PHYLOGENETIC AND MORPHOLOGICAL COMMUNITY …myweb.ttu.edu/richstev/pubs/Dissertations/PatrickLoreleiDissertation... · Dr. Luis Ruedas at the Portland State University Museum of Vertebrate

PHYLOGENETIC AND MORPHOLOGICAL COMMUNITY STRUCTURE OF

NORTH AMERICAN DESERT BATS

A Dissertation

Submitted to the Graduate Faculty of the

Louisiana State University and

Agricultural and Mechanical College

in partial fulfillment of the

requirements for the degree of

Doctor of Philosophy

in

The Department of Biological Sciences

by

Lorelei E. Patrick

B.S., Portland State University, 2003

M.S., Portland State University, 2007

August 2014

ii

ACKNOWLEDGEMENTS

I thank my advisor, Dr. Richard Stevens, for his guidance, suggestions, support, and

encouragement throughout my doctoral program. I also thank Drs. Kyle Harms, Fred Sheldon,

Michael Hellberg, and Alan Afton for serving on my committee and suggesting productive

research directions.

Several individuals and many museums provided data, without which my work would

have been severely hampered. Michael O’Farrell, Stacy Mantooth, Aimee Hart, and Jason

Williams generously shared previously unpublished bat capture and collection data. Angelo State

Natural History Collections, Museum of Southwestern Biology, Museum of Texas Tech

University, Louisiana State University Museum of Natural Science, Natural History Museum of

Los Angeles County, Texas Cooperative Wildlife Collection, and Portland State University

Museum of Vertebrate Biology provided tissues for genetic analyses. Dr. Mark Hafner at the

Louisiana State University Museum of Natural Science, Jeffrey Bradley at the Burke Museum,

Dr. Luis Ruedas at the Portland State University Museum of Vertebrate Biology, Dr. Joseph

Cook and Cindy Ramotnik at the Museum of Southwestern Biology, Dr. Robert Timm at the

University of Kansas, and Dr. Jim Dines at the Natural History Museum of Los Angeles County

allowed access to the collections under their care so that I could complete the morphological

component of my research.

Several institutions funded my work, without which this research may have been

impossible: American Museum of Natural History Theodore Roosevelt Memorial Grant, Society

of Systematic Biologists Graduate Student Award, American Society of Mammalogists Grants-

in-Aid of Research award, Louisiana Environmental Education Commission Research Grant, and

Louisiana State University BioGrads award BG11-38.

iii

Many thanks to Aimee Hart, Dr. Laurie Dizney, and my mother Lorelei F. Patrick for

help conducting fieldwork. Dr. Bryan Carstens invited me to do molecular work in his lab. I

would also like to thank Prissy Milligan, Chimene Boyd, and Charyl Thompson for their

knowledgeable assistance in navigating university bureaucracy. Lynnmarie Patrick, Cindy

Ramotnik and Dr. Mike Bogan, Dr. Yadeeh Sawyer, Dr. Kelly Grussendorf, and Jeanne Harris

opened their homes, spare bedrooms, and couches to this traveling graduate student; without

their help this research may not have been possible. Dr. Kyle Harms, Metha Klock, Katherine

Hovanes, and Sandra Galeano provided desk space and an environment conducive to writing

during my final semester at LSU, thank you! Drs. J. Sebas Tello, Eve McCulloch, Meche

Gavilanez, Sarah Hird, Noah Reid, Verity Mathis, and Jeremy Brown provided useful advice and

helpful comments. Garret Langlois, Tara Pelletier, Cassie Black, Danielle Jellison, Adriana

Dantin, Dr. Melissa Debiasse, Dr. Molly Fischer, Dr. John Hogan, Jeff Corkern and many others

provided much needed support in a multitude of other ways.

Finally, I thank my partner and my parents. Paul Robinson has been beside me offering

encouragement every step of the way, ensured that I never took myself too seriously, and saw to

it that there has never been a dull moment in our lives. My parents, Chris and Lorelei Patrick,

have always supported and encouraged me; without them, I would not be the person I am today.

Although my choice to study bats did elicit some quizzical looks, they always pushed me to

follow my dreams and do whatever I wanted with my life.

iv

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ............................................................................................................ ii

ABSTRACT ................................................................................................................................... vi

CHAPTER 1 INTRODUCTION .................................................................................................... 1

REFERENCES ............................................................................................................................ 4

CHAPTER 2 INVESTIGATING SENSITIVITY OF PHYLOGENETIC COMMUNITY

STRUCTURE METRICS USING NORTH AMERICAN DESERT BATS ................................. 6

INTRODUCTION ....................................................................................................................... 6

METHODS ................................................................................................................................ 10

Phylogeny .............................................................................................................................. 10

Communities .......................................................................................................................... 11

Phylogenetic community structure metrics ............................................................................ 14

Impact of phylogenetic tree on community structure metrics ............................................... 15

RESULTS .................................................................................................................................. 18

Phylogeny .............................................................................................................................. 18

Community delimitation method and PCS ............................................................................ 18

Impact of phylogenetic tree on analyses ................................................................................ 23

DISCUSSION ............................................................................................................................ 25

Phylogeny .............................................................................................................................. 25

Impact of community delimitation method on PCS metrics .................................................. 26

Impact of phylogeny on PCS metrics .................................................................................... 27

Phylogenetic community structure of desert bat communities .............................................. 29

ACKNOWLEDGEMENTS ...................................................................................................... 33

REFERENCES .......................................................................................................................... 34

CHAPTER 3 PHYLOGENETIC COMMUNITY STRUCTURE OF NORTH AMERICAN

DESERT BATS: INFLUENCE OF ENVIRONMENT AND ECOLOGICAL TRAITS AT

MULTIPLE SPATIAL AND TAXONOMIC SCALES ............................................................... 40

INTRODUCTION ..................................................................................................................... 40

METHODS ................................................................................................................................ 44

Phylogeny and community data ............................................................................................. 44

Species pools .......................................................................................................................... 45

Phylogenetic community structure metrics ............................................................................ 46

Functional traits and environmental data ............................................................................... 46

RESULTS .................................................................................................................................. 47

Phylogenetic community structure ........................................................................................ 47

Functional traits and environmental data ............................................................................... 49

DISCUSSION ............................................................................................................................ 50

Functional traits and environmental characteristics ............................................................... 50

Spatial and taxonomic scale ................................................................................................... 52

Phylogenetic community structure of desert bat communities .............................................. 53

v

CONCLUSIONS ....................................................................................................................... 57


REFERENCES .......................................................................................................................... 58

CHAPTER 4 MORPHOLOGICAL COMMUNITY STRUCTURE OF NORTH AMERICAN

DESERT BATS: ASSESSING PHYLOGENETIC SIGNAL IN MORPHOLOGICAL

TRAITS AND COMPARISON WITH PHYLOGENETIC COMMUNITY STRUCTURE .... 63

INTRODUCTION ..................................................................................................................... 63

METHODS ................................................................................................................................ 66

Community data ..................................................................................................................... 66

Morphological traits ............................................................................................................... 67

Species pools .......................................................................................................................... 67

Data analyses ......................................................................................................................... 69

RESULTS .................................................................................................................................. 70

Morphological community structure...................................................................................... 70

Correlation between morphological and phylogenetic community structure ........................ 72

Phylogenetic signal in morphological traits ........................................................................... 73

DISCUSSION ............................................................................................................................ 73

Phylogenetic signal in morphological traits ........................................................................... 77

Correlation between morphological and phylogenetic community structure ........................ 79

Community structure of North American desert bats ............................................................ 79


REFERENCES .......................................................................................................................... 82

CHAPTER 5 SUMMARY ............................................................................................................ 87

REFERENCES .......................................................................................................................... 90

APPENDIX I SEQUENCES IN THE REGIONAL POOL PHYLOGENY ................................ 91

APPENDIX II SEQUENCES IN FULL PHYLOGENY ............................................................ 100

APPENDIX III CHAPTER 2 SUPPLEMENTARY MATERIALS ........................................... 104

APPENDIX IV CHAPTER 3 SUPPLEMENTARY MATERIALS........................................... 120

APPENDIX V SPECIMENS EXAMINED IN THE MORPHOLOGICAL STUDY ................ 129

APPENDIX VI CHAPTER 4 SUPPLEMENTARY MATTERIALS ........................................ 161

VITA ........................................................................................................................................... 182

vi

ABSTRACT

Patterns of community structure may be examined using phylogenetic and morphological data;

these patterns can then be used to infer the processes that gave rise to these patterns.

Communities made up of similar species may be structured by habitat filtering, wherein only

species with traits necessary to survive in a particular location are found there. Communities

made up of dissimilar species may have been structured by competition, which reduces overlap

in resource use. I examined the sensitivity of phylogenetic community structure (PCS) metrics

to changes in phylogeny and community delimitation method, investigated patterns of PCS and

correlation to environmental variables at multiple spatial and taxonomic scales, and assessed

whether morphological data gave results similar to phylogenetic data using North American

desert bats as a model system. I found that PCS metrics were robust to moderate changes to

phylogeny and that these metrics also trend in the same direction regardless of delimitation

method. Bat communities tended to be made up of species that were significantly more closely

related than expected by chance, or phylogenetically clustered, at large spatial and taxonomic

scales; this tendency towards clustering decreases with decreasing scale. Phylogenetically

clustered communities also tended to occur in harsher environmental conditions than more

overdispersed communities, or those made up of species not closely related. From a

morphological perspective, communities were made up of species that were morphologically

clustered or not significantly different from random. Morphological community structure was

positively correlated with PCS, indicating that these different datasets gave similar results. These

results indicate that North American desert bat communities are made up of phylogenetically and

morphologically similar species and that environmental variables such as temperature and

vii

seasonality may influence community structure. This suggests that habitat filtering is playing a

predominant role in structuring these communities.

1

CHAPTER 1

INTRODUCTION

Ecologists seek to describe, explain, and predict patterns in the abundance, distribution, and

diversity of organisms. Such investigations can range in scale from populations of a single

species to biomes of many species, living in a single sample of soil or spread across a continent.

Community ecology focuses at intermediate scales, concentrating on how species interact with

each other and the abiotic environment to understand patterns of coexistence in a single

community or across multiple communities. To understand these patterns, temporal, spatial, and

taxonomic extent of a community must be defined, characteristics of species living in that

community quantified, and environmental conditions of the community measured. Although

many mechanisms can determine which species can co-occur, historically, habitat filtering and

density-dependent interactions have been the two main processes ecologists study (e.g., Webb et

al. 2002, Cavender-Bares et al. 2009, Vamosi et al. 2009). Habitat filtering occurs when species

are found in a particular place because they are capable of surviving the environmental

conditions or using available resources there (Webb et al. 2002, Ackerly et al. 2006). This

process prevails most often at large spatial and taxonomic scales because habitat heterogeneity

allows species to be sorted by different habitat types which tends to lead to co-occurrence of

similar, or clustered, species (Swenson et al. 2007, Cavender-Bares et al. 2009, Gómez et al.

2010). Density-dependent interactions include predation, mutualism, and parasitism, but

competition for resources has been the focus of most studies. Competitive interactions lead

species in a community to be dissimilar, or overdispersed, to reduce overlap in resource use. This

process tends to be most prevalent at smaller scales (Swenson et al. 2007, Cavender-Bares et al.

2009, Gómez et al. 2010).

2

Species similarity can be measured in multiple ways. Traditionally, morphological or

functional traits or physiological tolerances have been measured and analyzed to determine how

phenotypically similar co-occurring species are to each other. A newer approach is to determine

how phylogenetically similar co-occurring species are and using this similarity to determine how

the community is structured (e.g., Webb et al. 2002, Cavender-Bares et al. 2009, Vamosi et al.

2009). This approach usually assumes that closely related species are also phenotypically similar

based on niche conservatism (Wiens and Graham 2005, Losos 2008). Structure metrics of a

community can be compared to randomly assembled communities to determine if the observed

community’s structure is significantly different from random. These metrics can also be

compared to environmental variables, such as temperature or precipitation, to investigate how

abiotic factors influence species occurrence in communities.

I investigate phylogenetic and morphological community structure of North American

desert bats at multiple spatial and taxonomic scales in the four desert regions in North America:

the Great Basin, Mojave, Sonoran, and Chihuahuan. Each desert hosts unique floras (Shreve

1942) and potentially unique faunas. Bats are classified in the order Chiroptera, the second most

speciose order of mammals after rodents (Simmons 2005). Bats provide ecosystem services

wherever they occur (Jones et al. 2009); in these desert regions they consume economically

important insects and pollinate several species of plants (Jones et al. 2009). There are five

families, 28 genera, and 55 species of bats found in these four deserts. Most of these species are

insectivorous (insectivores occur in all five families), but some nectarivores, sanguivores, a

piscivore, and a frugivore occur in the Sonoran or Chihuahuan deserts.

In Chapter 2, I investigate how changes to the data used to calculate phylogenetic

community structure metrics influence results and interpretation. Calculating these metrics

3

requires a list of species occurring in communities in the region of interest and a phylogeny of

these species. I collated bat capture and collection data from a variety of sources to determine

species membership in each community. I then inferred a phylogeny of the regional species pool

using sequences available on GenBank as well as sequences I generated. Surprisingly little work

has been done to quantify how changes in community membership data and phylogeny influence

the metrics calculated from them. Accordingly, I delimited communities of bats in six ways to

determine if the method of defining a community affects community structure metrics. I then

introduced random changes to the phylogeny to determine how this influences community

structure metrics. Finally, I describe the community structure of bats found in all deserts and in

individual deserts.

In Chapter 3, I examine the impact of spatial and taxonomic scale on phylogenetic

community structure metrics in greater depth; as described above, phylogenetic clustering is

expected at larger scales due to habitat filtering, while phylogenetic overdispersion (the opposite

of clustering) is expected at smaller scales due to comptetition or other density-dependent

interactions. To do this, I investigated community structure of the most speciose family in the

region, Vespertilionidae, and the most speciose genus, Myotis, across all deserts and in each

individual desert. I also used ecological trait data to determine if phylogenetic proximity is a

useful proxy for phenotypic similarity. In addition, I determined if community structure is

correlated with environmental variables. Clustered communities are expected in harsher

environments, whereas overdispersed communities are expected in more favorable conditions.

Finally, in Chapter 4 I examined morphological community structure and assessed

whether it is correlated with phylogenetic structure. I collected morphological data from museum

specimens of all bat species present in North American deserts. I used these data to determine if

4

communities are made up of morphologically similar or dissimilar species. I then determined if

each trait was evolutionarily conserved, convergent, or random for each taxonomic scale. If a

trait is evolutionarily conserved, close relatives will be morphologically similar to each other and

distantly related species will be morphologically dissimilar, resulting in a positive correlation

between phylogenetic and morphological distances. Conversely, if species have undergone

convergent evolution, there will be a negative correlation between morphological and

phylogenetic distances because distantly related species will be more similar morphologically

than closely related species. Finally, I determined if community structure based on morphology

and phylogeny were correlated. Positive correlation means morphology and phylogeny are

congruent, whereas negative correlation indicates the datasets are producing different patterns of

community structure.

In Chapters 2-4 I use the personal pronouns “we” and “our” to refer to myself and my

advisor, Richard Stevens. At the time this dissertation was submitted to the LSU Graduate

School, all three chapters were in preparation to be submitted to journals for publication.

REFERENCES

Ackerly, D. D., D. W. Schwilk, and C. O. Webb. 2006. Niche evolution and adaptive radiation:

testing the order of trait divergence. Ecology 87:50-61.

Cavender-Bares, J., K. H. Kozak, P. V. A. Fine, and S. W. Kembel. 2009. The merging of

community ecology and phylogenetic biology. Ecology Letters 12:693-715.

Gómez, J. P., G. A. Bravo, R. T. Brumfield, J. G. Tello, and C. D. Cadena. 2010. A phylogenetic

approach to disentangling the role of competition and habitat filtering in community

assembly of Neotropical forest birds. Journal of Animal Ecology 79:1181-1192.

Jones, G., D. S. Jacobs, T. H. Kunz, M. R. Willig, and P. A. Racey. 2009. Carpe noctem: the

importance of bats as bioindicators. Endangered Species Research 8:93-115.

5

Losos, J. B. 2008. Phylogenetic niche conservatism, phylogenetic signal and the relationship

between phylogenetic relatedness and ecological similarity among species. Ecology

Letters 11:995-1003.

Shreve, F. 1942. The desert vegetation of North America. The Botanical Review 8:195-246.

Simmons, N. B. 2005. Order Chiroptera. Pages 312-529 in D. E. Wilson and D. M. Reeder,

editors. Mammal species of the World: a taxonomic and geographic reference. Johns

Hopkins University Press, Baltimore, Maryland, USA.

Swenson, N. G., B. J. Enquist, J. Thompson, and J. K. Zimmerman. 2007. The influence of

spatial and size scale on phylogenetic relatedness in tropical forest communities. Ecology

88:1770-1780.

Vamosi, S. M., S. B. Heard, J. C. Vamosi, and C. O. Webb. 2009. Emerging patterns in the

comparative analysis of phylogenetic community structure. Molecular Ecology 18:572-

592.

Webb, C. O., D. D. Ackerly, M. A. McPeek, and M. J. Donoghue. 2002. Phylogenies and

community ecology. Annual Review of Ecology and Systematics 33:475-505.

Wiens, J. J. and C. H. Graham. 2005. Niche conservatism: Integrating evolution, ecology, and

conservation biology. Annual Review of Ecology, Evolution, and Systematics 36:519-

539.

6

CHAPTER 2

INVESTIGATING SENSITIVITY OF PHYLOGENETIC COMMUNITY STRUCTURE

METRICS USING NORTH AMERICAN DESERT BATS

INTRODUCTION

Interpreting patterns of coexistence within communities and determining mechanisms involved

in community assembly have fascinated ecologists for decades. However, teasing apart complex

interactions of abiotic and biotic factors and how they ultimately affect community organization

has proven difficult. A fairly recent approach to investigating community structure that has

gained traction over the past decade is to combine information contained in the phylogeny of the

regional species pool with species membership in individual communities to make inferences

about processes involved in assembly (Emerson and Gillespie 2008, Cavender-Bares et al. 2009,

Vamosi et al. 2009); this approach is referred to as phylogenetic community structure (PCS).

Phylogeny is a hypothesis of evolutionary history of the clade of interest; since closely related

species tend to share similar traits (Wiens and Graham 2005), phylogenetic distance, which is

easily quantifiable, can be used as a proxy for ecological distance, which is often difficult to

quantify (Webb 2000, Cavender-Bares et al. 2009). It is important to note, however, that this

relationship can break down in cases of convergent evolution, divergent selection, and ecological

speciation (Emerson and Gillespie 2008). When species within a community are more closely

related to each other than expected by chance, or phylogenetically clustered, they are often

ecologically similar, indicating the possibility that the community may be structured by habitat

or environmental filtering, selecting species with traits necessary to persist in that particular

habitat (Webb et al. 2002, Ackerly et al. 2006). Alternatively, if members of a community are

phylogenetically overdispersed, or less closely related to each other than expected by chance,

then competition may have structured the community by excluding phenotypes that are too

7

similar (Ackerly et al. 2006, Webb et al. 2002; but see Mayfield and Levine, 2010 for an

alternative explanation). PCS has shown promise in providing insight into whether and how

evolutionary history influences community structure; examining roles of nonrandom assembly is

of interest in its own right because it identifies cases of deterministic structure that beg

explanation.

While the impacts of trait evolution (Kraft et al. 2007), species pool characteristics (e.g.,

Lessard et al. 2012), source data (González-Caro et al. 2012), null models (e.g., Kembel 2009),

and scale (Cavender-Bares et al. 2006, Swenson et al. 2006, Swenson et al. 2007, Gómez et al.

2010, Kraft and Ackerly 2010, González-Caro et al. 2012) on metrics of phylogenetic

community structure have been investigated in depth, little work has been done to investigate the

influence of the two most fundamental components of phylogenetic community structure

analyses: the phylogenetic tree from which all PCS metrics are calculated and how communities

themselves are delimited. Swenson (2009) investigated how phylogenetic resolution influences

commonly used metrics of structure, by simulating trees with polytomies either at the tips or

deeper in the phylogenies, then calculating structure metrics. He showed that overall PCS

metrics from unresolved trees are highly correlated with metrics calculated from the “true” tree,

but this relationship becomes weaker as tree resolution decreases. In addition, randomly creating

polytomies at internal nodes had a greater impact on metrics than did collapsing terminal nodes

(Swenson 2009). Similarly, no one has yet investigated if differences in how communities are

defined influences phylogenetic community structure. The appropriate spatial size of a

community will of course vary depending on the taxa of interest; however, it is important to

determine what impact this might have on interpretation of community structure. The focus of

8

the present study is to investigate how community delimitation methods and moderate changes to

phylogeny affect PCS metrics using North American desert bat communities.

North American desert bats are an ecologically and economically important group. In

North American deserts, there are five families, 28 genera, and 55 species. The majority of these

species are insectivorous, but a few members of the family Phyllostomidae in this region and are

nectarivorous, frugivorous, piscivorous, or sanguivorous. Bats are integral to ecosystem

functioning wherever they occur (Jones et al. 2009); in western North America they serve as

pollinators of several plant species and consume economically important insects (Jones et al.

2009).

We focused our study on the four large deserts of North America (Great Basin, Mojave,

Sonoran, and Chihuahuan; Figure 2.1, A), allowing us to investigate PCS within, between, and

across regions. These deserts were formed by a combination of rain-shadows from surrounding

mountains and cool ocean currents off the Pacific coast that limit precipitation (Axelrod 1983).

Although all four deserts experienced the effects of increasing aridity over time, the more

northern Great Basin and Mojave deserts are considered younger (~8000-10,000 years as deserts;

Axelrod 1983) than the two southern deserts, with the Chihuahuan Desert being the oldest

(~11,500 years old; Medellin-Leal 1982). In addition, these deserts cover a range of climatic

regimes, based on winter temperatures, from cold (Great Basin) and cool (Mojave) deserts,

which acquire most of their precipitation during the winter (Axelrod 1983), to the subtropical

Sonoran, with precipitation in both summer and winter (Crosswhite and Crosswhite 1982), to the

hot Chihuahuan, with most of its precipitation in summer (Medellin-Leal 1982). Because of

these differences in age and climatic regimes, each desert hosts distinct floral assemblages

(Shreve 1942). Similar mechanisms could have also led to different evolutionary histories of

9

.

Figure 2.1. (A) Map of desert regions of North America and bat collection/capture locations

showing only US and Mexican states containing biome13 of the World Wildlife Federation

terrestrial ecosystem layers. (B) Six methods for delimiting communities zoomed in to southern

California, southern Nevada, western Utah, and northwestern Arizona.

desert faunas as well. For this reason we might expect different patterns of community structure

among deserts. Conversely, similarly harsh conditions found in deserts could affect the bat

faunas found within them in similar ways leading to convergent patterns of community structure.

In the present study we infer a phylogeny and further Swenson’s (2009) work by

comparing metrics calculated on various trees generated from our data set to investigate how

changes in phylogenetic trees influence PCS metrics. We did not investigate how random trees

affect PCS, but instead concentrated on possible trees, as these are more likely to be used in this

type of analysis. In addition, we use six methods to delimit communities to determine if these

10

differences alter PCS metrics. Finally, we investigate if patterns of bat community structure

differ among North American deserts

METHODS

Phylogeny

Many phylogenetic trees include North American bats (e.g., Jones et al. 2002, Baker et al. 2003,

Hoofer and Van Den Bussche 2003, Stadelmann et al. 2007); however, none of them have all

genes or taxa in common, or they are poorly resolved for important taxa (i.e., Myotis), creating

the need to build a phylogenetic tree for species occurring in the four North American deserts.

Sequences were downloaded from GenBank when possible or generated from tissues preserved

in museum collections or collected in the field (see below and Appendix I).

DNA was extracted from organ or muscle tissues using a DNeasy Blood and Tissue Kit

(Qiagen). Mitochondrial cytochrome b and 12S-16S, as well as nuclear RAG2 were amplified

using previously published primers (Irwin et al. 1991, Baker et al. 2000, Teeling et al. 2000, Van

Den Bussche and Hoofer 2000, Ibanez et al. 2006, Stadelmann et al. 2007) as well as novel

primers designed for this study. Primer combinations and thermal-cycling profiles are given in

Appendix I. Amplifications were carried out either with pureTaq PCR beads (GE Healthcare) or

with 2.5 units of Taq polymerase, 10X buffer (Invitrogen), 1.5mM MgCl2, and 1 µM of each

primer. Resulting PCR products were sequenced using traditional Sanger techniques by

Beckman Coulter Genomics (Danvers, MA). Sequences were cleaned using Seqman (v.6.1) and

initially aligned in MegAlign (v.6.1); both are part of the DNA* Lasergene 6 package.

To improve accuracy of our phylogeny, 103 species not occurring in North American

desert regions were included in analyses (Appendix II). Two members of the family

Pteropodidae, Thoopterus nigrescens and Styloctenium wallacei, and 2 members of the family

11

Rhinolophidae, Rhinolophus luctus and R. celebensis, served as outgroups (Appendix I).

Sequences were aligned with the online version of MUSCLE (Edgar 2004) then converted to

NEXUS format with Phylogeny.fr (Dereeper et al. 2008). For some species, full length

sequences were not available for one or more genes, therefore data sets including and excluding

missing data were analyzed. In addition, trees were inferred including and excluding nuclear

RAG2.

Modeltest version 3.7 (Posada and Crandall 1998) was used to determine the most

appropriate models of evolution (parameters of nucleotide substitution) using Akaike

information criteria, for each gene including and excluding missing data (Appendix III: Table

S1). Genes were concatenated with SequenceMatrix (Vaidya et al. 2011). GARLI (Zwickl 2006)

was used to infer phylogenies using maximum likelihood for each partitioned dataset. GARLI

searches were run on each of the four nexus files (including and excluding missing data, with

and without RAG2) until several searches found identical best trees with similar scores. One

thousand bootstrap replicates were then performed on each tree. All trees produced in these

analyses have been submitted to the Dryad Digital Repository

(http://doi.org/10.5061/dryad.627ck).

Communities

Community composition was determined based on a GIS map of bat capture data. The majority

of these data were downloaded from MaNIS (http://manisnet.org) in addition to capture and

collection records from museums not affiliated with MaNIS (Angelo State University, Arizona

State University, Brigham Young University, Oregon State University, Sul Ross State

University, University of Arizona, University of California Davis, and University of Texas El

Paso), published studies and reports (O'Farrell and Bradley 1970, Steen et al. 1997, Williams et

12

al. 2006), our own fieldwork (Appendix III: S1), as well as collection records of other

researchers (Michael O’Farrell, Stacy Mantooth, and Jason Williams, personal communications),

as long as they collected at least some voucher specimens. Care was taken to ensure that records

from published accounts and museum specimens were not duplicated. Capture and collection

records were filtered to contain only bats collected/captured with geographic coordinates from

the desert regions (biome#13) as defined by the World Wildlife Federation’s (WWF) terrestrial

ecosystem layers (Olson et al. 2001) since 1950, when mist nets came into common use.

Scientific names for all bat records considered were standardized based on Simmons (2005). At

many geographic locations, species identification in the hand can be problematic due to cryptic

and/or phenotypically plastic species and alternate identifications could have had an impact on

PCS analyses; specific methods testing the impact of alternate identifications can be found in the

supplementary materials (Appendix III: S2). Individual specimen records were combined based

on identical geographic coordinates so that number of bats of each species was summed and

associated with each coordinate combination. Ecosystem types within biome#13 of the WWF

terrestrial ecosystem layers (Olson et al. 2001) were combined to approximately coincide with

Shreve’s (1942) Great Basin, Mojave, Sonoran, and Chihuahuan Desert designations.

Collection points were mapped using ArcGIS v. 9.3 (Figure 2.1, A). Currently, there is no

standardized method that defines how a community should be delimited. Therefore we devised

six methods of community delimitation (Figure 2.1, B). The first was to create 5 and 10km

radius buffers around each geographic point of collection or capture. If two or more of these

areas overlapped, then communities were formed by dissolving boundaries of touching buffers

and performing a spatial join (joining the data attributes of several points based on spatial

proximity) to sum number of individuals of each species captured/collected at each data point

13

within the combined buffer to the combined buffer layer. These communities are composed of

spatially clustered collection/capture localities; however there is no limit on how many points are

joined or the spatial extent of joined buffers. In addition, buffer communities could encompass

multiple microhabitats and elevations. The second method was to overlay a regular grid of 10-

by-10 and 50-by-50km cells (Ormsbee et al. 2006) on the map using ET GeoWizards version

10.0. Communities consisted of all collection points within each of these cells. This method

explicitly determines community spatial extent; however, it can split nearby collecting locations.

The final method was to subjectively place 50 and 100km diameter circles on the map to

encompass as many collection/capture points as possible (but at least four) without overlapping

circle boundaries. This method provides a spatially defined limit to the size of communities

while accounting for likely connectivity of nearby collecting locations.

PCS analyses are based on the assumption that differences in composition among

communities are not the result of incomplete sampling. In order to enhance likelihood that

resulting communities had been adequately sampled and could be statistically compared, Chao1

(Colwell 2009, Oksanen et al. 2010) was calculated for each community using the function

“estimateR” in the vegan package (Oksanen et al. 2010) of the R statistical platform. Chao1 uses

species abundance data, the number of species in the sample, and the number of species

represented by a singletons and doubletons to estimate the true number of species in an

assemblage (Colwell and Coddington 1994); this estimator has been shown to accurately

estimate true species richness (Hortal et al. 2006). Communities with three or more species were

considered adequately sampled if observed species richness fell within the 95% confidence

interval of the richness estimator. All community data matrices used in these analyses have been

submitted to the Dryad Digital Repository (http://doi.org/10.5061/dryad.627ck).

14

Phylogenetic community structure metrics

Delimitation of meaningful species pools is essential for constructing reasonable null models to

assess whether observed communities are significantly different from randomly generated

communities (e.g., Lessard et al. 2012). Species pools were established across two different

spatial scales: (1) all North American deserts and (2) species that occur in each of the large North

American deserts.

Mean pairwise distance (MPD) is a measure of phylogenetic dispersion of taxa within a

particular community; it is the average pairwise phylogenetic distance among all pairs of species

(Webb 2000, Webb et al. 2002). Mean nearest taxon distance (MNTD) measures how locally

clustered taxa are; it is the mean phylogenetic distance to the nearest taxon for all species in a

community (Webb 2000, Webb et al. 2002). These metrics were calculated in R using the picante

package (Kembel et al. 2010). In order to obtain standardized effect size (SES-) z-values and p-

values for each metric, empirical values of MPD and MNTD were compared to those calculated

for 10,000 communities randomly assembled from the appropriate species pool using the

independent swap null model. This null model randomizes the community data matrix while

maintaining species richness within samples and species occurrence frequency; it was chosen

because previous work has shown it to perform well in detecting community assembly processes

(Kembel 2009). When α=0.10, communities that are significantly phylogenetically overdispersed

have positive z-values and p-values >0.95 while phylogenetically clustered communities have

negative z-values and p-values <0.05. We chose this α because we wanted to acknowledge

communities in the upper and lower 5% of the tails as significantly different from randomly

assembled communities and the author of the package suggests this threshold (Kembel 2010).

Fisher’s test of combined probabilities (Sokal and Rohlf 1995) was calculated to determine

15

overall significance of SES-MPD and SES-MNTD for each community delimitation method for

each species pool. We assessed spatial structuring of results across the landscape by calculating

Moran’s I correlograms implemented in SAM v. 4.0 (Rangel et al. 2006).

Impact of phylogenetic tree on community structure metrics

All analyses described below used 5km buffer communities and the all-desert species pool. Four

different data sets were used to infer phylogenetic trees in this study: with and without missing

data and with and without nuclear RAG2. To see if these four trees influenced PCS results, SES-

MPD and SES-MNTD were calculated for each community from each tree. A MANOVA was

performed using the SES-MPD and SES-MNTD z-values (results not shown). There was no

significant difference among different trees, so for all PCS analyses the tree including missing

data and nuclear RAG2 was used (hereafter referred to as the best tree; Figure 2.2).

To determine if differences in phylogenetic trees influence PCS metrics, we calculated

SES-MPD and SES-MNTD for a population of bootstrap trees as well as randomized trees.

Twenty-one trees bootstrapped from the best tree, spanning the full range of maximum-

likelihood values (from the best bootstrap tree to the worst), were used to calculate SES-MPD

and SES-MNTD. The SES-MPD and SES-MNTD z-values were then compared to those

calculated from the best tree using a MANOVA.

In addition, to investigate reasonable alterations in the phylogeny, we created randomized

trees from the best tree using both nearest-neighbor interchange (NNI) as well as sub-tree prune

and re-graft (SPR) methods. NNI randomizations swap neighboring branches making smaller

changes to trees than SPR randomizations which remove a branch attached to a subtree then

inserts it somewhere else on the tree (Felsenstein 2004). The R package phangorn (Schliep 2011)

was used to make 10 trees that were each 10, 50, 100, 200, and 300 moves away from the best

16

Figure 2.2: Phylogenetic tree used in all phylogenetic community structure analyses (all three

genes, including missing data; referred to as “best tree” in text). Numbers at nodes are bootstrap

values. Bold species are found in the species pool; all other species were included with the

purpose of inferring an accurate phylogeny. In order to fit on the page, the tree has been cut in

half: the bold lines indicate where the upper (right) and lower (left) halves join.

17

tree for each randomization method for a total of 100 randomized trees. It is important to note

again that these trees are not truly randomized but have had randomly chosen branches or clades

rearranged a specified number of times. In addition to these random rearrangements, we

constructed a tree retaining the familial relationships of the best tree but unresolving all clades

below the family level, we refer to this tree as “Polytomy”; we also unresolved all of the clades

of the best tree, creating a tree we refer to as “Bush”. PCS metrics were then computed for each

of these trees. PCS results were compared using MANOVA as above; randomized trees were

compared to each other and to the best tree. Significant MANOVA results were further

investigated using ANOVA with Tukey’s Honestly Significant Difference (HSD) to assess

which metrics and trees were driving significance using the agricolae package (Mendiburu 2012)

in R. We performed a Mantel test in R between the distance matrix from the best tree and

distance matrices from some of the randomized and the unresolved Polytomy and Bush trees to

determine if resolution/randomization impacted the distance matrices from which PCS metrics

are calculated. For this analysis, we chose to use SPR300.2 because HSD showed it was the most

different from all other trees, then arbitrarily chose SPR50.2, NNI50.2, and NNI 300.2 to

represent minimally and maximally randomized trees. Robinson-Foulds distance between the

best tree and each bootstrap or randomized tree was calculated using phangorn (Schliep 2011) or

PAUP (Swofford 2000). Robinson-Foulds distance is computed by calculating the branch lengths

of all possible partitions for each tree then summing the absolute values of the differences

(Felsenstein 2004). Smaller distances indicate similar trees. Examples of these trees are

summarized in Appendix III: FigureS1 and all trees used in these analyses have been submitted

to the Dryad Digital Repository (http://doi.org/10.5061/dryad.627ck).

18

RESULTS

Phylogeny

In general, there were few differences in topologies of trees that included or excluded missing

data and included or excluded nuclear RAG2, although trees including missing data had higher

nodal support than those excluding missing data (Figure 2.2, Dryad). Familial relationships were

similar to those proposed by Teeling et al. (2005) and relationships among species within

families are similar to those in taxon specific phylogenies (e.g., Baker et al. 2003, Hoofer and

Van Den Bussche 2003, Stadelmann et al. 2007).

Community delimitation method and PCS

Adequately sampled communities based on Chao1 for each delimitation method are summarized

in Appendix III: Table S2 and Figure 2.3; these are communities used in PCS analyses. Visual

inspection of PCS results across the landscape revealed no discernible patterns (Appendix III:

Figure S2), however Moran’s I correlograms indicated that PCS metrics were positively and

significantly spatially autocorrelated at small distances and negatively and significantly

autocorrelated at large distances, but were not spatially autocorrelated at intermediate distances

(AppendixIII: Figure S3). Individual communities, regardless of spatial scale, run the gamut

from significantly clustered to significantly overdispersed (Table 2.1). Since we were more

interested in examining overall patterns of PCS, we will only discuss the results of Fisher’s

combined probability tests.

