the 45 species documented in the United States and Canada known to roost in trees (Lacki et al. 2007). Day-roosts are important because they provide bats with shelter from the elements, protection from predators, and a place to raise young (Barclay and Kurta

2007, Kunz and Lumsden 2003). Selection of day-roosts is well known to vary both within and among species. Within species, males and females often select different summer roosts based on their differing needs and limitations during this time (Dzal 18

and Brigham 2013, Encarnação et al 2012, Johnson and Lacki 2013). During summer, females require warm roosts to help keep energetic costs as low as possible during gestation and lactation (Olson and Barclay 2013, Pretzlaff et al. 2010, Willis and

Brigham 2007). In species that roost under bark or in tree cavities, females often select large trees that have space for large social groups that significantly increase roost temperatures and therefore provide energy savings (Willis and Brigham 2007,

Sedgeley 2001, Olson and Barclay 2013). However, solitary, foliage roosting species such as the eastern red bat, which hangs from branches and leaf petioles, lack the benefits of sociality (Kurta and Lehr 1995, Shump and Shump 1982). Because these bats roost in tree canopies exposed to weather, it has been suggested that eastern red bats select roosts based on temperature, but there are few studies examining the role of temperature and other habitat characteristics, especially considering the large geographic range of this species (Hutchinson and Lacki 2001).

Several characteristics of day-roosts are known to influence roost temperature.

Foliage roosting species such as red bats have been found selecting roosts that are taller compared to the height of the surrounding canopy, possibly increasing solar exposure and roost temperature (Hutchinson and Lacki 2000, Kalcounis-Ruppell et al.

2005, Lacki and Baker 2003). They have also been documented roosting in areas with more closed canopies than other species of tree roosting bats, possibly decreasing solar exposure and temperature (Hutchinson and Lacki 2000, Kalcounis-Ruppell et al. 2005,

Lacki and Baker 2003). These seemingly conflicting results may illustrate a trade-off between reducing energy expended on thermoregulation and avoiding predators by 19

roosting in denser canopies (Carter and Menzel 2007, Hutchinson and Lacki 2001).

Red bats are also known to select roosts based on characteristics that have little or no connection to temperature. For example, red bats are believed to be negatively affected by the amount of vegetative clutter in a stand, which can inhibit navigation in species with relatively high wing loading values (Loeb and O’Keefe 2006). Thus, day- roost selection in bats is influenced by several variables pertaining to habitat at different spatial scales.

Because red bats are influenced by forest structure and composition, different forest management regimes and silvicultural treatments are likely to affect habitat quality for this species. In intensively managed forests in the southern United States, red bats were typically found roosting in areas within or close to stands of closed canopy mature hardwoods (Elmore et al. 2004). In these areas, bats preferred older stands in areas of select thinning or unmanaged forests, as compared to younger, more intensively managed forests (Perry et al 2007). While these studies provide insight as to how bats utilize managed forests, many forests are not intensively managed, but are left to regenerate following disturbance. Even within relatively small geographic areas, regenerating forests with widely differing conditions can be found, and it is unclear how these conditions affect habitat use. This is true of many forests in the

Western Allegheny Plateau, where not only have forests been heavily disturbed, but also where there is currently no local information on roosting behaviors of the eastern red bat. Thus, land managers in this region lack the information necessary to 20

successfully promote and protect eastern red bat habitat to aid in the conservation of this declining species.

The purpose of this study was to compare day-roost selection by male and female eastern red bats in southeastern Ohio forests with markedly different land use histories and forest structures. We used an information theoretic approach to rank a suite of a priori hypotheses seeking to differentiate between day-roost selection at each site. We hypothesized that (i) eastern red bats would select day-roosts non- randomly, with different habitat characteristics used by males and females, and (ii) that selection patterns would differ between sites with different forest conditions.

Materials and Methods

Our study was conducted at two locations in Hocking County, Ohio, each with different land use histories and forest communities. Our first site, Crane Hollow

Nature Preserve (N 39.480089, W -82.584173), is approximately 768 hectares of state dedicated nature preserve devoted to promoting unique habitats, fostering education, and promoting research. Crane Hollow is a biologically diverse area with ≥10,000 documented floral and faunal species inhabiting the ridgetops and Blackhand

Sandstone gorges that run the length of the preserve (Coovert 2015). Elevation at this location ranges from 335 m on ridges to 228 m at the bottom of the gorges. The forest community along ridgetops is dominated by oak (Quercus sp.), maple (Acer sp.), hickory (Carya sp.), and elm species (Ulmus sp.), as well as American beech (Fagus grandifolia) and tulip poplar (Liriodendron tulipifera). Eastern hemlock (Tsuga canadensis), American sycamore (Platanus occidentalis), American beech (Fagus 21

grandifolia), birch (Betula sp.), and maple species (Acer sp.) dominate the steep slopes and bottoms of the gorges. Although extensive farming and timber harvesting has occurred across the preserve before the property became dedicated preserve in 1977, large stretches of Crane Hollow have maintained stands that are estimated to be over

100 years old.

Our second site, located on 1348 hectares owned by the Ohio Division of

Wildlife approximately 33 km northeast of Crane Hollow, is a section of O’Dowd

Wildlife Area locally referred to as Sunday Creek Coal (N 39.531083, W -82.198662).

From the late 1800’s to the 1950’s, the Sunday Creek property was the location of large coal mining operations that, along other industry activities, left the property deforested. Much of the land has naturally reforested since the closing of the final mine in the 1950’s, although disturbances such as timber harvests, oil and natural gas exploitation, and the creation of extensive ATV trails are common. The topography at

Sunday Creek is characterized by hillsides built up by surface mining. Slopes are not as precipitous as Crane Hollow and lacked exposed cliffs, but elevations were similar, ranging 222–323 m. The forest community consists of a variety of tree species commonly found in Ohio, such as oaks maple, tulip poplar, bigtooth aspen (Populus grandidentata), American beech, American sycamore, and elms species. Due to the recent mining operations, stands at Sunday Creek tended to be more densely stocked with smaller, younger stems compared to those at Crane Hollow (Table 2.1).

Capture and Radio-telemetry

We captured bats from late May through early August 2016–2019 using mist- nets (Avinet Inc, Portland, Maine) placed in foraging corridors such as roads, streams, or over open water sources. We opened nets shortly before sunset and left them open for 5 hours. All captured bats were identified to species, age (adult or young-of-the- year), sex, and reproductive condition. Right forearm length and weight were also measured. Eastern red bats weighing >6.6 g were fitted with 0.33 g transmitters

(model LB-2 and LB-2TX, Holohil Systems Ltd., Ontario, Canada) using surgical adhesive (Osto-bond, Quebec, Canada). We tracked radio-tagged bats to their day- roosts using handheld receivers (Biotracker Version, Lotek Wireless, Ontario, Canada) each day for the life of the transmitter or until the transmitter had fallen off. We recorded the UTM of each day-roost using a handheld GPS unit with a 3 m accuracy.

Day-roosting Habitat

To compare habitats used to those available, we collected a suite of tree-, stand-, and landscape-scape habitat measurements at each day-roost and at 100 random trees in each study area. Tree-scale habitat measurements included tree species, diameter at breast height (dbh), tree height, percent canopy cover, and slope aspect. Stand-scale measurements included canopy height, distance to nearest live tree, percent slope, and basal area (BA). Landscape-scale measurements included distance to roads, distance to water, distance to forest edge, and elevation. BA was measured using a 10-factor wedge prism, height was measured with a laser rangefinder/ hypsometer, percent canopy cover was measured using a spherical densitometer, and 23

dbh was measured using a dbh tape. Landscape-scale measurements were calculated using ArcMap (version 10.5.1, ESRI, Redlands, CA).

We generated random locations in each study area to measure habitat available for day-roosting. Before generating these locations, we determined the area covered by dominant habitat types within each study site and dispersed 100 points proportionally within each of those habitats. Dominant habitat types were provided by previous studies (Whittemore 2016) and from the Ohio Department of Natural Resources. At

Crane Hollow, these habitats were oak/mesic forests, hemlock forests, and forested floodplains. At Sunday Creek, the dominant habitat types were xeric forests, mesic forests, floodplain forests, and a small amount of unclassified habitat. Coordinates for random points were created using a random point generator in ArcMap. We navigated to these locations using a handheld GPS and collected the same suite of habitat variables described above for a tree >10 cm in diameter located closest to the random point.

To investigate if day-roost selection may be influenced by ambient air temperatures (Ta), we measured Ta across the elevational gradient present in each study area. We deployed temperature and relative humidity dataloggers (Onset

MX2301, Onset Computer Corp., Bourne, Massachusetts) inside solar radiation shields at three slope positions. These corresponded to ridgetops (Crane Hollow: 334 m; Sunday Creek Coal: 323 m), mid-slopes (288 and 270 m, respectively), and bottoms (228 and 222 m, respectively). Each logger was programmed to record Ta every ten minutes. Because we were interested in how temperatures varied across 24

slope positions during the day-roosting period, we summarized Ta recorded at each logger during the hours between sunrise and sunset into daily maximum and daily minimum temperature.