Phylogenetic community structure analyses for all deserts combined indicate that

communities were significantly phylogenetically clustered regardless of delimitation method or

PCS metric (Tables 2.1 and 2.2). For the Great Basin Desert, 10km buffer, 50km grid, and 50km

circle communities were significantly clustered for both SES-MPD and SES-MNTD while 5km

19

Figure 2.3: Adequately sampled communities with three or more taxa based on Chao1 used in

analyses. N refers to total number of communities for each delimitation method. (A) 5km buffer

(B) 10km buffer (C) 50km circles (D) 100 km circles (E) 10km grids (F) 50km grids. Maps show

only US and Mexican states containing biome13 of the World Wildlife Federation terrestrial

ecosystem layers.

buffer, 10km grid, and 100km circle communities also tended to be clustered (Tables 2.1 and

2.2). In the Mojave Desert, neither SES-MPD nor SES-MNTD was significantly different from

randomly assembled communities for any delimitation method (Tables 2.1 and 2.2). In the

Sonoran Desert, all delimitation methods were significantly clustered for both community

structure metrics except SES-MNTD for 5km buffer (tended toward clustering) and 50km circle

communities (not significantly different from random; Tables 2.1 and 2.2). Chihuahuan Desert

communities were significantly clustered or tended towards clustering except SES-MNTD for

50km grid communities and both metrics for 100km circle communities (Tables 2.1 and 2.2).

20

Tab

le 2

.1:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on P

CS

anal

yse

s fo

r ea

ch c

om

munit

y d

elim

itat

ion m

ethod f

or

each

des

ert.

Met

ric

Del

imit

atio

n

met

hod

Des

ert

Clu

ster

ed

com

munit

ies

Ran

dom

com

munit

ies

Over

dis

per

sed

com

munit

ies

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s df

MP

D

5km

buff

er

All

21

140

7

448.5

3

<0.0

01

clust

ered

336

Gre

at B

asin

6

48

5

148.8

9

0.0

287

clust

ered

118

Moja

ve

0

18

1

35.4

9

0.5

861

ns

38

Sonora

n

3

21

0

77.8

0.0

04

clust

ered

48

Chih

uah

uan

6

58

2

163

0.0

346

clust

ered

132

10km

buff

er

All

19

100

6

370.2

4

<0.0

01

clust

ered

250

Gre

at B

asin

5

39

5

123.8

2

0.0

401

clust

ered

98

Moja

ve

0

8

2

25.7

0.1

76

ns

20

Sonora

n

3

23

0

77.8

3

0.0

12

clust

ered

52

Chih

uah

uan

4

34

1

92.5

9

0.1

241

ns

78

10km

gri

d

All

26

182

7

576.3

5

<0.0

01

clust

ered

430

Gre

at B

asin

4

53

5

145.2

8

0.0

93

ns

124

Moja

ve

2

23

2

58.8

4

0.3

028

ns

54

Sonora

n

5

35

1

130.0

7

0.0

01

clust

ered

82

Chih

uah

uan

11

71

1

204.7

5

0.0

219

clust

ered

166

50km

gri

d

All

21

140

6

463.5

7

<0.0

01

clust

ered

334

Gre

at B

asin

4

50

5

148.5

8

0.0

298

clust

ered

118

Moja

ve

1

15

2

44.5

3

0.1

556

ns

36

Sonora

n

6

23

1

99.0

4

0.0

01

clust

ered

60

Chih

uah

uan

5

46

1

115.0

2

0.2

163

ns

104

50km

circ

les

All

16

83

8

344.4

5

<0.0

01

clust

ered

214

Gre

at B

asin

6

28

5

120.5

7

0.0

014

clust

ered

78

Moja

ve

1

13

2

32.8

8

0.4

238

ns

32

Sonora

n

3

20

1

73.2

0.0

11

clust

ered

48

(Tab

le 2

.1 c

onti

nued

)

21

Met

ric

Del

imit

atio

n

met

hod

Des

ert

Clu

ster

ed

com

munit

ies

Ran

dom

com

munit

ies

Over

dis

per

sed

com

munit

ies

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s df

Chih

uah

uan

3

24

1

72.0

4

0.0

731

ns

56

100km

circ

les

All

10

58

6

219.8

9

<0.0

01

clust

ered

148

Gre

at B

asin

3

26

3

78.5

1

0.1

048

ns

64

Moja

ve

0

9

1

22.6

5

0.3

063

ns

20

Sonora

n

1

8

0

29

0.0

48

clust

ered

18

Chih

uah

uan

1

19

3

50.4

9

0.3

007

ns

46

MN

TD

5km

buff

er

All

15

147

6

421.0

8

0.0

01

clust

ered

336

Gre

at B

asin

6

48

5

144.1

4

0.0

51

ns

118

Moja

ve

1

16

2

39.6

6

0.3

96

ns

38

Sonora

n

1

23

0

51.2

5

0.3

47

ns

48

Chih

uah

uan

4

60

2

153.2

0.1

ns

132

10km

buff

er

All

11

107

7

337.3

7

<0.0

01

clust

ered

250

Gre

at B

asin

5

40

4

122.7

2

0.0

46

clust

ered

98

Moja

ve

0

8

2

25.2

6

0.1

92

ns

20

Sonora

n

4

22

0

75.0

1

0.0

2

clust

ered

52

Chih

uah

uan

3

35

1

85.3

3

0.2

67

ns

78

10km

gri

d

All

22

189

4

561.8

5

<0.0

01

clust

ered

430

Gre

at B

asin

4

53

5

142.7

9

0.1

19

ns

124

Moja

ve

1

25

4

55.0

2

0.4

36

ns

54

Sonora

n

4

37

0

112.3

9

0.0

15

clust

ered

82

Chih

uah

uan

8

74

1

208.7

3

0.0

14

clust

ered

166

50km

gri

d

All

17

142

8

440.2

2

<0.0

01

clust

ered

334

Gre

at B

asin

5

51

3

147.5

6

0.0

34

clust

ered

118

Moja

ve

2

14

2

50.0

1

0.0

6

ns

36

(Tab

le 2

.1 c

onti

nued

)

22

Met

ric

Del

imit

atio

n

met

hod

Des

ert

Clu

ster

ed

com

munit

ies

Ran

dom

com

munit

ies

Over

dis

per

sed

com

munit

ies

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s df

Sonora

n

3

27

0

80.9

4

0.0

37

clust

ered

60

Chih

uah

uan

3

48

1

109.3

9

0.3

4

ns

104

50km

circ

les

All

9

94

4

316.7

6

<0.0

01

clust

ered

214

Gre

at B

asin

7

28

4

116.3

3

0.0

03

clust

ered

78

Moja

ve

3

10

3

32.5

9

0.4

38

ns

32

Sonora

n

3

21

0

62.6

2

0.0

76

ns

48

Chih

uah

uan

0

28

0

72.6

7

0.0

66

ns

56

100km

circ

les

All

8

63

3

204.8

1

0.0

01

clust

ered

148

Gre

at B

asin

3

29

0

80.3

6

0.0

81

ns

64

Moja

ve

1

8

1

25.5

8

0.1

8

ns

20

Sonora

n

2

7

0

33.4

2

0.0

15

clust

ered

18

Chih

uah

uan

1

21

1

44.0

8

0.5

53

ns

46

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

23

Table 2.2: All Fisher’s combined probability test p-values for all species pools and delimitation

methods, color-coded by significance.

Metric Delimitation method All deserts Great Basin Mojave Sonoran Chihuahuan

MPD

5km buffer <0.001 0.029 0.586 0.004 0.035

10km buffer <0.001 0.040 0.176 0.012 0.124

10km grid <0.001 0.093 0.303 0.001 0.022

50km grid <0.001 0.030 0.156 0.001 0.216

50km circle <0.001 0.001 0.424 0.011 0.073

100km circle <0.001 0.105 0.306 0.048 0.301

MNTD

5km buffer 0.001 0.051 0.396 0.347 0.100

10km buffer <0.001 0.046 0.192 0.020 0.267

10km grid <0.001 0.119 0.436 0.015 0.014

50km grid <0.001 0.034 0.060 0.037 0.340

50km circle <0.001 0.003 0.438 0.076 0.066

100km circle 0.001 0.081 0.180 0.015 0.553

Clustered (sig.; p-values <0.001-0.049)

Clustered (ns; p-values 0.05-0.29)

Not significant (p-values 0.30-0.69)

Impact of phylogenetic tree on analyses

As trees were randomized to increase branching differences from the best tree, Robinson-Foulds

distances increased (Appendix III: FigureS4); several of the SPR300 move trees were

themaximum possible distance from the best tree. Although differences in SES-MPD and SES-

MNTD z-values calculated from trees increasingly distant from the best tree were perceptible

upon visual inspection (Appendix III: FigureS5), these differences were not statistically

significant except for the most distant trees (Table 2.3). There were no significant differences in

PCS metrics between the best tree and bootstrap trees, or between any of the trees randomized

with the NNI method and the best tree (Table 2.3). Furthermore, we did not detect any

significant differences until we compared SPR300 trees to each other, to the best tree, and all

SPR trees to the best tree (Table 2.3). There were no significant differences in PCS z-values

between the best tree and the polytomy or bush trees (Table 2.3), although the p-value for the

24

Table 2.3: Results of multivariate analysis of variance (MANOVA) for SES-MPD and SES-

MNTD z-values for 5km buffer communities and the “all taxa” species pool. Significant p-values

are in bold. NNI refers to nearest-neighbor interchange randomizations while SPR refers to sub-

tree prune and re-graft randomizations. Numbers after NNI and SPR refer to the number of

randomization moves away from the best tree. Comparison refers to which trees are being

compared (NNI50 means all ten NNI 50 move trees are being compared to each other whereas

NNI50 and Best tree means all ten NNI 50 move trees and the best tree are being compared).

Bootstrap refers to the bootstrap trees, Polytomy refers to the tree with polytomies below the

family level, Bush refers to the tree with no bifurcating branches, and Best tree refers to the best

tree used in all other analyses (see methods).

Compairison approx F num Df den Df Pr(>F)

Bootstrap and Best tree 0.018068 42 7524 1

NNI50 0.20038 2 1717 0.818

NNI50 and Best tree 0.14216 20 3762 1

NNI100 0.035948 2 1717 0.965

NNI100 and Best tree 0.19458 20 3762 1

NNI200 0.004369 2 1717 0.996

NNI200 and Best tree 0.62127 20 3762 0.900

NNI300 0.030669 2 1717 0.970

NNI300 and Best tree 0.3497 20 3762 0.997

All NNI trees and Best 0.35662 80 14022 1

SPR10 0.7279 2 1717 0.483

SPR10 and Best tree 0.50733 20 3762 0.965

SPR50 0.47491 2 1717 0.622

SPR50 and Best tree 0.87546 20 3762 0.620

SPR100 1.8013 2 1717 0.165

SPR100 and Best tree 1.215 20 3762 0.230

SPR200 0.93034 2 1717 0.395

SPR200 and Best tree 1.3719 20 3762 0.124

SPR300 3.0903 2 1717 0.046

SPR300 and Best tree 1.7853 20 3762 0.017

All SPR Trees and Best 1.4418 100 17442 0.003

Polytomy and Best tree 0.93865 2 341 0.392

Bush and Best tree 2.7083 2 341 0.068

bush and best comparison was non-significant. ANOVA indicated that SES-MNTD was the

metric causing significant differences in the MANOVA in all three cases while SES-MPD was

also significant when all SPR trees were compared to the best tree (Table 2.4). Mantel tests

indicate significant correlations between distance matrices from the best tree and those from

25

SPR50, NNI300, SPR300, and polytomy trees, but not for distance matrices from NNI50

(although this approaches significance) or bush trees (Appendix III: Table S3).

Table 2.4: ANOVA results for PCS comparisons found to be significant with MANOVA.

Significant p-values are bolded. Terminology as in Table 2.3.

Comparison Metric num Df den Df Mean Sq F value Pr(>F)

SPR300 MPD.z 1 1718 1.5947 1.774 0.183

MNTD.z 1 1718 5.184 5.521 0.019

SPR300 and Best

tree

MPD.z 10 1881 1.3948 1.474 0.143

MNTD.z 10 1881 2.4147 2.537 0.005

All SPR Trees and

Best

MPD.z 50 8721 1.529 1.479 0.016

MNTD.z 50 8721 1.814 1.827 <0.001

DISCUSSION

We inferred a well resolved phylogenetic estimate using multiple genes and broad taxon

sampling that will be useful for a wide range of ecological and evolutionary studies. We then

used this tree to test the robustness of PCS metrics to community delimitation methods and

changes to the tree itself. We found that bat communities tend to be phylogenetically clustered

across deserts and within individual deserts regardless of community delimitation method. In

addition, we found that MPD and MNTD were robust to changes to the phylogeny from which

they were calculated.

Phylogeny

We estimated phylogenies in this study not to redefine evolutionary relationships, but to produce

a robust tree with which to test ecological hypotheses. Because of this, we focused our taxon

sampling on North American desert bats and species with sequences available on GenBank, not

on ensuring that all clades were equally represented. Our trees included sequences we produced

for several taxa that previously had little or no representation on GenBank (Eumops perotis,

Nyctinomops aurispinosus, N. femorosaccus, Leptonycteris nivalis, Myotis melanorhinus, M.

(evotis) milleri, and M. occultus); sequences for one or more genes were also made publicly

26

available for an additional 22 taxa (Appendix I). Our phylogenetic estimates were well resolved

and did not contain the numerous polytomies that pervade the vespertilionid clade of the bat

supertree (Jones et al. 2002), although resolution of this clade was not our explicit goal when

including taxa in this family. Because of their high resolution and dense taxonomic sampling,

these trees should prove useful to the broader scientific community to answer ecological and

evolutionary questions.

Impact of community delimitation method on PCS metrics

While there were some differences in PCS results between different community delimitation

methods (Tables 2.1 and 2.2), we did not find any overall pattern in these differences, making it

difficult to interpret results or recommend a particular delimitation method for bat communities.

All three methods have advantages and disadvantages. The buffer delimitation method has no

limit to how many buffers can be joined, which allows the spatial area of each community to

vary greatly (5km buffers: mean= 182.86 km2, range= 78.54-1280.89 km

2; 10km buffer: mean=

1063.31 km2, range= 314.16-15266.45 km

2). In contrast, both grid and circle drawing methods

are spatially consistent in their extent. Nonetheless, one drawback to the grid method is that

capture/collection locations may potentially be sufficiently close in proximity to share

individuals, yet be assigned to separate communities. The subjectivity of the circle drawing

method (circles are subjectively drawn around as many communities as possible but at least four)

could possibly introduce researcher bias regarding which communities are joined together.

Fortunately in our case, different delimitation methods tended to give results that at least trended

in the same general direction. Natural communities are not necessarily discrete entities.

Nonetheless, measurement requires discrete units. Ideally, congruence among delimitation

methods suggests unbiased pattern description. Such efforts are not always feasible. As long as a

27

researcher delimits communities consistently within a study, there is reasonable assurance that

whatever delimitation method is used, results should be comparable within the study and

accurately reflect trends in the data.

Impact of phylogeny on PCS metrics

PCS metrics are surprisingly robust even to substantial changes in phylogenetic tree topology.

Prior to this study only Swenson (2009) had investigated how randomly reducing tree resolution

affected PCS metrics and suggested that PCS metrics are sensitive to polytomies at basal nodes

of the phylogeny. We take Swenson’s work a step further by rearranging branches randomly

across the phylogeny. Bootstrap and NNI trees were not distant enough from the best tree to

make a significant difference in community structure metrics (Table 2.3 and Appendix III:

Figures S4-5). Trees must be almost as distant from the “true” tree as possible (maximum

Robinson-Foulds distance for a tree containing 56 taxa is 109; maximum distance achieved

through randomization was 108 SPR300 trees 2, 4, 6, 9, 10) before significant changes could be

detected in the PCS metrics, and even then it was only SES-MNTD that was consistently

affected (Table 2.4 and Appendix III: Figures S4-5). While most of the substantially randomized

trees produced quite different PCS metrics from those calculated from the best tree, a few

randomized trees still produced metrics very similar to the best tree metrics (Appendix III:

Figure S5 c-d).

These results suggest 2 possibilities: 1) that PCS metrics actually have little to do with

phylogeny or 2) that even a poorly inferred tree still offers useful evolutionary information that

can be used to describe patterns of species co-occurrence. We suggest that possibility 2 is the

case. Phylogenetic trees reflect evolutionary history, therefore ecology, of taxa within them. Our

randomization techniques moved clades and branches randomly on the best tree. Even

28

substantially randomized trees (bootstrap trees, all NNI trees, and SPR 10-200 trees, Appendix

III: Figure S1; all trees available from Dryad Digital Repository) retained some of the original

phylogenetic structure and produced PCS metrics that were similar to those calculated from the

best tree (Appendix III: Figure S5). In particular, familial relationships were retained so that

metrics calculated from these trees were statistically indistinguishable from those calculated

from the best tree (Table 2.3). This was also the case for metrics calculated from the polytomy

tree, which contained polytomies below the familial level, further strengthening this argument

(Table 2.3). These familial relationships reflect not only evolutionary history but also ecological

specialization, so that species membership in a community is dictated by ecology which is

reflected by PCS metrics. Maximally randomized (SPR300) trees retained essentially none of the

original evolutionary history exhibited in the best tree, accounting for the significant difference

between metrics calculated from these trees and those calculated from the best tree (Tables 2.3

and 2.4; Appendix III: Figures S1 and S5). While PCS metrics calculated from the bush tree

were not significantly different from those from the best tree, the relatively low p-values indicate

that there were substantial, albeit non-significant differences between the two trees (Table 2.3,

Appendix III: Figure S5).

Our goal for these analyses was not to produce truly random trees. It is essentially

unfathomable that with the data, methods, and programs available to researchers at this time, a

completely erroneous/random tree could be produced and used in PCS analyses. Instead our goal

was to investigate the impact of plausibly random trees on PCS metrics. Much more likely is the

possibility that a researcher would use a phylogeny in which some species relationships might be

incorrect while genera or at least familial relationships remain intact. In the majority of the trees

we produced the backbone remained intact while the clades were moved around in the tree

29

(Appendix III: Figure S1 and trees available on Dryad). In addition, branch lengths separating

species remain relatively unaffected by topological changes to the tree as evidenced by

significant correlation between distance matrices calculated from these trees with that from the

best tree in most cases (Appendix III: Table S3). Although branch lengths may differ based on

the data used to infer a tree, we suggest that misplacing one or a few species on a phylogeny is

not likely to significantly affect the branch lengths separating those species from others on the

tree and therefore likely will not greatly impact a distance matrix or PCS metrics calculated from

that phylogeny. Our results indicate that, in fact, MPD and to a slightly lesser extent MNTD are

robust to topological changes in a tree. These results should encourage ecologists that PCS

metrics do indeed reflect real processes acting at the community level and are not artifacts of

poorly inferred trees.

Phylogenetic community structure of desert bat communities

Spatial scale of the regional species pool has been shown to affect PCS; at large scales, habitat

filtering is expected to be most prevalent as species are filtered based on phylogenetically

conserved traits across a heterogeneous landscape (Cavender-Bares et al. 2009, Gómez et al.

2010). Habitat homogeneity at small spatial scales is expected to increase interspecific

interactions, such as competition, potentially leading to phylogenetic overdispersion (Cavender-

Bares et al. 2009, Gómez et al. 2010). Hence, community assembly should be influenced by

multiple factors acting at different scales with particular processes predominating at a given

scale. Although scale is not the focus of the present research, our study system allowed us to

examine how scale affects PCS by manipulating the species pools against which individual

communities were compared.

30

Deserts are unquestionably harsh environments. Precipitation in the four large North

American deserts is limited by the combination of cool Pacific ocean currents and rain-shadows

from surrounding mountains (Axelrod 1983). Desertification occurred over time with the

northern Great Basin and Mojave Deserts younger than the southern Sonoran and Chihuahuan

Deserts (Medellin-Leal 1982, Axelrod 1983) and each desert has its own climatic regime. These

differences in age and climate gave rise to distinct floral assemblages in each desert (Shreve

1942) and could present unique evolutionary and ecological histories to many taxa, bats

included. Indeed, while most desert bats are insectivores, given the species diversity of bats in

these desert regions (56 species and sub-species in all deserts, 25 in both the Great Basin and

Mojave deserts) it would be unsurprising to observe this pattern. Conversely, because deserts are

such harsh environments, we might expect to see habitat filtering, manifested as phylogenetic

clustering characterizing structure of desert communities. This latter pattern of predominant

phylogenetic clustering is in fact what we observe in North American desert bat communities

when all deserts are considered together as well as when each is considered separately with the

exception of the Mojave Desert.

The Mojave Desert departs from expectations of phylogenetic clustering: overall, Mojave

Desert communities are not significantly different from randomly assembled ones. Randomly

assembled communities indicate that processes such as competition or habitat filtering may not

play an important role in shaping community structure, that both may be acting simultaneously

thereby obscuring either process (Cavender-Bares et al. 2009, Vamosi et al. 2009), the traits on

which these processes are acting are not phylogenetically conserved, or some other process may

be of overriding importance. The Mojave is considered a cool desert (Axelrod 1983) and is also

the driest and most climatically unpredictable of the four desert areas considered here (Shreve

31

1942, Axelrod 1983). These conditions could foster communities that are composed of species

that can survive in such conditions (habitat filtering) and yet must compete for potentially

limiting resources, or could prevent species from reaching carrying capacity thereby preventing

competitive exclusion, or that environmental variability and unpredictability could prevent any

deterministic structure from forming.

Notwithstanding the exception outlined above, we found generally similar patterns of

phylogenetic clustering across desert regions (Table 2.2) suggesting greater importance of habitat

filtering over interspecific interactions in community assembly and indicating that desert bat

communities overall respond to the same ecological pressures in similar ways. These results

contrast with those of previous bat community structure studies using data on habitat use, diet,

morphology, and/or echolocation (e.g., Aldridge and Rautenbach 1987, Willig and Moulton

1989, Arita 1997, Stevens and Willig 1999, 2000, Campbell et al. 2007, Goncalves da Silva et al.

2008, Stevens and Amarilla-Stevens 2012) which have suggested that bat communities are

structured by competition limiting similarity of morphology or use of habitat or are made up of

species randomly drawn from the regional species pool. We should note an alternative

explanation for phylogenetic clustering put forth by Mayfield and Levine (2010): that

competition could give rise to phylogenetically clustered communities if competition for

resources limits community members to only those that possess phylogenetically conserved traits

that allow them to outcompete more distantly related species lacking such traits. This is a

plausible explanation for our observations but not one that is easily assessed given the difficulty

in determining which traits confer superior competitive ability. However, a recent study by

Riedinger et al. (2012) incorporating environmental data in PCS analyses found that overall

32

Bavarian bat communities were significantly phylogenetically clustered due to habitat filtering,

suggesting that at least in some cases bat communities are structured by environmental factors.

Our overall PCS results contrast with those for reptile and mammal communities in

Australian deserts (Lanier et al. 2013). PCS within and between taxonomic groups differed

within and between regions, indicating that taxon-specific communities respond differently to the

same ecological pressures. Our results also contrast with previous studies of mammalian PCS

(Cardillo et al. 2008, Cooper et al. 2008) which found overall tendencies for phylogenetic

overdispersion across several taxa; this dissimilarity may be due to differences in spatial and

taxonomic scale and geographic area between these studies and ours or to differing evolutionary

history and ecological responses of diverse taxa.

While our results conformed to expected patterns of overall phylogenetic clustering

(Table 2.2), individual communities actually run the gamut from significant phylogenetic

overdispersion to significant clustering regardless of scale or delimitation method (Table 2.1).

This pattern is observed in several other studies of mammalian community structure. Kamilar

and Guidi (2010) found that while continents differed in the relatedness of species within primate

communities, individual communities ranged between significantly clustered (very few

communities) to significantly overdispersed with the majority being not significantly different

from random. A similar pattern characterizes Mojave Desert rodent communities (Stevens et al.

2012) as well as bats in Bavaria (Riedinger et al. 2012) and such variation may be a general

result when numerous sites are examined simultaneously.

In conclusion, we found that PCS metrics are very robust to changes in the phylogenetic

tree used to calculate metrics. Phylogenetic trees had to be as distant from the “true” tree as

possible before differences in metrics could be detected. Such a poorly inferred phylogeny would

33

be unlikely to ever be considered for use in community structure studies, so as long as ecologists

use a reasonable tree they can be reasonably assured that trends in PCS are real. Community

delimitation method does impact PCS results, but there is no obvious pattern to these differences.

As long as a study uses the same method throughout, results should accurately reflect the same

underlying trend in the data. Finally, we found that overall, desert bat communities tend to be

phylogenetically clustered suggesting that bat communities may be responding to harsh desert

conditions in similar ways.

ACKNOWLEDGEMENTS

L.E.P. was funded by an American Museum of Natural History Theodore Roosevelt Memorial

Grant, Society of Systematic Biologists Graduate Student Award, American Society of

Mammalogists Grants-in-Aid of Research award, and Louisiana State University BioGrads

awards BG11-38 and BG14-12. Dr. Bryan Carstens allowed L.E.P. to do the molecular work in

his lab. Many thanks to Angelo State Natural History Collections, Museum of Southwestern

Biology, Museum of Texas Tech University, Louisiana State University Museum of Natural

Science, Natural History Museum of Los Angeles County, Texas Cooperative Wildlife

Collection, and Portland State University Museum of Vertebrate Zoology for providing tissues

for genetic analyses. We wish to thank Michael O’Farrell, Stacy Mantooth, and Jason Williams

for sharing their bat capture/collection datasets with us. We also thank J. S. Tello, E. S.

McCulloch, M. M. Gavilanez, S. M. Hird, N. M. Reid, J. M. Brown, and anonymous reviewers

for fruitful discussions during the development of this project and editing previous versions of

this manuscript.

34

REFERENCES



Aldridge, H. D. J. N. and I. L. Rautenbach. 1987. Morphology, echolocation, and resource

partitioning in insectivorous bats. Journal of Animal Ecology 56:763-778.

Arita, H. T. 1997. Species composition and morphological structure of the bat fauna of Yucatan,

Mexico. Journal of Animal Ecology 66:83-97.

Axelrod, D. I. 1983. Paleobotanical history of the western deserts. Pages 113-129 in S. G. Wells

and D. R. Haragan, editors. Origin and Evolution of Deserts. University of New Mexico

Press, Albuquerque, NM.

Baker, R. J., S. R. Hoofer, C. A. Porter, and R. A. Van Den Bussche. 2003. Diversification

among New World Leaf-nosed Bats: An Evolutionary Hypothesis and Classification

Inferred from Digenomic Congruence of DNA Sequence. Occasional Papers: Museum of

Texas Tech University 230:1-32.

Baker, R. J., C. A. Porter, J. C. Patton, and R. A. v. d. Bussche. 2000. Systematics of the family

Phyllostomidae based on RAG2 DNA sequences. Occasional Paters, Museum of Texas

Tech Universtiy 202:1-16.

Campbell, P., C. J. Schneider, A. Zubaid, A. M. Adnan, and T. H. Kunz. 2007. Morphological

and ecological correlates of coexistance in Maylasian fruit bats (Chiroptera:

Pteropodidae). Journal of Mammalogy 88:105-118.

Cardillo, M., J. L. Gittleman, and A. Purvis. 2008. Global patterns in the phylogenetic structure

of island mammal assemblages. Proceedings of the Royal Society B: Biological Sciences

275:1549-1556.

Cavender-Bares, J., A. Keen, and B. Miles. 2006. Phylogenetic structure of Floridian plant

communities depends on taxonomic and spatial scale. Ecology 87:109-122.



Colwell, R. K. 2009. EstimateS: Statistical Estimation of Species Richness and Shared Species

for Samples User's Guide. Storrs, CT.

Colwell, R. K. and J. A. Coddington. 1994. Estimating Terrestrial Biodiversity through

Extrapolation. Philosophical Transactions of the Royal Society of London. Series B:

Biological Sciences 345:101-118.

35

Cooper, N., J. Rodriguez, and A. Purvis. 2008. A common tendency for phylogenetic

overdispersion in mammalian assemblages. Proceedings of the Royal Society B:

Biological Sciences 275:2031-2037.

Crosswhite, F. S. and C. D. Crosswhite. 1982. The Sonoran Desert. Pages 163-295 in G. L.

Bender, editor. Reference Handbook on the Deserts of North America. Greenwood Press,

Westport, CT.

Dereeper, A., V. Guignon, G. Blanc, S. Audic, S. Buffet, F. Chevenet, J.-F. Dufayard, S.

Guindon, V. Lefort, M. Lescot, J.-M. Claverie, and O. Gascuel. 2008. Phylogeny.fr:

robust phylogenetic analysis for the non-specialist. Nucleic Acids Research 36:W465-

W469.

Edgar, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high

throughput. Nucleic Acids Research 32:1792-1797.

Emerson, B. C. and R. G. Gillespie. 2008. Phylogenetic analysis of community assembly and

structure over space and time. Trends in Ecology & Evolution 23:619-630.

Felsenstein, J. 2004. Inferring Phylogenies. Sinauer Associates, Inc, Sunderland, MA.




Goncalves da Silva, A., O. Gaona, and R. A. Medellin. 2008. Diet and trophic structure in a

community of fruit-eating bats in Lacandon Forest, Mexico. Journal of Mammalogy

89:43-49.

González-Caro, S., J. L. Parra, C. H. Graham, J. A. McGuire, and C. D. Cadena. 2012.

Sensitivity of Metrics of Phylogenetic Structure to Scale, Source of Data and Species

Pool of Hummingbird Assemblages along Elevational Gradients. PLoS ONE 7:e35472.

Hoofer, S. R. and R. A. Van Den Bussche. 2003. Molecular phylogenetics of the chiropteran

family Vespertilionidae. Acta Chiropterologica 5:1-63.

Hortal, J., P. A. V. Borges, and C. Gaspar. 2006. Evaluating the performance of species richness

estimators: sensitivity to sample grain size. Journal of Animal Ecology 75:274-287.

Ibanez, C., J. L. Garcia-Mudarra, M. Ruedi, B. Stadelmann, and J. Juste. 2006. The Iberian

contribution to cryptic diversity in European bats. Acta Chiropterologica 8:277-297.

Irwin, D. M., T. D. Kocher, and A. C. Wilson. 1991. Evolution of the Cytochrome b Gene of

Mammals. Journal of Molecular Evolution 32:128-144.

36



Jones, K. E., A. Purvis, A. MacLarnon, O. R. P. Bininda-Emonds, and N. B. Simmons. 2002. A

phylogenetic supertree of the bats (Mammalia: Chiroptera). Biological Reviews 77:223-

259.

Kamilar, J. M. and L. M. Guidi. 2010. The phylogenetic structure of primate communities:

variation within and across continents. Journal of Biogeography 37:801-813.

Kembel, S. W. 2009. Disentangling niche and neutral influences on community assembly:

assessing the performance of community phylogenetic structure tests. Ecology Letters

12:949-960.

Kembel, S. W. 2010. An introduction to the picante package. http://picante.r-forge.r-

project.org/picante-intro.pdf.

Kembel, S. W., P. D. Cowan, M. R. Helmus, W. K. Cornwell, H. Morlon, D. D. Ackerly, S. P.

Blomberg, and C. O. Webb. 2010. Picante: R tools for integrating phylogenies and

ecology. Bioinformatics 26:1463-1464.

Kraft, N. J. B. and D. D. Ackerly. 2010. Functional trait and phylogenetic tests of community

assembly across spatial scales in an Amazonian forest. Ecological Monographs 80:401-

422.

Kraft, N. J. B., W. K. Cornwell, C. O. Webb, and D. D. Ackerly. 2007. Trait evolution,

community assembly, and the phylogenetic structure of ecological communities.

American Naturalist 170:271-283.

Lanier, H. C., D. L. Edwards, and L. L. Knowles. 2013. Phylogenetic structure of vertebrate

communities across the Australian arid zone. Journal of Biogeography 40:1059-1070.

Lessard, J.-P., M. K. Borregaard, J. A. Fordyce, C. Rahbek, M. D. Weiser, R. R. Dunn, and N. J.

Sanders. 2012. Strong influence of regional species pools on continent-wide structuring

of local communities. Proceedings of the Royal Society B: Biological Sciences 279:266-

274.

Mayfield, M. M. and J. M. Levine. 2010. Opposing effects of competitive exclusion on the

phylogenetic structure of communities. Ecology Letters 13:1085-1093.

Medellin-Leal, F. 1982. The Chihuahuan Desert. Pages 321-372 in G. L. Bender, editor.

Reference Handbook on the Deserts of North America. Greenwood Press, Westport, CT.

Mendiburu, F. d. 2012. agricolae: Statistical Procedures for Agricultural Research.

37

O'Farrell, M. J. and W. G. Bradley. 1970. Activity patterns of bats over a desert spring. Journal

of Mammalogy 51:18-26.

Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, R. B. O'Hara, G. L. Simpson, P. Solymos, M.

Henry, H. Stevens, and H. Wagner. 2010. vegan: Community Ecology Package.

Olson, D. M., E. Dinerstein, E. D. Wikramanayake, N. D. Burgess, G. V. N. Powell, E. C.

Underwood, J. A. D'amico, I. Itoua, H. E. Strand, J. C. Morrison, C. J. Loucks, T. F.

Allnutt, T. H. Ricketts, Y. Kura, J. F. Lamoreux, W. W. Wettengel, P. Hedao, and K. R.

Kassem. 2001. Terrestrial Ecoregions of the World: A New Map of Life on Earth.

BioScience 51:933-938.

Ormsbee, P. C., J. M. Zinck, J. M. Szewczak, L. E. Patrick, and A. H. Hart. 2006. Benefits of a

standardized sampling frame: an update on the “Bat Grid". Bat Research News 47:4.

Posada, D. and K. A. Crandall. 1998. MODELTEST: testing the model of DNA substitution.

Bioinformatics 14:817-818.

Rangel, T. F., J. A. F. Diniz-Filho, and L. M. Bini. 2006. Towards an integrated computational

tool for spatial analysis in macroecology and biogeography. Global Ecology and

Biogeography 15:321-327.

Riedinger, V., J. Müller, J. Stadler, W. Ulrich, and R. Brandl. 2012. Assemblages of bats are

phylogenetically clustered on a regional scale. Basic and Applied Ecology.

Schliep, K. P. 2011. phangorn: phylogenetic analysis in R. Bioinformatics 27:592-593.


Simmons, N. B. 2005. Order Chiroptera. Pages 312-529 in D. E. Wilson and D. M. Reeder,

editors. Mammal species of the World: a taxonomic and geographic reference. Johns

Hopkins University Press, Baltimore, Maryland, USA.

Sokal, R. R. and F. J. Rohlf. 1995. Biometry: The Principles and Practice of Statistics in

Biological Research. W.H. Freeman and Co., New York, NY.

Stadelmann, B., L. K. Lin, T. H. Kunz, and M. Ruedi. 2007. Molecular phylogeny of New World

Myotis (Chiroptera, Vespertilionidae) inferred from mitochondrial and nuclear DNA

genes. Molecular Phylogenetics and Evolution 43:32-48.

Steen, D. C., D. B. Hall, P. D. Greger, and C. A. Willis. 1997. Distribution of the chuckwalla,

Western Burrowing owl, and six bat species on the Nevada Test Site. Pages 1-86 in D. o.

Energy, editor. Bechtel Nevada, Las Vegas.

38

Stevens, R. D. and H. Amarilla-Stevens. 2012. Seasonal environments, episodic density

compensation and dynamics of structure of chiropteran frugivore guilds in Paraguayan

Atlantic forest. Biodiversity and Conservation 21:267-279.

Stevens, R. D., M. M. Gavilanez, J. S. Tello, and D. A. Ray. 2012. Phylogenetic structure

illuminates the mechanistic role of environmental heterogeneity in community

organization. Journal of Animal Ecology 81:455-462.

Stevens, R. D. and M. R. Willig. 1999. Size assortment in New World Bat Communities. Journal


Stevens, R. D. and M. R. Willig. 2000. Density compensation in New World bat communities.

Oikos 89:367-377.

Swenson, N. G. 2009. Phylogenetic Resolution and Quantifying the Phylogenetic Diversity and

Dispersion of Communities. PLoS ONE 4:e4390.

Swenson, N. G., B. J. Enquist, J. Pither, J. Thompson, and J. K. Zimmerman. 2006. The problem

and promise of scale dependency in community phylogenetics. Ecology 87:2418-2424.



88:1770-1780.

Swofford, D. L. 2000. PAUP*. Phylogenetic analysis using parsimony (* and other methods).

Sinauer Associates, Inc., Publishers, Sunderland, Massachusetts.

Teeling, E. C., M. Scally, D. J. Kao, M. L. Romagnoli, M. S. Springer, and M. J. Stanhope.

2000. Molecular evidence regarding the origin of echolocation and flight in bats. Nature

403:188-192.

Teeling, E. C., M. S. Springer, O. Madsen, P. Bates, S. J. O'Brien, and W. J. Murphy. 2005. A

Molecular Phylogeny for Bats Illuminates Biogeography and the Fossil Record. Science

307:580-584.

Vaidya, G., D. J. Lohman, and R. Meier. 2011. SequenceMatrix: concatenation software for the

fast assembly of multi-gene datasets with character set and codon information. Cladistics

27:171-180.



592.

Van Den Bussche, R. A. and S. R. Hoofer. 2000. Further evidence for inclusion of the New

Zealand short-tailed bat (Mystacina tuberculata) within Noctilionidea. Journal of

Mammalogy 81:865-874.

39

Webb, C. O. 2000. Exploring the phylogenetic structure of ecological communities: an example

for rain forest trees. American Naturalist 156:145-155.





539.

Williams, J. A., M. J. O'Farrell, and B. R. Riddle. 2006. Habitat use by bats in a riparian corridor

of the Mojave Desert in southern Nevada. Journal of Mammalogy 87:1145-1153.

Willig, M. R. and M. P. Moulton. 1989. The role of stochastic and deterministic processes in

structuring neotropical bat communities. Journal of Mammalogy 70:323-329.

Zwickl, D. J. 2006. Genetic algorithm approaches for the phylogenetic analysis of large

biological sequence datasets under the maximum likelihood criterion. The University of

Texas at Austin.

40

CHAPTER 3

PHYLOGENETIC COMMUNITY STRUCTURE OF NORTH AMERICAN DESERT

BATS: INFLUENCE OF ENVIRONMENT AND ECOLOGICAL TRAITS AT

MULTIPLE SPATIAL AND TAXONOMIC SCALES

INTRODUCTION

Ecologists have long sought to understand mechanisms responsible for the structure of natural

communities. Complex interactions between abiotic and biotic factors, including environmental

heterogeneity, evolutionary history, dispersal ability, timing of colonization, and competition

(Cavender-Bares et al. 2009) have been proposed. Despite much research, some of the most

fundamental questions about how natural communities are assembled remain unanswered. A

relatively recent approach is to investigate phylogenetic community structure (PCS) which uses

phylogenetic information of a regional species pool to make inferences about processes

structuring local communities, thereby tying together ecological and evolutionary processes

(Webb et al. 2002, Cavender-Bares et al. 2009).

A common use of phylogenetic information is to study effects of competition and

environmental filters on community organization. When closely related species have similar

ecological characteristics (Wiens and Graham 2005), phylogenetic relationships among species

can be used to characterize their niches in order to infer the processes involved in community

assembly. Examining roles of nonrandom assembly is of interest in its own right because it

identifies cases of deterministic structure that warrant an explanation. For example, local

communities with species less related to each other than expected by chance (phylogenetic

overdispersion) could result from competition, because closely related species sharing similar

ecological phenotypes are absent and may have been eliminated by competitive exclusion

(Ackerly et al. 2006). In contrast, local communities composed of species that are more related to

41

each other than expected by chance (phylogenetic clustering) may have experienced

environmental filtering; abiotic and biotic factors can remove or filter species incapable of

surviving in a given habitat, leaving mainly closely related species that are similarly adapted

(Webb et al. 2002, Ackerly et al. 2006, but see Mayfield and Levine 2010 for alternative

expectations). In these respects, studying patterns of relatedness within local communities can

provide mechanistic insight into structure (Webb et al. 2002, Cavender-Bares et al. 2009,

Vamosi et al. 2009).

Processes governing which species are found at a given community also depend on scale

(Swenson et al. 2006, Swenson et al. 2007, Cavender-Bares et al. 2009). Biogeographic

processes such as speciation, extinction, and to some extent dispersal ability, may most readily

affect community organization at the largest spatial, temporal, and taxonomic scales (Cavender-

Bares et al. 2006, Swenson et al. 2006, Swenson et al. 2007, Cavender-Bares et al. 2009, Vamosi

et al. 2009). At intermediate scales, dispersal ability and environmental or habitat filtering may

influence community organization because habitat heterogeneity can cause species with similar

environmental requirements to sort out across habitat types (Swenson et al. 2007, Cavender-

Bares et al. 2009, Gómez et al. 2010). Finally, at small spatial, temporal, and taxonomic scales,

density-dependent interactions, such as competition and predation, may be most important in

influencing community structure due to habitat homogeneity and similar resource use among

closely related taxa (Cavender-Bares et al. 2006, Swenson et al. 2006, Cavender-Bares et al.

2009, Vamosi et al. 2009, Gómez et al. 2010). Therefore, numerous concurrent processes may

shape community membership, from large-scale biogeographic processes to small-scale density-

dependent processes (Cavender-Bares et al. 2009).

42

The predictions outlined above can be investigated by combining PCS results from

multiple spatial, temporal, or taxonomic scales with climatic data or information on life history

or functional traits. A significant correlation between PCS and climatic variables would suggest

that environment may strongly influence PCS (Cavender-Bares et al. 2009). Similarly, a strong

correlation between functional or life history traits and phylogeny would suggest that

phylogenetic distance approximates functional distance.

We investigate influences of climate and ecological traits at multiple spatial and

taxonomic scales using bat communities from the four deserts of North America (Figure 3.1).

The Great Basin, Mojave, Sonoran, and Chihuahuan deserts formed as cool Pacific coastal

currents and rain-shadow effects from surrounding mountains limited precipitation (Axelrod

1983). These deserts differ in their climatic regimes and the length of time they have

experienced desert conditions, with the northern Great Basin and Mojave deserts being colder

and the Mojave Desert being younger than the Sonoran and Chihuahuan deserts (Crosswhite and

Crosswhite 1982, Medellin-Leal 1982, Axelrod 1983). These differences have led to distinctive

floras in each desert (Axelrod 1983) and could potentially have led to unique evolutionary

histories for the fauna residing therein as well. Combined, these deserts host 55 species of bats

representing 28 genera and five families. Bats perform many ecosystem services; in North

American deserts, they feed on economically important insects, including crop pests, as well as

serving as pollinators for multiple plant species (Jones et al. 2009a).

Numerous studies have characterized community structure and resource partitioning of

bat communities using morphology, echolocation, habitat use, dietary data, or some combination

of these (e.g., Aldridge and Rautenbach 1987, Campbell et al. 2007, Goncalves da Silva et al.

2008, Stevens and Amarilla-Stevens 2012), with nearly all finding patterns suggesting that bat

43

Figure 3.1: Maps of desert regions of North America showing only US and Mexican states

containing biome13 of the World Wildlife Federation terrestrial ecosystem layers and variation

in climatic variables across the deserts. “Mean temp.” is BIO1, the mean annual temperature

represented as °C*10; “Temp. seas.” is BIO4, temperature seasonality; “Annual precip.” is

BIO12, annual precipitation in millimeters; and “Precip. seas.” is BIO15, precipitation

seasonality.

communities were structured by deterministic processes, such as competition. However, few

studies have compared observed patterns of bat community structure to those generated at

random by a null model. Although some studies using null models found that structure of

communities did not differ significantly from those assembled at random from a regional species

pool (Willig and Moulton 1989, Arita 1997), others found high levels of variability in how well

deterministic models fit data (Stevens and Willig 1999, 2000, Moreno et al. 2006). These

contrasting results suggest multiple mechanisms may be responsible for observed structure of bat

44

communities, and incorporating a comprehensive approach including phylogenetic information

could provide deeper insights and reconcile varying conclusions among studies.

Previously, we examined PCS of all bat taxa in each desert individually and all of these

deserts combined (Chapter 2). Overall we found significant phylogenetic clustering or a

tendency toward phylogenetic clustering in all deserts and in individual deserts except for the

Mojave, which was indistinguishable from randomly generated communities. This suggests that

bat species forming communities in individual deserts are responding to arid conditions in

similar ways. In the present study we focus on the influences of environmental and ecological

characteristics on PCS at two spatial scales (all deserts combined and each desert separately) and

three taxonomic scales (all bat taxa, members of the family Vespertilionidae, and members of the

genus Myotis). If ecological traits are significantly correlated with phylogeny, then phylogenetic

distance can be used as a proxy for ecological distance. Additionally, if communities respond to

harsh climates in similar ways, then the climatic variables correlated with PCS should be the

same across spatial and taxonomic scales.

METHODS

Phylogeny and community data

We used the “best tree” described in Chapter 2; it was inferred by maximum likelihood using

mitochondrial cytochrome b and 12S-16S and nuclear RAG2 gene sequences (to be available on

Dryad).We also used the same communities as Chapter 2. Briefly, bat capture and collection

records were mapped in ArcGIS and combined in three ways so that all bats captured/collected

within predetermined proximity constituted a community. Communities were defined by 1)

drawing 5- and 10-km buffers around all capture/collection locations and combining data from

points whose buffers overlapped; 2) overlaying the map with 10- and 50-km grid cells and

45

combining data from all capture/collection records within each grid cell; and 3) drawing 50- and

100-km circles around as many capture/collection points as possible without overlapping circle

boundaries and combining data from all points within each circle. While our previous study

indicated that different community delimitation methods did not greatly impact PCS results, for

the sake of completeness and direct comparability (Chapter 2), we include results for all

community delimitation methods in the current study.

Species pools

Species pools were established across two different spatial scales: (1) all North American deserts

and (2) species that occur in each of the large North American deserts. Both spatial scales were

sampled at three taxonomic levels: (1) all species (these results were originally reported in

Chapter 2 and are included here to present a more complete narrative), (2) all members of the

family Vespertilionidae and (3) all members of the genus Myotis. This family and genus were

chosen because they are the most species rich taxa within this region. The 15 species pools

created by these combinations of spatial and taxonomic extent and number of taxa in each pool

are summarized in Table 3.1.

Table 3.1: Spatial extents and taxonomic scales used to create species pools for PCS metrics.

Analyses at the level of all species at all scales are referred to as “all taxa”, while the Mojave

species pool contains only species known to occur in the Mojave Desert and is referred to as “MJ

taxa”, and Myotis from the Mojave are referred to as “MJ Myotis”. Numbers in parentheses

indicate the number of taxa in each pool. (“all”= all deserts combined, “GB”= Great Basin

Desert, “MJ”= Mojave Desert, “SN”= Sonoran Desert, “CH”= Chihuahuan Desert, “taxa”= all

taxa, “vesp”= members of the family Vespertilionidae, “Myotis”= members of the genus Myotis)

Spatial Extent

Taxonomic

scale All deserts Great Basin Mojave Sonoran Chihuahuan

All taxa

All taxa

(54) GB taxa (25) MJ taxa (25) SN taxa (39) CH taxa (46)

Vespertilionidae All vesp

(35) GB vesp (22)

MJ vesp

(21) SN vesp (23) CH vesp (30)

Myotis All Myotis

(17)

GB Myotis

(11)

MJ Myotis

(10)

SN Myotis

(11)

CH Myotis

(13)

46

Phylogenetic community structure metrics

We calculated mean pairwise distance (MPD) and mean nearest taxon distance (MNTD) as in

Chapter 2. We used the R package picante (Kembel et al. 2010) to calculate PCS metrics and

obtain standardized effect size (SES-) z-values and p-values for each metric by comparing

empirical values for each community to those from 10,000 communities randomly assembled

from the appropriate species pool (Table 3.1) using the independent swap null model. We

consider phylogenetically clustered communities to have p-values <0.05 and negative z-values

while overdispersed communities have positive z-values and p-values >0.95. To determine

overall significance of SES-MPD and SES-MNTD across communities, we used Fisher’s test of

combined probabilities (Sokal and Rohlf 1995) for each community delimitation method for each

species pool.

Functional traits and environmental data

Interpreting PCS results can be speculative. To aid our explanations, we collected ecological trait

data for species in the regional pool and environmental data for each community. Data on wing

aspect ratio, wing loading, mass, total length, head and body length, tail length, hind foot length,

ear length, diet, niche breadth, and litter size were gathered from PanTHERIA (Jones et al.

2009b), our own data, and other sources (to be available on Dryad). We performed a Mantel test

(based on Pearson’s product-moment correlation; significance based on 10,000 permutations)

between the phylogenetic distance matrix and a Euclidean distance matrix of log-transformed

ecological traits using the R package vegan (Oksanen et al. 2010) for each taxonomic scale.

Annual mean temperature (BIO1; represented as °C*10), temperature seasonality (BIO4;

standard deviation*100), annual precipitation (BIO12; in millimeters), and precipitation

seasonality (BIO15; coefficient of variation) were downloaded directly from WorldClim

47

(Figure3.1; Hijmans et al. 2005). Mean values for each community for each environmental

variable were calculated using the Zonal Statistics as Table function in ArcGIS v. 9.3. These data

will also be available on Dryad Digital Repository. Pearson’s product moment correlation

coefficients were calculated for SES-MPD and SES-MNTD and each environmental variable for

each community delimitation method for each species pool to determine if there was a significant

relationship between environment and PCS.

RESULTS

Phylogenetic community structure

Individual communities range from significantly phylogenetically overdispersed to significantly

phylogenetically clustered (Appendix IV: Tables S1-S5). Because we were more interested in

examining overall patterns of PCS, we will only discuss the results of Fisher’s combined

probability tests (Figure 3.2). The results for the largest taxonomic scale (all taxa; results

summarized in the upper portion of Figure 3.2) were described in detail in Chapter 2 but are

included here to facilitate comparisons.

Phylogenetic community structure analyses for vespertilionids in all deserts combined

(“all vespertilionids” species pool) indicate that only some delimitation methods (10km buffer,

10km grid, 50km circle) were significantly clustered for MPD while all other delimitation

methods and all MNTD communities exhibited structure that was not significantly different from

randomly assembled communities but tended toward phylogenetic clustering (Figure 3.2,

Appendix IV: Table S1). When just Myotis were considered, all metrics and delimitation

methods were non-significant except for MPD for 10km grid communities (Figure 3.2, Appendix

IV: Table S1).

48

(a) Tax

on

Del

imit

atio

n

met

ho

d

All

des

erts

Gre

at

Bas

in

Mo

jave

So

no

ran

Ch

ihu

ahu

an

All

5km

bu

ffer

0

.000

0.0

29

0.5

86

0.0

04

0.0

35

10

km

bu

ffer

0

.000

0.0

40

0.1

76

0.0

12

0.1

24

10

km

gri

d

0.0

00

0.0

93

0.3

03

0.0

01

0.0

22

50

km

gri

d

0.0

00

0.0

30

0.1

56

0.0

01

0.2

16

50

km

cir

cle

0.0

00

0.0

01

0.4

24

0.0

11

0.0

73

10

0km

circ

le

0.0

00

0.1

05

0.3

06

0.0

48

0.3

01

Ves

per

tili

on

idae

5km

bu

ffer

0

.088

0.1

35

0.5

90

0.3

08

0.1

66

10

km

bu

ffer

0

.041

0.1

55

0.1

90

0.3

29

0.2

85

10

km

gri

d

0.0

44

0.1

51

0.2

66

0.2

34

0.1

37

50

km

gri

d

0.0

72

0.2

05

0.2

59

0.1

20

0.2

97

50

km

cir

cle

0.0

30

0.0

37

0.2

37

0.5

37

0.3

89

10

0km

circ

le

0.4

09

0.5

77

0.3

01

0.2

99

0.3

95

Myo

tis

5km

bu

ffer

0

.056

0.9

63

0.4

80

0.0

39

0.0

28

10

km

bu

ffer

0

.061

0.6

80

0.8

85

0.0

36

0.0

31

10

km

gri

d

0.0

09

0.9

94

0.4

60

0.3

52

0.0

02

50

km

gri

d

0.1

59

0.4

91

0.9

48

0.4

58

0.0

29

50

km

cir

cle

0.2

09

0.7

63

0.8

56

0.3

95

0.0

69

10

0km

circ

le

0.1

91

0.5

31

0.9

29

0.2

39

0.0

76

b) T

axo

n

Del

imit

atio

n

met

ho

d

All

des

erts

Gre

at

Bas

in

Mo

jave

So

no

ran

Ch

ihu

ahu

an

All

5km

bu

ffer

0

.001

0.0

51

0.3

96

0.3

47

0.1

00

10

km

bu

ffer

0

.000

0.0

46

0.1

92

0.0

20

0.2

67

10

km

gri

d

0.0

00

0.1

19

0.4

36

0.0

15

0.0

14

50

km

gri

d

0.0

00

0.0

34

0.0

60

0.0

37

0.3

40

50

km

cir

cle

0.0

00

0.0

03

0.4

38

0.0

76

0.0

66

10

0km

circ

le

0.0

01

0.0

81

0.1

80

0.0

15

0.5

53

Ves

per

tili

on

idae

5km

bu

ffer

0

.167

0.1

58

0.5

02

0.4

61

0.4

36

10

km

bu

ffer

0

.079

0.1

22

0.2

30

0.2

08

0.2

91

10

km

gri

d

0.0

69

0.1

82

0.3

49

0.1

96

0.2

15

50

km

gri

d

0.1

21

0.2

22

0.1

67

0.1

63

0.4

86

50

km

cir

cle

0.0

50

0.0

43

0.1

68

0.5

80

0.3

53

10

0km

circ

le

0.5

55

0.5

15

0.3

28

0.2

40

0.7

67

Myo

tis

5km

bu

ffer

0

.414

0.9

40

0.4

63

0.5

48

0.6

69

10

km

bu

ffer

0

.205

0.7

49

0.5

82

0.1

86

0.0

59

10

km

gri

d

0.1

58

0.9

57

0.3

49

0.1

33

0.7

97

50

km

gri

d

0.6

87

0.4

71

0.8

77

0.5

31

0.5

34

50

km

cir

cle

0.4

95

0.8

00

0.6

93

0.4

06

0.1

26

10

0km

circ

le

0.6

97

0.5

83

0.6

74

0.4

18

0.5

24

C

lust

ered

(si

g.;

p-v

alues

<0.0

01-0

.049)

C

lust

ered

(ns;

p-v

alues

0.0

5-0

.29)

N

ot

signif

ican

t (p

-val

ues

0.3

0-0

.69)

O

ver

dis

per

sed (

ns;

p-v

alues

0.7

0-0

.94)

O

ver

dis

per

sed (

sig.;

p-v

alues

0.9

5-1

.0)

Fig

ure

3.2

: A

ll F

isher

’s c

om

bin

ed p

robab

ilit

y t

est

p-v

alues

for

all

spec

ies

pools

and d

elim

itat

ion m

ethods,

colo

r-co

ded

by s

ignif

ican

ce.

(a)

SE

S-M

PD

res

ult

s. (

b)

SE

S-M

NT

D r

esult

s.

49

Communities of vespertilionids in each desert were not significantly different from

random but tended to be phylogenetically clustered (Figure 3.2, Appendix IV: Tables S1-S5).

Patterns of PCS for Myotis differed between deserts. Both metrics for Great Basin Desert Myotis

indicated communities tended toward or were significantly overdispersed for most delimitation

methods. Results for Sonoran and Chihuahuan Desert Myotis indicate phylogenetic clustering for

SES-MPD but were not significantly different from random for SES-MNTD (Figure 3.2,

Appendix IV: Tables S5-S8). Mojave Desert Myotis communities were not significantly different

from random (Figure 3.2, Appendix IV: Table S3).

Functional traits and environmental data

Ecological traits and phylogenetic distance were positively correlated for the three taxonomic

scales considered (Table 3.2). PCS was positively correlated with mean annual temperature

(BIO1) at the largest spatial scale regardless of taxonomic scale. PCS metrics also increase with

increasing temperature for Great Basin taxa and vespertilionids. Additionally, Sonoran SES-

MPD increases with increasing temperature for all three taxonomic scales but was not significant

for all delimitation methods (Appendix IV: Table S6). Temperature seasonality (BIO4) was

significantly negatively correlated with PCS for all three taxonomic scales in all deserts and for

Sonoran Myotis (Appendix IV: Table S7). Annual precipitation (BIO12) was not consistently

correlated with SES-MPD or SES-MNTD (Appendix IV: Table S8). Mean precipitation

seasonality (BIO15) was positively correlated with PCS metrics at the largest spatial scale but

there was no discernible pattern for individual deserts Appendix IV: (Table S9).

Table 3.2: Results of Mantel tests between phylogenetic distance and ecological traits.

Taxon Mantel statistic p-value

All taxa 0.598 <0.001

Vespertilionidae 0.417 <0.001

Myotis 0.544 0.002

50

DISCUSSION

We found that bat communities tend to be phylogenetically clustered, or made up of closely

related species, to a greater degree at the largest spatial scale and to a lesser degree at smaller

spatial and taxonomic scales. This nonrandom pattern suggests that deterministic processes were

involved in community assembly. This is further supported by strong correlation between

phylogeny and ecological traits as well as between PCS metrics and temperature and, to a lesser

extent, seasonality.

Previous work has shown that there is often a relationship between environmental

variables and PCS metrics but the importance of individual environmental variables varies by

taxon and region. For example, precipitation is correlated with PCS in Australian honeyeaters

(Miller et al. 2013); temperature is related to PCS in ants (Machac et al. 2011), Australian

vertebrates (Lanier et al. 2013), and Himalayan leaf warblers (Ghosh-Harihar 2014); light for

Minnesota plants (Willis et al. 2010); and a suite of environmental factors were significantly

correlated with PCS for alpine tundra plants (Spasojevic and Suding 2012), grassland plants

(Soliveres et al. 2012), and antbirds (Gómez et al. 2010).

Functional traits and environmental characteristics

In this study system, phylogeny reflects not only evolutionary history of our focal taxa, but also

ecology (Table 3.2). This means that phylogenetically clustered communities were made up of

species that had similar ecological traits while phylogenetically overdispersed communities

contained species that had dissimilar ecological traits. These results suggest that traditional

interpretations of PCS are applicable to North American desert bats. A more thorough

examination of morphological community structure in this system is forthcoming.

51

In this study system, some environmental factors were significantly correlated with PCS

metrics at particular spatial and taxonomic scales, suggesting their importance in community

assembly. Significant positive correlations between PCS metrics and the climatic variables mean

annual temperature (BIO1; all deserts, the Sonoran desert, and Great Basin taxa and

vespertilionids) and precipitation seasonality (BIO15; all deserts) indicates that more clustered

communities tend to occur where annual temperatures and precipitation seasonality are lower

while more overdispersed communities tend to occur where annual temperatures and

precipitation seasonality are higher (Figure 3.3, Appendix IV: Tables S6 and S9). Significant

negative correlations between PCS and temperature seasonality (BIO4; all deserts and

Chihuahuan Myotis) indicated that more clustered communities occurred in areas of high

temperature seasonality while more overdispersed communities tend to occur where

temperatures are more constant (Figure 3.3, Appendix IV: Table S7). Annual precipitation is not

significantly correlated with PCS, suggesting minimal importance in community structure

(Appendix IV: Table S8). Previous studies have found similar significant correlations between

PCS and environmental variables suggesting a general pattern of phylogenetic clustering in harsh

conditions and overdispersion in milder conditions (e. g., Anderson et al. 2011, Spasojevic and

Suding 2012, Miller et al. 2013, Stevens and Gavilanez in review). “Harsh” conditions would be

those posing greater physiological challenges to maintaining homeostasis for the focal taxa

whereas “mild” conditions would allow homeostasis to be more easily maintained.

It is somewhat surprising that annual precipitation and precipitation seasonality were not

significantly correlated with PCS as other authors (Patten 2004, McCain 2007) have shown them

to predict bat species richness. While species richness does not necessarily predict PCS, one

explanation for the difference between our results and Patten’s (2004) is that vespertilionids

Figure 3.3: Schematic diagram illustrating positive (solid line) and negative (dashed line)

correlations between PCS metrics and environmental variables.

(species richness driven by temperature; Patten 2004)

is driven by precipitation; Patten 2004)

Spatial and taxonomic scale

Numerous studies have shown that spatial, temporal, and taxonomic scale can

results and interpretation (e. g., Cavender

Ackerly 2010, Cardillo 2011), which is why we were explicit about the spatial and taxonomic

scales used in this study. Habitat filtering is expected

habitat requirements across the landscape, resulting in a tendency for phylogenetic clustering

(Cavender-Bares et al. 2009, Gómez et al. 2010)

overdispersion is expected, as interspecific interactio

number with more habitat homogeneity at smaller spatial scales and niche conservatism at small

taxonomic scales (Cavender-Bares et al. 2009, Gómez et al. 2010)

52

3: Schematic diagram illustrating positive (solid line) and negative (dashed line)

correlations between PCS metrics and environmental variables.

(species richness driven by temperature; Patten 2004) outnumber phyllostomids (species richness

is driven by precipitation; Patten 2004) in our study region and may be driving this trend.

Numerous studies have shown that spatial, temporal, and taxonomic scale can influence

(e. g., Cavender-Bares et al. 2006, Swenson et al. 2007, Kraft and

which is why we were explicit about the spatial and taxonomic

abitat filtering is expected at large scales, as species are


Bares et al. 2009, Gómez et al. 2010). At smaller scales, a tendency for phylogenetic

as interspecific interactions are thought to increase in intensity and

with more habitat homogeneity at smaller spatial scales and niche conservatism at small

Bares et al. 2009, Gómez et al. 2010). Overall, our findings were

3: Schematic diagram illustrating positive (solid line) and negative (dashed line)

(species richness

in our study region and may be driving this trend.

influence PCS

Bares et al. 2006, Swenson et al. 2007, Kraft and

which is why we were explicit about the spatial and taxonomic

as species are sorted by


. At smaller scales, a tendency for phylogenetic

in intensity and

with more habitat homogeneity at smaller spatial scales and niche conservatism at small

. Overall, our findings were

53

consistent with these expectations: at the largest spatial and taxonomic scales, there was a

tendency for phylogenetic clustering while this tendency decreased at smaller scales (Figure 3.2,

Appendix IV: Tables S1-5, Chapter 2). Interestingly, climatic variables correlated with PCS also

varied with spatial and taxonomic scale (Appendix IV: Tables S6-9) indicating that environment

may have a great impact on community structure at some scales but not at others.

Phylogenetic community structure of desert bat communities

At the largest spatial scale (i.e., all deserts) there is an overall tendency for phylogenetic

clustering at all taxonomic scales (Figure 3.2). The traditional explanation for this pattern is that

habitat filtering is important for structuring communities. Based on ecological traits and

environmental data, we suggest that this is the case. The correlation between PCS and

temperature (BIO1, positive correlation, Appendix IV: Table S6) and temperature seasonality

(BIO4, negative correlation, Appendix IV: Table S7) indicates that communities in colder areas

with more seasonal temperature variation tend to be phylogenetically clustered while

communities in warmer, more thermally stable areas tend to be less phylogenetically clustered

(Figure 3.3). Precipitation seasonality (BIO15) was also positively correlated with PCS metrics

contrary to our expectations; clustered communities tend to experience less variable precipitation

(Figure 3.3, Appendix IV: Table S9). That annual precipitation was not correlated with PCS

metrics may suggest that overall, temperature is of overriding importance of the variables we

examined. This perhaps is not surprising given that thermoregulation is important for bats with

many species resorting to torpor or migration given unfavorable weather conditions. Species that

may not typically utilize these strategies, such as members of the family Phyllostomidae, may

effectively be excluded from habitats in unsuitable climates.

54

Interestingly, while individual deserts (except the Mojave) tend to share similar patterns

of community structure for all taxa and vespertilionids, each have unique correlation patterns in

relation to the climatic variables we examined. Great Basin taxa and vespertilionids tended to be

phylogenetically clustered and their PCS metrics were significantly positively related to

temperature (BIO1; Appendix IV: Table S6); this is perhaps unsurprising given that there is a

much broader range of temperatures in this desert compared to the others (Figure 3.1). However,

the genus Myotis tended to be phylogenetically overdispersed and was uncorrelated with any

environmental factor (Figure 3.2, Appendix IV: Tables S6-S9). This suggests the possibility that

interspecific interactions may be more important than habitat filtering in structuring these

taxonomically restricted communities.

Mojave vespertilionids and Myotis communities were not significantly different from

randomly assembled communities and were not correlated with any climatic variables examined

(Figure 3.2, Appendix IV: Tables S6-S9; Chapter 2). These observations defy ready explanation,

but one possibility is that this smallest desert has relatively uniform, albeit harsh, conditions

(Fig.3.1) such that individual communities were assembled randomly from a species pool that

had already been filtered to contain only those species physiologically capable of surviving the

area.

Sonoran communities at all taxonomic scales tend to be phylogenetically clustered

(although this tendency is lessened in Myotis, Figure 3.2) and are significantly positively

correlated with mean temperature (BIO1, Appendix IV: Table S6) again suggesting that

phylogenetically clustered communities tend to be in harsher environments while less clustered

communities tend to be found in less harsh habitats (Figure 3.3).

55

Similar to the Sonoran Desert, Chihuahuan communities at all three taxonomic scales

tend to be phylogenetically clustered (Figure 3.2). However, none of the environmental variables

examined (Appendix IV: Tables S6-S9) were significantly correlated with PCS with the

exception of temperature seasonality which is significantly negatively correlated in Myotis

(BIO4, Appendix IV: Table S7). While the Chihuahuan Desert exhibits the most variation in

temperature seasonality of the four deserts examined here with the northern reaches much more

thermally variable than the southern region (Figure 3.1), communities containing three or more

Myotis species were almost exclusively restricted to the northern reaches of this desert (data

available on the Dryad Digital Repository). More southerly communities contained Myotis

species, which were included in the Chihuahuan Myotis species pool (Table 3.1), but were not

actually present in most Chihuahuan Myotis communities, thereby inflating the species pool

against which these communities were compared, so potentially leading to the observed pattern

of significant phylogenetic clustering. Even with these caveats, clustered communities of Myotis

tend to be found in areas of high temperature seasonality while less clustered communities tend

to occur in areas of low temperature seasonality (Figure 3.3).

Our results are the first to show that North American bat communities are indeed

significantly structured over a large area, not just random assortments of species from the

regional pool. Based on echolocation call frequency, wing shape, diet, habitat preferences, and

temporal activity, previous work has found evidence that bats can partition similar resources by

making use of slightly different foraging strategies, thereby reducing interspecific competition

(e.g. Black 1974, Findley 1976, Findley and Black 1983, Aldridge and Rautenbach 1987,

Findley 1993). Others studies have suggested that bat communities are assembled randomly

56

from the regional species pool or that simplistic models fit these data poorly (e.g. Schum 1984,

Willig and Moulton 1989, Arita 1997, Stevens and Willig 1999, Cardillo et al. 2008).

Phylogenetic clustering, the pattern we observed most frequently in this study system, has

typically been interpreted as indicating habitat filtering has structured communities if traits

important to coexistence are phylogenetically conserved (Emerson and Gillespie 2008). Mayfield

and Levine (2010) proposed, alternatively, that phylogenetic clustering may indicate competition

if species possess phylogenetically conserved traits that allow them to out compete more

distantly related species lacking those traits. Another alternative explanation is that traits

important to coexistence are not phylogenetically conserved such that phylogenetically clustered

communities are morphologically overdispersed, potentially indicative of competition structuring

communities. Here, Myotis is suspected to have undergone convergent evolution multiple times

(Ruedi and Mayer 2001, Stadelmann et al. 2007), potentially undermining traditional

interpretations of PCS patterns. However, since ecological traits are significantly correlated with

phylogenetic distance regardless of taxonomic scale (Table 3.2), the former alternative

explanation may be more plausible than the latter; we are examining these questions more

directly in an upcoming contribution. Despite these alternative explanations, we suggest that the

conventional interpretation of phylogenetic clustering is applicable in this study system based on

significant correlations between PCS and environmental variables: overall, desert bat

communities seem to be structured predominantly by habitat filtering. A previous study of bat

PCS in Bavaria has also observed phylogenetically clustered communities with bat species being

filtered by anthropogenic habitat disturbance (Riedinger et al. 2012).

Our results also suggest that while overall patterns of PCS were similar among deserts,

the environmental factors driving these patterns differed by taxon and desert (Figures 3.2 and

57

3.3, Appendix IV: Tables S6-S9). These deserts, while all harsh, do not necessarily each present

organisms with the same ecological challenges. This is evidenced by the plant communities

unique to each desert (Shreve 1942) but has not been demonstrated in bat communities before.

This indicates that ecological pressures impacting the bat communities in these regions differ

although the resultant PCS patterns appear to be the same. These differences in environmental

patterns among deserts may perhaps stem from constriction of desert ecosystems in western

North America during the last glacial maximum (Adams 1997). This would likely have

concomitantly restricted the ranges of desert bat communities in disparate geographical areas

with potentially unique climatic regimes influencing the species residing there. As deserts

expanded after the last glacial maximum to their current extents (Adams 1997), so too would

desert bats expand across the landscape. We may be observing the influence of these differing

climatic refugia on the bat communities now living in these expanded deserts. We would not

have been able to tease apart the climatic drivers of community structure had we not examined

multiple spatial and taxonomic scales.

CONCLUSIONS

Patterns of PCS at different spatial and taxonomic scales previously described by other authors,

namely tendency towards clustering at large scales which decreases at small scales, were

observed in our study system. The overall consistency of these patterns across deserts suggests

that bat communities may respond similarly to ecological pressures and indicates that habitat

filtering is important in community assembly. In this study system, phylogeny was a good proxy

for ecological traits and phylogenetically clustered communities tended to occur in harsher

habitats. However, climatic variables did not impact communities at different spatial and

58

taxonomic scales in the same ways suggesting that while the observed patterns of PCS were

similar, the evolutionary and ecological routes may be different.

ACKNOWLEDGEMENTS

L.E.P. was funded by an American Museum of Natural History Theodore Roosevelt Memorial

Grant, Society of Systematic Biologists Graduate Student Award, American Society of

Mammalogists Grants-in-Aid of Research award, and Louisiana State University BioGrads

awards BG11-38. We also thank J. S. Tello, E. McCulloch, M. M. Gavilanez, S. M. Hird, and

anonymous reviewers for fruitful discussions during the development of this project and editing

previous versions of this manuscript.

REFERENCES



Adams, J. M. 1997. Global land environments since the last interglacial. Oak Ridge National

Laboratory, TN, USA.



Anderson, T. M., J. Shaw, and H. Olff. 2011. Ecology’s cruel dilemma, phylogenetic trait

evolution and the assembly of Serengeti plant communities. Journal of Ecology 99:797-

806.

Arita, H. T. 1997. Species composition and morphological structure of the bat fauna of Yucatan,

Mexico. Journal of Animal Ecology 66:83-97.




Black, H. L. 1974. A North Temperate Bat Community: Structure and Prey Populations. Journal


59




Cardillo, M. 2011. Phylogenetic structure of mammal assemblages at large geographical scales:

linking phylogenetic community ecology with macroecology. Philosophical Transactions

of the Royal Society B: Biological Sciences 366:2545-2553.

Cardillo, M., J. L. Gittleman, and A. Purvis. 2008. Global patterns in the phylogenetic structure

of island mammal assemblages. Proceedings of the Royal Society B: Biological Sciences

275:1549-1556.

Cavender-Bares, J., A. Keen, and B. Miles. 2006. Phylogenetic structure of Floridian plant

communities depends on taxonomic and spatial scale. Ecology 87:109-122.



Crosswhite, F. S. and C. D. Crosswhite. 1982. The Sonoran Desert. Pages 163-295 in G. L.

Bender, editor. Reference Handbook on the Deserts of North America. Greenwood Press,

Westport, CT.



Findley, J. S. 1976. The structure of bat communities. The American Naturalist 110:120-139.

Findley, J. S. 1993. Bats: A community perspective. Cambridge University Press, Cambridge.

Findley, J. S. and H. Black. 1983. Morphological and dietary structuring of a Zambian

Insectivorous bat community. Ecology 64:625-630.

Ghosh-Harihar, M. 2014. Phylogenetic and ecomorphological structure of assemblages of

breeding leaf warblers (Phylloscopidae) along Himalayan elevational gradients. Journal

of Biogeography:n/a-n/a.




Goncalves da Silva, A., O. Gaona, and R. A. Medellin. 2008. Diet and trophic structure in a

community of fruit-eating bats in Lacandon Forest, Mexico. Journal of Mammalogy

89:43-49.

Hijmans, R. J., S. E. Cameron, J. L. Parra, P. G. Jones, and A. Jarvis. 2005. Very high resolution

interpolated climate surfaces for global land areas. International Journal of Climatology

25:1965-1978.

60

Jones, G., D. S. Jacobs, T. H. Kunz, M. R. Willig, and P. A. Racey. 2009a. Carpe noctem: the


Jones, K. E., J. Bielby, M. Cardillo, S. A. Fritz, J. O'Dell, C. D. L. Orme, K. Safi, W. Sechrest,

E. H. Boakes, C. Carbone, C. Connolly, M. J. Cutts, J. K. Foster, R. Grenyer, M. Habib,

C. A. Plaster, S. A. Price, E. A. Rigby, J. Rist, A. Teacher, O. R. P. Bininda-Emonds, J.

L. Gittleman, G. M. Mace, A. Purvis, and W. K. Michener. 2009b. PanTHERIA: a

species-level database of life history, ecology, and geography of extant and recently

extinct mammals. Ecology 90:2648-2648.