Data Analysis

Male and female eastern red bat capture rates were calculated for each site by dividing the total number of each sex captured by capture effort (nights of netting) to provide a measure of capture success. We did not statistically compare capture rates as mist-netting is not a reliable sampling technique, but we do present capture rates as a qualitative measure of capture success (MacCarthy et al. 2006). To test our two hypotheses, we compared the habitat characteristics of male roosts, female roosts, and random trees at each site using multinomial logistic regression and a model selection approach. Prior to analysis, we tested for multicollinearity in RStudio (RStudio Team

2016) using a correlation matrix and did not combine variables in a model if they were

> 70% correlated. We ran 20 multinomial logistic regression models for each study area, with each model representing a distinct, biologically informed a priori model developed from a review of existing literature (Appendix 2.1). Our dependent variable was sex, which was organized into three categories male, female, and random trees to compare roosts of each sex to available habitat. To assess our hypothesis that day- roost selection would differ between the two areas, we ranked models for each study area separately using Akaike’s Information Criterion corrected for small sample sizes

(AICc). Models with ΔAICc < 2 were considered competitive as a top model. We assessed the fit of our top models using Hosmer-Lemeshow goodness of fit tests and 25

determined if we had false positive variables by referencing our odds ratios to see which variables were driving our top models. Confidence ratios for our odds ratios were also assessed and any variable with confidence intervals crossing 1 were not considered to be a strong predictor.

Results

Capture and Radio-telemetry

We captured bats on 69 nights, including 38 nights at Crane Hollow and 31 nights at Sunday Creek. Adult males were more commonly captured than adult females at both sites, and adult females were rarely captured at Sunday Creek compared to Crane Hollow. We captured 1.1 adult male and 0.6 adult female red bats/night at Crane Hollow (n = 43 male, 23 female), and 0.9 adult male and 0.2 adult female bats/night captured at Sunday Creek (n = 28 male, 7 female) (Figure 2.1). We radio-tagged 17 adult females and 14 adult males at Crane Hollow, tracking these bats to 87 roosts (male: n = 29, female: n = 60). At Sunday Creek, we radio-tagged 7 adult females and 16 adult males, tracking them to 51 roosts (male: n = 34, female: n = 17).

Overall, bats were tracked for an average of 6.5 days (range = 1–19) to an average of

3.3 day-roosts (range = 1–11) per bat.

Day-roosting Habitat

Habitat measurements from our random trees indicated differing forest structure and composition between Crane Hollow and Sunday Creek. Basal area at

Sunday Creek (푥̅ = 16.3 ± 0.86 m2/ha) was lower than Crane Hollow (푥̅ = 23.0 ± 1.1 m2/ha), but average tree diameter (푥̅ = 30.4 ± 1.2 and 푥̅ = 37.8 ± 5.2 cm, respectively) 26

and height (푥̅ = 18.7 ± 0.47 and 푥̅ = 20.5 ± 0.67 cm) was lower at Sunday Creek. In addition, the number of live stems within 20 m for Sunday Creek (푥̅ = 72.0 ± 4.6) was higher than for Crane Hollow (푥̅ = 64.0 ± 3.0), indicating that Sunday Creek has a higher density of smaller stems than Crane Hollow. The species composition of the

100 random trees sampled at each site reflect differing forest communities between the two sites, but roost tree species composition was similar at the two locations (Table

2.1). At Crane Hollow, the tree species bats most frequently chose for day-roosting included sugar maple (n = 29), oak species (n = 20), and tulip poplar (n = 14). Females were found most frequently in sugar maples, while males were found most frequently in oaks. At Sunday Creek, day-roost species included tulip poplar (n = 13), oak species (n = 10), sugar maple (n = 8), and American beech (n = 8). Females were most frequently located in sugar maples and tulip poplar, while males most used oak.

The best supported model from Crane Hollow was the Tree Diameter and

Elevation Model, which received 69% of the overall model support (Table 2.2). No other model had ΔAICc < 2. Odds ratios of variables in this model show that the variable with the largest impact on probability of a tree being used by females was elevation (odds ratio = 1.151, CI = 1.151–1.153), but less so for males (odds ratio =

0.986, CI = 0.98–0.99). These ratios show that for every meter increase in elevation, the likelihood that a bat would occupy a roost increased by 15% for females but decreased by 1% for males. Females roosted within a narrow range of elevations (푥̅ =

330 ± 0.07 m, range = 319–334 m) while males (푥̅ = 319 ± 0.41 m, range = 251–331) and random trees (푥̅ = 321 ± 1.46 m, range = 282–335) were found across a wider 27

range (Figure 2.2). For males, the variable with the largest effect on probability of a tree being used was dbh (odds ratio = 1.05, CI = 1.027–1.083). This indicates that for every centimeter increase in diameter of a tree, the odds that a male will use that tree for a roosting location increases by 5%. However, diameter was not as informative as elevation in driving female selection (odds ratio = 1.02, CI = 1.002–1.05). Mean diameter of male (푥̅ = 52 ± 0.52 cm), female (푥̅ = 44 ± 0.30 cm), and random trees (푥̅

= 38 ± 1.55 cm) reflect these odds ratios (Figure 2.3).

In contrast to Crane Hollow, the best supported model for Sunday Creek was the Elevation and Distance to Water Model, which received 72% of the overall model support (Table 2.3). No other model had ΔAICc < 2. Odds ratios of variables in this model show that elevation was the variable with the largest impact on the probability of tree being used by both males (odds ratio = 1.05, CI = 1. 02–1.077) and females

(odds ratio = 1.034, CI = 1.025–1.077). Thus, each meter increase in elevation had a lesser effect probability of a tree being used by females compared to Crane Hollow

(3% versus 15% increase) but a greater effect for males (1% decrease versus 5% increase). Resultingly, males (푥̅ = 297 ± 0.72 m, range = 285–304) and females (푥̅ =

294 ± 0.59 m, range = 251–331) occupied similar ranges of elevations at Sunday

Creek. Random trees were found across a wider range of elevations (푥̅ = 283 ± 1.71 m, range = 242–319) (Figure 2.2). Distance to water was an explanatory variable in the model for both males (odds ratio = 1.005, CI = 1.0002–1.0009) and females (odds ratio = 1.009, CI = 1.003–1.015), with both sexes roosting farther from water sources 28

(females: 푥̅ = 204 ± 3.22 m, males: 푥̅ = 164 ± 3.66 m) compared to random trees (푥̅ =

120 ± 9.40 m) (Figure 2.4).

Ambient temperatures varied across the elevational gradient at both sites.

Ridgetops had slightly warmer daytime temperatures from June–August at Crane

Hollow (푥̅ = 24.0 ± 0.25° C) and Sunday Creek (푥̅ = 22.6 ± 0.22° C) compared to lower elevations (푥̅ = 21.0 ± 0.16° C and 푥̅ = 20.8 C ± 0.15° C, respectively). Similar elevations between the two sites varied slightly in temperature with the nature preserve being slightly warmer on average than the previously mined site (Figure 2.5).

Discussion

We found that daytime habitat selection by eastern red bats in southeast Ohio varied between sexes and between study areas with different forest conditions. In a forest characterized by larger, less densely stocked trees, males and females selected day-roosts differently based on tree diameter and elevation. Selection of day-roosts was influenced by diameter and elevation in both sexes, but female use was more positively affected by elevation while male use was more positively affected by tree diameter and male roosts tended to be located at lower elevations. However, in a forest with a similar elevational gradient characterized by a higher density of smaller trees, tree diameter was not important in roost selection, and male and female roosts overlapped more in elevation. These data show that patterns of habitat use by bats can vary between areas located in the same part of a species’ range, likely in response to forest structure. This highlights the need to study habitat use across a range of environmental conditions to untangle local effects on habitat use. 29

The possibility that heterogeneity in environmental conditions can lead to differing effect sizes, and even directions, for variables influencing habitat selection in bats was recently explored by Fabianek and colleagues (2015). Focusing solely on cavity and bark roosting species, the authors found both consistent and inconsistent trends for the influence of variables measured in the present study, and concluded that summer temperatures moderate some, but not all, of this variability. The importance of temperature for small, insectivorous bats is well-known, as these small mammals lose heat quickly, and reproductive females benefit from selecting day-roosts that minimize this heat loss (Pretzlaff et al. 2010. Speakman 1999, Sedgeley 2001). Thus, a frequently appearing narrative is that reproductive females roost in trees located at lower elevations where conditions are warmer (Angell et al. 2013, Cryan et al 2000,

Encarnação et al 2005). Our results point to different effect of elevation, that eastern red bats, especially reproductive females, roost at higher elevations, but with the same inference: that female bats often chose to roost in warmer areas on the landscape.

Unlike the studies cited above, lower elevations were associated with lower temperatures in our study area. This difference in the thermal gradient likely stems from the scale of topography between our sites in southeastern Ohio, where elevation only varied by approximately 100 m, often in the form of narrow, steep gorges.

Despite the subtle difference in elevation our study, temperatures differed by ~3° C from ridges to bottoms, a biologically meaningful difference from a thermoregulatory perspective (Boyles et al. 2011, Willis and Brigham 2003). 30

It is notable that at Crane Hollow females were more likely to select roosts at higher elevations in comparison to males. Sexual segregation in bats has been observed by others, where again, females occupied lower, warmer elevations, than males during pregnancy and lactation (Angell et al. 2013, Cryan et al. 2000). Males are less energetically constrained than females, and likely experience fewer costs associated with roosting at lower temperatures (Grinevitch et al. 1995, Johnson and

Lacki 2012). However, lesser energetic constraints in males does not fully explain why males at Crane Hollow roosted at a different range of elevations than females, as roosting at colder elevations would increase the cost of maintaining normothermy for males. This finding suggests either that lower elevations provided males with a benefit such as the ability to use deeper torpor, the opportunity to avoid competition with bats in higher quality (warmer) microclimates, or that roosting at higher elevations comes with a cost that males are less willing to incur than females.