Kraft, N. J. B. and D. D. Ackerly. 2010. Functional trait and phylogenetic tests of community

assembly across spatial scales in an Amazonian forest. Ecological Monographs 80:401-

422.

Lanier, H. C., D. L. Edwards, and L. L. Knowles. 2013. Phylogenetic structure of vertebrate

communities across the Australian arid zone. Journal of Biogeography 40:1059-1070.

Machac, A., M. Janda, R. R. Dunn, and N. J. Sanders. 2011. Elevational gradients in

phylogenetic structure of ant communities reveal the interplay of biotic and abiotic

constraints on diversity. Ecography 34:364-371.



McCain, C. M. 2007. Could temperature and water availability drive elevational species richness

patterns? A global case study for bats. Global Ecology and Biogeography 16:1-13.

Medellin-Leal, F. 1982. The Chihuahuan Desert. Pages 321-372 in G. L. Bender, editor.

Reference Handbook on the Deserts of North America. Greenwood Press, Westport, CT.

Miller, E. T., A. E. Zanne, and R. E. Ricklefs. 2013. Niche conservatism constrains Australian

honeyeater assemblages in stressful environments. Ecology Letters 16:1186-1194.

Moreno, C. E., H. T. Arita, and L. Solis. 2006. Morphological assembly mechanisms in

Neotropical bat assemblages and ensembles within a landscape. Oecologia 149:133-140.


Henry, H. Stevens, and H. Wagner. 2010. vegan: Community Ecology Package.

Patten, M. A. 2004. Correlates of species richness in North American bat families. Journal of

Biogeography 31:975-985.

61



Ruedi, M. and F. Mayer. 2001. Molecular systematics of bats of the genus Myotis

(Vespertilionidae) suggests deterministic ecomorphological convergences. Molecular

Phylogenetics and Evolution 21:436-448.

Schum, M. 1984. Phenetic structure and species richness in North and Central American bat

faunas. Ecology 65:1315-1324.




Soliveres, S., R. Torices, and F. T. Maestre. 2012. Environmental conditions and biotic

interactions acting together promote phylogenetic randomness in semi-arid plant

communities: new methods help to avoid misleading conclusions. Journal of Vegetation

Science 23:822-836.

Spasojevic, M. J. and K. N. Suding. 2012. Inferring community assembly mechanisms from

functional diversity patterns: the importance of multiple assembly processes. Journal of

Ecology 100:652-661.




Stevens, R. D. and H. Amarilla-Stevens. 2012. Seasonal environments, episodic density

compensation and dynamics of structure of chiropteran frugivore guilds in Paraguayan

Atlantic forest. Biodiversity and Conservation 21:267-279.

Stevens, R. D. and M. M. Gavilanez. in review. Dimensionality of community structure:

phylogenetic, phenetic and functional perspectives along biodiversity and environmental

gradients (n.d.).



Stevens, R. D. and M. R. Willig. 2000. Density compensation in New World bat communities.

Oikos 89:367-377.

Swenson, N. G., B. J. Enquist, J. Pither, J. Thompson, and J. K. Zimmerman. 2006. The problem

and promise of scale dependency in community phylogenetics. Ecology 87:2418-2424.

62



88:1770-1780.



592.





539.

Willig, M. R. and M. P. Moulton. 1989. The role of stochastic and deterministic processes in

structuring neotropical bat communities. Journal of Mammalogy 70:323-329.

Willis, C. G., M. Halina, C. Lehman, P. B. Reich, A. Keen, S. McCarthy, and J. Cavender-Bares.

2010. Phylogenetic community structure in Minnesota oak savanna is influenced by

spatial extent and environmental variation. Ecography 33:565-577.

63

CHAPTER 4

MORPHOLOGICAL COMMUNITY STRUCTURE OF NORTH AMERICAN DESERT

BATS: ASSESSING PHYLOGENETIC SIGNAL IN MORPHOLOGICAL TRAITS AND

COMPARISON WITH PHYLOGENETIC COMMUNITY STRUCTURE

INTRODUCTION

Ecomorphology was traditionally used by ecologists to understand how organisms function

within their niche (Ricklefs and Miles 1994). Using functional traits or morphology to elucidate

community assembly processes has enjoyed a revival of sorts over the past decade (McGill et al.

2006). Morphological distance can be used to estimate resource partitioning (Hespenheide 1973),

so investigating how morphologically similar members of a community are to each other can

provide insights into the ecological processes structuring that community. However, this

provides relatively little information on evolutionary patterns that may also be structuring a

community. Phylogeny can represent a general estimate of overall phenotype that includes

information such as life history, behavior, and environmental tolerances (Webb 2000). Well

resolved phylogenies for an increasing number of taxa used in combination with trait data and

null models have allowed researchers to explicitly test ecological and evolutionary hypotheses.

One approach is to investigate phylogenetic community structure (PCS) to determine if species

found in a community are more (phylogenetically clustered) or less (phylogenetically

overdispersed) related to each other than expected by chance. These patterns may suggest that

environmental filtering or competition, respectively, have structured such communities.

Communities made up of morphologically clustered species may be experiencing

environmental filtering while morphologically overdispersed communities may be structured by

competition. For organisms with evolutionarily labile phenotypes, such as Anolis lizards (Losos

et al. 1998) and Myotis bats (Ruedi and Mayer 2001, Stadelmann et al. 2007), phylogeny and

morphology do not completely correspond, that is, morphological or ecological traits lack

64

phylogenetic signal, which is the tendency for closely related species to possess characteristics

more similar to each other than more distantly related species (Losos 2008). Evolutionarily

labile phenotypes and convergent evolution could lead to communities that are morphologically

overdispersed and phylogenetically clustered, or morphologically clustered and phylogenetically

overdispersed (Webb et al. 2002, Emerson and Gillespie 2008, Losos 2008). However

phylogenetic signal is often assumed of clades without substantiation (Webb 2000, Losos 2008).

This assumption has been recently challenged (Losos 2008). To accurately infer the processes

producing patterns of community structure, ecological traits, including morphology, should be

examined in conjunction with phylogenetic distance.

We focus on bats occurring in the four desert regions of North America. The Great Basin,

Mojave, Sonoran, and Chihuahuan deserts differ in age and floral assemblages (Shreve 1942) but

were all formed by the combined forces of rain shadow effects of surrounding mountains and

cool Pacific ocean currents (Axelrod 1983). These deserts host 56 species or subspecies of bats

in 28 genera and five families. Bats have the most diverse feeding habits of all mammals and

perform many ecosystem functions (Jones et al. 2009); in these deserts the majority of species

are insectivorous, feeding on economically important insects (Cleveland et al. 2006, Jones et al.

2009), or nectarivorous (4 species), that pollinate several important plant species (Jones et al.

2009). In addition, a frugivore, a piscivore, and two sanguivores have been found infrequently in

the southern reaches of the Chihuahuan and Sonoran Deserts. This study system is ideal to test

how PCS corresponds to morphological community structure (MCS); of the 55 species of bats in

this region, 33 belong to the family Vespertilionidae, and 14 of these vespertilionid species

belong to the genus Myotis. As previously mentioned, Myotis is suspected to have gone through

convergent evolution multiple times; by examining communities at multiple taxonomic scales,

65

we can determine what if any impact potential convergence has on traditional interpretations of

PCS and MCS.

Morphological investigations of animal communities typically focus on structures used in

locomotion or feeding (Hespenheide 1973). For bats, these structures are the wing and skull

(Swartz et al. 2003). Wing morphology plays a large role in determining how bats forage. Bats

with short broad wings are highly maneuverable (similar to fighter jets), whereas bats with long

narrow wings are efficient long-distance fliers (similar to passenger airplanes; Norberg and

Rayner 1987). Skull morphology varies with diet. For example, bats that eat hard foods (such as

beetles) tend to have more robust, stronger cranial structures than bats that eat softer foods (e.g.,

moths; Freeman 1981a, b, Gannon and Racz 2006). Most studies of morphological similarity

among co-occurring bats have described patterns suggesting that competitive interactions

structure these communities (e.g. Findley 1976, Findley and Black 1983, Aldridge and

Rautenbach 1987, Barlow et al. 1997, Aguirre et al. 2002, Rhodes 2002, Campbell et al. 2007,

York and Papes 2007). However, none have interpreted their results in the context of

phylogenetic community structure.

Previous work on PCS of North American desert bats has indicated a general tendency

for phylogenetic clustering at the largest spatial or taxonomic scales that becomes less

pronounced at smaller scales (Chapters 2 and 3). We have also demonstrated that PCS metrics

are correlated with climatic variables, indicating that communities in environmentally harsher

areas are made up of species that are more closely related to each other than those in

communities in less harsh environments (Chapter 3). These results suggest that habitat filtering

may structure these communities, but without investigating the assumption of phylogenetic

66

signal, this interpretation of observed pattern may be suspect (Emerson and Gillespie 2008,

Losos 2008).

Our objective in this study is to investigate whether North American desert bat

communities are made up of morphologically similar species (indicating habitat filtering),

morphologically dissimilar species (competition), or are not different from communities

assembled randomly from the local spceis pool. To date, nearly all studies of bat community

structure have found either no structure or evidence for competition; only 3 have suggested

habitat filtering (Riedinger et al. 2012, Chapters 2 and 3). We test whether MCS corresponds

with PCS and quantify the strength of phylogenetic signal in the traits we measured. A strong

positive correlation between phylogeny and morphology would indicate that morphological traits

exhibit phylogenetic signal while a strong negative correlation would indicate that morphological

traits are convergent (Davis 2005, Losos 2008). At the largest taxonomic scale (all bats), we

expect that morphology will exhibit more phylogenetic signal than at the smallest taxonomic

scale (the genus Myotis) because this genus is suspected to have undergone convergent evolution

multiple times (Ruedi and Mayer 2001, Stadelmann et al. 2007).

METHODS

Community data

Communities were those used in Chapter 2; data will be submitted to Dryad. Capture and

collection records were obtained from a variety of sources and mapped using GIS. Data with

identical coordinates within deserts as defined by the World Wildlife Federation’s terrestrial

ecosystem layers (biome13; Olson et al. 2001) were combined and used to delimit communities

in 3 ways: 1) buffers with radii of 5 and 10km were drawn around each capture/collection

location; when buffer boundaries touched, data from all touching buffers were combined to

67

create a community. 2) Grids with cells 10x10 and 50x50km were overlaid on the map and data

from capture/collection locations that fell within the same cell were combined. 3) Finally, circles

with diameters of 50 and 100km were placed on the map to encompass as many

capture/collection locations as possible, but at least four; data from capture/collection locations

falling within the same circle were combined. The richness estimator Chao1 (Colwell 2009,

Oksanen et al. 2010b) was calculated for each community using the R package vegan (Oksanen

et al. 2010a) to ensure that only adequately sampled communities were used in subsequent

analyses; communities containing three or more species were considered adequately sampled if

observed species richness fell within the 95% confidence interval of the estimator.

Morphological traits

Ten males and ten females (or all available specimens if fewer than 20 were available) of each

bat species occurring in North American deserts were measured with digital calipers.

Measurements taken from each specimen are illustrated in Figure 4.1 and specimens examined

can be found in Appendix V. When possible, we measured specimens collected from desert

regions to account for the possible effects of morphological plasticity. The log-transformed mean

of each trait for each species was used in the following analyses.

Species pools

We delimited several species pools to determine if observed community values differed from

values generated by randomly assembling communities from species found in the appropriate

pool. We used pools identical to those in Chapters 2 and 3 except that Eumops underwoodi was

included in the present study (excluded previously due to lack of genetic material). We delimited

pools that differed in spatial and taxonomic extent: 1) all North American deserts combined and

Figure 4.1: Skull and wing measurements taken from each specimen.

each measurement can be found in [1] Freeman

(2013), [3] van Zyll de Jong (1979)

[1], 2=greatest skull length maxilla [2], 3=rostrum length premaxilla (this study; from cribiform

plate anteriormost point of maxillary bone)

canine [1], 6=length of temporal fossa [2], 7=height of braincase [1], 8=breadth at mastoids [1],

9=breadth of braincase [1], 10=rostrum width [1], 11=postorbital width [1], 12=width at upper

canines [1] , 13=length of maxillary toothrow [1], 14=length of upper molar row [2], 15=length

of M3 [1], 16=length M2 [2], 17=width of M3 [1], 18=width M2 [2], 19=intermolar breadth [2],

20=palatal length (premaxillary) [1], 21=palatal length maxilla (this study; f

border of the hard palate to the anterior border of the maxillary bone) , 22=zygomatic breadth

[1], 23=width at anterior pterygoids [1], 24=width at posterior pterygoids [1], 25=condylocanine

length [1], 26=dentary length [1], 27=height o

29=length of lower tooth row [1], 30=dentary thickness [2], 31=length of condyle to M1 [1],

32=rostral width immediately posterior to canines [3], 33=palatal width at P2 [3], 34=basal width

of upper canine at the cingulum [3], 35=height of coronoid process [4], 36=distance from

angular process to coronoid [4], 37=distance from articular process to angular process [4],

38=distance from coronoid process to articular process [4], 39=total toothrow length [4],

40=condylobasal length [2], 41=len

43=length of third metacarpal first phalanx [1], 44=length of third metacarpal second phalanx

[1], 45=length of third metacarpal third phalanx and tip [1], 46=length of fo

47=length of fourth metacarpal first phalanx [1], 48=length of fourth metacarpal second phalanx

[1], 49=length of fifth metacarpal [1], 50=length of fifth metacarpal first phalanx [1], 51=length

of fifth metacarpal second phalanx [1].

b=digit 4 [1], c=digit 5 [1], d=aspect ratio [1],

and g=jaw closure ratio [4].

68

Skull and wing measurements taken from each specimen. Detailed descriptions of

each measurement can be found in [1] Freeman (1981b), [2] Patrick, McCulloch, and Ruedas

(1979), or [4] Gannon and Racz (2006). 1=greatest length of skull


plate anteriormost point of maxillary bone), 4=rostrum length maxilla [2], 5=height of the upper



] , 13=length of maxillary toothrow [1], 14=length of upper molar row [2], 15=length


20=palatal length (premaxillary) [1], 21=palatal length maxilla (this study; from the posterior



length [1], 26=dentary length [1], 27=height of coronoid [1], 28=height of lower canine [1],



the cingulum [3], 35=height of coronoid process [4], 36=distance from



ondylobasal length [2], 41=length of forearm [1], 42=length of third metacarpal [1],


[1], 45=length of third metacarpal third phalanx and tip [1], 46=length of fourth metacarpal [1],



of fifth metacarpal second phalanx [1]. Other variables not shown on the figure are

aspect ratio [1], e=tip index [1], f=digit 3 divided by digit 5 [1],

Detailed descriptions of

Culloch, and Ruedas

1=greatest length of skull


, 4=rostrum length maxilla [2], 5=height of the upper



] , 13=length of maxillary toothrow [1], 14=length of upper molar row [2], 15=length


rom the posterior



f coronoid [1], 28=height of lower canine [1],



the cingulum [3], 35=height of coronoid process [4], 36=distance from



gth of forearm [1], 42=length of third metacarpal [1],


urth metacarpal [1],



Other variables not shown on the figure are a=digit 3 [1],

digit 3 divided by digit 5 [1],

69

2) individual deserts as well as a) all bat taxa in North American deserts, b) only members of the

family Vespertilionidae, and c) only members of the genus Myotis.

Data analyses

To characterize the distribution of species in morphological space, we calculated mean pairwise

distance (MPD; Webb 2000, Webb et al. 2002) and mean nearest taxon distance (MNTD; Webb

2000, Webb et al. 2002) for each community. MPD is the mean distance between all pairs of

species in a particular community while MNTD is the mean distance to the nearest species

within a particular community. These metrics were calculated in the R package picante (Kembel

et al. 2010) using a Euclidean distance matrix derived from the morphological data. These

observed values were then compared to 10,000 communities generated randomly using the

independent swap null model drawing species from the appropriate pool in order to obtain

standardized effect size (SES) z- and p-values. Kembel (2009) has shown that this null model

performs well in distinguishing community assembly processes. Communities that have positive

z-values and p-values >0.95 are considered to be significantly overdispersed while communities

with negative z-values and p-values <0.05 are considered to be significantly clustered when

α=0.10. MPD and MNTD were calculated from distance matrices containing both skull and wing

data combined (hereafter referred to as “both”), just skull data, and just wing data. In order to

assess overall trends, Fisher’s test of combined probability (Sokal and Rohlf 1995) was

calculated for SES-MPD and SES-MNTD for each community delimitation method for each

species pool.

To determine if our MCS results were correlated with PCS results presented in Chapters

2 and 3, we calculated Pearson’s correlation coefficients between the SES-MPD and SES-

70

MNTD z-values, respectively, from the phylogenetic and morphological datasets for the

appropriate pools and community delimitation method.

Finally, to determine if the traits we measured exhibited phylogenetic signal we

performed Mantel tests (based on Pearson’s product-moment correlation; significance based on

10000 permutations) between phylogenetic and morphological distance matrices using the

package vegan (Oksanen et al. 2010a), which tests for phylogenetic signal in suites of traits as

well as individual traits (Hardy and Pavoine 2012). The Euclidean distance matrices described

above for “both”, skull, and wing datasets were analyzed with the phylogenetic distance matrix

for each spatial and taxonomic scale. A significant positive test statistic indicates that the trait(s)

exhibit phylogenetic signal indicating closely related species are morphologically similar, while

no significant correlation indicates that the traits in question lack signal, indicating no

relationship between phylogenetic distance and morphological distance. Plotting phylogenetic

distance against morphological distance allows for visual exploration of these relationships

(Losos 2008, Hardy and Pavoine 2012); we created distograms for species pools of interest

consisting of all pairwise morphological distances among species plotted against all pairwise

phylogenetic distances among species.

RESULTS

Morphological community structure

Across all combinations of spatial and taxonomic species pools and community delimitation

methods, individual communities ranged from significantly clustered (more morphologically

similar than expected by chance) to significantly overdispersed (less morphologically similar

than expected by chance), although the majority of individual communities were not

significantly different from randomly generated communities (Appendix VI: Tables S1-15).

71

Since we are more interested in the overall patterns of morphological community structure, we

will focus on the results of the Fisher’s test of combined probabilities. In order to increase clarity

and brevity, we only present results for 10km grid communities in the main text; results for all

community delimitation methods can be found in Appendix VI (Tables S1-18, Figures S1-3).

When individual communities were compared to random ones assembled from all taxa

occurring in all deserts SES-MPD and SES-MNTD were significantly clustered for skull, wing,

and “both” datasets as well as for SES-MPD for Myotis for “both” and skull datasets (Figure

4.2). Communities consisting of vespertilionids from all deserts were not significantly different

from randomly assembled communities, nor were Myotis SES-MNTD communities (Figure 4.2).

Great Basin taxa were significantly clustered for “both” data, but were not significant when the

data were parsed into wing or skull measurements or by taxa (Figure 4.2). There was a trend

toward overdispersion in the Great Basin measured by SES-MPD at all taxonomic levels for

skull data, for vespertilionids and Myotis for “both”, and for SES-MNTD for vespertilionids and

Myotis skull data (Figure 4.2). Mojave Desert communities were not significantly different from

random no matter the dataset or taxonomic scale, although skull and “both” datasets tended

toward overdispersion for SES-MPD for Myotis and toward clustering for SES-MNTD for all

taxa and vespertilionids (Figure 4.2). Overall, Sonoran bat communities tended to be clustered

but not significantly except for SES-MPD and SES-MNTD for all taxa for wing and “both”

datasets and SES-MNTD for skull for all taxa and wing for vespertilionids (Figure 4.2).

Chihuahuan desert communities were not significantly different from random communities.

However, there was a tendency toward clustering for SES-MPD for all taxa for all datasets, for

skull data for Myotis, and for SES-MNTD for wing data for all taxa (Figure 4.2). In addition,

72

a)

Data Taxon

All

deserts

Great

Basin Mojave Sonoran Chihuahuan

Skull and

wing

All 0.008 0.000 0.635 0.034 0.165

Vespertilionidae 0.472 0.802 0.613 0.244 0.616

Myotis 0.041 0.705 0.808 0.127 0.419

Skull

All 0.011 0.843 0.431 0.136 0.223


Myotis 0.026 0.786 0.720 0.145 0.255

Wing

All 0.003 0.263 0.618 0.003 0.115


Myotis 0.217 0.264 0.506 0.057 0.855

b)

Data Taxon

All

deserts

Great

Basin Mojave Sonoran Chihuahuan

Skull and

wing

All 0.004 0.000 0.267 0.003 0.416


Myotis 0.326 0.504 0.605 0.329 0.808

Skull

All 0.005 0.639 0.172 0.007 0.378


Myotis 0.174 0.731 0.491 0.355 0.464

Wing

All 0.002 0.096 0.394 0.006 0.197


Myotis 0.402 0.084 0.423 0.146 0.702

Clustered

(sig.)

Clustered

(ns)

Not

significant

Overdispersed

(ns)

Overdispersed

(sig.)

Figure 4.2: Fisher’s combined probability test p-values for all species pools and 10km grid

communities for all three morphological datasets, color-coded by significance. Clustered

communities contain morphologically similar species while overdispersed communities consist

of morphologically dissimilar species. (a) SES-MPD results. (b) SES-MNTD results.

there was a tendency toward overdispersion for SES-MNTD in vespertilionids and Myotis for

“both” data as well as for SES-MPD and SES-MNTD for wing data (Figure 4.2).

Correlation between morphological and phylogenetic community structure

Overall, PCS and MCS metrics were positively correlated (Table 4.1). In all deserts, all metrics

were correlated for all taxa and datasets save SES-MNTD for Myotis (Table 4.1). In both the

73

Great Basin and Mojave Deserts, all metrics and datasets were significantly correlated except for

Myotis (Table 4.1). Sonoran Desert vespertilionid MCS and PCS were significantly correlated

for all datasets and both metrics while Myotis wing MCS was correlated with PCS for both

metrics, as was SES-MNTD for Myotis “both” data (Table 4.1). Chihuahuan taxa and

vespertilionid MCS were significantly correlated with PCS for all datasets, as were SES-MPD

for Myotis skull and combined datasets (Table 4.1).

Phylogenetic signal in morphological traits

Overall, we found evidence of phylogenetic signal in “both”, skull, and wing datasets at all taxa

and vespertilionid taxonomic scales and all spatial scales as indicated by significantly positive

Mantel test statistics (Table 4.2) and positive relationship between pairwise phylogenetic and

morphological distances (Figure 4.3, a-b). Sonoran, Chihuahuan, and all-desert Myotis also

exhibited significant phylogenetic signal for all datasets (Table 4.2, Figure 4.3, c), while Great

Basin and Mojave Myotis did not. However, removing M. vivesi from the all-desert Myotis

“both” data matrix resulted a lack of phylogenetic signal (Mantel’s r= 0.1409; p-value= 0.154;

Figure 4.3, d).

When analyzed separately in Mantel tests, most individual traits exhibited phylogenetic

signal (Table 4.3). Two traits did not exhibit signal for all taxa, twelve traits did not exhibit

signal in vespertilionids, and three traits showed no phylogenetic signal for Myotis (Table 4.3).

DISCUSSION

Overall we find little evidence that communities were made up of species that are more or less

morphologically similar than expected by chance. PCS correlated with MCS for all taxa and

vespertilionids, but Myotis showed no consistent pattern. Phylogenetic signal was present for

suites of morphological traits and individual traits at nearly all examined scales.

74

Tab

le 4

.1:

Pea

rson p

rodu

ct-m

om

ent

corr

elat

ion c

oef

fici

ents

fo

r P

CS

and M

CS

for

all

thre

e dat

aset

s. G

ray c

ells

indic

ate

signif

ican

t

corr

elat

ion w

ith p

-val

ue

<0.0

5.

MP

D

MN

TD

Dat

a T

axon

All

des

erts

Gre

at

Bas

in

Moja

v

e S

onora

n

Chih

uah

ua

n

All

des

ert

s

Gre

at

Bas

in

Moja

v

e S

onora

n

Chih

uah

ua

n

“Both

”

All

0.607

0.597

0.613

0.068

0.682

0.632

0.811

0.650

0.183

0.578

Ves

per

tili

onid

ae

0.658

0.641

0.749

0.700

0.489

0.639

0.812

0.667

0.679

0.546

Myo

tis

0.505

0.011

0.281

0.740

0.822

0.133

0.056

-0.025

0.836

-0.011

Skull

All

0.594

0.468

0.545

0.076

0.673

0.594

0.695

0.600

0.248

0.503

Ves

per

tili

onid

ae

0.621

0.533

0.674

0.651

0.471

0.579

0.723

0.673

0.615

0.450

Myo

tis

0.539

0.165

0.634

0.694

0.843

0.191

0.115

0.307

0.775

-0.059

Win

g

All

0.555

0.762

0.729

0.036

0.553

0.630

0.803

0.666

0.063

0.635

Ves

per

tili

onid

ae

0.673

0.784

0.785

0.655

0.463

0.666

0.796

0.667

0.665

0.644

Myo

tis

0.274

-0.257

-0.302

0.994

0.417

0.120

-0.142

0.448

0.955

0.063

Tab

le 4

.2:

Man

tel

test

r-c

oef

fici

ents

indic

atin

g t

he

stre

ngth

and d

irec

tion o

f th

e re

lati

onsh

ip b

etw

een p

hylo

gen

etic

dis

tance

an

d

morp

holo

gic

al d

ista

nce

for

each

suit

e o

f m

orp

holo

gic

al t

rait

s. G

ray c

ells

indic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5.

Dat

a T

axon

All

des

erts

G

reat

Bas

in

Moja

ve

Sonora

n

Chih

uah

uan

"Both

”

All

0.515

0.596

0.570

0.466

0.486

Ves

per

tili

onid

ae

0.388

0.574

0.510

0.355

0.332

Myoti

s 0.538

0.2

30

0.2

37

0.608

0.570

Skull

All

0.495

0.561

0.573

0.452

0.471

Ves

per

tili

onid

ae

0.365

0.531

0.476

0.348

0.320

Myoti

s 0.523

0.2

01

0.2

42

0.599

0.570

Win

g

All

0.458

0.578

0.447

0.386

0.415

Ves

per

tili

onid

ae

0.380

0.567

0.495

0.317

0.306

Myoti

s 0.545

0.2

60

0.1

78

0.601

0.554

75

a)

b)

c)

d)

Figure 4.3: Distograms of pairwise phylogenetic distances and pairwise morphological distances

using the “both” dataset for all species pairs occurring in a given species pool: a) All-desert

taxa; b) All-desert vespertilionids; c) All-desert Myotis; d) All-desert Myotis excluding the

species Myotis vivesi.

0

2

4

6

8

0 0.5 1 1.5 2Pairwise morp

hological

distance

Pairwise phylogenetic distance

0

1

2

3

4

5

6

0 0.2 0.4 0.6 0.8 1Pairwise morp

hological

distance


0

1

2

3

4

5

0 0.1 0.2 0.3Pairwise morp

hological

distance


0

0.5

1

1.5

2

0 0.05 0.1 0.15 0.2 0.25Pairwise morp

hological

distance


76

Table 4.3: Results of Mantel tests performed using distance matrices for individual traits at each

taxonomic level examined at the all-desert spatial scale. Trait numbers correspond to traits in

Figure 4.1. Shaded cells highlight p-values >0.05, indicating non-significant phylogenetic signal.

All taxa All vespertilionids All Myotis

Trait Mantel's r p-value Mantel's r p-value Mantel's r p-value

1 0.107 0.037 0.277 0.002 0.495 0.019

2 0.173 0.002 0.243 0.003 0.492 0.018

3 0.401 0.000 0.179 0.035 0.567 0.003

4 0.275 0.000 0.213 0.026 0.627 0.000

5 0.263 0.000 0.202 0.025 0.562 0.002

6 0.265 0.000 0.325 0.000 0.051 0.299

7 0.102 0.066 0.297 0.001 0.491 0.019

8 0.388 0.000 0.103 0.166 0.478 0.013

9 0.358 0.000 0.212 0.018 0.537 0.008

10 0.385 0.000 0.138 0.079 0.486 0.011

11 0.464 0.000 0.185 0.023 0.507 0.006

12 0.424 0.000 0.189 0.034 0.547 0.006

13 0.318 0.000 0.242 0.005 0.576 0.001

14 0.469 0.000 0.412 0.000 0.459 0.030

15 0.148 0.013 0.339 0.000 0.434 0.025

16 0.149 0.013 0.353 0.001 0.548 0.002

17 0.285 0.000 0.285 0.001 0.461 0.034

18 0.142 0.013 0.300 0.001 0.582 0.001

19 0.459 0.000 0.195 0.018 0.504 0.009

20 0.301 0.000 0.308 0.001 0.500 0.017

21 0.421 0.000 0.112 0.141 0.451 0.031

22 0.179 0.003 0.193 0.025 0.528 0.005

23 0.166 0.011 -0.024 0.592 0.351 0.031

24 0.433 0.000 0.106 0.156 0.542 0.004

25 0.373 0.000 0.137 0.084 0.516 0.006

26 0.223 0.000 0.205 0.014 0.513 0.014

27 0.424 0.000 0.229 0.010 0.591 0.002

28 0.242 0.001 0.174 0.050 0.484 0.022

29 0.293 0.000 0.198 0.023 0.497 0.007

30 0.365 0.000 0.125 0.105 0.515 0.007

31 0.321 0.000 0.157 0.043 0.508 0.003

32 0.385 0.000 0.178 0.033 0.487 0.011

33 0.229 0.000 0.132 0.096 0.490 0.006

34 0.035 0.281 0.153 0.063 0.091 0.259

35 0.304 0.000 0.185 0.039 0.491 0.006

36 0.238 0.001 0.205 0.021 0.520 0.003

37 0.161 0.004 0.259 0.003 0.530 0.003

39 0.140 0.014 0.357 0.000 0.506 0.005

Table 4.3 continued

77

All taxa All vespertilionids All Myotis

Trait Mantel's r p-value Mantel's r p-value Mantel's r p-value

40 0.390 0.000 0.127 0.061 0.428 0.013

41 0.169 0.002 0.215 0.009 0.495 0.015

42 0.129 0.020 0.374 0.374 0.502 0.006

43 0.446 0.000 0.421 0.000 0.478 0.021

44 0.142 0.012 0.378 0.000 0.517 0.015

45 0.189 0.001 0.351 0.000 0.317 0.061

46 0.604 0.000 0.235 0.009 0.510 0.014

47 0.107 0.037 0.302 0.001 0.508 0.015

48 0.116 0.039 0.359 0.000 0.501 0.028

49 0.317 0.000 0.143 0.074 0.530 0.003

50 0.262 0.000 0.361 0.000 0.525 0.018

51 0.297 0.000 0.110 0.158 0.521 0.006

a 0.202 0.001 0.167 0.039 0.505 0.009

b 0.173 0.001 0.294 0.001 0.507 0.026

c 0.175 0.011 0.456 0.000 0.469 0.011

d 0.159 0.026 0.432 0.000 0.350 0.050

e 0.360 0.000 0.371 0.000 0.565 0.003

f 0.140 0.023 0.181 0.035 0.600 0.000

g 0.478 0.000 0.254 0.005 0.524 0.005

Phylogenetic signal in morphological traits

For all taxa together and vespertilionids, as phylogenetic distance increased, so too did

morphological distance (Tables 4.2 and 4.3, Figure 4.3, a-b). This means that traditional

interpretations of PCS (Webb 2000) can be used for the two largest taxonomic scales in this

study system; closely related taxa were more morphologically similar to each other than to

distantly related taxa.

The patterns of phylogenetic signal for the genus Myotis, however, were quite different.

When Myotis were present in the appropriate species pools, significant phylogenetic signal was

observed in most cases for suites of traits (Table 4.2 and Figure 4.3, c) and individual traits

(Table 4.3). However, these correlations were being driven solely by a single species, Myotis

vivesi. M. vivesi is a morphologically distinctive member of the genus, specialized for catching

small marine fish and invertebrates at the water’s surface. This species is so specialized that it

78

has been assigned to a separate genus, Pizonyx, by various authors (Stadelmann et al. 2004);

molecular work has confirmed its placement within the Neotropical clade of Myotis (Stadelmann

et al. 2007, Chapter 2). When this species is removed from the species pool, no significant

correlation remains among the suites of characters (Figure 4.3, d), and only nine of 58 individual

traits still exhibit phylogenetic signal (results not shown). This indicates that closely related

species are neither more nor less morphologically similar to each other than distantly related

species, which is a pattern also observed in Sylvia warblers (Brohning-Gaese et al. 2003).

However, we had expected to see evidence of convergence. Myotis was split into three

subgenera or morphotypes by Findley (1972); these subgenera were later found to be

polyphyletic, suggesting that the genus had gone through convergent evolution multiple times

(Ruedi and Mayer 2001, Stadelmann et al. 2007), which is the reason we examined the genus by

itself and in combination with other taxa. Significant negative Mantel coefficients would be

indicative of convergent evolution because closely related species would be morphologically

dissimilar whereas distantly related species would be morphologically similar. This is not the

pattern we see, however; no evolutionary pattern is observed (Figure 4.3, d). Perhaps broader

taxon sampling, not limited to desert bats, might reveal evidence of convergent evolution. Had

we not investigated the results in the absence of M. vivesi, we may not have made these

observations; this finding shows that investigation of phylogenetic signal should proceed

cautiously as a single species may drive observed patterns. However, as M. vivesi is a member of

several species pools, we will discuss the remainder of the results with it present in the

appropriate dataset.

79

Correlation between morphological and phylogenetic community structure

Overall, we see less evidence of significant community structure using morphological data

(Figure 4.2, Appendix VI: Table S1-15), regardless of spatial or taxonomic scale, than with

phylogenetic data (Chapters 2 and 3). Results from both types of data did trend in the same

direction therefore phylogeny and morphology give similar results. This is unsurprising given

that morphological traits had significant phylogenetic signal (Tables 4.2 and 4.3). Sonoran

communities consisting of all bat taxa were made up of species that were significantly

morphologically clustered (Figure 4.2) and were also significantly phylogenetically clustered

(Chapter 2); however, the datasets were not correlated (Table 4.1). This suggests that the

communities that were morphologically clustered were not necessarily the same communities

that were phylogenetically clustered.

There was little correspondence among community structure metrics in the genus Myotis;

MCS and PCS metrics were not strongly correlated for most deserts and datasets (Table 4.1). As

described above, there was phylogenetic signal in the traits measured for this taxonomic scale

only when the full species pool was used; otherwise there was no pattern between morphological

distance and phylogenetic distance (Figure 4.3, d). This has likely given rise to the observed

pattern between MCS and PCS.

Community structure of North American desert bats

Overall, communities of desert bats tend to be made up of species that are not morphologically

different from species drawn randomly from the species pool (Figure 4.2). This suggests that

species may not be partitioning resources or, if they are, they are not doing so in a manner that

affects the morphological make up of the community.

80

Some communities (all-desert taxa and Myotis, Great Basin and Sonoran taxa) were

significantly morphologically clustered, meaning that species in these communities were

morphologically more similar than expected by chance (Figure 4.2). The traditional explanation

of clustering is that habitat filtering structures these communities: only species with the

appropriate morphotypes to survive in these areas are found there. At the all-desert spatial scale

this is unsurprising, as species are filtered by functional traits across heterogeneous regions.

Alternatively, species are similar to each other because they have traits that increase their

competitive ability for limited resources thereby excluding species lacking such traits (Mayfield

and Levine 2010). Both explanations are reasonable, however more research is needed to

determine to what extent bats partition available resources in order to determine which

interpretation is most plausible and to exclude the possibility that both are occurring

simultaneously.

No communities were significantly overdispersed overall (Figure 4.2). This means that

North American desert bat communities were not made up of species less morphologically

similar than chance, which would indicate competition. This is somewhat surprising given that

many authors have found evidence of competitive interactions in bat communities (e.g. Black

1974, Findley 1976, Findley and Black 1983, Aldridge and Rautenbach 1987, Barlow et al. 1997,

Aguirre et al. 2002, Rhodes 2002, Campbell et al. 2007, York and Papes 2007). For competition

or other density-dependent interactions to occur, populations must at least approach carrying

capacity (e.g., Stevens and Willig 1999, Stevens and Willig 2000). Deserts are unpredictable

environments (Shreve 1942) and this environmental instability may prevent density-dependent

interactions from occurring (e.g., Stevens and Willig 1999, Stevens and Willig 2000). Desert bat

populations may not reach carrying capacity, thereby influencing competition for food and

81

allowing a random set of morphologies to co-exist. Previous work has shown that all members of

a desert bat community responded similarly to experimentally manipulated insect densities,

suggesting that bats were not competing for food resources (Bell 1980).

The lack of strong evidence for competitive interactions does not necessarily mean that

they do not occur, we just might not be able to quantify these interactions using our data. For

example, some bats partition resources temporally by feeding or drinking at different times,

potentially minimizing competitive interactions (e. g., Black 1974) but such behavioral

modifications may not be apparent in morphological data. Likewise, other bats partition

resources spatially to minimize competitive interactions (Aldridge and Rautenbach 1987) with

some species preferentially foraging in more cluttered habitats while others utilize more open

habitats, although desert bats may have few such options because arid habitats tend to exhibit

little structural complexity (Shreve 1942). Additionally, broadening of dietary resources may not

be accompanied by morphological changes and thus may not be detectable using morphological

methods. For example, pallid bats (Antrozous pallidus) in the Sonoran desert may feed on the

fruits of columnar catci (Howell 1980) but have also been observed feeding on nectar from these

cacti as well, proving to be better pollinators than specialized nectarivorous bats (Frick et al.

2013).

Previous work on this study system has revealed that communities tend to be made up of

species that are more closely related to each other than expected by chance (Chapters 2 and 3)

and clustered communities tend to be found in harsher habitats than more overdispersed

communities (Chapter 3) which suggests that habitat filtering may be structuring these

communities. This finding is supported by the current study because we found significant

morphological clustering in a few cases and little evidence for competition overall. This suggests

82

to us that in some cases, phylogeny and environmental data may be more useful to investigations

of community structure than morphology, even when convergent evolution is suspected.

ACKNOWLEDGEMENTS

L.E.P. was funded by the American Society of Mammalogists Grants-in-Aid of Research award

and Louisiana Environmental Education Commission Research Grant. Many thanks to the

museums, curators, and collections managers that allowed access to their collections: Dr. Mark

Hafner at the Louisiana State University Museum of Natural Science, Jeffrey Bradley at the

Burke Museum, Dr. Luis Ruedas and Dr. Jan Zinck at the Portland State University Museum of

Vertebrate Biology, Dr. Joseph Cook and Cindy Ramotnik at the Museum of Southwestern

Biology, Dr. Robert Timm at the University of Kansas, and Dr. Jim Dines at the Natural History

Museum of Los Angeles County. L.E.P would especially like to thank Lynnmarie Patrick, Cindy

Ramotnik and Dr. Mike Bogan, Yadeeh Sawyer, Dr. Kelly Grussendorf, and Jeanne Harris for

opening their homes, spare bedrooms, and couches to a traveling graduate student; without their

help this research may not have been possible.

REFERENCES

Aguirre, L. F., A. Herrel, R. van Damme, and E. Matthysen. 2002. Ecomorphological analysis of

trophic niche partitioning in a tropical savannah bat community. Proceedings of the

Royal Society of London Series B-Biological Sciences 269:1271-1278.






Barlow, K. E., G. Jones, and E. M. Barratt. 1997. Can skull morphology be used to predict

ecological relationships between bat species? A test using two cryptic species of

83

pipistrelle. Proceedings of the Royal Society of London Series B-Biological Sciences

264:1695-1700.

Bell, G. P. 1980. Habitat use and response to patches of prey by desert insectivorous bats.

Canadian Journal of Zoology 58:1876-1883.

Black, H. L. 1974. A North Temperate Bat Community: Structure and Prey Populations. Journal


Brohning-Gaese, K., M. D. Schuda, and A. J. Helbig. 2003. Weak phylogenetic effects on

ecological niches of Sylvia warblers. Journal of Evolutionary Biology 16:956-965.




Cleveland, C. J., M. Betke, P. Federico, J. D. Frank, T. G. Hallam, J. Horn, J. D. Lopez, G. F.

McCracken, R. A. Medellin, A. Moreno-Valdez, C. G. Sansone, J. K. Westbrook, and T.

H. Kunz. 2006. Economic value of the pest control service provided by Brazilian free-

tailed bats in south-central Texas. Frontiers in Ecology and the Environment 5:238-243.

Colwell, R. K. 2009. EstimateS: Statistical Estimation of Species Richness and Shared Species

for Samples User's Guide. Storrs, CT.

Davis, E. B. 2005. Comparison of climate space and phylogeny of Marmota (Mammalia:

Rodentia) indicates a connection between evolutionary history and climate preference.

Proceedings of the Royal Society B: Biological Sciences 272:519-526.



Findley, J. S. 1972. Phenetic relationships among bats of the genus Myotis. Systematic Zoology

21:31-52.

Findley, J. S. 1976. The structure of bat communities. The American Naturalist 110:120-139.

Findley, J. S. and H. Black. 1983. Morphological and dietary structuring of a Zambian

Insectivorous bat community. Ecology 64:625-630.

Freeman, P. W. 1981a. Correspondence of food habits and morphology in insectivorous bats.

Journal of Mammalogy 62:168-173.

Freeman, P. W. 1981b. A multivariate study of the Family Molossidae (Mammalia: Chiroptera):

Morphology, ecology, and evolution. Fieldiana Zoology:1-173.

84

Frick, W. F., R. D. Price, P. A. H. Iii, and K. M. Kay. 2013. Insectivorous bat pollinates

columnar cactus more fffectively per visit than specialized nectar bat. The American

Naturalist 181:137-144.

Gannon, W. L. and G. R. Racz. 2006. Character displacement and ecomorphological analysis of

two long-eared Myotis (M. auriculus and M. evotis). Journal of Mammalogy 87:171-179.

Hardy, O. J. and S. Pavoine. 2012. Assessing phylogenetic signal with measurement error: a

comparison of Mantel tests, Blomberg et al.'s K, and phylogenetic distograms. Evolution

66:2614-2621.

Hespenheide, H. A. 1973. Ecological inferences from morphological data. Annual Review of

Ecology and Systematics 4:213-229.

Howell, D. J. 1980. Adaptive Variation in Diets of Desert Bats Has Implications for Evolution of

Feeding Strategies. Journal of Mammalogy 61:730-733.



Kembel, S. W. 2009. Disentangling niche and neutral influences on community assembly:

assessing the performance of community phylogenetic structure tests. Ecology Letters

12:949-960.




Losos, J. B. 2008. Phylogenetic niche conservatism, phylogenetic signal and the relationship

between phylogenetic relatedness and ecological similarity among species. Ecology

Letters 11:995-1003.

Losos, J. B., T. R. Jackman, A. Larson, K. de Queiroz, and L. Rodriguez-Schettino. 1998.

Contingency and determinism in replicated adaptive radiations of island lizards. Science

279:2115-2118.



McGill, B. J., B. J. Enquist, E. Weiher, and M. Westoby. 2006. Rebuilding community ecology

from functional traits. Trends in Ecology & Evolution 21:178-185.

Norberg, U. M. and J. M. Rayner. 1987. Ecological morphology and flight in bats (Mammalia;

Chiroptera): Wing adaptations, flight performance, foraging strategy, and echolocation.

Philosophical Transactions of tne Royal Society of London B 316:335-427.

85


Henry, H. Stevens, and H. Wagner. 2010a. vegan: Community Ecology Package.

Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, R. B. O'Hara, G. L. Simpson, P. Solymos, H.

Stevens, and H. Wagner. 2010b. Community Ecology Package 'vegan' User's Guide.

Olson, D. M., E. Dinerstein, E. D. Wikramanayake, N. D. Burgess, G. V. N. Powell, E. C.

Underwood, J. A. D'amico, I. Itoua, H. E. Strand, J. C. Morrison, C. J. Loucks, T. F.

Allnutt, T. H. Ricketts, Y. Kura, J. F. Lamoreux, W. W. Wettengel, P. Hedao, and K. R.

Kassem. 2001. Terrestrial Ecoregions of the World: A New Map of Life on Earth.

BioScience 51:933-938.

Patrick, L. E., E. S. McCulloch, and L. A. Ruedas. 2013. Systematics and biogeography of the

arcuate horseshoe bat species complex (Chiroptera, Rhinolophidae). Zoologica Scripta

42:553-590.

Rhodes, M. P. 2002. Assessment of sources of variance and patterns of overlap in

microchiropteran wing morphology in southeast Queensland, Australia. Canadian Journal

of Zoology-Revue Canadienne De Zoologie 80:450-460.

Ricklefs, R. E. and D. B. Miles. 1994. Ecological and evolutionary inferences from morphology:

An ecological perspective. Pages 13-41 in P. C. Wainwright and S. M. Reilly, editors.

Ecological Morphology. The University of Chicago Press, Chicago.



Ruedi, M. and F. Mayer. 2001. Molecular systematics of bats of the genus Myotis

(Vespertilionidae) suggests deterministic ecomorphological convergences. Molecular

Phylogenetics and Evolution 21:436-448.




Stadelmann, B., L. G. Herrera, J. Arroyo-Cabrales, J. J. Flores-Martinez, B. P. May, and M.

Ruedi. 2004. Molecular systematics of the fishing bat Myotis (Pizonyx) vivesi. Journal of

Mammalogy 85:133-139.






86

Stevens, R. D. and M. R. Willig. 2000. Community structure, abundance, and morphology.

Oikos 88:48-56.

Swartz, S. M., P. W. Freeman, and E. F. Stockwell. 2003. Ecomorphology of bats: Comparative

and experimental approaches relating structural design to ecology. Pages 257-300 in T.

H. Kunz and M. B. Fenton, editors. Bat Ecology. The University of Chicago Press,

Chicago.

van Zyll de Jong, C. G. 1979. Distribution and systematic relationship of long-eared Myotis in

western Canada. Canadian Journal of Zoology 57:987-994.

Webb, C. O. 2000. Exploring the phylogenetic structure of ecological communities: an example

for rain forest trees. American Naturalist 156:145-155.



York, H. A. and M. Papes. 2007. Limiting similarity and species assemblages in the short-tailed

fruit bats. Journal of Zoology 273:249-256.

87

CHAPTER 5

SUMMARY

In this study, I examined whether commonly used community structure metrics were greatly

influenced by changes in the data used to calculate them. I described patterns of North American

desert bat community structure at multiple spatial and taxonomic scales. I then investigated if

biotic and abiotic factors were strongly correlated with these patterns in an effort to understand

which ecological and evolutionary processes may be contributing to these results.

I first defined bat communities and inferred a well supported phylogeny that included

several species with poor representation on public databases. These datasets form the foundation

for all three chapters of this dissertation. Using these data, I found that community delimitation

method did influence community structure metrics, but the results at least trended in the same

direction. This suggests that as long as a researcher is consistent with the method used to delimit

communities within a particular study, the phylogenetic community structure results should not

be greatly affected. Community structure metrics were also robust to moderate changes to the

phylogeny from which they were calculated. These findings demonstrate that phylogenetic

community structure results are due to actual patterns in the data and not to poorly inferred trees.

I also found that bat communities in all deserts combined, the Great Basin Desert, and Sonoran

Desert were significantly clustered, meaning they contain species that were more closely related

to each other than expected by chance. The Chihuahuan Desert was made up of communities that

tended to be clustered, while the Mojave Desert’s communities were indistinguishable from

random.

Next, I investigated phylogenetic community structure at different spatial and taxonomic

scales. I found that at the largest scales, communities were significantly clustered, as expected.

At smaller taxonomic scales, overdispersion of species was expected; I did not observe

88

significant overdispersion, except in Great Basin Myotis, but did find less clustering than at the

larger scales. Ecological traits were significantly correlated with phylogeny; closely related

species tended to have similar phenotypes. This suggests that phylogeny is a good proxy for

ecology. In addition, bat community structure was significantly correlated with temperature,

temperature seasonality, and precipitation seasonality, although the environmental variables that

were significant differed by taxon and desert. Phylogenetically clustered communities were

found in harsher environmental conditions than more overdispersed communities, which tended

to occur in less harsh conditions. Based on these results, desert bat communities tend to be made

up of similar species that can survive the environmental conditions in the area. This suggests that

desert bat communities are structured mainly by habitat filtering. Furthermore, although deserts

are harsh environments, they are not all the same, differing in the environmental hardships they

pose to taxa residing in them. This suggests that while bat communities are responding to harsh

conditions in a similar way, the environmental conditions likely driving these patterns differ

based on spatial and taxonomic scale.

Interpreting phylogenetic community structure results requires the assumption that

closely related species are also phenotypically similar to each other. While ecological traits

suggested this was the case with desert bats, I tested this assumption more thoroughly by

measuring a suite of skull and wing characteristics to determine if morphological community

structure reflected phylogenetic community structure. I found that in most cases, bat

communities were made up of species with morphologies randomly drawn from the available

pool; those that were not random had species that were more similar to each other (i.e.

morphologically clustered) than expected by chance. None of the communities were

overdispersed, which is what we would expect to see if competition were structuring

89

communities. This suggests that bats may be feeding on the same insect species. In most cases,

phylogenetic community structure was significantly positively correlated with morphological

community structure, indicating that these methods provide similar results. In addition, I found

that at the largest taxonomic scales, closely related species tend to be morphologically similar to

each other while distantly related species tend to be more morphologically distant; this pattern is

termed phylogenetic signal. At the smallest taxonomic scale, the genus Myotis, there tended to be

less or no phylogenetic signal in the traits I analyzed meaning that closely related species were

not necessarily morphologically similar.

Overall, these results suggest that desert bat communities are predominantly structured by

habitat filtering, that is, the bats that coexist in the same community are those that can tolerate

the environmental conditions and make use of the resources available at a given location. These

results are counter to many previous studies of bat communities, most of which have found

evidence for competition playing a dominant role in structuring communities. In order to gain a

more complete understanding of desert bat community structure, I suggest that in depth studies

of bat diet be undertaken to determine if bats are partitioning prey resources or if all bat species

are eating all available insect species. My results suggest that, at least in North American deserts,

bat populations have not been able to reach sizes that would allow density-dependent

interactions, such as competition, to greatly impact the species that are found in particular

communities. From a conservation standpoint, these results may imply that North American

desert bat species range shifts due to climate change (e. g., Humphries et al. 2004) may not be

impacted by interspecific competitive interactions.

90

REFERENCES

Humphries, M. M., J. Umbanhowar, and K. S. McCann. 2004. Bioenergetic Prediction of Climate Change

Impacts on Northern Mammals. Integrative and Comparative Biology 44:152-162.

91

APPENDIX

I

SEQUENCES IN THE REGIO

NAL POOL PHYLOGENY

Inst

ituti

ons

that

len

t ti

ssues

, G

enB

ank a

cces

sion n

um

ber

s w

ith s

pec

imen

num

ber

s in

par

enth

eses

, pri

mer

s use

d t

o s

ucc

essf

ull

y a

mpli

fy g

enes

(pri

mer

seq

uen

ces

des

igned

for

this

stu

dy a

re l

iste

d b

elow

tab

le),

and P

CR

pro

file

s use

d (

full

pro

file

s ar

e li

sted

bel

ow

tab

le).

Gen

Ban

k a

cces

sion

num

ber

s in

bold

wer

e gen

erat

ed b

y t

his

stu

dy.

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Molo

ssid

ae

Eum

ops per

otis

An

gel

o S

tate

Nat

ura

l H

isto

ry

Coll

ecti

on

s KC747672

(AS

K6

27

1)

Molc

itF

an

d

H1

591

5

LP

CY

TB

KC747652

(AS

K6

27

1)

12cd

an

d

12

gg;

12

a

and

16q;

16j

and

16

t;

16p

and

16

t

LP

12

S;

LP

16

S;

LP

16

S;

LP

16

S

KC747701

(A

SK

62

71

)

179

F

and

1458

R

LP

RA

G2

Molo

ssid

ae

Nyc

tinom

ops

aurisp

inosu

s

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747674

(NK

92

34

)

Molc

itF

and

H

1591

5

LP

CY

TB

KC747654

(NK

92

34

)

12c

and

1

2g;

12

e

and

16q;

16j

and

16

t;

16p

and

16

t

LP

12

S;

LP

16

S;

LP

16

S;

LP

16

S;

KC747703

(N

K9

23

4)

179

F

and

1

458

R

LP

RA

G2

Molo

ssid

ae

Nyc

tinom

ops

fem

erosa

ccus

Mu

seu

m o

f

Tex

as T

ech

Un

iver

sity

KC747675

(T

K1

955

2)

Molc

itF

and

H1

591

5

LP

CY

TB

AY

49

5458

(T

K

1955

2;

TT

U

3773

1)

n/a

n

/a

KC747704

(T

K1

955

2)

RA

G2

-

F2

20

an

d

RA

G2

-

R995

LP

RA

G2

Molo

ssid

ae

Nyc

tinom

ops

macr

otis

Mu

seu

m o

f T

exas

Tec

h

Un

iver

sity

KC747676

(T

K7

890

8);

KC747673

(T

K4

860

5)

L1

47

24

and

B

SV

ES

26

8

H

LE

PC

YT

B5

AF

26

3217

(TK

78

90

8);

KC747653

(T

K4

860

5)

12cd

an

d

12

gg;

12

c

and

12

g;

12h

and

1

6t;

16

j an

d

16t

LP

12

S;

LP

12

S;

LP

16

S;

LP

16

S

AY

14

1018

(T

K

7890

8);

KC747702

(T

K4

860

5)

179

F

and

1458

R

LP

RA

G2

Molo

ssid

ae

Tadarida

bra

silien

sis

JF4

891

29

(MC

-2)

n/a

n

/a

AF

26

3219

(OK

43

0)

n/a

n

/a

AY

14

1019

(O

K

430

) n

/a

n/a

Morm

oop

idae

M

orm

oops

meg

alo

phyl

a

A

F3

30

808

(T

K2

764

0)

n/a

n

/a

AF

40

7174

(T

K2

764

0)

n/a

n

/a

AF

33

0818

(T

K2

764

0)

n/a

n

/a

Morm

oop

idae

Pte

ronotu

s

davy

i

AF

33

8671

(TK

25

12

7)

n/a

n

/a

AF

40

7176

(TK

25

12

7)

n/a

n

/a

AF

33

8692

(TK

25

12

7),

n

/a

n/a

Morm

oop

idae

Pte

ronotu

s

parn

ellii

AF

33

0807

(TK

17

95

3)

n/a

n

/a

AF

40

7180

(TK

17

95

3)

n/a

n

/a

AF

33

0817

(TK

17

95

3)

n/a

n

/a

92

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Nat

alid

ae

Nata

lus

stra

min

eus

AY

62

1015

(TT

U 3

14

58

) n

/a

n/a

AF

34

5924

(TK

15

66

0)

n/a

n

/a

AY

14

1024

(T

K

1566

0)

n/a

n

/a

Ph

yll

ost

om

idae

C

hoer

onyc

teris

mex

icana

Mu

seu

m o

f

Tex

as T

ech

U

niv

ersi

ty

KC747677

(T

K2

050

1)

L1

47

24

and

B

SV

ES

26

8

H;

Molc

itF

and

H

1591

5

LE

PC

YT

B5;

LP

CY

TB

A

Y3

95

808

(T

K2

050

1)

n

/a

n/a

AF

31

6441

(TK

20

50

1;

TT

U3

6118

) n

/a

n/a

Ph

yll

ost

om

idae

Des

modus

rotu

ndus

DQ

077

398

(TK

40

36

8)

n/a

n

/a

AF

26

3228

(TK

47

64

) n

/a

n/a

AF

31

6444

(TK

47

64

) n

/a

n/a

Ph

yll

ost

om

idae

Glo

ssophaga

sorici

na

AF

38

2831

(TK

15

15

4)

n/a

n

/a

AY

39

5840

(TK

70

46

1)

n/a

n

/a

AF

31

6452

(TK

15

31

1)

n/a

n

/a

Ph

yll

ost

om

idae

Lep

tonyc

teri

s

niv

alis

New

Mex

ico

Mu

seu

m o

f

Nat

ura

l H

isto

ry a

nd

Sci

ence

KC747678

(D

JH3

59

5)

Molc

itF

an

d

H1

591

5

LP

CY

TB

KC747655

(DJH

359

5)

12c

and

1

2s;

12

a

and

12

g;

12h

and

1

6q

; 16j

and

16t

LP

12

S;

LP

12

S;

LP

16

S;

LP

16

S

KC74770

(DJH

359

5)

RA

G2

-

F1

and

R

AG

2-

R2

RA

G2

B

Ph

yll

ost

om

idae

Lep

tonyc

teri

s

yerb

abuen

ae

AF

38

2889

(TK

45

10

7)

n/a

n

/a

AY

39

5814

(TK

45

10

8)

n/a

n

/a

AF

31

6454

(T

K4

510

8;

TT

UU

AM

-1)

n/a

n

/a

Ph

yll

ost

om

idae

D

iphyl

la

ecaudata

FJ1

554

76

(TT

U4

7509

; T

K1

350

8)

n/a

n

/a

AF

41

1533

n/a

n

/a

AF

31

6447

n/a

n

/a

Ph

yll

ost

om

idae

Macr

otu

s

califo

rnic

us

Mu

seu

m o

f

Tex

as T

ech

Un

iver

sity

AY

38

0744

(TK

28

96

2)

n/a

n

/a

KC747656

(T

K2

896

7)

12cd

an

d

12

gg;

12

e

and

16q;

LP

12

S;

LP

16

S

AF

31

6459

(TK

28

96

2)

n/a

n

/a

Ph

yll

ost

om

idae

Macr

otu

s

wate

rhousii

AY

38

0745

(TK

27

88

9)

n/a

n

/a

AF

26

3229

(TK

32

02

1)

n/a

n

/a

AF

31

6460

(T

K

2853

5)

n/a

n

/a

Ph

yll

ost

om

idae

Stu

rnira liliu

m

Lou

isia

na

Sta

te

Un

iver

sity

Mu

seu

m o

f

Nat

ura

l S

cien

ce

AF

18

7035

(T

K2

263

1)

n/a

n

/a

KC747657

(M

959

5)

12c

and

12

g;

12h

an

d 1

6t,

seq

uen

ced

wit

h 1

6q

, 1

6p

, 16

j,

LP

12

S;

LP

16

S

AF

31

6488

(T

K2

516

3)

n/a

n

/a

93

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

and

16n

Ves

per

tili

on

idae

Antrozo

us

pallid

us

EF

22

238

2

(AK

21

090

) n

/a

n/a

AF

32

6088

(T

K

4964

6;

TT

U

7110

1)

n/a

n

/a

GU

32

8047

(T

TU

7110

1)

n/a

n

/a

Ves

per

tili

on

idae

C

ory

norh

inus

mex

icanus

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747679

(N

K4

52

9)

Molc

itF

and

H

1591

5

LP

CY

TB

A

F3

26

090

n/a

n

/a

GU

32

8053

(T

K

4584

9)

n/a

n

/a

Ves

per

tili

on

idae

C

ory

norh

inus

tow

nse

ndii

Lou

isia

na

Sta

te

Un

iver

sity

Mu

seu

m o

f

Nat

ura

l S

cien

ce

KC747680

(L

EP

132

)

L1

47

24

and

BS

VE

S26

8H

L

EP

CY

TB

5

AF

26

3238

(T

K8

318

2)

n/a

n

/a

AY

14

1029

(T

K

8318

2)

n/a

n

/a

Ves

per

tili

on

idae

Epte

sicu

s

fusc

us

AF

37

6835

(MV

Z

1486

81

) n

/a

n/a

AF

32

6092

(SP

844

) n

/a

n/a

EU

7869

13

(Efs

1590

) n

/a

n/a

Ves

per

tili

on

idae

Euder

ma

macu

latu

m

JF4

891

25

(AM

-8)

n/a

n

/a

AF

32

6093

(N

K

3626

0)

n/a

n

/a

GU

32

8060

(MS

B 1

2137

3)

n/a

n

/a

Ves

per

tili

on

idae

Id

ionyc

teris

phyl

lotis

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747681

(N

K3

98

71

)

L1

47

24

and

BS

VE

S26

8H

L

EP

CY

TB

5

AF

32

6094

(N

K

3612

2)

n/a

n

/a

GU

32

8063

(A

CU

73

6)

n/a

n

/a

Ves

per

tili

on

idae

Lasionyc

teris

noct

ivagans

Lou

isia

na

Sta

te

Un

iver

sity

Mu

seu

m o

f N

atu

ral

Sci

ence

KC747682

(2

2Ju

l09

-03

-

AH

H)

Las

iuru

smi

dd

leF

1 a

nd

H1

591

5;

Myo-7

L

and

Myo-

16;

Molc

itF

an

d

H1

591

5

LP

CY

TB

AF

32

6095

(T

K

2421

6)

n/a

n

/a

GU

32

8065

(T

TU

5625

5)

n/a

n

/a

Ves

per

tili

on

idae

Lasiuru

s blo

ssev

illii

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747683

(N

K2

12

90

)

Myo-7

L

and

Myo-

16;

Las

iuru

smi

dd

leF

1 a

nd

LP

CY

TB

;

LP

CY

TB

; C

TB

50

AY

49

5479

(F

3813

3;

RO

M

1042

85

) n

/a

n/a

HM

561

636

(F3

81

33

(R

OM

1042

85

))

n/a

n

/a

94

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

H1

591

5;

Myo-7

L

and

Las

iuru

sR1

Cytb

1

2S

-16

S

RA

G2

Fam

ily

Sp

ecie

s M

use

um

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Ves

per

tili

on

idae

Lasiuru

s ci

ner

eus

Mu

seu

m o

f

Tex

as T

ech

U

niv

ersi

ty

KC747685

(T

K7

892

6)

Myo-7

L

and

H

1591

5;

Las

iuru

sF1

and

H

1591

5;

Las

iuru

sF2

and

H

1591

5

CT

B50

;

LP

CY

TB

; L

PC

YT

B

AY

49

5482

(T

K

7892

6;

TT

U)

n/a

n

/a

HM

561

638

(T

K7

892

6)

n/a

n

/a

Ves

per

tili

on

idae

Lasiuru

s eg

a

Mu

seu

m o

f T

exas

Tec

h

Un

iver

sity

KC747686

(TK

78

70

4)

Myo-7

L

and

Las

iuru

sR1

TD

CT

B

AY

49

5483

(T

K

4313

2)

n/a

n

/a

HM

561

639

(TK

43

13

2)

n/a

n

/a

Ves

per

tili

on

idae

Lasiuru

s in

term

ediu

s

Lou

isia

na

Sta

te

Un

iver

sity

Mu

seu

m o

f

Nat

ura

l S

cien

ce

KC747687

(M

352

)

Las

iuru

sF1

and

L

asiu

rusR

1

CT

B50

HM

561

627

(TT

U8

0739

(T

K8

451

0))

n

/a

n/a

HM

561

640

(TK

20

51

3

(TT

U3

6631

))

n/a

n

/a

95

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Ves

per

tili

on

idae

Lasiuru

s

sem

inolu

s

Lou

isia

na

Sta

te

Un

iver

sity

M

use

um

of

Nat

ura

l

Sci

ence

KC747688

(M

897

0)

Myo-7

L

and

Las

iuru

sR1

LP

CY

TB

AY

49

5484

(T

K

9068

6)

n

/a

n/a

HM

561

641

(TK

90

68

6

(TT

U8

0699

))

n/a

n

/a

Ves

per

tili

on

idae

Lasiuru

s

xanth

inus

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747689

(N

K3

64

5)

Myo-7

L

and

Myo-

16;

Las

iuru

smi

dd

leF

1 a

nd

Las

iuru

sR1

LP

CY

TB

;

LP

CY

TB

AY

49

5485

(T

K

7870

4;

TT

U

7829

6)

n/a

n

/a

HM

561

642

(TK

78

70

4

(TT

U7

8296

))

n/a

n

/a

Ves

per

tili

on

idae

Myo

tis

auricu

lus

Mu

seu

m o

f S

ou

thw

este

rn

Bio

log

y

AM

261

884

(C

DR

3288

(In

stit

uto

Poli

tecn

ico

Nac

ional

in

Mex

ico))

n

/a

n/a

KC747658

(N

K4

27

00

)

12c

and

1

2s;

12

a

and

12

g;

12a

and

1

6q

; 16j

and

16t

LP

12

S;

LP

12

S;

LP

16

S;

LP

16

S

AM

265

641

(C

DR

3288

(In

stit

uto

Poli

tecn

ico

Nac

ional

Mex

ico))

n

/a

n/a

Ves

per

tili

on

idae

Myo

tis

califo

rnic

us

AM

261

887

(C

DR

3276

(In

stit

uto

Poli

tecn

ico

Nac

ional

in

Mex

ico))

n

/a

n/a

AY

49

5495

(T

K

7879

7;

TT

U

7932

5)

n/a

n

/a

AM

265

649

(C

DR

3276

(In

stit

uto

Poli

tecn

ico

Nac

ional

Mex

ico))

n

/a

n/a

Ves

per

tili

on

idae

Myo

tis

ciliola

bru

m

AM

261

889

(C

DR

3172

(In

stit

uto

Poli

tecn

ico

Nac

ional

in

Mex

ico))

n

/a

n/a

AY

49

5497

(T

K

8315

5;

TT

U

7852

0)

n/a

n

/a

GU

32

8080

(T

TU

7852

0)

n/a

n

/a

Ves

per

tili

on

idae

M

yotis ev

otis

Lou

isia

na

Sta

te

Un

iver

sity

M

use

um

of

Nat

ura

l

Sci

ence

AJ8

419

49

(n

o

vou

cher

) n

/a

n/a

KC747659

(L

EP

121

)

12c

and

1

2g;

12h

and

16t,

seq

uen

ced

wit

h 1

6q

,

16p

, 16

j,

and

16n

LP

12

S;

LP

16

S

AM

265

657

(n

o

vou

cher

) n

/a

n/a

96

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Ves

per

tili

on

idae

Myo

tis

fortid

ens

Nat

ura

l

His

tory

Mu

seu

m o

f L

os

An

gel

es

Coun

ty

KC747690

(L

AF

00

30

)

Molc

itF

an

d

H1

591

5

LP

CY

TB

AY

49

5502

(T

K

4318

6)

n/a

n

/a

GU

32

8082

(T

K

4318

6)

n/a

n

/a

Ves

per

tili

on

idae

Myo

tis

luci

fugus

ala

censis

Lou

isia

na

Sta

te

Un

iver

sity

Mu

seu

m o

f N

atu

ral

Sci

ence

KC747691

(1

5JU

L0

9-0

1-

LE

P)

Molc

itF

an

d

H1

591

5

LP

CY

TB

KC747660

(1

5JU

L0

9-0

1-

LE

P)

12c

and

1

2g;

12h

and

16q;

16j

and

1

6n

; 16p

and

16t

LP

12

S;

LP

16

S;

LP

16

S;

LP

16

S

KC747706

(1

5JU

L0

9-0

1-

LE

P)

179

F

and

1458

R

LP

RA

G2

Ves

per

tili

on

idae

Myo

tis

luci

fugus

cariss

ima

Mu

seu

m o

f S

ou

thw

este

rn

Bio

log

y

KC747692

(N

K3

42

4)

Molc

itF

an

d

H1

591

5

LP

CY

TB

KC747661

(N

K3

42

4)

12c

and

12

g;

12h

an

d 1

6q;

16p

and

16

t

LP

12

S;

LP

16

S;

LP

16

S

KC747707

(N

K3

42

4)

179

F

and

1458

R

LP

RA

G2

Ves

per

tili

on

idae

Myo

tis

luci

fugus

relict

us

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747693

(N

K6

53

)

Molc

itF

and

H1

591

5

LP

CY

TB

KC747662

(N

K6

53

)

12c

and

1

2g;

12h

and

16t,

seq

uen

ced

wit

h 1

6q

,

16p

, 16

j,

and

16n

LP

12

S;

LP

16

S

KC747708

(N

K6

53

)

179

F

and

1458

R

LP

RA

G2

Ves

per

tili

on

idae

Myo

tis

mel

anorh

inus

Lou

isia

na

Sta

te

Un

iver

sity

M

use

um

of

Nat

ura

l

Sci

ence

KC747694

(M

893

7)

Molc

itF

and

H1

591

5

LP

CY

TB

KC747663

(M

893

7)

12c

and

12

s; 1

2a

and

12

gg;

12

e an

d

16n

; 16j

and

16

l;

LP

12

S;

LP

12

S;

LP

16

S;

LP

16

S

KC747709

(M

893

7)

179

F

and

1458

R

LP

RA

G2

Ves

per

tili

on

idae

M

yotis m

ille

ri

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747695

(N

K8

13

3)

Molc

itF

and

H1

591

5

LP

CY

TB

KC747664

(N

K8

13

3)

12c

and

12

g w

ith

12a

and

1

2s;

12

h

and

16t,

seq

uen

ced

wit

h 1

6q

,

16p

, 16

j,

and

16n

LP

12

S;

LP

12

S;

LP

16

S

KC747710

(N

K8

13

3)

179

F

and

1458

R

LP

RA

G2

Ves

per

tili

on

idae

Myo

tis

nig

rica

ns

AF

37

6864

(MV

Z A

D5

0)

n/a

n

/a

AF

32

6099

(FM

NH

12

921

0)

n/a

n

/a

GU

32

8088

(FM

NH

12

921

0)

n/a

n

/a

97

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Ves

per

tili

on

idae

M

yotis occ

ultus

Mu

seu

m o

f S

ou

thw

este

rn

Bio

log

y

KC747696

(N

K4

08

01

)

Molc

itF

an

d

H1

591

5

LP

CY

TB

KC747665

(N

K4

08

01

)

12c

and

12

s; 1

2a

and

12

g;

12

e an

d

16q

; 12h

an

d 1

6q;

16j

and

16

t

LP

12

S;

LP

12

S;

LP

16

S;

LP

16

S;

LP

16

S

KC747711

(N

K4

08

01

)

179

F

and

1458

R

LP

RA

G2

Ves

per

tili

on

idae

M

yotis

thys

anodes

AF

37

6869

(T

K 7

8796

) n

/a

n/a

A

F3

26

100

(T

K

7880

0)

n/a

n

/a

AM

265

693

(T

K7

879

6)

n/a

n

/a

Ves

per

tili

on

idae

M

yotis ve

life

r

AF

37

6870

(M

VZ

1467

66

) n

/a

n/a

AY

49

5509

(T

K

1192

9;

TT

U

4640

5)

n/a

n

/a

AM

265

695

(MV

Z 1

467

66

) n

/a

n/a

Ves

per

tili

on

idae

M

yotis vi

vesi

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

AJ5

044

06

n/a

n

/a

KC747666

(N

K5

68

8)

12c

and

1

2g;

12h

and

16t,

seq

uen

ced

wit

h 1

6q

,

16p

, 16

j,

and

16n

LP

12

S;

LP

16

S

AM

265

696

n/a

n

/a

Ves

per

tili

on

idae

M

yotis vo

lans

AF

37

6871

(TK

78

980

) n

/a

n/a

AY

49

5510

(T

K

7898

0;

TT

U

7954

5)

n/a

n

/a

GU

32

8092

(T

TU

7954

5)

n/a

n

/a

Ves

per

tili

on

idae

Myo

tis

yum

anen

sis

AF

37

6875

(MV

Z 1

558

5)

n/a

n

/a

AY

49

5512

(T

K

2875

3;

TT

U

4320

0)

n/a

n

/a

AM

265

700

(MV

Z 1

558

53

)

n/a

n

/a

Ves

per

tili

on

idae

Nyc

tice

ius

hum

eralis

Lou

isia

na

Sta

te

Un

iver

sity

Mu

seu

m o

f N

atu

ral

Sci

ence

KC747697

(M

890

1)

L1

47

24

and

B

SV

ES

26

8

H

LE

PC

YT

B5

AF

32

6102

(T

K

9064

9;

TT

U

8066

4)

n/a

n

/a

GU

32

8096

(T

TU

4953

6)

n/a

n

/a

Ves

per

tili

on

idae

Para

stre

llus

hes

per

us

Mu

seu

m o

f

Sou

thw

este

rn

Bio

log

y

KC747698

(N

K3

22

23

)

Molc

itF

and

H

1591

5

LP

CY

TB

AY

49

5522

(T

K

7870

3;

TT

U

7926

9)

n/a

n

/a

GU

32

8099

(T

TU

7

926

9)

n/a

n

/a

98

Fam

ily

Sp

ecie

s M

use

um

Cytb

1

2S

-16

S

RA

G2

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Gen

Ban

k

acce

ssio

n

nu

mb

ers

(sp

ecim

en

nu

mb

ers)

P

rim

ers

PC

R

pro

file

Ves

per

tili

on

idae

Per

imyo

tis

subflavu

s

AJ5

044

49

(T

K 9

0671

) n

/a

n/a

AY

49

5523

(T

K

9067

1;

TT

U

8068

4)

n/a

n

/a

GU

32

8103

(T

TU

8

068

4)

n/a

n

/a

Ves

per

tili

on

idae

Rhogee

ssa

gra

cilis

Tex

as

Coop

erat

ive

Wil

dli

fe

Coll

ecti

on

EF

22

236

2

(AK

11

059

) n

/a

n/a

KC747667

(A

K1

10

59

)

12c

and

12

s; 1

2a

and

12

g;

12h

and

16q

; 16j

and

16n;

16p

and

16

t

LP

12

S;

LP

12

S;

LP

16

S;

LP

16

S

KC747712

(A

K1

10

59

)

179

F

and

1458

R

LP

RA

G2

Rhin

olo

phid

ae

Rhin

olo

phus

luct

us

Port

lan

d S

tate

Un

iver

sity

Mu

seu

m o

f V

erte

bra

te

Zoo

log

y

JN1

062

64

(MY

7)

Molc

itF

an

d

H1

591

5

LP

CY

TB

KC747668

(M

Y7

)

12c

and

12

g;

12h

an

d 1

6t,

seq

uen

ced

wit

h 1

6q

, 1

6p

, 16

j,

and

16n

LP

12

S;

LP

16

S

KC747713

(M

Y7

)

RA

G2

-

F1

and

R

AG

2-

R2

RA

G2

B

Rhin

olo

phid

ae

Rhin

olo

phus

cele

ben

sis

Port

lan

d S

tate

Un

iver

sity

M

use

um

of

Ver

teb

rate

Zoo

log

y

JN1

062

65

(PD

X4

1)

Molc

itF

and

H1

591

5

LP

CY

TB

KC747669

(P

DX

41

)

12c

and

1

2g;

12h

and

16t,

seq

uen

ced

wit

h 1

6q

,

16p

, 16

j,

and

16n

LP

12

S;

LP

16

S

KC747714

(P

DX

41

)

RA

G2

-F

1 a

nd

RA

G2

-

R2

RA

G2

B

Pte

rop

od

idae

Thoopte

rus

nig

resc

ens

Port

lan

d S

tate

Un

iver

sity

M

use

um

of

Ver

teb

rate

Zoo

log

y

KC747699

(PD

X3

5)

Molc

itF

and

H1

591

5

LP

CY

TB

KC747670

(P

DX

35

)

12c

and

1

2g;

12h

and

16t,

seq

uen

ced

wit

h 1

6q

,

16p

, 16

j,

and

16n

LP

12

S;

LP

16

S

KC747715

(P

DX

35

)

RA

G2

-F

1 a

nd

RA

G2

-

R2

RA

G2

B

Pte

rop

od

idae

Sty

loct

eniu

m

wallace

i

Port

lan

d S

tate

U

niv

ersi

ty

Mu

seu

m o

f

Ver

teb

rate

Z

oo

log

y

KC747700

(PD

X4

0)

Molc

itF

and

H

1591

5

LP

CY

TB

KC747671

(P

DX

40

)

12c

and

12

g;

12h

and

16t,

se

qu

ence

d

wit

h 1

6q

,

16p

, 16

j,

and

16n

LP

12

S;

LP

16

S

KC747716

(P

DX

40

)

RA

G2

-

F1

and

RA

G2

-R

2

RA

G2

B

99

Primers designed for this study:

Las

iuru

sF1 5

’-3’:

AW

AY

CC

AC

GA

CY

AR

TG

AC

AC

G

Las

iuru

sF2 5

’-3’:

GC

CC

CT

TC

AA

AC

AT

CT

CC

T

Las

iuru

smid

dle

F1 5

’-3’:

AY

AT

AA

TY

CC

HT

TC

CA

YC

CY

TA

Las

iuru

sR1 5

’-3’:

AT

TA

GG

CT

GG

CG

AG

TG

GT

AT

PCR profiles:

LP

CY

TB

: 94°C

for

3:0

0 m

in.;

40 c

ycl

es o

f 94°C

for

0:4

5, 45°C

for

0:4

5, 72°C

for

1:3

0;

72°C

for

5:0

0 m

in.