Meta-analyses and reviews of day-roost selection consistently find that tree diameter is a primary predictor of whether a tree is used by bats (Fabianek et al. 2015,

Lacki and Baker 2003, Kalcounis-Ruppell et al. 2005). Larger diameter trees are thought to benefit bark and cavity roosting species because these trees provide more room for roosting bats (Olson and Barclay 2013, Barclay and Kurta 2007). Larger trees have also been hypothesized to have thicker bark and are sometimes located in more open areas, making the tree a better thermal environment for roosting (Vonhof and Barclay 1996, Waldien et al. 2000, Barclay and Kurta 2007). These explanations are unlikely to explain why red bats preferred larger diameter trees at the preserve, as 31

increased space for social roosting would not benefit solitary red bats, and roost trees were not located in open habitats. It is also difficult to explain why males would roost in larger trees than females. Tree diameter at this site was not correlated with elevation, and we saw no evidence that larger trees were available at lower elevations, which could have otherwise helped to explain this difference. Thus, while it is apparent that several aspects of larger trees can benefit other bat species, it’s unclear why foliage roosting species would prefer larger diameter trees.

Although our results don’t provide insight into why tree diameter is important for red bats, it is clear from our results that sexes selected roosts differently when larger diameter trees were available. However, at Sunday Creek, where we caught fewer red bats, especially females, we found no evidence of sexual segregation. We hypothesize that this difference was the result of the different forest conditions present. We propose that young, dense forests with little diversity in forest structure are marginal roosting habitats for eastern red bats. In such marginal conditions, males and females either fail to segregate where they otherwise would, or female populations are so low as to eliminate the need for segregation, although further studies are needed in this area. Regardless, our results provide clear support of previous studies showing that forest structure is an important aspect of day-roosting habitat for the eastern red bat in southeastern Ohio (Hutchinson and Lacki 2000, Mager and Nelson 2001, Perry et al. 2007).

Finally, distance to water was an important aspect of roost selection at Sunday

Creek. Proximity to water is important for many bats, which often drink and forage in 32

the area surrounding their roosts upon leaving their evening emergence (Lacki et al.

2007, Fabianek et al. 2015, Hayes and Waldien 2001). However, we observed the opposite trend, with both male and female eastern red bats roosting farther from permanent water sources than available trees. Because it is unlikely that bats seek to roost far from water, it is likely that this metric either failed to properly capture water availability or was correlated with another, unmeasured variable. Ephemeral water sources regularly used by bats, such as puddles in forest roads or vernal pools, were found at higher elevations but were not be accounted for in our study. Regardless, the average distance measured between the water sources and roosts was 200 m, a distance easily traversed by bats, and unlikely to impact them negatively (Kurta et al.

2002, Lumsden et al. 2002).

The underlying motivation for many studies of bat habitat selection is to guide management practices (Brigham 2007). Over the last several decades, a large body of work has emerged describing the habitat of different species in different regions, with some clear trends emerging (Fabianek et al. 2015, Lacki et al. 2010). Unfortunately, many of these trends are aggregated across species and are biased towards studies of cavity- and bark-roosting species. However, foliage roosting species are also experiencing significant population declines, and more research is needed to guide management for these species as well. While our netting data cannot be used to directly compare relative abundance of each sex in our study areas, it is notable that we only captured 7 female red bats at Sunday Creek over four years of effort. It is therefore likely that females are rare in this area. Although it is not uncommon for 33

surveys documenting male-biased sex ratios to conclude that some areas unsuitable for female reproduction, data are often insufficient to make clear conclusions as to why

(Angell et al. 2013, Cryan et al 2000, Eencarnação et al. 2005). We encourage researchers to study habitat use within different forest contexts, beyond different silvicultural treatments. Investigations of where female reproduction does and does not occur in reference to these characteristics across a species’ range can lead to better understanding of the factors necessary for reproduction, driving habitat use, and helping guide management and conservation efforts.

In the absence of these broad scale-scale studies, our work highlights the importance of maintaining diversity in forest structure over elevational gradients for eastern red bats in southeastern Ohio. In particular, we suggest maintaining un-even aged forests that include mature forests with many trees > 42 cm dbh (based on our overall average roost diameter at Crane Hollow) across an elevational gradient. To improve habitat in areas such as Sunday Creek, single tree selective thinning or group selection thinning could be employed to open the understory and release remaining trees from competition. We recommended retaining species such as maple, tulip poplar, and oaks, over diverse elevational gradients, as these trees appear to be used at rates greater than their availability on the landscape. While the specific effects of the variables we report must be viewed in the context of our study areas, where the topography created cooler temperatures at the bottom of gorges and riparian areas, the inference that temperature has a strong influence on habitat selection by eastern red bats may broadly apply throughout the species’ range, and we urge land managers to 34

provide roosting in habitat in relatively warm microclimates where thermal conditions vary on the local conditions.

Tables and Figures

Table 2.1 Number of trees, followed by percent of total in parentheses, of various species used as day-roosts by male and female eastern red bats, along with available habitat (randomly sampled trees), in two study areas in Ohio from 2016–2019. Study Area & Tree Species Females Males Available Crane Hollow Nature Preserve Acer saccharum 23 (38%) 6 (21%) 15 (15%) Quercus sp. 12 (20%) 8 (28%) 12 (12%) Tsuga canadensis 0 0 20 (20%) Liriodendron tulipifera 8 (13%) 6 (21%) 12 (12%) Acer rubrum 3 (5%) 3 (10%) 17 (17%) Ulmus sp. 3 (5%) 2 (7%) 2 (2%) Betula lenta 1 (2%) 2 (7%) 3 (3%) Fagus grandifolia 1 (2%) 0 5 (5%) Carya sp. 3 (5%) 0 2 (2%) Pinus virginiana 0 0 4 (4%) Juglans nigra 2 (3%) 0 0 Fraxinus sp. 1 (2%) 1 (3%) 1 (1%) Betula alleghaniensis 1 (2%) 1 (3%) 0 Sassafras albidum 1 (2%) 0 0 Platanus occidentalis 1 (2%) 0 2 (2%) Prunus serotina 0 0 2 (2%) Robinia pseudoacacia 0 0 1 (1%) Oxydendrum arboretum 0 0 1 (1%) Sunday Creek Coal Company Liriodendron tulipifera 5 (29%) 7 (21%) 16 (16%) Acer saccharum 5 (29%) 3 (9%) 11 (11%) Quercus sp. 1 (6%) 9 (26%) 20 (20%) Fagus grandifolia 1 (6%) 7 (21%) 10 (10%) Acer rubrum 1 (6%) 0 8 (8%) Platanus occidentalis 0 0 8 (8%) Populus grandidentata 0 2 (6%) 4 (4%) Carya sp. 1 (6%) 1 (3%) 4 (4%) Nyssa sylvatica 1 (6%) 1 (3%) 2 (2%) Ostrya virginiana 1 (6%) 1 (3%) 1 (1%) Ulmus sp. 0 2 (6%) 3 (3%) Betula lenta 1 (6%) 0 0 Prunus serotina 0 0 4 (4%) Robinia pseudoacacia 0 1 (3%) 2 (2%) Fraxinus sp. 0 0 2 (2%) Acer saccharinum 0 0 1 (1%) Liquidambar styraciflua 0 0 1 (1%) Ailanthus altissima 0 0 1 (1%) Pinus virginiana 0 0 1 (1%) 36

Table 2.2 Top 4 models for day-roost selection by male and female eastern red bats, along with the number of parameters, AICc score, ΔAICc score, and model weight for each, for Crane Hollow Preserve. Model K AICc Δi wi

Diameter + Elevation 4 334.15 0 0.686 Elevation + Stand 6 336.97 2.81 0.245 Tree Size Hypothesis 1 + Elevation 5 337.71 3.55 0.116 Elevation + Distance to Water 4 340.62 6.46 0.027

Table 2.3 Top 4 models for day-roost selection by male and female eastern red bats, along with the number of parameters, AICc score, ΔAICc score, and model weight for each, for Sunday Creek Coal. Model K AICc Δi wi

Elevation + Distance to Water 4 242.04 0 0.720 Tree Size Hypothesis 1 + Elevation 5 245.75 3.71 0.113 Slope Aspect + Elevation 5 246.67 4.64 0.070 DBH + Elevation 4 248.07 6.03 0.035

Figure 2.1 Capture rates of male and female eastern red bats at Crane Hollow Nature Preserve and Sunday Creek Coal study areas in southeastern Ohio, 2016–2018. Female capture rates are represented by the light grey bars and male capture rates are represented by the dark grey bars.

Figure 2.2 Male (light grey) and female (white) eastern red bats roosted at different elevation than randomly sampled trees (dark grey) at Crane Hollow Preserve and Sunday Creek Coal study areas in southeast Ohio, 2016–2018. Females roosted at higher elevations than males at the preserve but sexes roosted at similar elevations at the reforested mine. Medians are depicted by the solid line, means are represented by the X’s, and outliers are depicted as black dots. Quartiles 2 and 3 are represented by the box and quartiles 1 and 4 are represented by whiskers.