LE

PC

YT

B5:

94°C

for

3:0

0 m

in.;

39 c

ycl

es o

f 94°C

for

0:4

5, 50°C

for

0:4

5, 72°C

for

1:3

0;

72°C

for

5:0

0 m

in.

CT

B50:

94°C

for

4:0

0 m

in.;

35 c

ycl

es o

f 94°C

for

0:4

0, 50°C

for

0:4

0, 72°C

for

1:0

0;

72°C

for

10:0

0 m

in.

TD

CT

B:

94°C

for

3:0

0 m

in.;

18 c

ycl

es o

f 94°C

for

0:4

5, 60°C

for

0:4

5 (

low

er b

y 1

° per

cycl

e), 72°C

for

1:0

0;

22 c

ycl

es o

f 94°C

for

0:4

5, 42°C

for

0:4

5, 72°C

for

1:3

0;

72°C

for

5:0

0 m

in.

LP

12S

: 35 c

ycl

es o

f 94°C

for

0:4

0, 42°C

for

2:0

0, 72°C

for

3:0

0;

72°C

for

30:0

0 m

in.

LP

16S

: 35 c

ycl

es o

f 94°C

for

0:4

0, 52°C

for

1:0

0, 72°C

for

2:0

0;

72°C

for

30:0

0 m

in.

LP

RA

G2:

94°C

for

3:0

0 m

in.;

39 c

ycl

es o

f 94°C

for

0:4

5, 60°C

for

0:4

5, 72°C

for

1:3

0;

72°C

for

5:0

0 m

in.

RA

G2B

: 95°C

for

2:0

0 m

in.;

35 c

ycl

es o

f 95°C

for

0:3

0, 65°C

for

0:3

0, 72°C

for

2:0

0;

72°C

for

10:0

0 m

in.

100

APPENDIX

II

SEQUENCES IN FULL PHYLOGENY

Non-s

pec

ies

pool

Gen

Ban

k s

equen

ces

use

d t

o i

nfe

r phylo

gen

y. M

use

um

nu

mber

s of

spec

imen

s ar

e giv

en i

n p

aren

thes

es.

Family

Species

Cytb

12S-16S

RAG2

Molo

ssid

ae

Oto

mops m

artie

nss

eni

AY

59

1536

(unp

ub

lish

ed)

AY

49

5459

(F

MN

H 1

3763

3)

GU

32

8097

(F

MN

H 1

3763

3)

Morm

oop

idae

M

orm

oops bla

invi

llii

AF

33

8685

(T

K3

216

6)

AF

40

7172

(T

K3

216

6)

AY

02

8169

(T

K3

216

6)

Morm

oop

idae

Pte

ronotu

s gym

nonotu

s A

F3

38

674

(T

K2

284

5)

AF

40

7177

(T

K2

284

5)

AF

33

8694

(C

N10

426

5)

Morm

oop

idae

Pte

ronotu

s m

acl

eayi

i A

F3

38

683

(T

K3

217

1)

AF

40

7178

(T

K3

216

2)

AF

33

8700

(T

K3

216

2)

Morm

oop

idae

Pte

ronotu

s per

sonatu

s A

F3

38

680

(T

K1

204

3)

AF

40

7182

(T

K1

204

3)

AF

33

8699

(T

K1

204

3)

Morm

oop

idae

Pte

ronotu

s quadriden

s A

F3

38

681

(T

K3

217

1)

AF

40

7179

(T

K3

217

1)

AF

33

8695

(T

K3

217

1)

Nat

alid

ae

Nata

lus m

icro

pus

AY

62

1026

(A

MN

H 2

7463

1)

AF

34

5925

(T

K1

945

4)

AY

14

1023

(T

K 9

454

)

Ph

yll

ost

om

idae

Bra

chyp

hyl

la c

ave

rnaru

m

AY

62

0457

(T

K 2

1807

) A

Y3

95

806

(T

K2

180

7)

AF

31

6436

(T

K1

563

0)

Ph

yll

ost

om

idae

D

iaem

us yo

ungi

FJ1

554

75

(T

K3

4625

) A

F4

11

534

A

F3

16

445

(T

K3

462

5)

Ph

yll

ost

om

idae

M

onophyl

lus re

dm

ani

AF

38

2888

(T

K2

769

4)

AY

39

5824

(T

K2

770

8)

AF

31

6473

(T

K2

765

4)

Ph

yll

ost

om

idae

Lio

nyc

teris sp

urr

elli

AF

42

3100

(T

K2

254

8)

AY

39

5815

(T

K2

254

1)

AF

31

6455

(T

K1

016

4)

Ph

yll

ost

om

idae

Lonch

ophyl

la thom

asi

AF

18

7034

(T

K1

717

7)

AY

39

5842

(T

K5

532

1)

AF

31

6456

(T

K1

717

7)

Ph

yll

ost

om

idae

Ero

phyl

la sez

ekorn

i A

Y6

20

439

(A

MC

C 1

02

699

) A

Y3

95

839

(T

K9

416

) A

F3

16

450

(T

K9

416

)

Ph

yll

ost

om

idae

C

hro

topte

rus auritu

s F

J15

54

81

(T

K2

1039

) A

F4

11

538

(T

K7

045

7)

AF

31

6442

(T

K1

710

4)

Ph

yll

ost

om

idae

G

lyphonyc

teris davi

esi

AY

38

0747

(T

K1

637

0)

AY

39

5812

(R

OM

1040

422

) A

F3

16

464

Ph

yll

ost

om

idae

G

lyphonyc

teris sy

lves

tris

AY

38

0746

(T

K1

637

4)

AY

39

5841

(T

K1

046

1)

AF

31

6471

(T

K1

045

3)

Ph

yll

ost

om

idae

Lonch

orh

ina a

urita

F

J15

54

94

(T

K2

0560

) A

Y3

95

843

(T

K2

056

0)

AF

31

6457

(T

K2

056

0)

Ph

yll

ost

om

idae

Lophostom

a silvi

cola

F

J15

54

93

(T

K5

6716

) A

F4

42

092

(T

K5

671

6)

AF

44

2081

(T

K5

671

6)

Ph

yll

ost

om

idae

Macr

ophyl

lum

macr

ophyl

lum

F

J15

54

84

(C

MN

H7

828

9)

AF

41

1540

(T

K1

9119

) A

F3

16

458

(T

K1

911

9)

Ph

yll

ost

om

idae

M

icro

nyc

teris bra

chyo

tis

AY

38

0748

(T

K2

523

9)

AF

41

1536

(T

K2

523

8)

AF

31

6463

(T

K2

523

8)

Ph

yll

ost

om

idae

M

icro

nyc

teris hirsu

ta

AY

38

0751

(T

K2

504

1)

AY

39

5819

(T

K2

504

1)

AF

31

6465

(T

K 2

5041

)

Ph

yll

ost

om

idae

M

icro

nyc

teris m

egalo

tis

AY

38

0758

(T

K1

707

1)

AY

39

5821

(T

K1

707

1)

AF

31

6467

(T

K1

878

5)

Ph

yll

ost

om

idae

M

icro

nyc

teris m

inuta

A

Y3

80

752

(T

K1

637

1)

AY

39

5823

(T

K1

978

1)

AF

31

6468

(T

K1

787

7)

Ph

yll

ost

om

idae

M

icro

nyc

teris nic

efori

AY

38

0749

(T

K1

518

9)

AY

39

5830

(T

K1

5189

) A

F3

16

469

Ph

yll

ost

om

idae

Mic

ronyc

teris

schm

idto

rum

A

Y3

80

753

(T

K4

044

7)

AF

41

1535

(T

K7

044

7)

AF

31

6470

(T

K7

044

7)

101

Family

Species

Cytb

12S-16S

RAG2

Ph

yll

ost

om

idae

M

imon c

renula

tum

F

J15

54

78

(C

MN

H2

523

0)

AF

41

1543

(T

K2

523

0)

AF

31

6472

(T

K1

512

1)

Ph

yll

ost

om

idae

Phyl

loder

ma ste

nops

FJ1

554

80

(T

K8

6685

) A

F4

11

542

(T

K1

020

1)

AF

31

6480

(T

K1

020

1)

Ph

yll

ost

om

idae

Phyl

lostom

us hastatu

s F

J15

54

79

(C

MN

H78

333

) A

F4

11

541

(T

K1

928

9)

AF

31

6479

(T

K1

924

3)

Ph

yll

ost

om

idae

Tonatia b

iden

s F

J15

54

90

(M

VZ

1856

73

) A

F4

42

091

(T

K5

651

9, M

VZ

185

673

) A

F4

42

088

(M

VZ

185

673

; T

K5

65

19

)

Ph

yll

ost

om

idae

Tonatia sauro

phila

FJ1

554

88

(R

OM

10

321

0)

AF

41

1530

(R

10

340

1)

AF

44

2084

(R

OM

1032

10

)

Ph

yll

ost

om

idae

Tra

chops ci

rrhosu

s D

Q2

33

669

(T

K 1

8829

) A

F4

11

539

(T

K1

882

9)

AF

31

6490

(T

K1

882

9)

Ph

yll

ost

om

idae

Vam

pyr

um

spec

trum

F

J15

54

82

(T

TU

61

070

) A

F4

11

537

(T

K4

037

0)

AF

31

6495

(T

K4

037

0)

Ph

yll

ost

om

idae

Am

etrida c

entu

rio

AY

60

4446

(A

MC

C11

032

4)

AY

39

5802

(T

K1

774

1)

AF

31

6430

(T

K1

881

0)

Ph

yll

ost

om

idae

Ard

ops nic

hollsi

AY

57

2337

AY

39

5803

(T

K1

560

2)

AF

31

6434

(T

K1

560

2)

Ph

yll

ost

om

idae

Ari

teus flave

scen

s A

Y6

04

436

(A

MC

C10

276

1)

AY

39

5804

(T

K2

769

6)

AF

31

6435

(T

K2

769

6)

Ph

yll

ost

om

idae

Art

ibeu

s ci

ner

eus

AC

U6

6511

(T

K 1

8790

AM

NH

267

19

7)

AY

39

5810

(T

K1

879

0)

AF

31

6443

(T

K1

879

0)

Ph

yll

ost

om

idae

Ench

isth

enes

hartii

AH

U6

65

17

(T

K 2

2690

) A

Y3

95

838

(T

K5

533

1)

AF

31

6449

(T

K2

269

0)

Ph

yll

ost

om

idae

Art

ibeu

s ja

maic

ensis

DQ

869

480

(T

K2

768

2)

AF

26

3225

(T

K2

679

98

) F

N6

41

674

(A

jam

198

0)

Ph

yll

ost

om

idae

C

entu

rio sen

ex

AY

60

4441

(T

K1

311

0)

AF

26

3227

(T

K1

353

7)

AF

31

6438

(T

K1

311

0)

Ph

yll

ost

om

idae

Ect

ophyl

la a

lba

AY

15

7033

(T

K1

639

5)

AY

39

5811

(T

K1

351

4)

AF

31

6448

(T

K1

639

5)

Ph

yll

ost

om

idae

M

esophyl

la m

acc

onnel

li

AY

15

7035

(T

K5

531

6)

AY

39

5818

(T

K7

049

1)

AF

31

6462

(T

K5

531

6)

Ph

yll

ost

om

idae

Pyg

oder

ma b

ilabia

tum

A

Y6

04

438

(M

VZ

185

904

) A

Y3

95

826

(T

K1

268

2)

AF

31

6483

(T

K1

268

2)

Ph

yll

ost

om

idae

Ste

noder

ma rufu

m

DQ

312

400

(T

K2

851

5)

AY

39

5829

(T

K2

178

6)

AF

31

6487

(T

K2

179

0)

Ph

yll

ost

om

idae

U

roder

ma b

ilobatu

m

AY

16

9913

(tk

3496

3)

AY

39

5831

(T

K4

600

6)

AF

31

6491

(T

K3

492

6)

Ph

yll

ost

om

idae

Vam

pyr

essa

bid

ens

AY

15

7045

(T

K5

532

2)

AY

39

5833

(T

K7

045

1)

AF

31

6492

(T

K5

532

2)

Ph

yll

ost

om

idae

Vam

pyr

essa

pusilla

AY

15

7050

(T

K7

053

3)

AY

39

5832

(T

K7

045

4)

AF

31

6493

(T

K7

053

3)

Ph

yll

ost

om

idae

Vam

pyr

odes

cara

ccio

li

AY

15

7034

(T

K2

508

3)

AY

39

5846

(T

K7

054

0)

AF

31

6494

(T

K2

508

3)

Ves

per

tili

on

idae

Antrozo

us dubia

quer

cus

EF

22

238

1 (

SP

1259

8)

AY

39

5863

(R

OM

9771

9)

GU

32

8050

(R

OM

97

719

)

Ves

per

tili

on

idae

C

istu

go sea

bra

e A

J84

19

62

(M

R-M

977

) G

U3

28

039

(M

977

) G

U3

28

052

(M

977

)

Ves

per

tili

on

idae

Epte

sicu

s dim

inutu

s A

F3

76

833

(M

VZ

AD

49

6)

AY

49

5465

(T

K 1

5033

; T

TU

4815

4)

GU

32

8056

(T

TU

481

54

)

Ves

per

tili

on

idae

Epte

sicu

s hotten

totu

s A

J84

19

63

(M

R-M

984

) A

Y4

95

466

(C

M 8

9000

; T

K 3

301

3)

GU

32

8059

(C

M 8

9000

)

Ves

per

tili

on

idae

Epte

sicu

s se

rotinus

AF

37

6837

(E

R 6

59

) A

Y4

95

467

(T

K 4

0897

; T

TU

7094

7)

HM

561

651

(T

K4

08

97;

TT

U7

0947

)

Ves

per

tili

on

idae

K

eriv

oula

hard

wic

kii

GU

58

5655

(T

K 1

5241

0)

AF

34

5928

(R

OM

1108

29

) A

Y1

41

034

(R

OM

11

082

9)

102

Family

Species

Cytb

12S-16S

RAG2

Ves

per

tili

on

idae

K

eriv

oula

papillo

sa

GU

58

5663

(T

K 1

5240

3)

AF

34

5927

(R

OM

1108

50

) A

Y1

41

035

(R

OM

11

085

0)

Ves

per

tili

on

idae

K

eriv

oula

pel

luci

da

EU

1887

88

(T

K1

5205

5)

AY

49

5476

(F

3598

7, R

OM

10

217

7)

GU

32

8064

(R

OM

10

217

7)

Ves

per

tili

on

idae

Laep

hotis nam

iben

sis

EU

7974

42

(S

P4

160

) A

Y4

95

477

(S

P 4

097

, T

M 3

75

47

) H

M561

668

(S

P416

0,

CM

9318

7)

Ves

per

tili

on

idae

M

inio

pte

rus frate

rculu

s A

J84

19

75

(M

R-M

988

) A

Y4

95

486

(C

M 9

8058

, T

K 3

31

32

) G

U3

28

067

(C

M 9

8058

)

Ves

per

tili

on

idae

M

inio

pte

rus sc

hre

iber

sii

AY

20

8140

GU

32

8042

(M

HN

G 1

805

.01

3)

GU

32

8069

(M

HN

G 1

805

.01

3)

Ves

per

tili

on

idae

M

yotis alb

esce

ns

AF

37

6839

(F

MN

H 1

470

67

) A

Y4

95

492

(C

M 7

7691

; T

K 1

793

2)

GU

32

8076

(C

M 7

7691

)

Ves

per

tili

on

idae

M

yotis austro

riparius

AM

261

885

(T

HK

00

2-F

BF

-13

) A

Y4

95

493

(M

LK

40

79

, U

M 1

6629

) A

M2

65

642

(T

HK

00

2-F

BF

-13

)

Ves

per

tili

on

idae

M

yotis boca

gei

A

J50

44

08

AF

32

6096

GU

32

8077

(F

MN

H 1

5007

5)

Ves

per

tili

on

idae

M

yotis ca

pacc

inii

AF

37

6845

AY

49

5494

(T

K 2

5610

, T

TU

4055

4)

GU

32

8079

(T

TU

4055

4)

Ves

per

tili

on

idae

M

yotis dauben

tonii

AF

37

6847

(E

R 1

44

) A

Y4

95

498

(IZ

EA

26

92

, M

HN

G 1

805

.054

) F

N6

41

679

(M

dau

21

08

)

Ves

per

tili

on

idae

M

yotis dom

inic

ensis

AF

37

6848

(T

K 1

5613

) A

Y4

95

500

(T

K 1

5613

) A

M2

65

654

(T

K 1

561

3)

Ves

per

tili

on

idae

M

yotis el

egans

AM

261

891

(301

1)

AY

49

5501

(F

3547

1, R

OM

10

129

3)

AM

265

655

(301

1)

Ves

per

tili

on

idae

M

yotis ke

ays

i A

F3

76

852

(T

K 1

3532

) A

Y4

95

503

(T

K 1

3532

) A

M2

65

668

(T

K 1

353

2)

Ves

per

tili

on

idae

M

yotis la

tiro

stris

AM

262

330

(M

R-6

08

) G

U9

52

769

(M

606

) G

U3

28

084

(M

606

)

Ves

per

tili

on

idae

M

yotis le

vis

AF

37

6853

(F

MN

H 1

41

600

) A

F3

26

097

(F

MN

H 1

41

600

) G

U3

28

085

(F

MN

H 1

4160

0)

Ves

per

tili

on

idae

M

yotis m

yotis

AF

37

6860

(E

R 1

312

) A

F3

26

098

(IZ

EA

37

90

) G

U3

28

087

(M

HN

G 1

805

.06

2)

Ves

per

tili

on

idae

M

yotis riparius

AF

37

6866

(M

VZ

AD

11

9)

AF

26

3236

(A

MN

H2

685

91

) A

Y1

41

032

(A

MN

H 2

6859

1)

Ves

per

tili

on

idae

M

yotis ru

ber

A

F3

76

867

(M

VZ

AD

47

2)

AY

49

5506

(F

4440

9, R

OM

11

111

0)

AM

265

688

(M

VZ

1859

99

)

Ves

per

tili

on

idae

M

yotis wel

witsc

hii

AF

37

6873

(F

MN

H 1

44

313

) A

Y4

95

511

(F

MN

H 1

4431

3)

GU

32

8093

(F

MN

H 1

4431

3)

Ves

per

tili

on

idae

N

eoro

mic

ia n

anus

EU

7974

28

(A

K2

116

1)

AY

49

5474

(C

M 9

8003

, T

K 3

33

78

) G

U3

28

062

(D

M 7

54

2)

Ves

per

tili

on

idae

N

ycta

lus le

isle

ri

AF

37

6832

(IZ

EA

26

39

) A

Y4

95

517

(F

MN

H 1

4037

4)

HM

561

657

(F

MN

H1

403

74

)

Ves

per

tili

on

idae

N

ycta

lus noct

ula

A

J84

19

67

(u

nvouch

ered

) A

Y4

95

518

(N

HM

B 2

09/8

7)

HM

561

658

(N

HM

B 2

09

/87

)

Ves

per

tili

on

idae

O

tonyc

teris hem

prich

i H

M030

844

(N

MP

92

667

, un

pub

lish

ed)

AF

32

6103

(S

P 7

782

) G

U3

28

098

(S

P 7

882

)

Ves

per

tili

on

idae

Pip

istrel

lus hes

per

idus

AJ8

419

68

(M

R-M

987

) H

M561

628

(D

M80

13

) H

M561

659

(D

M80

13

)

Ves

per

tili

on

idae

Pip

istrel

lus nath

usii

AJ5

044

46

(u

nvouch

ered

) A

F3

26

104

(IZ

EA

28

30

) H

M561

660

(IZ

EA

28

30

, M

HN

G1

806

.003

)

Ves

per

tili

on

idae

Pip

istrel

lus pip

istr

ellu

s D

Q6

30

431

HM

561

630

(M

HN

G1

956

.031

, M

14

39

) H

M561

662

(M

14

39

, M

HN

G1

956

.031

)

Ves

per

tili

on

idae

Ple

cotu

s auritu

s A

B0

8573

4

AF

32

6106

(IZ

EA

26

94

) G

U3

28

100

(M

HN

G 1

806

.04

7)

Ves

per

tili

on

idae

C

ory

norh

inus ra

fines

quii

AY

78

1725

(813

) A

F3

26

091

(T

K 5

959

) G

U3

28

055

(T

TU

453

80

)

Ves

per

tili

on

idae

Rhogee

ssa a

eneu

s E

F2

2236

4 (

TK

20

712

) A

Y4

95

530

(T

K 2

0712

, T

TU

4001

2)

HM

561

633

(T

K2

07

12

, T

TU

40

01

)

Ves

per

tili

on

idae

Rhogee

ssa m

ira

EF

22

233

6 (

TK

45

014

) A

Y4

95

531

(T

K 4

5014

, U

NA

M)

HM

561

634

(T

K4

50

14

)

103

Family

Species

Cytb

12S-16S

RAG2

Ves

per

tili

on

idae

Rhogee

ssa p

arv

ula

E

F2

2234

6 (

TK

20

653

) A

F3

26

109

(T

K 2

0653

) G

U3

28

108

(T

TU

366

33

)

Ves

per

tili

on

idae

Rhogee

ssa tum

ida

EF

22

235

0 (

TK

40

186

) A

F3

26

110

(T

K 4

0186

) G

U3

28

109

(T

TU

612

31

)

Ves

per

tili

on

idae

Sco

tom

anes

orn

atu

s D

Q4

35

069

AY

49

5537

(F

4256

8, R

OM

10

759

4)

HM

561

656

(F

425

68

, R

OM

107

594

)

Ves

per

tili

on

idae

Sco

tophilus din

ganii

EU

7509

95

(F

52

131

) A

Y4

95

533

(F

MN

H 1

4723

5)

GU

32

8111

(F

MN

H 1

4723

5)

Ves

per

tili

on

idae

Sco

tophilus hea

thi

EU

7509

44

(R

OM

107

786

) A

Y4

95

534

(F

4276

9, R

OM

10

778

6)

GU

32

8112

(R

OM

10

778

6 )

Ves

per

tili

on

idae

Sco

tophilus ku

hlii

EU

7509

31

(M

VZ

18

642

1)

AF

32

6111

(F

MN

H 1

45

684

) G

U3

28

113

(F

MN

H 1

4568

4)

Ves

per

tili

on

idae

Sco

tophilus le

uco

gaster

E

U7

509

40

(S

P 1

0136

) A

Y3

95

867

(T

K3

335

9)

GU

32

8114

(C

M 9

0854

)

Ves

per

tili

on

idae

Sco

tophilus nux

EU

7509

38

(T

K 3

34

85

) A

Y4

95

535

(T

K 3

3484

) G

U3

28

115

(T

K 3

3484

)

Ves

per

tili

on

idae

Sco

tophilus vi

ridus

EU

7509

91

(S

P 5

500

) A

F3

26

112

(F

MN

H 1

50

084

) G

U3

28

117

(F

MN

H 1

5008

4)

Ves

per

tili

on

idae

Tyl

onyc

teri

s pach

ypus

EF

51

731

3 (

CT

P1

) A

Y4

95

538

(F

3844

2, R

OM

10

616

4)

HM

561

672

(F

384

42

, R

OM

106

164

)

Ves

per

tili

on

idae

Ves

per

tilio m

urinus

AF

37

6834

(IZ

EA

35

99

) A

Y3

95

866

(IZ

EA

35

99

) H

M561

676

(IZ

EA

35

99

, M

HN

G1

808

.017

)

Noct

ilio

nid

ae

Noct

ilio

alb

iven

tris

AF

33

0806

(T

K2

284

9)

AF

26

3223

(T

K4

600

4)

AF

31

6476

(T

K4

600

4)

Noct

ilio

nid

ae

Noct

ilio

lep

orinus

AF

33

0797

(T

K1

870

0)

AF

26

3224

(T

K1

851

5)

AF

33

0816

(T

K1

870

0)

Fu

rip

teri

dae

Furipte

rus horr

ens

AY

62

1004

(A

MC

C 1

09

523

) A

F3

45

922

(R

OM

1002

02

) A

Y1

41

016

(R

OM

10

020

2)

Th

yro

pte

rid

ae

Thyr

opte

ra trico

lor

AY

62

1005

(A

MC

C 1

10

107

) A

F2

63

233

(A

MN

H2

685

77

) A

Y1

41

028

(A

MN

H 2

6857

7)

Myst

acin

idae

M

ysta

cina tuber

cula

ta

NC

_0

069

25

(m

ayb

e)

AF

26

3222

(U

WZ

MM

2702

7)

AY

14

1021

(U

WZ

M-M

27

027

)

Myzo

pod

idae

M

yzopoda a

urita

D

Q1

78

334

AF

34

5926

AY

14

1022

(O

K4

24

6)

Em

bal

onu

rid

ae

Rhyn

chonyc

teris naso

E

F5

8419

2 (

RO

M 1

07

891

) A

Y3

95

851

(A

MN

H2

673

73

) A

Y8

34

662

Em

bal

onu

rid

ae

Sacc

opte

ryx

bilin

eata

E

F5

8420

2 (

RO

M 1

15

534

) A

F2

63

213

(A

MN

H 2

6784

2)

AY

14

1015

(A

MN

H 2

6784

2)

Meg

ader

mat

idae

M

egader

ma lyr

a

DQ

888

678

(A

9)

AF

06

9538

AF

20

3767

104

APPENDIX III

CHAPTER 2 SUPPLEMENTARY MATERIALS

S1. Fieldwork was conducted in July and August, 2010 in Oregon (permit #117-10) and Utah

(permit #1COLL8463) with the approval of Louisiana State University’s institutional animal

care and use committee (protocol #09-012) and following ASM’s guidelines for research on

small mammals (Sikes et al. 2011). The Bat Grid protocol of Ormsbee et al. (2006) was

implemented when performing surveys. Bats were captured using mist nets set over water and in

fly-ways; nets were open for the first 3.5 hours after sunset and checked at least every 15

minutes. Specimens were prepared using standard museum techniques; heart, liver, kidney, and

muscle tissues were preserved in ethanol. Specimens and tissues will be deposited at the

Louisiana State University Museum of Natural Science.

S2. Twenty-eight sites (1.7% of all sites) had individuals identified only as “Myotis sp.”; since

there was no way to know if one or more than one species were included in this identification,

this taxon was deleted from the community matrix. Myotis planiceps occurs in some of the

collection/capture locations included in this study but genetic data were unavailable so M.

planiceps was removed from the data matrix. One individual identified as Eptesicus sp. was

assigned to E. fuscus since no other Eptesicus species occur where this specimen was collected.

A few sites in the Sonoran and Chihuahuan Deserts had Eumops underwoodi, however the

sequences included in our phylogeny for E. underwoodi were in fact from Nyctinomops

macrotis. Therefore, sites with E. underwoodi were deleted from the data matrix.

In one location both Leptonycteris nivalis and L. yerbabuenae were collected along with

several individuals identified as Leptonycteris sp. Myotis californicus and M. ciliolabrum are

105

sister species that are difficult to differentiate and at several sites individuals were identified as

M. californicus/ciliolabrum. Myotis lucifugus and M. yumanensis are not sister species but are

almost impossible to differentiate in parts of their ranges without acoustic or genetic data and

individuals at several sites were identified as M. lucifugus/yumanensis. In addition, some Myotis

lucifugus subspecies may warrant elevation to specific status (Dewey 2006, Carstens and Dewey

2010).

To test whether alternate species identifications changed the outcome of PCS results,

SES-MPD and SES-MNTD were calculated using 5km buffer communities for all combinations

of Leptonycteris sp. assigned to L. nivalis or L. yerbabuenae; M. californicus/ciliolabrum

assigned to either californicus or ciliolabrum; M. lucifugus/yumanensis assigned to either

lucifugus or yumanensis; and finally, M. lucifugus subspecies assigned to respective subspecies

based on where the individuals were captured/collected or all M. lucifugus individuals assigned

to a single subspecies. A MANOVA was performed using the SES-MPD and SES-MNTD z-

values. There were no significant differences in SES-MPD and SES-MNTD z-values between

different combinations of species identifications (MANOVA, F=0.0262, 8576, p-value=1), so for all

subsequent analyses, Leptonycteris sp. were assigned to L. nivalis, M. californicus/ciliolabrum

were assigned to M. californicus, M. lucifugus/yumanensis were assigned to M. yumanensis, and

M. lucifugus subspecies were assigned to respective subspecies based on where individuals were

captured/collected.

106

Table S1: Models chosen by Modeltest for each gene

Gene Missing data Model

12S to 16S No GTR+I+G

12S to 16S Yes GTR+I+G

Cytb No TVM+I+G

Cytb Yes TVM+I+G

RAG2 No TVMef+I+G

RAG2 Yes GTR+I+G

Table S2: Number of communities with three or more species determined to be adequately sampled based

on Chao1 for each delimitation method in each desert.

Delimitation method

Desert

5km buffer 10km buffer 10km grid 50km grid 50km circle 100km circle

Great Basin 59 49 62 59 39 32

Mojave 19 10 27 18 16 10

Sonoran 24 26 41 30 24 9

Chihuahuan 66 39 83 52 28 23

Table S3: Results of Mantel tests between the distance matrix from the Best tree and other trees (listed in

the “Tree” column).

Tree Mantel statistic p-value

NNI50.2 0.090 0.087

SPR50.2 0.237 <0.001

NNI300.2 0.176 0.006

SPR300.2 0.099 0.021

Polytomy 0.414 <0.001

Bush 0.017 0.420

107

A.

B.

Fig

ure

S1. E

xam

ple

s of

pru

ned

tre

es u

sed i

n a

nal

yse

s. A

. B

est

tree

. B

. B

oots

trap

tre

e 807;

Robin

son-F

ould

s (R

F)

dis

tan

ce=

22.

C.

Nea

rest

-nei

ghbor

inte

rch

ange

(NN

I) 5

0 m

ove

tree

2 (

NN

I50.2

); R

F d

ista

nce

=20. D

. S

ub-t

ree

pru

ne

and r

egra

ft (

SP

R)

50 m

ove

tree

2

(SP

R50.2

); R

F d

ista

nce

=92. E

. N

NI3

00.2

; R

F d

ista

nce

=62. F

. S

PR

300.2

; R

F d

ista

nce

=108. G

. P

oly

tom

ies

tree

in w

hic

h a

ll c

lades

bel

ow

the

level

of

fam

ily w

ere

mad

e in

to p

oly

tom

ies;

RF

dis

tance

=94.

H.

Bush

tre

e in

whic

h a

ll c

lades

wer

e unre

solv

ed;

RF

dis

tance

=106. S

pec

ies

abbre

via

tions

are

the

firs

t 2

let

ters

of

the

gen

us

nam

e an

d f

irst

2 l

ette

rs o

f th

e sp

ecie

s nam

e as

found i

n

Appen

dix

I, fo

llow

ed b

y t

he

firs

t 2 l

ette

rs o

f th

e su

bsp

ecie

s nam

e fo

r M

yotis lu

cifu

gus.

M. dom

inic

ensi

s w

as a

pla

ce-h

old

er f

or M

.

pla

nic

eps

but

was

rem

ov

ed f

or

anal

yse

s su

mm

ariz

ed i

n T

able

3 a

nd A

pp

endix

III

as

was

Eum

ops under

woodi

(Euun)

whic

h w

as i

n

fact

Nyc

tinom

ops m

acr

otis.

108

C.

(Fig

ure

S1 c

onti

nued

)

D.

109

E.

(Fig

ure

S1 c

onti

nued

)

F.

110

G.

(Fig

ure

S1 c

onti

nued

)

H.

111

Figure S2. Distribution of MPD p-values across the study area for the all-deserts species pool for

A. 50km grid and B. 5km circles.

112

A.

B.

C.

D.

Figure S3. Moran’s I correllograms for the all-desert species pool for 5km buffer communities A.

MPD and B. MNTD; 10km buffer communities C. MPD and D. MNTD; 10km grid communities

E. MPD and F. MNTD; 50km grid communities G. MPD and H. MNTD; 50km circle

communities I. MPD and J. MNTD; and 100km circle communities K. MPD and L. MNTD.

Distance units on the X-axes are in meters.

113

E.

F.

G.

H.

(Figure S3 continued)

114

I.

J.

K.

L.


115

Figure S4: Mean and standard deviation of Robinson-Foulds (RF) distance between the best tree

and bootstrap (mean of 21 trees, shown on the x-axis at 350 moves), nearest-neighbor

interchange (NNI; mean of 10 trees per number of moves), and sub-tree prune and re-graft (SPR;

mean of 10 trees per number of moves) trees, both unpruned (NNI and SPR) and pruned to

include only species in the “all taxa” species pool. The maximum RF distance between the best

tree and unpruned trees (162 leaves or taxa) is 318 while the maximum distance between pruned

trees (56 leaves or taxa) is 109.