Figure 2.3 Male eastern red bats (light grey) roosted in larger diameter trees than females (white) and randomly sampled trees (dark grey) at Crane Hollow Preserve, but not at the Sunday Creek Coal study area, where forest structure was less diverse, in southeast Ohio, 2016–2018. Medians are depicted by the solid line, means are represented by the X’s, and outliers are depicted as black dots. Quartiles 2 and 3 are represented by the box and quartiles 1 and 4 are represented by whiskers.

Figure 2.4 Male (light grey) and female (white) eastern red bats roosted farther from available water sources than randomly sampled trees (dark grey) at the Sunday Creek Coal, but not Crane Hollow Preserve, study area in southeast Ohio, 2016–2018. Medians are depicted by the solid line, means are represented by the X’s, and outliers are depicted as black dots. Quartiles 2 and 3 are represented by the box and quartiles 1 and 4 are represented by whiskers.

(a)

27 C) ° 25

Temperature Temperature ( 19

15 5-Jul 4-Aug 9-Aug 10-Jul 15-Jul 20-Jul 25-Jul 30-Jul 20-Jun 25-Jun 30-Jun 14-Aug 19-Aug 24-Aug 29-Aug

Low Mid High

(b)

27 C) ° 25

Temperature Temperature ( 19

15 5-Jul 4-Aug 9-Aug 10-Jul 15-Jul 20-Jul 25-Jul 30-Jul 20-Jun 25-Jun 30-Jun 14-Aug 19-Aug 24-Aug 29-Aug

Low Mid High

Figure 2.5 Average daily ambient temperatures at three different elevations at (a) Crane Hollow Preserve and (b) Sunday Creek Coal field sites in Southeast Ohio from 20 June–29 August, 2018.

Appendix

Appendix 2.1 Biologically informed a priori models used for our information theoretic approach and multinomial logistic regressions. Superscripts provide citations justifying the variables in each model. Model Variables Tree Size Model 1 Canopy DifferenceF+ DBHA+E Tree Size Model 2 Canopy DifferenceF + DBHA+E + Tree SpeciesD Airspace Model Canopy DifferenceF + Airspace* Tree Species Model Tree SpeciesD Tree Species Elevation Model Tree SpeciesD + ElevationC+D Basal Area Model Basal AreaE Canopy + Basal Area Model Canopy CoverF + Basal AreaE Large Open Stand Model Canopy DifferenceF + DBHA+F + Basal AreaE Tree Elevation Model ElevationC+D + DBHA+E + Tree SpeciesD Elevation Model ElevationC+D Distance to Water Model Distance to WaterF+B Distance to Edge Model Distance to EdgeG+E Elevation + Distance to Water Model Distance to WaterF+B + Elevation C+D Distance to Edge + Water Model Distance to EdgeI+J + Distance to WaterF+B Elevation + Basal Area Model ElevationC+D + Basal AreaE Canopy Closure Model Canopy CoverA + Canopy DifferenceF +DBHA+E Diameter of Roost Tree Model DBHA+E Tree Diameter + Elevation Model DBHA+E + ElevationC+D Slope Aspect + Elevation Model Slope AspectD + ElevationC+D Slope Aspect Model Slope AspectD *Airspace = distance (m) from lower canopy to the ground AKalcounis-Ruppell et al. 2005 BLimpert et al. 2007 CCryan et al. 2000 DHutchinson and Lacki 2000 ECarter and Menzel 2007 FMenzel et al. 1998 GO’Keefe et al. 2009

CHAPTER 3: ROOST ELEVATION AND DAILY TEMPERATURE, BUT NOT SEX, PREDICT PATTERNS OF SUMMER HETEROTHERMY IN EASTERN RED BATS (LASIURUS BOREALIS) IN SOUTHERN OHIO

Abstract

Temperate bat species can respond to environmental temperatures outside their thermoneutral zone through physiological responses, habitat selection, and social roosting. While there have been several studies examining these factors in bats that form maternity colonies in structures such as tree cavities, underneath the bark of snags, or manmade structures, less is known about bats that roost solitarily in tree foliage. To better understand the thermal environment of a foliage-roosting bat and how they respond to thermal stimuli, we used temperature-sensitive radio-telemetry and environmental dataloggers to study thermoregulation in the eastern red bat

(Lasiurus borealis) in southeast Ohio. We successfully collected skin temperature data from 9 males and 10 reproductive females and found that minimum ambient temperature, maximum ambient temperature, but not sex, predicted heterothermic responses in these bats. Microclimate dataloggers placed within the tree canopy showed that trees located at higher elevations had warmer canopies than those at lower elevations. These results illustrate how roosting ecology effects thermoregulatory strategies in bats, and that solitary, foliage-roosting species face stress from cool temperatures during summer. These results have conservation implications, indicating that areas on the landscape with warmer conditions are important for reproduction. In southeastern Ohio, this includes upland forests, but our data suggest that temperatures at these slope positions, and not the elevation itself, is the important habitat feature for eastern red bats. 44

Introduction

Bat species inhabiting temperate zones must often cope with environmental temperatures that are below their thermoneutral zone during summer. These temperatures increase the cost of maintaining euthermic body temperatures at a time when many individuals, especially reproductive females, have energy budgets that are difficult to meet (Becker et al. 2013, Kurta et al. 1989, Willis and Bigham 2007). To cope with this energetic stress, bats frequently use torpor, although its use comes with physiological and ecological costs (Dzal and Brigham 2013, Johnson and Lacki 2014,

Rintoul and Brigham 2014). For example, torpor delays gestation and reduces milk production, thereby slowing the development of young (Racey and Swift 1981, Wilde et al. 1999). Because the length of time that juvenile bats have to grow and prepare for autumn migration and winter hibernation influences their survival (Ransome 1989,

Frick et al. 2010), it was long presumed that females should avoid using torpor during reproduction. This dialogue has since shifted towards an understanding that torpor is often necessary in times of low ambient temperatures, food, and water availability, and may be an adaptation for delaying birth during unfavorable conditions to increase survival of both mother and offspring (Johnson and Lacki 2014, Geiser and Brigham

2012, Willis et al. 2006).

Bats also respond to unfavorable environmental conditions through habitat selection. Many species roost in structures such as the space underneath exfoliating tree bark, tree cavities, and manmade structures; thermal insulators that reduce thermoregulatory costs of normothermy during cold periods (Kurta 1985, Lacki et al. 45

2013, Sedgeley 2001). Females of some species can reduce the cost of thermoregulation and the need to use torpor by roosting in maternity colonies (Barclay and Kurta 2007, Willis and Brigham 2007). For example, social groups of big brown bats (Eptesicus fuscus) were found to increase temperatures inside roosts by up to 7°

C, reducing daily energy budgets by an average of 9% (Willis and Brigham 2007).

Thus, selecting a roost where other bats are present may often result in greater energetic savings than can be accomplished by selecting roosts with warmer microclimates. And together, social roosting and habitat selection can significantly protect bats from unfavorable thermal conditions.

Nevertheless, not all species roost in social groups or within the insulated structures described above. The eastern red bat (Lasiurus borealis) is a solitary, foliage-roosting species that is more vulnerable to ambient temperatures due to the exposed nature of their roosting sites (Hutchinson and Lacki 2001). Red bats have a furred uropatagium that increases their insulation by 15% when wrapped around the body, helping them to cope with some of these additional thermal challenges (Shump and Shump 1980). However, this insulation alone may not offset the costs associated with maintaining euthermy while roosting in foliage on cold days. Furthermore, although the energetic needs of this species during reproduction have not been studied, they are likely to be quite high relative to other North American species, as red bats are notable for having litters of up to four pups (Shump and Shump 1982). Indeed, a recent study found that reproductive female red bats roost at higher elevations than 46

males, where conditions were warmer than at lower slope positions, suggesting that this species experiences thermal stress during summer (Chapter 2).

The goal of this study was to investigate summer roost temperatures and thermoregulation in eastern red bats in southeastern Ohio to better understand the thermal environment this species roosts in and how bats respond to these conditions.

We hypothesized (1) that males would thermoregulate less precisely than females during the maternity season, as reflected by lower daily average and minimum skin temperatures (Tsk); (2) that average and minimum Tsk would be correlated with daily temperatures in both sexes; and (3) that females would select roosts with microclimates that are warmer than those of trees randomly sampled on the landscape.

Methods

Study Area

We conducted our study at two locations in southeastern Ohio, Crane Hollow

Nature Preserve and the Sunday Creek Coal Property, a recent addition to the O’Dowd

Wildlife Area. Crane Hollow is a state dedicated nature preserve located in Ohio’s

Hocking Hills and is approximately 768 hectares. Due to the Blackhand Standstone gorges that run through the region, Crane Hollow is topographically diverse, supporting a diversity of forest communities ranging from 241–335 m in elevation.

Sunday Creek, located approximately 33 km to the east, is approximately 1348 hectares of land naturally reforesting following extensive coal mine operations that ceased in the 1950s. Although it lacks the sandstone gorges of Crane Hollow, Sunday

Creek also includes forest communities across a similar range of elevations, 222–323 47

m. A more detailed description of each area is found in Chapter 2. Ambient temperatures, measured with environmental dataloggers placed within solar-radiation shields (Onset MX2301, Onset Computer Corp., Bourne, Massachusetts) were comparable between study areas, with warmer conditions present at higher elevations.

Capture and Temperature-sensitive Radio-telemetry

Bats were captured from late-May through early-August of 2016–2019 using mist-nets (Avinet Inc, Portland, Maine) placed in foraging corridors and over water.

Captured bats were identified to species, age, sex, and reproductive condition. Right forearm length and weight were also recorded. Bats weighing >6.6 g were fitted with

0.33 g radio-transmitters (model LB-2 and LB-2TX, Holohil Systems Ltd., Ontario,

Canada) using surgical adhesive (Osto-bond, Quebec, Canada) placed between the scapulae. Bats were radio-tracked to daytime roosting locations each day until transmitter battery failure or the transmitter was shed by the bat. We recorded the GPS coordinates of each roost using a handheld GPS unit with 3-m accuracy and collected a suite of habitat measurements (Chapter 2).

Tsk data were collected using a datalogging receiver programmed to record transmitter pulse rates at 6-min intervals (SRX800-D VHF Radio Receiver, Lotek

Wireless, Ontario, Canada). Each transmitter was calibrated by the manufacturer, who provided a unique polynomial equation allowing pulse rates to be converted into Tsk.

Tsk has been shown to be a close approximation of body temperature for small mammals such as bats and is widely used in research (Besler and Broder 2019, Willis et al. 2004). We elected not to define torpid versus euthermic temperatures using a Tsk 48

cutoff, or threshold, value because the transition to torpor is difficult to recognize using temperature (Boyles et al. 2011, Brigham et al. 2011, Johnson and Lacki 2014).

Instead, we use minimum Tsk as a measure of torpor depth, and average Tsk as a course measure of an individual’s thermoregulation during the entire day. We only summarized average and minimum Tsk for days where these data were successfully measured and recorded between sunrise and sunset. For bats where Tsk data were collected during the night, we limited our summaries to include only data collected from a half hour before sunrise to a half hour after sunset. An exception to this is when bats were in an obvious torpor bout (Tsk < 30° C) that began during the night and lasted into the day. In these instances, Tsk data for that day were summarized from the start of the nocturnal torpor bout.

Roost Microclimates

To investigate roost microclimates, we deployed temperature dataloggers

(HOBO MX2202 and HOBO Pendant Series, Onset Computer Corp., Bourne,

Massachusetts) in the canopies of 25 female roost trees and 15 random trees at Crane

Hollow. Dataloggers were affixed to high test fishing line and launched over branches in the canopy using a hunting slingshot. The line was then tied to stable vegetation for later removal. To the extent possible, dataloggers were placed in the outer canopies of trees in order to replicate areas where red bats have been observed roosting. Random trees were selected using a random point generator in ArcMap (version 10.5.1, ESRI,

Redlands, CA) and evenly distributed across different habitat types at the preserve.

Due to limited equipment, we did not sample temperature of male roosts, and limited 49

data collection to Crane Hollow. Dataloggers were programmed to record canopy temperature (Tcan) at 10-min intervals. Although dataloggers remained in the tree for at least a year, we only summarized Tcan data collected during the summer (June–

August) of 2018. We further limited these data to temperatures collected during daytime hours to reflect the temperatures while day-roosting, and calculated minimum and average daytime Tcan for each day.

Data Analyses

We used linear mixed-effect models (LMEs) and a model selection approach to test our hypothesis that males would exhibit lower minimum and average Tsk than females. Each day of Tsk data was a unique datapoint in this analysis, with the elevation of a bat’s day-roost, the bat’s sex, minimum ambient temperature (Ta min) and maximum ambient temperature (Ta max) as independent variables. We used the ambient temperature collected from the environmental datalogger located on ridgetops (Crane

Hollow: 334 m; Sunday Creek: 323m), using the temperature that corresponds to the study area that each bat inhabited. These independent variables were used to create a set of 13 a priori models, all of which used bat identity as a random factor (Tables 3.1 and 3.2). Female reproductive class was not included as an independent variable due to low sample sizes for each group, and because an evaluation of minimum and average Tsk of pregnant, lactating, and post-lactating females showed similar trends for each group. Prior to running the models, all fixed factors were standardized and were checked for multicollinearity using variance inflation factors (VIFs). VIFs were < 3 in all cases, suggesting that no multicollinearity existed between fixed variables. 50

Our LMEs were created using the lmer function in the lme4 package (Bates et al. 2014) in R (RStudio Team 2016). Models were ranked using AICc (Akaike’s information criterion with correction for small sample sizes) in order to determine the models of best fit and models with a ΔAICc < 2 were considered a candidate model.

To determine the amount of variation explained by each model and quantify model fit, we calculated both the marginal R² and conditional R² using the r squared GLMM function in the MuMIn package (Barton, 2018). To assess the which variables were most positively affecting the models, parameter coefficients and t-values were assessed. The number of degrees of freedom were determined using the Satterthwaite approximation.

We also employed an LME analysis and model selection approach to determine what factors influenced canopy microclimate (Tcan). Similar to Tsk, all variables were standardized before analyses and tested for multicollinearity using

VIFs. VIF’s were < 3 in all cases, indicating that there was no multicollinearity between our variables. We used all combinations of our independent variables, which included tree type (roost versus random) and tree elevation, to create a set of 4 candidate models (Table 3.3). In all models, tree identity was used as a random effect.

All models were ranked using AICc and models with a ΔAICc < 2 were considered a candidate model. Variation best explained by each model was calculated using both the marginal R² and conditional R². Similar to the skin temperature models, to assess the which variables were most positively affecting the models, parameter coefficients 51

and t-values were assessed. The number of degrees of freedom were determined using the Satterthwaite approximation.

Results

Capture and Temperature Sensitive Radio Telemetry

We radio-tagged 54 eastern red bats from 2016–2019, including 40 bats tagged with temperature-sensitive transmitters (n = 19 males; n = 17 females). We collected

107 days (males: n = 42, females: n = 65) of usable Tsk information from 9 males and

10 females (6 pregnant, 3 lactating, and 1 post-lactating) for analysis. The daily average Tsk of females was similar to males (females: 푥̅ = 32.1 ± 0.46 males: 푥̅ = 30.8

± 0.52; Figure 3.1.), as was daily minimum Tsk (females: 푥̅ = 24.6 ± 0.75 males: 푥̅ =

23.2 ± 1.01; Figure 3.2.). Both sexes used deep torpor bouts throughout the summer, with Tsk measured as low as 11° C for females (Figure 3.3a) and 14° C for males

(Figure 3.3b). Tsk fell below 25° C on 19 days (68 %) in males and on 30 days in females (56%).

Two LME models were competitive in explaining average Tsk (Table 3.1).

Model 1 contained elevation (β = 1.0, SE = 0.545, t(17.32) = 1.86, p = 0.08), Ta min (β =

0.99, SE = 0.33, t(97.7) = 3.004, p < 0.05), and Ta max (β = 2.1, SE = 0.35, t(106.6) = 6.11, p

< 0.05) as predictor variables. These variables explained approximately 42% of the variation in average Tsk (marginal R²: 0.42) and the overall model (including the random effect of individual) explained 66% of the variation in average Tsk (conditional

R²: 0.66). Model 2 contained Ta max (β = 2.2, SE = 0.35, t(105.2) = 6.20, p < 0.05) and Ta min (β = 0.94, SE = 0.33, t(99.9) = 2.86, p = 0.0052). These variables explained 41% of 52

variation in average Tsk (marginal R²: 0.41) and the overall model explained 69% of the variation in average Tsk (conditional R²: 0.69). While these models differ slightly, parameter coefficients and t-values of the fixed effects in both indicate that Ta max best explained average Tsk (Table 3.1).

There were three competing top models for minimum Tsk (Table 3.2). Model 1 contained Ta max (β = 1.9, S.E.= 0.62, t(105.2) = 2.85, p = 0.005) and Ta min (β = 3.6, SE =

0.57, t(99.7) = 6.33, p < 0.05). These two variables explained 40% of the variation in minimum Tsk (marginal R²: 0.40) and the overall model explained 68% of the variation

(conditional R²: 0.68). Model 2 contained elevation (β = 1.2, SE = 1.0, t(16.8) = 1.21, p

= 0.24), Ta min (β = 3.7, SE = 0.59, t(96.5) = 6.39, p <0.05), and Ta max (β = 1.7, SE= 0.62, t(105.9) = 2.81, p = 0.006). These variables explained approximately 38% of the variation in average Tsk (marginal R²: 0.38) and the overall model explained 66% of the variation in average Tsk (conditional R²: 0.67). The last competing model contained sex (β = -1.4, SE = 2.1, t(15.5) = -0.672, p = 0.51), Ta max (β= 3.6, SE = 0.57, t(105.4) =

2.849, p = 0.005), and Ta min (β= 1.9, SE = 0.62, t(99.6) = 6.313, p < 0.05). These variables explained approximately 39% of the variation (marginal R²: 0.39) and the overall model explained 67% of the variation (conditional R²: 0.67). While these models differed over inclusion of sex and elevation, Ta min and Ta max were in all competing models, and the parameter coefficients and t-values indicated that Ta min best explained variation in minimum Tsk.