0

50

100

150

200

250

300

350

0 50 100 150 200 250 300 350

Mean RF Distance

Number of Moves

NNI ± StDev

Pruned NNI ± StDev

SPR ± StDev

Pruned SPR ± StDev

Bootstrap ± StDev

Bootstrap

116

A.

B.

Figure S5: Graphs showing distribution of PCS metrics calculated from all trees in relation to RF

distance. “None” indicates metrics calculated from the best tree, “Polytomy” refers to the tree

containing polytomies below the family level, “Bush” indicates the completely unresolved

phylogeny, while the remaining data labels are explained in Figure S2’s legend above. All 5km

buffer communities are represented in (A) for MPD and (B) for MNTD. Since trends for

individual communities are difficult to distinguish, Sites 2 (C and D) and 192 (E and F) were

arbitrarily chosen as exemplars of changes to PCS metrics with differences in tree distance.

-7

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

SES-M

PD

RF distance

None

Bootstrap

NNI

SPR

Polytomy

Bush

-7

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

SES-M

NTD

RF distance

None

Bootstrap

NNI

SPR

Polytomy

Bush

117

C.

D.


-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

SES-M

PD

RF distance

None

Bootstrap

NNI

SPR

Polytomy

Bush

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

SES-M

NTD

RF distance

None

Bootstrap

NNI

SPR

Polytomy

Bush

118

E.

F.


-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

SES-M

PD

RF distance

None

Bootstrap

NNI

SPR

Polytomy

Bush

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120

SES-M

NTD

RF distance

None

Bootstrap

NNI

SPR

Polytomy

Bush

119

REFERENCES

Carstens, B. C. and T. A. Dewey. 2010. Species Delimitation Using a Combined Coalescent and

Information-Theoretic Approach: An Example from North American Myotis Bats.

Systematic Biology 59:400-414.

Dewey, T. A. 2006. Systematics and Phylogeography of North American Myotis (Chiroptera:

Vespertilionidae). University of Michigan, Ann Arbor.

Ormsbee, P. C., J. M. Zinck, J. M. Szewczak, L. E. Patrick, and A. H. Hart. 2006. Benefits of a

standardized sampling frame: an update on the “Bat Grid". Bat Research News 47:4.

Sikes, R. S., W. L. Gannon, and A. C. a. U. C. o. t. A. S. o. Mammalogists. 2011. Guidelines of

the American Society of Mammalogists for the use of wild mammals in research. Journal


120

APPENDIX

IV

CHAPTER 3 SUPPLEMENTARY M

ATERIA

LS

Tab

le S

1:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on P

CS

anal

yse

s fo

r al

l des

erts

com

bin

ed f

or

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“all

ves

per

tili

onid

s” a

nd “

all M

yotis”

. M

PD

M

NT

D

Del

imit

atio

n

met

hod

Tax

on

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s d

f

5k

m b

uff

er

Ves

per

tili

on

idae

1

3

122

9

320

.97

0.0

88

ns

9

127

8

311

.09

0.1

67

ns

288

Myo

tis

4

61

1

158

.79

0.0

56

ns

3

62

1

134

.89

0.4

14

ns

132

10k

m b

uff

er

Ves

per

tili

on

idae

1

0

94

8

262

.10

0.0

41

clu

ster

ed

10

95

7

254

.49

0.0

79

ns

224

Myo

tis

3

53

3

142

.61

0.0

61

ns

4

51

4

130

.42

0.2

05

ns

118

10k

m g

rid

Ves

per

tili

on

idae

1

2

150

12

394

.2

0.0

44

clu

ster

ed

9

156

9

387

.98

0.0

69

ns

348

Myo

tis

3

55

2

159

.76

0.0

09

clu

ster

ed

5

53

2

135

.53

0.1

58

ns

120

50k

m g

rid

Ves

per

tili

on

idae

1

3

126

10

334

.33

0.0

72

ns

12

131

6

326

.85

0.1

21

ns

298

Myo

tis

5

64

6

167

.28

0.1

59

ns

4

67

4

141

.07

0.6

87

ns

150

50k

m c

ircl

e V

esp

erti

lion

idae

7

87

7

241

.43

0.0

30

clu

ster

ed

7

87

7

236

.23

0.0

50

ns

202

Myo

tis

3

61

3

147

.00

0.2

09

ns

2

61

4

133

.56

0.4

95

ns

134

100k

m

circ

le

Ves

per

tili

on

idae

4

65

7

155

.38

0.4

09

ns

2

71

3

148

.92

0.5

55

ns

152

Myo

tis

4

49

4

127

.02

0.1

91

ns

3

50

4

105

.76

0.6

97

ns

114

df=

2*(n

um

ber

of

com

mun

itie

s)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

121

Tab

le S

2:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on P

CS

anal

yse

s fo

r th

e G

reat

Bas

in D

eser

t fo

r ea

ch c

om

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“GB

ves

per

tili

onid

s” a

nd “

GB

Myo

tis”

. M

PD

M

NT

D

Del

imit

atio

n

met

hod

Tax

on

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c p

-val

ue

Res

ult

s

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s d

f

5k

m b

uff

er

Ves

per

tili

on

idae

4

50

2

128

.63

0.1

347

ns

4

51

1

126

.95

0.1

583

ns

112

Myo

tis

0

27

2

40.3

3

0.9

626

over

dis

per

sed

0

27

2

42.2

6

0.9

401

ns

58

10k

m b

uff

er

Ves

per

tili

on

idae

5

39

3

107

.88

0.1

552

ns

5

37

5

110

.14

0.1

222

ns

94

Myo

tis

1

27

3

54.4

0

0.6

797

ns

1

29

0

52.3

2

0.7

492

ns

60

10k

m g

rid

Ves

per

tili

on

idae

4

54

3

127

.45

0.1

510

ns

3

55

3

125

.42

0.1

821

ns

122

Myo

tis

0

26

1

31.2

5

0.9

944

over

dis

per

sed

0

26

1

37.5

5

0.9

567

over

dis

per

sed

5

4

50k

m g

rid

Ves

per

tili

on

idae

3

53

3

130

.41

0.2

049

ns

3

53

3

129

.46

0.2

218

ns

118

Myo

tis

0

28

2

59.5

8

0.4

910

ns

2

27

1

60.1

3

0.4

708

ns

60

50k

m c

ircl

e V

esp

erti

lion

idae

6

29

3

99.3

8

0.0

373

clu

ster

ed

6

27

5

98.4

7

0.0

426

clu

ster

ed

76

Myo

tis

0

27

2

50.0

2

0.7

627

ns

1

27

1

48.8

0

0.7

998

ns

58

100k

m

circ

le

Ves

per

tili

on

idae

2

27

2

59.2

0

0.5

773

ns

2

27

2

60.9

2

0.5

15

ns

62

Myo

tis

0

22

3

48.5

6

0.5

312

ns

3

19

3

47.2

8

0.5

831

ns

50

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

122

Tab

le S

3:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on P

CS

anal

yse

s fo

r th

e M

oja

ve

Des

ert

for

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“MJ

ves

per

tili

onid

s” a

nd “

MJ M

yotis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

Tax

on

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s d

f

5k

m b

uff

er

Ves

per

tili

on

idae

0

19

0

35.4

1

0.5

899

ns

0

18

1

37.2

9

0.5

022

ns

38

Myo

tis

1

10

0

21.6

7

0.4

796

ns

1

10

0

21.9

5

0.4

626

ns

22

10k

m b

uff

er

Ves

per

tili

on

idae

1

7

2

25.3

0

0.1

902

ns

1

7

2

24.2

9

0.2

300

ns

20

Myo

tis

0

6

1

8.0

8

0.8

849

ns

0

7

0

12.3

0

0.5

822

ns

14

10k

m g

rid

Ves

per

tili

on

idae

3

21

1

55.8

0

0.2

661

ns

2

22

1

53.2

8

0.3

493

ns

50

Myo

tis

1

8

0

17.9

4

0.4

598

ns

1

8

0

19.7

3

0.3

485

ns

18

50k

m g

rid

Ves

per

tili

on

idae

1

16

1

41.0

4

0.2

593

ns

1

16

1

44.0

6

0.1

674

ns

36

Myo

tis

0

8

2

10.9

2

0.9

483

ns

0

9

1

13.0

2

0.8

766

ns

20

50k

m c

ircl

e V

esp

erti

lion

idae

0

14

2

37.3

4

0.2

371

ns

1

12

3

39.5

6

0.1

681

ns

32

Myo

tis

0

11

1

16.8

2

0.8

563

ns

0

12

0

20.0

7

0.6

929

ns

24

100k

m

circ

le

Ves

per

tili

on

idae

0

9

1

22.7

5

0.3

014

ns

0

9

1

22.2

4

0.3

279

ns

20

Myo

tis

1

4

2

7.1

4

0.9

293

ns

0

7

0

11.1

5

0.6

744

ns

14

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

123

Tab

le S

4:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on P

CS

anal

yse

s fo

r th

e S

onora

n D

eser

t fo

r ea

ch c

om

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“SN

ves

per

tili

onid

s” a

nd “

SN

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

Tax

on

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c p

-val

ue

Res

ult

s

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c p

-val

ue

Res

ult

s d

f

5k

m b

uff

er

Ves

per

tili

on

idae

1

16

0

37.6

0

0.3

08

ns

1

14

2

34.1

4

0.4

61

ns

34

Myo

tis

1

4

0

19.0

8

0.0

39

clu

ster

ed

0

5

0

8.8

4

0.5

48

ns

10

10k

m b

uff

er

Ves

per

tili

on

idae

1

20

1

47.6

0

0.3

29

ns

1

21

0

51.3

6

0.2

08

ns

44

Myo

tis

2

5

0

24.8

8

0.0

36

clu

ster

ed

1

6

0

18.4

7

0.1

86

ns

14

10k

m g

rid

Ves

per

tili

on

idae

2

23

0

56.9

1

0.2

34

ns

3

21

1

58.3

4

0.1

96

ns

50

Myo

tis

0

6

0

13.2

4

0.3

52

ns

1

5

0

17.4

7

0.1

33

ns

12

50k

m g

rid

Ves

per

tili

on

idae

3

20

2

61.9

3

0.1

20

ns

3

21

1

59.7

4

0.1

63

ns

50

Myo

tis

1

10

1

24.0

6

0.4

58

ns

0

11

1

22.8

1

0.5

31

ns

24

50k

m c

ircl

e V

esp

erti

lion

idae

0

17

1

34.5

5

0.5

37

ns

0

18

0

33.6

7

0.5

80

ns

36

Myo

tis

0

8

0

16.8

5

0.3

95

ns

0

8

0

16.6

9

0.4

06

ns

16

100k

m

circ

le

Ves

per

tili

on

idae

1

11

0

27.1

2

0.2

99

ns

1

11

0

28.4

9

0.2

40

ns

24

Myo

tis

1

6

1

19.6

0

0.2

39

ns

0

7

1

16.5

1

0.4

18

ns

16

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

124

Tab

le S

5:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on P

CS

anal

yse

s fo

r th

e C

hih

uah

uan

Des

ert

for

each

com

munit

y d

elim

itat

ion m

ethod.

Spec

ies

pools

use

d w

ere

“CH

ves

per

tili

onid

s” a

nd

“C

H M

yotis”

. M

PD

M

NT

D

Del

imit

atio

n

met

hod

Tax

on

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s

Clu

ster

ed

com

mun

itie

s

Ran

dom

com

mun

itie

s

Over

dis

per

sed

com

mun

itie

s

Tes

t

stat

isti

c

p-

val

ue

Res

ult

s d

f

5k

m b

uff

er

Ves

per

tili

on

idae

4

44

4

117

.91

0.1

66

ns

2

46

4

105

.68

0.4

36

ns

104

Myo

tis

4

17

0

61.1

6

0.0

28

clu

ster

ed

0

21

0

37.4

8

0.6

69

ns

42

10k

m b

uff

er

Ves

per

tili

on

idae

2

29

2

67.8

4

0.2

85

ns

2

29

0

67.6

4

0.2

91

ns

62

Myo

tis

2

10

1

40.9

9

0.0

31

clu

ster

ed

1

12

0

38.1

6

0.0

59

ns

26

10k

m g

rid

Ves

per

tili

on

idae

4

51

6

139

.20

0.1

37

ns

4

53

4

134

.05

0.2

15

ns

122

Myo

tis

6

11

0

63.6

3

0.0

02

clu

ster

ed

0

17

0

27.0

1

0.7

97

ns

34

50k

m g

rid

Ves

per

tili

on

idae

3

37

2

90.3

9

0.2

97

ns

1

39

2

83.7

8

0.4

86

ns

84

Myo

tis

4

16

1

60.9

7

0.0

29

clu

ster

ed

2

18

1

40.5

8

0.5

34

ns

42

50k

m c

ircl

e V

esp

erti

lion

idae

1

27

1

60.4

2

0.3

89

ns

1

25

3

61.4

5

0.3

53

ns

58

Myo

tis

3

14

1

49.3

1

0.0

69

ns

1

16

1

45.8

6

0.1

26

ns

36

100k

m c

ircl

e V

esp

erti

lion

idae

2

18

3

47.9

1

0.3

95

ns

0

21

2

38.7

5

0.7

67

ns

46

Myo

tis

2

13

2

46.4

1

0.0

76

ns

0

16

1

32.8

6

0.5

24

ns

34

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

125

Tab

le S

6:

Pea

rson p

rodu

ct-m

om

ent

corr

elat

ion c

oef

fici

ents

fo

r m

ean a

nnual

tem

per

ature

(B

IO1)

and

(a)

SE

S-M

PD

and (

b)

SE

S-M

NT

D. G

ray c

ells

indic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5.

a)

Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

ua

All

5k

m b

uff

er

0.618

0.575

0.276

0.176

0.301

10k

m b

uff

er

0.661

0.512

0.669

0.502

0.101

10k

m g

rid

0.584

0.372

0.471

0.062

0.108

50k

m g

rid

0.624

0.405

0.053

0.521

0.054

50k

m c

ircl

e 0.699

0.495

0.469

0.512

0.121

100k

m

circ

le

0.722

0.440

0.425

0.792

0.246

Ves

per

tili

on

idae

5k

m b

uff

er

0.408

0.591

0.096

0.029

0.185

10k

m b

uff

er

0.509

0.559

0.515

0.694

0.065

10k

m g

rid

0.350

0.268

0.401

0.164

0.051

50k

m g

rid

0.370

0.441

-0.074

0.616

0.133

50k

m c

ircl

e 0.446

0.437

0.287

0.518

-0.091

100k

m

circ

le

0.464

0.432

0.154

0.426

0.349

Myo

tis

5k

m b

uff

er

0.529

0.279

0.539

0.671

-0.151

10k

m b

uff

er

0.679

0.234

0.463

0.800

-0.228

10k

m g

rid

0.434

0.080

0.299

-0.092

-0.323

50k

m g

rid

0.578

0.091

0.524

0.731

-0.061

50k

m c

ircl

e 0.633

0.249

0.084

0.593

-0.147

100k

m

circ

le

0.711

0.274

0.408

0.798

0.121

b) Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

ua

All

5k

m b

uff

er

0.533

0.618

0.270

-0.050

0.250

10k

m b

uff

er

0.564

0.587

0.546

0.593

0.143

10k

m g

rid

0.471

0.384

0.432

0.116

0.052

50k

m g

rid

0.516

0.423

-0.034

0.552

-0.030

50k

m c

ircl

e 0.527

0.463

0.323

0.272

0.007

100k

m

circ

le

0.530

0.374

0.077

0.597

0.263

Ves

per

tili

on

idae

5k

m b

uff

er

0.369

0.611

-0.195

-0.111

0.128

10k

m b

uff

er

0.400

0.613

0.427

0.536

0.026

10k

m g

rid

0.330

0.357

0.360

0.162

0.043

50k

m g

rid

0.333

0.451

-0.066

0.400

-0.035

50k

m c

ircl

e 0.320

0.468

0.188

0.217

-0.048

100k

m

circ

le

0.272

0.461

-0.097

0.249

0.339

Myo

tis

5k

m b

uff

er

0.343

0.165

0.356

-0.100

-0.361

10k

m b

uff

er

0.494

0.199

-0.590

0.870

-0.021

10k

m g

rid

0.253

-0.046

0.264

-0.100

-0.115

50k

m g

rid

0.388

-0.012

0.382

0.511

-0.357

50k

m c

ircl

e 0.326

0.161

0.024

0.505

-0.218

100k

m

circ

le

0.460

0.295

-0.200

0.641

-0.085

126

Tab

le S

7:

Pea

rson p

rodu

ct-m

om

ent

corr

elat

ion c

oef

fici

ents

fo

r m

ean t

emp

erat

ure

sea

son

alit

y (

BIO

4)

and (

a) S

ES

-MP

D a

nd (

b)

SE

S-M

NT

D. G

ray

cell

s in

dic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5.

a)

Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

ua

All

5k

m b

uff

er

-0.412

0.072

0.277

-0.188

-0.214

10k

m b

uff

er

-0.488

0.006

0.498

-0.169

-0.343

10k

m g

rid

-0.372

0.120

0.414

-0.268

-0.073

50k

m g

rid

-0.413

0.047

0.315

-0.012

-0.183

50k

m c

ircl

e -0.462

-0.003

0.423

-0.040

-0.332

100k

m c

ircl

e -0.549

0.048

0.160

-0.168

-0.608

Ves

per

tili

on

idae

5k

m b

uff

er

-0.268

0.017

0.081

-0.384

-0.189

10k

m b

uff

er

-0.295

-0.092

0.477

-0.205

-0.127

10k

m g

rid

-0.223

0.069

0.544

-0.422

-0.106

50k

m g

rid

-0.178

0.028

0.088

-0.178

0.018

50k

m c

ircl

e -0.207

-0.081

0.357

-0.102

0.076

100k

m c

ircl

e -0.174

-0.004

0.126

-0.357

0.018

Myo

tis

5k

m b

uff

er

-0.269

0.296

0.591

-0.357

-0.253

10k

m b

uff

er

-0.543

0.019

-0.137

-0.278

-0.612

10k

m g

rid

-0.234

0.134

0.668

-0.563

-0.024

50k

m g

rid

-0.407

0.113

0.610

-0.549

-0.460

50k

m c

ircl

e -0.478

0.144

0.332

-0.405

-0.542

100k

m c

ircl

e -0.634

-0.124

-0.075

-0.620

-0.742

b) Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

ua

All

5k

m b

uff

er

-0.308

0.012

0.439

-0.320

-0.121

10k

m b

uff

er

-0.384

-0.071

0.510

-0.409

-0.135

10k

m g

rid

-0.271

0.057

0.462

-0.439

-0.049

50k

m g

rid

-0.293

0.018

0.409

-0.201

-0.021

50k

m c

ircl

e -0.241

-0.017

0.508

0.050

-0.139

100k

m c

ircl

e -0.348

-0.032

0.400

-0.238

-0.293

Ves

per

tili

on

idae

5k

m b

uff

er

-0.211

-0.016

0.531

-0.206

-0.178

10k

m b

uff

er

-0.188

-0.080

0.490

-0.233

-0.044

10k

m g

rid

-0.161

0.044

0.517

-0.287

-0.105

50k

m g

rid

-0.128

0.031

0.287

-0.110

0.083

50k

m c

ircl

e -0.093

-0.042

0.453

0.144

-0.011

100k

m c

ircl

e -0.021

-0.036

0.125

-0.191

0.163

Myo

tis

5k

m b

uff

er

-0.254

0.172

0.374

-0.629

-0.155

10k

m b

uff

er

-0.452

-0.114

0.168

-0.326

-0.568

10k

m g

rid

-0.158

0.106

0.617

-0.469

0.185

50k

m g

rid

-0.363

0.011

0.575

-0.553

-0.350

50k

m c

ircl

e -0.296

0.235

0.270

-0.338

-0.501

100k

m c

ircl

e -0.539

-0.101

-0.200

-0.611

-0.755

127

Tab

le S

8:

Pea

rson p

rodu

ct-m

om

ent

corr

elat

ion c

oef

fici

ents

fo

r an

nual

pre

cipit

atio

n (

BIO

12)

and (

a) S

ES

-MP

D a

nd (

b)

SE

S-M

NT

D. G

ray c

ells

indic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5.

a)

Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.086

-0.230

-0.211

-0.103

0.175

10k

m b

uff

er

0.054

-0.074

-0.588

-0.410

0.285

10k

m g

rid

0.020

-0.121

-0.323

-0.089

0.123

50k

m g

rid

0.096

-0.092

-0.091

-0.312

0.180

50k

m c

ircl

e -0.022

-0.140

-0.467

-0.422

0.292

100k

m

circ

le

0.191

-0.079

-0.130

-0.459

0.655

Ves

per

tili

on

idae

5k

m b

uff

er

0.043

-0.238

-0.211

0.106

0.174

10k

m b

uff

er

-0.040

-0.138

-0.542

-0.529

0.118

10k

m g

rid

0.010

-0.140

-0.338

0.156

0.135

50k

m g

rid

-0.056

-0.179

-0.019

-0.198

-0.002

50k

m c

ircl

e -0.148

-0.155

-0.364

-0.355

-0.172

100k

m

circ

le

-0.139

-0.204

-0.078

-0.218

0.017

Myo

tis

5k

m b

uff

er

-0.097

-0.213

-0.570

-0.661

0.408

10k

m b

uff

er

0.038

-0.218

0.305

-0.611

0.694

10k

m g

rid

-0.155

-0.057

-0.523

-0.009

0.029

50k

m g

rid

0.000

0.026

-0.633

-0.185

0.325

50k

m c

ircl

e 0.124

-0.192

0.005

-0.430

0.488

100k

m

circ

le

0.031

-0.355

0.181

-0.560

0.453

b) Tax

on

D

elim

itat

ion

met

hod

All

des

erts

G

reat

B

asin

M

oja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.007

-0.288

-0.311

0.065

0.131

10k

m b

uff

er

-0.057

-0.162

-0.531

-0.594

0.120

10k

m g

rid

-0.033

-0.154

-0.335

0.011

0.105

50k

m g

rid

-0.036

-0.170

-0.172

-0.456

0.076

50k

m c

ircl

e -0.146

-0.202

-0.429

-0.184

0.152

100k

m

circ

le

0.048

-0.090

-0.238

-0.343

0.358

Ves

per

tili

on

idae

5k

m b

uff

er

-0.017

-0.275

-0.288

0.208

0.146

10k

m b

uff

er

-0.122

-0.209

-0.483

-0.511

0.063

10k

m g

rid

-0.037

-0.167

-0.340

0.115

0.101

50k

m g

rid

-0.103

-0.206

-0.164

-0.217

-0.068

50k

m c

ircl

e -0.134

-0.226

-0.356

-0.111

-0.048

100k

m

circ

le

-0.168

-0.181

0.050

-0.227

-0.062

Myo

tis

5k

m b

uff

er

0.028

-0.237

-0.447

0.126

0.607

10k

m b

uff

er

0.092

-0.231

0.001

-0.757

0.728

10k

m g

rid

-0.060

-0.004

-0.526

-0.012

0.216

50k

m g

rid

0.087

0.059

-0.647

-0.077

0.412

50k

m c

ircl

e 0.144

-0.226

0.041

-0.256

0.533

100k

m

circ

le

0.075

-0.297

0.412

-0.586

0.456

128

Tab

le S

9:

Pea

rson p

rodu

ct-m

om

ent

corr

elat

ion c

oef

fici

ents

fo

r m

ean p

reci

pit

atio

n s

easo

nal

ity (

BIO

15)

and (

a) S

ES

-MP

D a

nd (

b)

SE

S-M

NT

D. G

ray

cell

s in

dic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5.

a)

Tax

on

D

elim

itat

ion

met

hod

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.382

0.034

-0.180

-0.061

-0.138

10k

m b

uff

er

0.453

0.023

-0.279

0.134

-0.050

10k

m g

rid

0.404

0.018

-0.094

0.182

-0.241

50k

m g

rid

0.421

-0.041

-0.270

0.034

0.026

50k

m c

ircl

e 0.454

0.094

-0.254

-0.039

-0.315

100k

m

circ

le

0.552

0.067

0.062

0.336

0.045

Ves

per

tili

on

idae

5k

m b

uff

er

0.303

0.161

0.041

0.330

0.029

10k

m b

uff

er

0.379

0.135

-0.283

0.371

0.169

10k

m g

rid

0.246

0.045

-0.170

0.425

-0.216

50k

m g

rid

0.239

0.026

-0.041

0.355

0.069

50k

m c

ircl

e 0.302

0.111

-0.203

0.280

0.095

100k

m

circ

le

0.300

0.076

0.120

0.368

-0.099

Myo

tis

5k

m b

uff

er

0.419

0.201

-0.403

0.414

-0.285

10k

m b

uff

er

0.626

0.385

0.401

0.335

-0.001

10k

m g

rid

0.373

0.300

-0.507

0.692

-0.289

50k

m g

rid

0.473

0.146

-0.385

0.637

0.186

50k

m c

ircl

e 0.562

0.142

-0.350

0.587

0.038

100k

m

circ

le

0.635

0.318

0.330

0.550

-0.069

b) Tax

on

D

elim

itat

ion

met

hod

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.289

0.109

-0.296

0.043

-0.088

10k

m b

uff

er

0.385

0.118

-0.313

0.332

-0.020

10k

m g

rid

0.304

0.109

-0.181

0.281

-0.152

50k

m g

rid

0.289

-0.043

-0.314

0.146

0.006

50k

m c

ircl

e 0.241

0.077

-0.313

-0.100

-0.305

100k

m

circ

le

0.355

0.080

-0.222

0.339

-0.013

Ves

per

tili

on

idae

5k

m b

uff

er

0.221

0.213

-0.499

0.091

0.030

10k

m b

uff

er

0.253

0.170

-0.366

0.301

0.193

10k

m g

rid

0.176

0.137

-0.241

0.223

-0.163

50k

m g

rid

0.139

-0.032

-0.188

0.165

0.019

50k

m c

ircl

e 0.153

0.111

-0.264

-0.022

0.089

100k

m

circ

le

0.065

0.116

0.070

0.108

-0.228

Myo

tis

5k

m b

uff

er

0.358

0.159

-0.356

0.812

0.018

10k

m b

uff

er

0.473

0.363

-0.108

0.396

0.078

10k

m g

rid

0.278

0.232

-0.510

0.580

-0.129

50k

m g

rid

0.394

0.015

-0.434

0.610

0.233

50k

m c

ircl

e 0.314

-0.119

-0.379

0.494

0.215

100k

m

circ

le

0.511

0.196

0.343

0.463

0.216

129

APPENDIX V

SPECIMENS EXAMINED IN THE MORPHOLOGICAL STUDY

Specimens examined in the morphological study. LSUMNS= Louisiana State University

Museum of Natural Science; MSB= Museum of Southwestern Biology; RDS= specimens in Dr.

Richard Stevens collection; KU= University of Kansas; Burke= Burke Museum; LACM= Los

Angeles County Museum of Natural History; PSUMVB= Portland State University Museum of

Vertebrate Biology.

Museum

Specimen

number Genus Species Subspecies sex

State/

Province

County/

District

LSUMNS L10426 Antrozous pallidus f Arizona Cochise

LSUMNS LEP186 Antrozous pallidus f Oregon Lake

MSB M11187 Antrozous pallidus f

New

Mexico Bernalillo

MSB M11189 Antrozous pallidus f

New

Mexico Bernalillo

MSB M18808 Antrozous pallidus f Mexico Sonora

MSB M18809 Antrozous pallidus f Mexico Sonora

MSB M70870 Antrozous pallidus f Mexico

Baja

California

Sur

MSB M70873 Antrozous pallidus f Mexico

Baja

California

Sur

RDS RDS8093 Antrozous pallidus f Arizona Mojave

RDS RDS8098 Antrozous pallidus f Arizona Mojave

LSUMNS L10427 Antrozous pallidus m Arizona Yuma

LSUMNS LEP129 Antrozous pallidus m Oregon Lake

MSB M116546 Antrozous pallidus pallidus m Utah Garfield

MSB M120013 Antrozous pallidus pallidus m Utah Garfield

MSB M12946 Antrozous pallidus pallidus m

New

Mexico Bernalillo

MSB M12947 Antrozous pallidus pallidus m

New

Mexico Bernalillo

MSB M18323 Antrozous pallidus m Mexico Sonora

MSB M42580 Antrozous pallidus pallidus m Mexico Sonora

MSB M43110 Antrozous pallidus pacificus m Mexico

Baja

California

MSB M43839 Antrozous pallidus pacificus m Mexico

Baja

California

MSB M18328 Artibeus hirsutus f Mexico Sonora










130

Museum

Specimen


State/

Province

County/

District

MSB M18381 Artibeus hirsutus m Mexico Sonora










LSUMNS L3885 Choernycteris mexicana m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M160648 Choeronycteris mexicana f Arizona Cochise




MSB M1741 Choeronycteris mexicana f

New

Mexico Hidalgo


New

Mexico Hidalgo

MSB M18306 Choeronycteris mexicana f Mexico Sonora

MSB M18324 Choeronycteris mexicana f Mexico Sonora


New

Mexico Hidalgo


New

Mexico Hidalgo

KU K102082 Choeronycteris mexicana m Arizona Cochise

MSB M160649 Choeronycteris mexicana m Arizona Cochise




MSB M160703 Choeronycteris mexicana m Arizona Santa Cruz

Burke B62750 Corynorhinus mexicanus f Mexico





KU K143770 Corynorhinus mexicanus f Mexico Mexico




KU K29906 Corynorhinus mexicanus f Mexico Veracruz

131

Museum

Specimen


State/

Province

County/

District

Burke B62752 Corynorhinus mexicanus m Mexico



KU K143773 Corynorhinus mexicanus m Mexico Mexico

KU K29888 Corynorhinus mexicanus m Mexico Veracruz





KU K73591 Corynorhinus mexicanus m Mexico Chihuahua

KU K7131 Corynorhinus townsendii pallescens f Idaho Bannock

LSUMNS L10130 Corynorhinus townsendii f Arizona Cochise

LSUMNS L10420 Corynorhinus townsendii f Arizona Pima

LSUMNS L11197 Corynorhinus townsendii f Colorado Conejos

LSUMNS L1121 Corynorhinus townsendii f California

San

Bernadino

LSUMNS L1199 Corynorhinus townsendii f California

San

Bernadino

LSUMNS L1875 Corynorhinus townsendii pallescens f California Riverside

LSUMNS L20915 Corynorhinus townsendii f Washington Spokane

LSUMNS L20916 Corynorhinus townsendii f Washington Spokane

MSB M11573 Corynorhinus townsendii f

New

Mexico Bernalillo

LSUMNS L11195 Corynorhinus townsendii m Colorado Conejos

LSUMNS L11196 Corynorhinus townsendii m Colorado Conejos

LSUMNS L1876 Corynorhinus townsendii pallescens m California Riverside

LSUMNS LEP114 Corynorhinus townsendii m ? ?

LSUMNS LEP124 Corynorhinus townsendii m Oregon Lake

MSB M114799 Corynorhinus townsendii m Utah Garfield


MSB M11571 Corynorhinus townsendii m

New

Mexico Bernalillo

MSB M11572 Corynorhinus townsendii m

New

Mexico Bernalillo


LSUMNS L11051 Desmodus rotundus f Mexico Colima

LSUMNS L3942 Desmodus rotundus f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

132

Museum

Specimen


State/

Province

County/

District

LSUMNS L8399 Desmodus rotundus f Mexico Tabasco




LSUMNS L2828 Desmodus rotundus m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L8398 Desmodus rotundus m Mexico Tabasco





LSUMNS L3988 Diphylla ecaudata centralis f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L2829 Diphylla ecaudata centralis m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

133

Museum

Specimen


State/

Province

County/

District


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

Burke B13724 Eptesicus fuscus f Washington Grant

Burke B62196 Eptesicus fuscus f Oregon Deschutes

LSUMNS L10128 Eptesicus fuscus f Arizona Pima

LSUMNS L10419 Eptesicus fuscus f Arizona Pima

LSUMNS L11932 Eptesicus fuscus f Mexico Oaxaca

LSUMNS L2780 Eptesicus fuscus f Mexico

San Luis

Potosi

LSUMNS LEP024 Eptesicus fuscus f Washington Klickitat

LSUMNS LEP025 Eptesicus fuscus f Washington Klickitat

LSUMNS LEP149 Eptesicus fuscus f Utah Juab

LSUMNS LEP159 Eptesicus fuscus f Utah Juab

Burke B33268 Eptesicus fuscus m Washington Douglas

Burke B38245 Eptesicus fuscus m Arizona Coconino

Burke B62167 Eptesicus fuscus m California Napa

LSUMNS L10129 Eptesicus fuscus m Arizona Pima

LSUMNS L22024 Eptesicus fuscus m

New

Mexico Socorro


New

Mexico Socorro


New

Mexico Socorro

LSUMNS L4039 Eptesicus fuscus m Mexico

San Luis

Potosi

LSUMNS L4932 Eptesicus fuscus m Mexico

San Luis

Potosi

LSUMNS LEP150 Eptesicus fuscus m Utah Juab

KU K119275 Euderma maculatum f Texas Brewster

MSB M107557 Euderma maculatum f Colorado Moffat

MSB M114512 Euderma maculatum f Wyoming Big Horn

MSB M114513 Euderma maculatum f Wyoming Big Horn

MSB M17285 Euderma maculatum f

New

Mexico Catron

134

Museum

Specimen


State/

Province

County/

District


New

Mexico Sandoval


New

Mexico Sandoval


New

Mexico Socorro


New

Mexico Socorro


New

Mexico Catron

LSUMNS L17652 Euderma maculatum m Texas Brewster

MSB M112056 Euderma maculatum m Colorado Moffat




MSB M116740 Euderma maculatum m Utah Wayne

MSB M121373 Euderma maculatum m Utah San Juan

MSB M25000 Euderma maculatum m

New

Mexico Socorro


New

Mexico Sandoval


New

Mexico Rio Arriba

KU K#6 Eumops perotis californicus f Texas Brewster

KU K#9 Eumops perotis californicus f Texas Brewster

KU K150208 Eumops perotis californicus f California Los Angeles

KU K160270 Eumops perotis californicus f California Los Angeles

LSUMNS L10468 Eumops perotis californicus f California Kern

LSUMNS L1870 Eumops perotis californicus f California Los Angeles

LACM LA9326 Eumops perotis californicus f California Los Angeles

MSB M160472 Eumops perotis californicus f California Los Angeles

MSB M160473 Eumops perotis californicus f California Los Angeles

MSB M160477 Eumops perotis californicus f California

San

Bernadino

KU K9420 Eumops perotis californicus m California

San

Bernadino

LSUMNS L1869 Eumops perotis californicus m California Los Angeles

LACM LA13075 Eumops perotis californicus m Arizona Pima

LACM LA34328 Eumops perotis californicus m Mexico Zacatecas

LACM LA37576 Eumops perotis californicus m California Los Angeles

135

Museum

Specimen


State/

Province

County/

District



MSB M160470 Eumops perotis californicus m Arizona Pima

MSB M160471 Eumops perotis californicus m California Los Angeles

MSB M4300 Eumops perotis m Arizona Pima

KU K100404 Eumops underwoodi underwoodi f Mexico Jalisco

KU K68795 Eumops underwoodi underwoodi f Mexico Oaxaca

KU K92952 Eumops underwoodi underwoodi f Mexico Jalisco

LSUMNS L10428 Eumops underwoodi f Arizona Pima

LSUMNS L11054 Eumops underwoodi f Mexico Colima

LSUMNS L8431 Eumops underwoodi f Mexico Tabasco

LACM LA11603 Eumops underwoodi sonoriensis f Mexico Chihuahua

LACM LA13199 Eumops underwoodi sonoriensis f Mexico Sonora

LACM LA13200 Eumops underwoodi sonoriensis f Mexico Sonora

LACM LA29162 Eumops underwoodi underwoodi f Mexico Colima

KU K#1998 Eumops underwoodi m Arizona Pima

KU K59092 Eumops underwoodi sonoriensis m Arizona Pima

KU K92955 Eumops underwoodi underwoodi m Mexico Jalisco

LSUMNS L8428 Eumops underwoodi m Mexico Tabasco

LSUMNS L8429 Eumops underwoodi m Mexico Tabasco

LACM LA29163 Eumops underwoodi underwoodi m Mexico Colima



MSB M160478 Eumops underwoodi sonoriensis m Arizona Pima

MSB M160479 Eumops underwoodi sonoriensis m Arizona Pima

LSUMNS L3852 Glossophaga soricina f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L8161 Glossophaga soricina f Mexico Tabasco