Roost Microclimates

From our 25 female roost trees and 15 random trees, we recorded and analyzed

2,523 days of Tcan. Two models competed for explaining variation in Tcan (Table 3.3).

Model 1 included only elevation (β = 0.13, S.E.= 0.07, t(29.4) = 2.37, p < 0.05), explaining only 2% of the variation in roost microclimate (marginal R²: 0.02), and with the overall model accounting for 9.3% of the variation (conditional R²: 0.093).

Model 2 included elevation (β = 0.11, S.E.= 0.06, t(29.5) = 1.5, p = 0.14) and tree type

(β = 0.04, SE= 0.15, t(29.5) = 0.27, p = 0.79). These variables explained 2% of the variation in TR (marginal R²: 0.02) and the overall model accounted for 9.3% of the variation (conditional R²: 0.09). Comparison of these models show that while neither variable explained much of the variation in the models, elevation had a consistent positive effect on Tcan (Table 3.3, Figure 3.4).

Discussion

Contrary to our hypothesis, and with most studies comparing male to reproductive female bats, we found that sex drove little of the observed variation in average and minimum Tsk in eastern red bats. Instead, we found that daily temperature patterns influenced thermoregulation similarly in both sexes. Specifically, daytime minimum temperature had the most influence on minimum Tsk, while daytime maximum temperature had the most influence on average Tsk. Although we did not define Tsk as torpid versus euthermic, Tsk frequently fell below 25 C, indicating that both sexes frequently used torpor. Use of torpor was reduced, and was shallower when used, on days with warmer daytime temperatures and when bats roosted at higher 54

elevations. These results were consistent with data from our roost microclimate dataloggers, which found that tree elevation most influenced Tcan. Overall, these results suggest that both sexes employ torpor at similar depths and frequencies throughout the summer regardless of reproductive condition.

In other species, differences in thermoregulatory strategies between sexes are commonly reported, although there have been no field studies comparing thermoregulation between male and female Lasiurines, and no studies of thermoregulation in red bats until now. Cryan and Wolf (2003) investigated thermoregulation of male and female hoary bats (Lasiurus cinereus) captured and held in metabolic chambers during their spring migration, finding that females more often defended normothermy when exposed to cold temperatures than males. Field studies of other species have met with similar conclusions, finding that males are less likely to defend normothermy than reproductive females (Grinevitch et al. 1995, Johnson and

Lacki 2014). However, our results show that male and reproductive female eastern red bats used torpor similarly during periodically occurring cool summer temperatures.

Recent research has made it apparent that females often use torpor during gestation and lactation, with several studies finding females are less likely to do so during pregnancy (Dzal and Brigham 2013, Johnson and Lacki 2014, Rintoul and

Brigham 2014). Although we did not include reproductive status of females as a variable in our analyses due to low sample size, we documented numerous examples of prolonged torpor in pregnant females, including one pregnant female that maintained a Tsk ≤ 25° C for 31 hours (figure 3.5). In addition, the lowest Tsk we 55

recorded was that of a pregnant female (11° C). Other Lasiurine species have been found to use deep torpor during pregnancy and lactation, and are thought to be under more thermal stress due to the exposed nature of their roosts compared to cavity- or bark-roosting species (Klug and Barclay 2013, Klug et al. 2012, Willis and Bigham

2005). Willis and Brigham (2005) hypothesized that pregnant hoary bats benefit from extended torpor bouts during cold springs, as torpor would delay gestation until conditions are more favorable, increasing the likelihood of offspring and mother survival. Our data emphasize that not only do Lasiurines use extended torpor during early pregnancy but continue to do so during late pregnancy and lactation.

The solitary, foliage roosting nature of eastern red bats likely requires a different thermoregulatory strategy than what is used by cavity roosting bats, and may explain why females use similar thermoregulatory strategies as males in this species.

Microclimate of roosting locations have been found to be important for both mother bats and their offspring (Klug et al. 2012, Johnson et al. 2014, Sedgeley 2001). Tree foliage is exposed to rain, wind, and other inclement weather, so we would expect that foliage more exposed to ambient temperature fluctuations than other roost types commonly used by other bat species. It is therefore unsurprising that we found Ta min and Ta max highly influenced Tsk for both sexes of eastern red bats. However, contrary to our hypothesis, whether or not a tree was a roost or a randomly sampled tree did little to explain variation in roost microclimates. Other studies have found that roosting locations were more thermally stable than a random point selected in the surrounding foliage (Hutchinson and Lacki 2001). However, random locations 56

selected by Hutchinson and Lacki (2001) were recorded within the same tree as the roost, whereas we sampled distinct random trees. More similar to our study, Willis and

Brigham (2005) compared microclimates in random trees to roosts of hoary bats and found no difference in temperatures. This lack of difference is an indication that describing trees as “used” or “unused” was not a fine enough measure to understand variation in microclimates in our study. Other variables, such as solar or wind exposure, may be greater contributors to variation, but were not sampled in our study.

In addition, we only compared daytime temperatures between our roost and random trees. However, based on our Tsk data, we know that bats were using roosts at night.

Therefore, nighttime temperatures may need to be included in the analysis to detect differences between the roosts and random microclimates.

Even though we did not see a temperature difference between roost and random trees, elevation consistently appeared in our models explaining microclimate differences. Even though this variable did not explain much of the total variation, trees at higher elevations were warmer and therefore offer opportunities for energy savings for roosting bats while normothermic. In Chapter 2, our results indicated that male and female eastern red bats selected roosts differently at Crane Hollow, with females being more influenced by higher elevations, where temperatures were warmer, than males.

The results of the present study link ambient air temperatures along that elevational gradient to roost microclimates. Females preference for roosting at elevations with warmer ambient temperatures has been documented in several species, although lower elevations are often associated warmer conditions (Angell et al. 2013, Cryan et al 57

2000, Encarnação et al 2012). Although our model-selection approach suggests that sex explains little of the variation in Tsk, it is possible that some of the variation between sexes was attributable to roosting at different elevations at Crane Hollow.

These results show that roosting ecology has significant effects on the thermoregulatory strategies, and that patterns of thermoregulation are not consistent across species. We provide evidence that foliage roosting species deal with a different set of challenges than other tree roosting bats and use torpor differently to cope with those stressors. These results also highlight the importance that microclimates can have in aiding conservation efforts. We have documented that warmer elevations within forests have an important impact on habitat selection and physiological response to thermal stress of eastern red bats. In southeastern Ohio, higher elevations are associated with warmer temperatures, but they are also areas that areas that are more subject to human disturbance. Therefore, protecting forested habitats at warmer elevations, will be important to the continue the success of this species and potentially other species of foliage roosting bats.

Tables and Figures

Table 3.1 Rankings of linear mixed effect models of average skin temperatures of eastern red bats in response to elevation, daytime minimum temperature (Ta min), daytime maximum temperature (Ta min) in southeastern Ohio. Model Number and Variables K AICc ∆i wi 1) Elevation + Ta min + Ta max 6 535.63 0 0.40 2) Ta max + Ta min 5 536.46 0.83 0.27 3) Sex + Ta min + Ta max 6 537.66 2.03 0.15 4) Sex + Elevation + Ta min + Ta max 7 537.87 2.24 0.13 5) Ta min 4 541.62 5.99 0.02 6) Ta max + Elevation 5 541.76 6.13 0.02 7) Ta max + Sex 5 542.62 6.99 0.01 8) Elevation + Ta min 5 562.45 26.82 0.00 9) Ta min 4 565.08 29.45 0.00 10) Ta min + Sex 5 566.03 30.40 0.00 11) Null 3 580.47 44.84 0.00 12) Elevation 4 580.84 45.71 0.00 13) Sex 4 581.34 46.23 0.00 14) Sex + Elevation 5 582.69 47.06 0.00

Table 3.2 Rankings of linear mixed effect models of minimum skin temperatures of eastern red bats in response to elevation, daytime minimum temperature (Ta min), daytime maximum temperature (Ta max) in southeastern Ohio. Model Number and Variables K AICc ∆i wi 1) Ta max + Ta min 5 655.50 0 0.41 2) Elevation + Ta min + Ta max 6 656.37 0.87 0.27 3) Sex + Ta max + Ta min 6 657.31 1.81 0.17 4) Sex + Elevation + Ta min + Ta max 7 658.65 3.15 0.09 5) Ta min 4 660.80 5.30 0.03 6) Elevation + Ta min 5 661.28 5.78 0.02 7) Ta min + Sex 5 662.54 7.04 0.01 8) Ta max 4 684.02 28.52 0.00 9) Ta max + Sex 5 685.61 30.11 0.00 10) Ta max + Elevation 5 686.10 30.60 0.00 11) Null 3 699.17 43.68 0.00 12) Sex 4 700.80 45.30 0.00 13) Elevation 4 701.32 45.82 0.00 14) Sex + Elevation 5 702.91 47.41 0.00

Table 3.3 Rankings of linear mixed effect models of canopy temperatures in response to elevation and roost/random in southeastern Ohio. Model K AICc ∆i wi 1) Elevation 4 8708.85 0 0.52 2) Elevation + Roost/Random 5 8710.78 1.93 0.20 3) Roost/Random 4 8710.95 2.11 0.18 4) Null 3 8711.98 3.14 0.11

Figure 3.1 Male and female eastern red bats exhibit similar average skin temperatures as measured by temperature-sensitive radio-telemetry in southeastern Ohio between May and August 2017–2019

Figure 3.2 Male and female eastern red bats exhibit similar minimum skin temperatures, as measured by temperature-sensitive radio-telemetry in southeastern Ohio between May and August 2017–2019

(a)

(b)

Figure 3.3 Skin temperatures (dark circles) of a pregnant female (a) and male (b) eastern red bat compared to ambient temperature (grey line) collected in southeastern Ohio. Shaded areas represent nighttime hours.

Figure 3.4 Average daytime summer temperature and elevation of female roost and random trees that contained microclimate dataloggers at Crane Hollow Nature Preserve in southeast Ohio.

30 C) °

20 Temperature Temperature (

0 6/20/18 0:00 6/20/18 12:00 6/21/18 0:00 6/21/18 12:00 6/22/18 0:00 6/22/18 12:00 6/23/18 0:00 Figure 3.5 Skin temperature trace of a pregnant eastern red bat using extended and deep bouts of torpor in southeastern Ohio.

LITERATURE CITED

Angell, R. L., Butlin, R. K., and Altringham, J. D. (2013). Sexual Segregation and Flexible Mating Patterns in Temperate Bats. PLOS ONE, 8(1), e54194. Arnett, E. B., Brown, W. K., Erickson, W. P., Fiedler, J. K., Hamilton, B. L., Henry, T. H., …Tankersley, R. D. (2008). Patterns of Bat Fatalities at Wind Energy Facilities in North America. Journal of Wildlife Management, 72(1), 61–78. Arnett, E. B., and Baerwald, E. F. (2013). Impacts of Wind Energy Development on Bats: Implications for Conservation. In R. A. Adams & S. C. Pedersen (Eds.), Bat Evolution, Ecology, and Conservation (pp. 435–456). New York, NY: Springer. Balmford, A., Green, R. E., and Jenkins, M. (2003). Measuring the changing state of nature. Trends in Ecology & Evolution, 18(7), 326–330. Barclay, R. M., Kurta, A. (2007). Ecology and behavior of bats roosting in tree cavities and under bark. In Lacki, M. J., Hayes, J. P., Kurta, A., (Eds.), Bats in forests: conservation and management (pp. 17–59). Baltimore, Maryland: Johns Hopkins University Press. Betts, M. G., Fahrig, L., Hadley, A. S., Halstead, K. E., Bowman, J., Robinson, W. D., Wiens, J.A., Lindenmayer, D. B. (2014). A species-centered approach for uncovering generalities in organism responses to habitat loss and fragmentation. Ecography, 37(6), 517–527. Becker, N. I., Tschapka, M., Kalko, E. K. V., & Encarnação, J. A. (2013). Balancing the Energy Budget in Free-Ranging Male Myotis daubentonii Bats. Physiological and Biochemical Zoology, 86(3), 361–369. Besler, N. K., and Broders, H. G. (2019). Combinations of reproductive, individual, and weather effects best explain torpor patterns among female little brown bats (Myotis lucifugus). Ecology and Evolution, 9(9), 5158–5171. Boyles, J. G., Smit, B., and McKechnie, A. E. (2011). A New Comparative Metric for Estimating Heterothermy in Endotherms. Physiological and Biochemical Zoology, 84(1), 115–123. Brigham, R. M. (2007) Bats in Forests: What we know and what we need to learn. In Lacki, M. L., Hayes, J. P., Kurta A (Eds.) Bats in forests: conservation and management (pp. 1-15). Baltimore, Maryland: John Hopkins University Press. Carter, T. C., M. A., Menzel, and D. A. Saugey. (2003). Population trends of solitary foliage-roosting bats. In T. J. O’Shea and M. A. Bogan (Eds.), Monitoring trends in bat populations of the United States and territories: problems and prospects (pp. 41–47). U.S. Geological Survey, Biological Resources Discipline, Information and Technology Report, USGS/BRD/ITR–2003–0003: 1–274. Carter, C. C., and Menzel, J. M., (2007). Behavior and Day-Roosting Ecology of North American Foliage-Roosting Bats. In Lacki, M. J., Hayes, J. P., & Kurta, A. (Eds.), Bats in Forests Conservation and Management (pp. 60-81). Baltimore, Maryland: The John Hopkins University Press. Coovert, G.A. (2015). Crane Hollow: all taxa biodiversity inventory, in possession of 66

Crane Hollow, Inc. Cryan, P. M., Bogan, M. A., and Altenbach, J. S. (2000). Effect of Elevation on Distribution of Female Bats in the Black Hills, South Dakota. Journal of Mammalogy, 81(3), 719–725. Cryan, P. M., and Wolf, B. O. (2003). Sex differences in the thermoregulation and evaporative water loss of a heterothermic bat, Lasiurus cinereus, during its spring migration. Journal of Experimental Biology, 206(19), 3381–3390. Duchamp, J. E., Arnett, E. B., Larson, A. M., et al (2007). Ecological considerations for landscape-level management of bats. In Lacki, M. J., Hayes, J. P., Kurta, A., (Eds.), Bats in forests, Conservation and management (pp. 237–261) Baltimore, Maryland: The John Hopkins University Press. Duchamp, J. E., and Swihart, R. K. (2008). Shifts in bat community structure related to evolved traits and features of human-altered landscapes. Landscape Ecology, 23(7), 849–860. Dunbar, M. B., and Tomasi, T. E. (2006). Arousal Patterns, Metabolic Rate, and an Energy Budget of Eastern Red Bats (Lasiurus borealis) in Winter. Journal of Mammalogy, 87(6), 1096–1102. Dzal, Y. A., and Brigham, R. M. (2013). The tradeoff between torpor use and reproduction in little brown bats (Myotis lucifugus). Journal of Comparative Physiology B, 183(2), 279–288. Elmore, L. W., Miller, D. A., and Vilella, F. J. (2004). Selection of diurnal roosts by red bats (Lasiurus borealis) in an intensively managed pine forest in Mississippi. Forest Ecology and Management, 199(1), 11–20. Encarnação, J. A., Kierdorf, U., Holweg, D., Jasnoch, U., and Wolters, V. (2005). Sex- related differences in roost-site selection by Daubenton’s bats Myotis daubentonii during the nursery period. Mammal Review, 35(3–4), 285–294. Encarnação, J. A. (2012). Spatiotemporal pattern of local sexual segregation in a tree- dwelling temperate bat Myotis daubentonii. Journal of Ethology, 30(2), 271– 278. Fabianek, F., Simard, M. A., and Desrochers, A. (2015). Exploring Regional Variation in Roost Selection by Bats: Evidence from a Meta-Analysis. PLOS ONE, 10(9), e0139126. Frick, W. F., Reynolds, D. S., and Kunz, T. H. (2010). Influence of climate and reproductive timing on demography of little brown myotis Myotis lucifugus. Journal of Animal Ecology, 79(1), 128–136. Frick W. F., Puechmaille, S. J., and Willis, C. K. (2016) White-nose syndrome in bats. In Voigt, C.C., Kingston, T. (Eds.), Bats in the Anthropocene: Conservation of Bats in a Changing World (pp. 245–262). Cham, Switzerland: Springer. Frick, W. F., Baerwald, E. F., Pollock, J. F., Barclay, R. M. R., Szymanski, J. A., Weller, T. J., … McGuire, L. P. (2017). Fatalities at wind turbines may threaten population viability of a migratory bat. Biological Conservation, 209, 172–177. Geiser, F. (2004). Metabolic Rate and Body Temperature Reduction During Hibernation and Daily Torpor. Annual Review of Physiology, 66(1), 239–274. 67

Geiser, F., and Brigham, R. M. (2012). The Other Functions of Torpor. In T. Ruf, C. Bieber, Arnold, W. and Millesi, E. (Eds.), Living in a Seasonal World: Thermoregulatory and Metabolic Adaptations (pp. 109–121). Berlin, Heidelberg: Springer. Grinevitch, L., Holroyd, S. L., and Barclay, R. M. R. (1995). Sex differences in the use of daily torpor and foraging time by big brown bats (Eptesicus fuscus) during the reproductive season. Journal of Zoology, 235(2), 301–309. Hayes, J. P, and Loeb, S. C. (2007) The influences of forest management on bats in North America. In Lacki, M. L., Hayes, J. P., and Kurta, A. (Eds.), Bats in forests: conservation and management (pp. 207–235). Baltimore, Maryland: John Hopkins University Press. Hayes, J. P., and Waldien, D. L. (2001). Activity areas of female long-eared myotis in coniferous forests in Western Oregon, 64(3), 785-796. Hutchinson, J. T., and Lacki, M. J. (2000). Selection of day roosts by red bats in mixed mesophytic forests. The Journal of Wildlife Management, 64(1), 87– 94. Hutchinson J. T., and Lacki, M. J. (2001). Possible microclimate benefits of roost site selection in the red bat, Lasiurus Borealis, in mixed mesophytic forests of Kentucky. The Canadian Field Naturalist, 115(2), 205-209. Johnson, J. S. and Lacki, M. J. (2012). Summer heterothermy in Rafinesque’s big- eared bats (Corynorhinus rafinesquii) roosting in tree cavities in bottomland hardwood forests. Journal of Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology, 183, 709-721. Johnson, J. S., and Lacki, M. J. (2013). Habitat associations of Rafinesque’s big-eared bats (Corynorhinus rafinesquii) and their lepidopteran prey in bottomland hardwood forests. Canadian Journal of Zoology, 91(2), 94–101. Johnson, J. S. and Lacki, M. J. (2014). Effects of reproductive condition, roost microclimate, and weather patterns on summer torpor use by a vespertilionid bat. Ecology and Evolution, 4(2),157-166. Kalcounis‐Rüppell, M. C., Psyllakis, J. M., and Brigham, R. M. (2005). Tree roost selection by bats: An empirical synthesis using meta-analysis. Wildlife Society Bulletin, 33(3), 1123–1132. Klug, B. J., Goldsmith, D. A., and Barclay, R. M. R. (2012). Roost selection by the solitary, foliage-roosting hoary bat (Lasiurus cinereus) during lactation. Canadian Journal of Zoology, 90(3), 329–336. Kunz T. H. and Lumsden L. F. (2003). Ecology of cavity and foliage roosting bats. In Kunz T. H., and Fenton M. B. (Eds.), Bat ecology (pp. 3–89). Chicago, Illinois: University of Chicago Press. Kurta, A. (1985). External insulation available to a non-nesting mammal, the little brown bat (Myotis lucifugus). Comparative Biochemistry and Physiology Part A: Physiology, 82(2), 413–420. Kurta, A., Bell, G. P., Nagy, K. A., and Kunz, T. H. (1989). Energetics of Pregnancy and Lactation in Freeranging Little Brown Bats (Myotis lucifugus). Physiological Zoology, 62(3), 804–818. 68

Kurta, A., and Lehr, G. C. (1995). Lasiurus ega. Mammalian Species, (515), 1–7. Lacki, M. J., and Baker, M. D. (2003). A Prospective Power Analysis and Review of Habitat Characteristics Used in Studies of Tree-Roosting Bats. Acta Chiropterologica, 5(2), 199–208. Lacki, M.J., Hayes, J.P., and Kurta, A., (2007). Bats in Forests: Conservation and Management. Baltimore, Maryland: The Hopkins University Press. Lacki, M. J., Baker, M. D., and Johnson, J. S. (2010). Geographic Variation in Roost- Site Selection of Long-Legged Myotis in the Pacific Northwest. The Journal of Wildlife Management, 74(6), 1218–1228. Lacki, M. J., Johnson, J. S., and Baker, M. D. (2013). Temperatures beneath bark of dead trees used as roosts by Myotis volans in forests of the Pacific Northwest, USA. Acta Chiropterologica, 15(1): 143-151. Loeb, S. C., and O’Keefe, J. M. (2006). Habitat Use by Forest Bats in South Carolina in Relation to Local, Stand, and Landscape Characteristics. Journal of Wildlife Management, 70(5), 1210–1218. Lumsden, L. F., Bennett, A. F., and Silins, J. E. (2002). Location of roosts of the lesser long-eared bat Nyctophilus geoffroyi and Gould’s wattled bat Chalinolobus gouldii in a fragmented landscape in south-eastern Australia. Biological Conservation, 106(2), 237–249. Mayor, S. J., Schneider, D. C., Schaefer, J. A., and Mahoney, S. P. (2009). Habitat selection at multiple scales. Écoscience, 16(2), 238–247. MacCarthy, K. A., Carter, T. C., Steffen, B. J., and Feldhamer, G. A. (2006). Efficacy of the Mist-net Protocol for Indiana Bats: A Video Analysis. Northeastern Naturalist, 13(1), 25–28. Mager, K. J., and Nelson, T. A. (2001). Roost-site Selection by Eastern Red Bats (Lasiurus borealis). The American Midland Naturalist, 145(1), 120–126. Menzel, M. A., Carter, T. C., Chapman, B. R., and Laerm, J. (1998). Quantitative comparison of tree roosts used by red bats (Lasiurus borealis) and Seminole bats (L. seminolus). Canadian Journal of Zoology, 76(4), 630–634. Miller, D., Arnett, E., and Lacki, M. J. (2003). Habitat Management for Forest- Roosting Bats of North America: A Critical Review of Habitat Studies. Wildlife Society Bulletin, 31(1), 30-44. Morrison, M. L., Marcot, B., and Mannan, W. (2012). Wildlife-Habitat Relationships: Concepts and Applications. Island Press. O’Keefe, J. M., Loeb, S. C., Lanham, J. D., and Hill, H. S. (2009). Macrohabitat factors affect day roost selection by eastern red bats and eastern pipistrelles in the southern Appalachian Mountains, USA. Forest Ecology and Management, 257(8), 1757–1763. Olson, C. R., and Barclay, R. M. R. (2013). Concurrent changes in group size and roost use by reproductive female little brown bats (Myotis lucifugus). Canadian Journal of Zoology, 91(3), 149–155. Pettit, J. L., and O’Keefe, J. M. (2017). Impacts of White-Nose Syndrome Observed During Long-Term Monitoring of a Midwestern Bat Community. Journal of Fish and Wildlife Management, 8(1), 69–78. 69

Perry, R. W., Thill, R. E., and Carter, S. A. (2007). Sex-specific roost selection by adult red bats in a diverse forested landscape. Forest Ecology and Management, 253(1), 48–55. Pretzlaff, I., Kerth, G., and Dausmann, K. H. (2010). Communally breeding bats use physiological and behavioural adjustments to optimise daily energy expenditure. Naturwissenschaften, 97(4), 353–363. Racey, P. A., and Swift, S. M. (1981). Variations in gestation length in a colony of pipistrelle bats (Pipistrellus pipistrellus) from year to year. Reproduction, 61(1), 123–129. Ransome, R. D. (1989). Population changes of Greater horseshoe bats studied near Bristol over the past twenty-six years. Biological Journal of the Linnean Society, 38(1), 71–82. Rawat, U. S., and Agarwal, N. K. (2015). Biodiversity: Concepts, threats, and conservation. Environment Conservation Journal, 16(3), 19-28. Rintoul, J. L. P., and Brigham, R. M. (2014). The influence of reproductive condition and concurrent environmental factors on torpor and foraging patterns in female big brown bats (Eptesicus fuscus). Journal of Comparative Physiology B, 184(6), 777–787. Ruhl, P., Kellner, K., Pierce, J., Riegel, J., Swihart, R., Saunders, M., and Dunning, J. (2018). Characterization of Worm-eating Warbler (Helmitheros vermivorum) breeding habitat at the landscape level and nest scale. Avian Conservation and Ecology, 13(1). Russo, D., Cistrone, L., Garonna, A. P., and Jones, G. (2010). Reconsidering the importance of harvested forests for the conservation of tree-dwelling bats. Biodiversity and Conservation, 19(9), 2501–2515. Sedgeley, J. A. (2001). Quality of cavity microclimate as a factor influencing selection of maternity roosts by a tree-dwelling bat, Chalinolobus tuberculatus, in New Zealand. Journal of Applied Ecology, 38(2), 425–438. Shump, K. A., and Shump, A. U. (1980). Comparative insulation in vespertilionid bats. Comparative Biochemistry and Physiology Part A: Physiology, 66(2), 351–354. Shump, K. A., and Shump, A. U. (1982). Lasiurus borealis. Mammalian Species, 183: 1–6. Speakman, J. R. (1999). The Cost of Living: Field Metabolic Rates of Small Mammals. In A. H. Fitter & D. G. Raffaelli (Eds.), Advances in Ecological Research, 30:177–297. Waldien, D. L., Hayes, J. P., and Arnett, E. B. (2000). Day-roosts of female long- eared myotis in western Oregon. Journal of Wildlife Management, 64:785-796. Wilde, C. J., Knight, C. H., and Racey, P. A. (1999). Influence of torpor on milk protein composition and secretion in lactating bats. J Exp Zool 284:35–41 Whittemore, M. (2016). Mapping and Assessment of Plant Communities at Crane Hollow Nature Preserve in Hocking County, Ohio (Unpublished Master’s Thesis). Ohio University, Athens, USA. Willis, C. K. R., and Brigham, R. M. R. (2003). Defining torpor in free-ranging bats: 70

Experimental evaluation of external temperature-sensitive radiotransmitters and the concept of active temperature. Journal of Comparative Physiology B, 173(5), 379–389. Willis, C. K. R., Brigham, R. M. R., and Geiser, F. (2006). Deep, prolonged torpor by pregnant, freeranging bats. Naturwissenschaften, 93(2), 80–83. Willis, C. K. R., and Brigham, R. M. R. (2007). Social thermoregulation exerts more influence than microclimate on forest roost preferences by a cavity-dwelling bat. Behavioral Ecology and Sociobiology, 62(1), 97–108. Winhold, L. A., Kurta, A., and Foster, R. (2008). Long-term change in an assemblage of North American bats: are eastern red bats declining? Acta Chiropterologica 10(2), 359–366. Vonhof, M. J., and Barclay, R. M. R. (1996). Roost-site selection and roosting ecology of forest-dwelling bats in southern British Columbia. Canadian Journal of Zoology, 74(10), 1797–1805. Voigt, C. C., and Kingston, T. (Eds.). (2016). Bats in the Anthropocene: Conservation of Bats in a Changing World. Cham, Switzerland: Springer.

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

Thesis and Dissertation Services ! !