136

Museum

Specimen


State/

Province

County/

District

LSUMNS L3855 Glossophaga soricina m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L8158 Glossophaga soricina m Mexico Tabasco



LSUMNS L11427 Idionycteris phyllotis f

New

Mexico Catron

MSB M116207 Idionycteris phyllotis f Utah San Juan


MSB M13014 Idionycteris phyllotis f

New

Mexico Catron


New

Mexico Catron


New

Mexico Catron

MSB M161533 Idionycteris phyllotis f Arizona Coconino




KU K73594 Idionycteris phyllotis m Mexico Chihuahua

LSUMNS L22032 Idionycteris phyllotis m

New

Mexico Socorro

MSB M116206 Idionycteris phyllotis m Utah San Juan

MSB M120921 Idionycteris phyllotis m Utah Kane

MSB M13013 Idionycteris phyllotis m

New

Mexico Catron

MSB M161231 Idionycteris phyllotis m Arizona Gila




New

Mexico Catron


New

Mexico Catron

Burke B63089 Lasionycteris noctivagans f Washington Walla Walla

Burke B76226 Lasionycteris noctivagans f Oregon Douglas

137

Museum

Specimen


State/

Province

County/

District

Burke B77915 Lasionycteris noctivagans f Washington Walla Walla

Burke B78216 Lasionycteris noctivagans f Washington Columbia

MSB M114627 Lasionycteris noctivagans f Utah Wayne

MSB M13025 Lasionycteris noctivagans f

New

Mexico Catron

MSB M13026 Lasionycteris noctivagans f

New

Mexico Catron

MSB M161545 Lasionycteris noctivagans f Arizona Apache

MSB M161547 Lasionycteris noctivagans f Arizona Cochise

MSB M37376 Lasionycteris noctivagans f California Mariposa

Burke B35496 Lasionycteris noctivagans m Oregon Jackson

Burke B39182 Lasionycteris noctivagans m Washington Ferry

Burke B78230 Lasionycteris noctivagans m Washington Columbia

Burke B78261 Lasionycteris noctivagans m Washington Columbia

MSB M109190 Lasionycteris noctivagans m Utah Uintah

MSB M161546 Lasionycteris noctivagans m Arizona Cochise

MSB M161548 Lasionycteris noctivagans m Arizona Cochise

MSB M40651 Lasionycteris noctivagans m California El Dorado

MSB M9583 Lasionycteris noctivagans m

New

Mexico Catron

MSB M9584 Lasionycteris noctivagans m

New

Mexico Catron

MSB M10516 Lasiurus blossevillii f

New

Mexico Catron

MSB M161560 Lasiurus blossevillii f Arizona Cochise

MSB M161563 Lasiurus blossevillii f Arizona Graham

MSB M16855 Lasiurus blossevillii f Mexico Nayarit


New

Mexico Hidalgo

MSB M37377 Lasiurus blossevillii f California Mariposa


New

Mexico Hidalgo


New

Mexico Catron


New

Mexico Catron


New

Mexico Catron

KU K107491 Lasiurus blossevillii teliotis m Mexico Jalisco




MSB M161561 Lasiurus blossevillii m Arizona Coconino

138

Museum

Specimen


State/

Province

County/

District

MSB M161562 Lasiurus blossevillii m Arizona Gila

MSB M18588 Lasiurus blossevillii m Mexico Sonora

MSB M42502 Lasiurus blossevillii m

New

Mexico Hidalgo

MSB M54937 Lasiurus blossevillii m Mexico Sonora

MSB M68581 Lasiurus blossevillii m

New

Mexico Eddy

KU K48304 Lasiurus borealis borealis f Mexico Coahuila

LSUMNS L10557 Lasiurus borealis f Louisiana

East Baton

Rouge


East Baton

Rouge

LSUMNS L11737 Lasiurus borealis f Texas Franklin

LSUMNS L11739 Lasiurus borealis f Texas Franklin


East Baton

Rouge


East Baton

Rouge


East Baton

Rouge

LSUMNS L8557 Lasiurus borealis borealis f Louisiana

East Baton

Rouge

LSUMNS L8731 Lasiurus borealis borealis f Louisiana

East Baton

Rouge

LSUMNS L11734 Lasiurus borealis m Texas Franklin




LSUMNS L1706 Lasiurus borealis borealis m Louisiana

East Baton

Rouge

LSUMNS L25088 Lasiurus borealis m Louisiana

Gulf of

Mexico

LSUMNS L25408 Lasiurus borealis m Louisiana Grant


East Baton

Rouge


East Baton

Rouge

LSUMNS L6784 Lasiurus borealis m Louisiana

East

Feliciana

Burke B48272 Lasiurus cinereus f Washington Walla Walla

Burke B9531 Lasiurus cinereus f Washington King

LSUMNS L20919 Lasiurus cinereus f California Contra Costa

LSUMNS L29131 Lasiurus cinereus f California

San

Bernadino

139

Museum

Specimen


State/

Province

County/

District

MSB M12816 Lasiurus cinereus cinereus f

New

Mexico Bernalillo


New

Mexico Bernalillo


New

Mexico Bernalillo


New

Mexico Bernalillo

MSB M18305 Lasiurus cinereus f Mexico Sonora

MSB M19031 Lasiurus cinereus f Mexico Sonora

Burke B32565 Lasiurus cinereus m California Yolo

Burke B39474 Lasiurus cinereus m Washington Yakima

Burke B9219 Lasiurus cinereus m Washington Snohomish

LSUMNS L10511 Lasiurus cinereus m

New

Mexico Rio Arriba


New

Mexico Socorro


New

Mexico Socorro

LSUMNS L25095 Lasiurus cinereus m Mexico Michoacan

LSUMNS L25098 Lasiurus cinereus m Mexico Michoacan

LSUMNS L4043 Lasiurus cinereus m Mexico

San Luis

Potosi

LSUMNS L4958 Lasiurus cinereus m Mexico

San Luis

Potosi

KU K55318 Lasiurus ega f Mexico Tamaulipas



LSUMNS L11929 Lasiurus ega f Mexico Chiapas

LSUMNS L12986 Lasiurus ega f Costa Rica San Jose

LSUMNS L12987 Lasiurus ega f Costa Rica San Jose

LSUMNS L4044 Lasiurus ega f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LACM LA18680 Lasiurus ega f Mexico Chiapas

KU K100399 Lasiurus ega m Mexico Nuevo Leon

KU K55316 Lasiurus ega m Mexico Tamaulipas



LSUMNS L4059 Lasiurus ega m Mexico

San Luis

Potosi

140

Museum

Specimen


State/

Province

County/

District


San Luis

Potosi


San Luis

Potosi

LACM LA73717 Lasiurus ega m Mexico Guerrero



Burke B62345 Lasiurus intermedius f Florida Hillsborough

KU K67549 Lasiurus intermedius intermedius f Mexico Veracruz

KU K67550 Lasiurus intermedius intermedius f Mexico Veracruz

LSUMNS L11053 Lasiurus intermedius f Mexico Colima

LACM LA12034 Lasiurus intermedius f Mexico Colima

LACM LA12530 Lasiurus intermedius f Mexico Nayarit




LACM LA8818 Lasiurus intermedius f Texas Cameron

KU K100400 Lasiurus intermedius intermedius m Mexico Jalisco

KU K55317 Lasiurus intermedius intermedius m Mexico Tamaulipas






LSUMNS L11928 Lasiurus intermedius m Mexico Chiapas

LSUMNS L25096 Lasiurus intermedius m Mexico Michoacan

LSUMNS L25097 Lasiurus intermedius m Mexico Michoacan

LSUMNS L11613 Lasiurus seminolus f Louisiana Natchitoches

LSUMNS L11614 Lasiurus seminolus f Louisiana Sabine

LSUMNS L11618 Lasiurus seminolus f Louisiana Ascension

LSUMNS L25416 Lasiurus seminolus f Louisiana Grant

LSUMNS L3680 Lasiurus seminolus f Louisiana Rapides

LSUMNS L6158 Lasiurus seminolus f Louisiana

East Baton

Rouge

LSUMNS L747 Lasiurus seminolus f Louisiana

East Baton

Rouge

LACM LA5997 Lasiurus seminolus f Texas Harris

LACM LA8898 Lasiurus seminolus f Louisiana Natchitoches

LACM LA9406 Lasiurus seminolus f Florida Alachua

LSUMNS L25409 Lasiurus seminolus m Louisiana Grant

LSUMNS L26733 Lasiurus seminolus m Louisiana

East Baton

Rouge

LSUMNS L30054 Lasiurus seminolus m Louisiana Lafourche

LSUMNS L3308 Lasiurus seminolus m Louisiana Grant

LSUMNS L3309 Lasiurus seminolus m Louisiana Rapides

141

Museum

Specimen


State/

Province

County/

District

LSUMNS L3310 Lasiurus seminolus m Louisiana Washington


East Baton

Rouge

LSUMNS L9232 Lasiurus seminolus m Louisiana Washington


East Baton

Rouge

LACM LA8897 Lasiurus seminolus m Texas Harris

KU K94336 Lasiurus xanthinus f Mexico

Baja

California

del Sur


Baja

California

del Sur


Baja

California

del Sur


Baja

California

del Sur


Baja

California

del Sur


Baja

California

del Sur

MSB M14505 Lasiurus xanthinus f

New

Mexico Hidalgo

MSB M161590 Lasiurus xanthinus f Arizona Maricopa

MSB M26861 Lasiurus xanthinus f Mexico Sonora

MSB M45881 Lasiurus xanthinus f

New

Mexico Hidalgo

MSB M16856 Lasiurus xanthinus m Mexico Nayarit

MSB M18302 Lasiurus xanthinus m Mexico Sonora



MSB M25038 Lasiurus xanthinus m

New

Mexico Hidalgo


New

Mexico Hidalgo

MSB M42840 Lasiurus xanthinus m Mexico

Baja

California



New

Mexico Hidalgo


New

Mexico Hidalgo

142

Museum

Specimen


State/

Province

County/

District

KU K33068 Leptonycteris nivalis nivalis f Mexico Coahuila









MSB M28913 Leptonycteris nivalis f Texas Brewster

KU K98370 Leptonycteris nivalis m Mexico Nuevo Leon










MSB M160734 Leptonycteris yerbabuenae f Arizona Cochise



MSB M25048 Leptonycteris yerbabuenae f

New

Mexico Hidalgo


New

Mexico Hidalgo


New

Mexico Hidalgo

MSB M29521 Leptonycteris yerbabuenae f Mexico Sonora




MSB M160737 Leptonycteris yerbabuenae m Arizona Cochise





143

Museum

Specimen


State/

Province

County/

District


MSB M25047 Leptonycteris yerbabuenae m

New

Mexico Hidalgo

MSB M31559 Leptonycteris yerbabuenae m Mexico Sonora

MSB M31563 Leptonycteris yerbabuenae m Mexico Sonora

MSB M43836 Leptonycteris yerbabuenae m Mexico

Baja

California

LSUMNS L1197 Macrotus californicus f California Riverside

LSUMNS L1871 Macrotus californicus f California Riverside

MSB M160899 Macrotus californicus f Arizona Pinal

MSB M160901 Macrotus californicus f Arizona Pinal

MSB M18346 Macrotus californicus f Mexico Sonora




MSB M38744 Macrotus californicus f Arizona Pima

MSB M38748 Macrotus californicus f Arizona Pima

LSUMNS L1873 Macrotus californicus m California Riverside

MSB M103122 Macrotus californicus m Arizona Pima

MSB M18589 Macrotus californicus m Mexico Sonora








KU K103399 Macrotus waterhousii bulleri f Mexico Jalisco



KU K29412 Macrotus waterhousii mexicanus f Mexico Oaxaca

KU K29415 Macrotus waterhousii mexicanus f Mexico Oaxaca

KU K85611 Macrotus waterhousii bulleri f Mexico Sinaloa


LSUMNS L11010 Macrotus waterhousii mexicanus f Mexico Colima

LSUMNS L11011 Macrotus waterhousii mexicanus f Mexico Colima

MSB M27549 Macrotus waterhousii mexicanus f Mexico Oaxaca

KU K29414 Macrotus waterhousii mexicanus m Mexico Oaxaca




KU K67351 Macrotus waterhousii bulleri m Mexico Sinaloa

LSUMNS L11008 Macrotus waterhousii mexicanus m Mexico Colima





KU K10733 Mormoops megalophylla megalophylla f Mexico Sonora

144

Museum

Specimen


State/

Province

County/

District

KU K10734 Mormoops megalophylla megalophylla f Mexico Sonora

KU K142865 Mormoops megalophylla megalophylla f Texas Uvalde

KU K142866 Mormoops megalophylla megalophylla f Texas Uvalde

KU K85590 Mormoops megalophylla megalophylla f Mexico Sinaloa

KU K85591 Mormoops megalophylla megalophylla f Mexico Sinaloa

LSUMNS L11002 Mormoops megalophylla f Mexico Colima



MSB M32645 Mormoops megalophylla f Mexico Guerrero

KU K85608 Mormoops megalophylla megalophylla m Mexico Sinaloa



LSUMNS L11005 Mormoops megalophylla m Mexico Colima

LSUMNS L11006 Mormoops megalophylla m Mexico Colima

LSUMNS L11962 Mormoops megalophylla m Mexico Yucatan

LSUMNS L4828 Mormoops megalophylla m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M70876 Mormoops megalophylla m Mexico

Baja

California

LSUMNS L10421 Myotis auriculus apache f Arizona Pima



MSB M11160 Myotis auriculus f

New

Mexico Bernalillo


New

Mexico Bernalillo

MSB M122085 Myotis auriculus apache f

New

Mexico Torrance


New

Mexico Socorro


New

Mexico Sierra

MSB M45887 Myotis auriculus apache f

New

Mexico Hidalgo


New

Mexico Bernalillo

Burke B62578 Myotis auriculus m

New

Mexico Socorro

LSUMNS L10424 Myotis auriculus apache m Arizona Pima

MSB M10856 Myotis auriculus m

New

Mexico Bernalillo


New

Mexico Bernalillo

145

Museum

Specimen


State/

Province

County/

District


New

Mexico Socorro


New

Mexico Socorro

MSB M23676 Myotis auriculus apache m

New

Mexico Cibola


New

Mexico Sandoval


New

Mexico Sandoval

MSB M45882 Myotis auriculus apache m

New

Mexico Hidalgo

LSUMNS L1201 Myotis californicus pallidus f California Inyo

LSUMNS L1883 Myotis californicus calfornicus f California Kern

LSUMNS L4012 Myotis californicus mexicanus f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M108579 Myotis californicus stephensi f Arizona Coconino

MSB M108580 Myotis californicus stephensi f Arizona Coconino

MSB M123025 Myotis californicus stephensi f Utah Garfield

MSB M42610 Myotis californicus californicus f Mexico Sonora

MSB M42613 Myotis californicus californicus f Mexico Sonora

LSUMNS L4014 Myotis californicus mexicanus m Mexico

San Luis

Potosi

LSUMNS L4027 Myotis californicus mexicanus m Mexico

San Luis

Potosi

MSB M108003 Myotis californicus stephensi m Arizona Coconino

MSB M108005 Myotis californicus stephensi m Arizona Coconino

MSB M122827 Myotis californicus stephensi m Utah Garfield



MSB M42612 Myotis californicus californicus m Mexico Sonora

MSB M83887 Myotis californicus m Mexico Sonora

MSB M83889 Myotis californicus m Mexico Sonora

LSUMNS LEP156 Myotis ciliolabrum f Utah Juab

LSUMNS LEP162 Myotis ciliolabrum f Utah Juab

MSB M103545 Myotis ciliolabrum f Montana Big Horn

MSB M114480 Myotis ciliolabrum ciliolabrum f Wyoming Fremont

MSB M119942 Myotis ciliolabrum ciliolabrum f Montana Carbon

MSB M122266 Myotis ciliolabrum ciliolabrum f Wyoming Weston

MSB M24970 Myotis ciliolabrum f Washington Douglas

MSB M32049 Myotis ciliolabrum f

New

Mexico Hidalgo

146

Museum

Specimen


State/

Province

County/

District


New

Mexico Hidalgo


New

Mexico Hidalgo

LSUMNS

20JUL09-

02-AHH Myotis ciliolabrum m Washington Douglas

LSUMNS LEP151 Myotis ciliolabrum m Utah Juab

LSUMNS LEP157 Myotis ciliolabrum m Utah Juab

LSUMNS LEP171 Myotis ciliolabrum m Oregon Lake

LSUMNS LEP173 Myotis ciliolabrum m Oregon Lake

MSB M114514 Myotis ciliolabrum ciliolabrum m Wyoming Big Horn




MSB M119941 Myotis ciliolabrum ciliolabrum m Montana Carbon

Burke B33270 Myotis evotis f Washington Douglas

Burke B39176 Myotis evotis f Washington Douglas

Burke B60938 Myotis evotis f Washington Pierce

Burke B62477 Myotis evotis f California Napa

Burke B78233 Myotis evotis f Washington Columbia

MSB M135307 Myotis evotis pacificus f

New

Mexico San Juan

MSB M18883 Myotis evotis f

New

Mexico Taos

MSB M40673 Myotis evotis pacificus f California Humboldt

PSUMVB P3052 Myotis evotis f Oregon Crook

PSUMVB P709 Myotis evotis f Oregon

Burke B62476 Myotis evotis m California Napa

Burke B62577 Myotis evotis m California San Diego

Burke B76162 Myotis evotis m Oregon Josephine

Burke B78234 Myotis evotis m Washington Columbia

MSB M107928 Myotis evotis evotis m Colorado Garfield

MSB M11634 Myotis evotis m

New

Mexico Catron

MSB M122462 Myotis evotis m Utah Washington

MSB M53786 Myotis evotis evotis m

New

Mexico Socorro

PSUMVB P2165 Myotis evotis m Oregon Harney

PSUMVB P3055 Myotis evotis m Oregon Wasco

LSUMNS L11052 Myotis fortidens f Mexico Colima

LSUMNS L8409 Myotis fortidens f Mexico Tabasco


147

Museum

Specimen


State/

Province

County/

District








LSUMNS L8420 Myotis fortidens m Mexico Tabasco

MSB M13134 Myotis fortidens m Mexico Nayarit

MSB M18292 Myotis fortidens m Mexico Sonora





MSB M27559 Myotis fortidens m Mexico Oaxaca

MSB M54941 Myotis fortidens sonoriensis m Mexico Sonora


LSUMNS LEP142 Myotis lucifugus alascensis f Oregon Lake







LACM LA9933 Myotis lucifugus alascensis f Canada

British

Columbia


British

Columbia


British

Columbia

LSUMNS

15JUL09-

01-LEP Myotis lucifugus alascensis m Washington Whatcom

LSUMNS LEP126 Myotis lucifugus alascensis m Oregon Lake









LSUMNS L11368 Myotis lucifugus carissima f Wyoming Teton

LSUMNS L11369 Myotis lucifugus carissima f Wyoming Teton

MSB M104453 Myotis lucifugus carissima f Colorado Montezuma


148

Museum

Specimen


State/

Province

County/

District



MSB M46654 Myotis lucifugus carissima f Wyoming

Yellowstone

NP


Yellowstone

NP


Yellowstone

NP


Yellowstone

NP


Yellowstone

NP


Yellowstone

NP

MSB M102854 Myotis lucifugus carissima m Colorado Rio Blanco



MSB M104456 Myotis lucifugus carissima m Colorado Montezuma

MSB M104457 Myotis lucifugus carissima m Colorado Montezuma

MSB M114499 Myotis lucifugus carissima m Wyoming Carbon



MSB M115341 Myotis lucifugus carissima m Colorado Moffat

MSB M115342 Myotis lucifugus carissima m Colorado Moffat

Burke B79220 Myotis lucifugus relictus f Washington Walla Walla

LSUMNS LEP117 Myotis lucifugus relictus f Washington Snohomish

LSUMNS LEP134 Myotis lucifugus relictus f Oregon Lake





MSB M40674 Myotis lucifugus relictus f Washington King

MSB M46574 Myotis lucifugus relictus f Oregon Deschutes

MSB M46633 Myotis lucifugus relictus f Washington Grant

LSUMNS LEP131 Myotis lucifugus relictus m Oregon Lake






149

Museum

Specimen


State/

Province

County/

District





LSUMNS LEP027 Myotis melanorhinus melanorhinus f Texas Jeff Davis

LSUMNS LEP029 Myotis melanorhinus melanorhinus f Texas Jeff Davis

MSB M120242 Myotis melanorhinus melanorhinus f

New

Mexico Dona Ana


New

Mexico Dona Ana


New

Mexico Los Alamos


New

Mexico Los Alamos


New

Mexico Los Alamos


New

Mexico Lincoln

RDS RDS8145 Myotis melanorhinus melanorhinus f Texas Jeff Davis

RDS RDS8148 Myotis melanorhinus melanorhinus f Texas Jeff Davis

LSUMNS LEP028 Myotis melanorhinus melanorhinus m Texas Jeff Davis

MSB M108804 Myotis melanorhinus melanorhinus m Colorado Moffat

MSB M110949 Myotis melanorhinus melanorhinus m Colorado Moffat

MSB M116745 Myotis melanorhinus melanorhinus m Utah Wayne

MSB M116746 Myotis melanorhinus melanorhinus m Utah Wayne

MSB M45900 Myotis melanorhinus melanorhinus m

New

Mexico Hidalgo


New

Mexico Hidalgo


New

Mexico Hidalgo

RDS MS024 Myotis melanorhinus melanorhinus m Texas Jeff Davis

RDS RDS8135 Myotis melanorhinus melanorhinus m Texas Jeff Davis

MSB M47322 Myotis milleri f

Baja

California

LACM LA91061 Myotis milleri m Mexico

Baja

California

del Norte

MSB M43021 Myotis milleri m Mexico

Baja

California


Baja

California

MSB M47321 Myotis milleri m

Baja

California


Baja

California

Norte

KU K23457 Myotis nigricans nigricans f Mexico Veracruz

150

Museum

Specimen


State/

Province

County/

District










KU K17840 Myotis nigricans nigricans m Mexico Veracruz








KU K58844 Myotis nigricans nigricans m Mexico Tamaulipas

KU K58847 Myotis nigricans nigricans m Mexico Tamaulipas

LSUMNS L10508 Myotis occultus f

New

Mexico Socorro

LSUMNS L10509 Myotis occultus f

New

Mexico Socorro

MSB M121943 Myotis occultus f

New

Mexico Socorro


New

Mexico Catron


New

Mexico Cibola

MSB M122030 Myotis occultus f Colorado Las Animas


New

Mexico Catron


New

Mexico Otero

MSB M27750 Myotis occultus f Mexico Chihuahua


New

Mexico Socorro

MSB M121949 Myotis occultus m

New

Mexico Socorro


New

Mexico Catron


New

Mexico Cibola

MSB M122031 Myotis occultus m Colorado Las Animas


New

Mexico Socorro

151

Museum

Specimen


State/

Province

County/

District


New

Mexico Catron


New

Mexico Otero

MSB M161792 Myotis occultus m Arizona Greenlee


New

Mexico Grant

MSB M3483 Myotis occultus m Arizona Apache

LSUMNS L4010 Myotis thysanodes thysanodes f Mexico

San Luis

Potosi

MSB M117100 Myotis thysanodes thysanodes f Colorado Montezuma

MSB M117102 Myotis thysanodes thysanodes f Colorado Montezuma

MSB M11984 Myotis thysanodes thysanodes f

New

Mexico Bernalillo

MSB M123247 Myotis thysanodes pahasapensis f Nebraska Scotts Bluff

MSB M161890 Myotis thysanodes thysanodes f Arizona Yavapai

MSB M161891 Myotis thysanodes thysanodes f Arizona Yavapai


New

Mexico Socorro


New

Mexico Socorro

MSB M69417 Myotis thysanodes f

New

Mexico Cibola

Burke B62590 Myotis thysanodes m California Napa

Burke B62592 Myotis thysanodes m California Riverside

MSB M120969 Myotis thysanodes thysanodes m Utah San Juan

MSB M140930 Myotis thysanodes m

New

Mexico Catron

MSB M37382 Myotis thysanodes m California Santa Clara

MSB M45907 Myotis thysanodes thysanodes m

New

Mexico Hidalgo

MSB M69416 Myotis thysanodes m

New

Mexico Cibola

MSB M99300 Myotis thysanodes m Texas Jeff Davis

PSUMVB P1360 Myotis thysanodes m Arizona Pima

PSUMVB P2985 Myotis thysanodes m Oregon Union

LSUMNS L10408 Myotis velifer f Arizona Pima

LSUMNS L10536 Myotis velifer f Texas Comal

LSUMNS L10543 Myotis velifer f Texas Comal

MSB M21866 Myotis velifer f

New

Mexico Lincoln

MSB M23053 Myotis velifer f Texas Presidio

MSB M25013 Myotis velifer f Arizona Cochise

MSB M30638 Myotis velifer f Oklahoma Harmon

MSB M61110 Myotis velifer f Sonora

152

Museum

Specimen


State/

Province

County/

District


Baja

California

Sur


New

Mexico Chaves

MSB M21867 Myotis velifer m

New

Mexico Lincoln

MSB M23054 Myotis velifer m Texas Presidio

MSB M30637 Myotis velifer m Oklahoma Harmon

MSB M41617 Myotis velifer incautus m

New

Mexico Chaves

MSB M41618 Myotis velifer incautus m

New

Mexico Chaves

MSB M41659 Myotis velifer velifer (brevis) m Arizona Cochise

MSB M45910 Myotis velifer brevis m

New

Mexico Hidalgo

MSB M45911 Myotis velifer brevis m

New

Mexico Hidalgo

MSB M53789 Myotis velifer velifer m Sonora

PSUMVB P2248 Myotis velifer m Kansas Barber C

LSUMNS L1191 Myotis vivesi f Mexico

"Lower

California"

MSB M42643 Myotis vivesi f Mexico Sonora









KU K80184 Myotis vivesi m Mexico Sonora

KU K80188 Myotis vivesi m Mexico Sonora

MSB M42642 Myotis vivesi m Mexico Sonora








Burke B33269 Myotis volans f Washington Douglas

Burke B79348 Myotis volans f Washington Ferry

LSUMNS L10413 Myotis volans f Arizona Cochise

LSUMNS L11188 Myotis volans f Colorado Rio Grande

153

Museum

Specimen


State/

Province

County/

District

MSB M41261 Myotis volans interior f

New

Mexico Taos

MSB M42514 Myotis volans interior f

New

Mexico Hidalgo

MSB M69408 Myotis volans f

New

Mexico Cibola

PSUMVB P2942 Myotis volans f Oregon Lane

PSUMVB P3001 Myotis volans f Oregon Malheur

PSUMVB P3280 Myotis volans f Oregon Wallowa

Burke B62627 Myotis volans m Oregon Deschutes

Burke B6547 Myotis volans longicrus m Washington Columbia

Burke B79423 Myotis volans m Washington Spokane

MSB M13800 Myotis volans interior m Utah Garfield

MSB M41262 Myotis volans interior m

New

Mexico Taos

MSB M45217 Myotis volans interior m

New

Mexico Hidalgo

PSUMVB P2936 Myotis volans m Oregon Marion

PSUMVB P3000 Myotis volans m Oregon Baker

PSUMVB P3049 Myotis volans m Oregon Wheeler

PSUMVB P3054 Myotis volans m Oregon Wasco

LSUMNS L1154 Myotis yumanensis f Nevada

"Pyramid

Lake"

LSUMNS L4903 Myotis yumanensis f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M17902 Myotis yumanensis f

New

Mexico Taos

MSB M29883 Myotis yumanensis f Utah Uintah

MSB M40575 Myotis yumanensis saturatus f California Madera

MSB M41616 Myotis yumanensis yumanensis f

New

Mexico Socorro

MSB M46571 Myotis yumanensis sociabilis f California Lassen

MSB M53793 Myotis yumanensis yumanensis f Sonora

LSUMNS L4925 Myotis yumanensis m Mexico

San Luis

Potosi

MSB M13270 Myotis yumanensis yumanensis m

New

Mexico Catron

MSB M14294 Myotis yumanensis m

New

Mexico Taos


New

Mexico Union

MSB M29880 Myotis yumanensis m Utah Uintah

154

Museum

Specimen


State/

Province

County/

District


New

Mexico Socorro

RDS RDS8118 Myotis yumanensis m Texas Brewster




LSUMNS L20883 Natalus stramineus f Mexico Sonora

MSB M11050 Natalus stramineus mexicanus f Mexico Sonora

MSB M19090 Natalus stramineus f Mexico Sonora








MSB M19084 Natalus stramineus m Mexico Sonora










KU K44754 Nycticeius humeralis mexicanus f Mexico Coahuila

LSUMNS L4874 Nycticeius humeralis f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M162361 Nycticeius humeralis humeralis f Texas San Patricio




KU K48316 Nycticeius humeralis mexicanus m Mexico Coahuila

LSUMNS L4884 Nycticeius humeralis m Mexico

San Luis

Potosi

155

Museum

Specimen


State/

Province

County/

District


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M162358 Nycticeius humeralis humeralis m Texas San Patricio




LSUMNS L11060 Nyctinomops aurispinosus f Mexico Colima





LACM LA14176 Nyctinomops aurispinosus f Mexico Chiapas

MSB M22657 Nyctinomops aurispinosus f Mexico Sinaloa



MSB M55459 Nyctinomops aurispinosus f Mexico Sonora

LSUMNS L11057 Nyctinomops aurispinosus m Mexico Colima





MSB M22648 Nyctinomops aurispinosus m Mexico Sinaloa





MSB M160480 Nyctinomops femorosaccus mexicana f Arizona Pima

MSB M19313 Nyctinomops femorosaccus f

New

Mexico Hidalgo


New

Mexico Hidalgo


New

Mexico Hidalgo

MSB M42857 Nyctinomops femorosaccus f Mexico

Baja

California


Baja

California

156

Museum

Specimen


State/

Province

County/

District


Baja

California

MSB M60881 Nyctinomops femorosaccus f Mexico Sonora



MSB M18579 Nyctinomops femorosaccus m Mexico Sonora

MSB M19315 Nyctinomops femorosaccus m

New

Mexico Hidalgo


MSB M42858 Nyctinomops femorosaccus m Mexico

Baja

California


Baja

California


Baja

California


Baja

California




Burke B50616 Nyctinomops macrotis f Utah

LSUMNS L8079 Nyctinomops macrotis f

New

Mexico Rio Arriba

MSB M116461 Nyctinomops macrotis f Utah Grand

MSB M116462 Nyctinomops macrotis f Utah Grand

MSB M160481 Nyctinomops macrotis f Arizona Mohave

MSB M30647 Nyctinomops macrotis f Texas Brewster

MSB M30648 Nyctinomops macrotis f Texas Brewster

MSB M4552 Nyctinomops macrotis f

New

Mexico Bernalillo

MSB M53842 Nyctinomops macrotis f Mexico Sonora

MSB M53843 Nyctinomops macrotis f Mexico Sonora

KU K97087 Nyctinomops macrotis m Mexico Sinaloa



MSB M122221 Nyctinomops macrotis m Wyoming Teton

MSB M16595 Nyctinomops macrotis m

New

Mexico Bernalillo

MSB M36884 Nyctinomops macrotis m

New

Mexico Valencia

MSB M53840 Nyctinomops macrotis m Mexico Sonora




LSUMNS L10126 Parastrellus hesperus f Arizona Pima

LSUMNS L10127 Parastrellus hesperus f Arizona Pima

LSUMNS L1152 Parastrellus hesperus f California Inyo

LSUMNS L1200 Parastrellus hesperus f California Inyo

157

Museum

Specimen


State/

Province

County/

District

LSUMNS L4021 Parastrellus hesperus f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M116757 Parastrellus hesperus hesperus f Utah Wayne

LSUMNS L10430 Parastrellus hesperus m Arizona Pima

LSUMNS L1888 Parastrellus hesperus m California Kern

LSUMNS L22041 Parastrellus hesperus m

New

Mexico Socorro

LSUMNS L4023 Parastrellus hesperus m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

MSB M118695 Parastrellus hesperus hesperus m Utah Wayne

MSB M118696 Parastrellus hesperus hesperus m Utah Wayne

MSB M162636 Parastrellus hesperus m Arizona Yuma

MSB M162637 Parastrellus hesperus m Arizona Yuma

KU K29872 Perimyotis subflavus veraecrucis f Mexico Veracruz



KU K48263 Perimyotis subflavus clarus f Mexico Coahuila

KU K48267 Perimyotis subflavus clarus f Mexico Coahuila

MSB M162706 Perimyotis subflavus f Texas Comal



MSB M162710 Perimyotis subflavus subflavus f Texas Shelby

MSB M162711 Perimyotis subflavus subflavus f Texas Shelby

KU K29875 Perimyotis subflavus veraecrucis m Mexico Veracruz







KU K48272 Perimyotis subflavus clarus m Mexico Coahuila

KU K58849 Perimyotis subflavus subflavus m Mexico Tamaulipas

MSB M162709 Perimyotis subflavus m Texas Comal

Burke B62784 Pteronotus davyi f Mexico Morelos



158

Museum

Specimen


State/

Province

County/

District

Burke B63205 Pteronotus davyi f Mexico

San Luis

Potosi

LSUMNS L4785 Pteronotus davyi f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

Burke B62785 Pteronotus davyi m Mexico Morelos









Burke B62796 Pteronotus davyi m Mexico

San Luis

Potosi

LSUMNS L10977 Pteronotus parnellii f Mexico Colima




LSUMNS L4811 Pteronotus parnellii f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L8156 Pteronotus parnellii f Mexico Tabasco

LSUMNS L8167 Pteronotus parnellii f Mexico Tabasco

LSUMNS L11963 Pteronotus parnellii m Mexico Oaxaca

LSUMNS L4812 Pteronotus parnellii m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

159

Museum

Specimen


State/

Province

County/

District

LSUMNS L7373 Pteronotus parnellii m Mexico Tabasco





KU K97050 Rhogeessa gracilis f Mexico Jalisco

KU K108976 Rhogeessa gracilis m Mexico Jalisco

KU K92951 Rhogeessa gracilis m Mexico Jalisco

LSUMNS L11030 Sturnira lilium f Mexico Colima

LSUMNS L11036 Sturnira lilium f Mexico Colima

LSUMNS L4836 Sturnira lilium parvidens f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L8210 Sturnira lilium f Mexico Tabasco




Burke B50724 Sturnira lilium m Mexico Veracruz

Burke B63218 Sturnira lilium m Mexico

San Luis

Potosi

LSUMNS L11028 Sturnira lilium m Mexico Colima




LSUMNS L4839 Sturnira lilium parvidens m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L8237 Sturnira lilium m Mexico Tabasco

LSUMNS L10133 Tadarida brasiliensis f Arizona Pima

LSUMNS L10134 Tadarida brasiliensis f Arizona Pima

LSUMNS L11198 Tadarida brasiliensis f Colorado Rio Grande

LSUMNS L1880 Tadarida brasiliensis f California Los Angeles

LSUMNS L4050 Tadarida brasiliensis f Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

160

Museum

Specimen


State/

Province

County/

District


San Luis

Potosi


San Luis

Potosi

LSUMNS L429 Tadarida brasiliensis f California Yolo

LSUMNS L10132 Tadarida brasiliensis m Arizona Pima

LSUMNS L11917 Tadarida brasiliensis m Mexico Oaxoca

LSUMNS L1879 Tadarida brasiliensis m California Los Angeles

LSUMNS L1881 Tadarida brasiliensis m California Los Angeles

LSUMNS L19808 Tadarida brasiliensis m Texas Gregg

LSUMNS L2840 Tadarida brasiliensis m Mexico

San Luis

Potosi


San Luis

Potosi


San Luis

Potosi


San Luis

Potosi

LSUMNS L8736 Tadarida brasiliensis m Texas Hidalgo

161

APPENDIX

VI

CHAPTER 4 SUPPLEMENTARY M

ATTERIA

LS

Tab

le S

1:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

and w

ing d

ata

for

all

des

erts

com

bin

ed f

or

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“all

tax

a”, “a

ll v

esper

tili

onid

s”, an

d “

all M

yotis”

. M

PD

M

NT

D

Del

imit

atio

n

met

hod

T

axo

n

clust

ere

d

com

munit

ies

ran

do

m

com

munit

ies

overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

resu

lts

clust

ere

d

com

munit

ies

rando

m

com

munit

ies

overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

resu

lts

df

5k

m b

uff

er

All

13

148

11

405.5

9

0.0

12

clust

ere

d

14

146

12

423.6

8

0.0

02

clust

ere

d

344

Ves

per

tili

onid

ae

7

130

7

282.1

8

0.5

86

ns

8

126

10

300.6

1

0.2

93

ns

288

Myo

tis

4

55

7

166.4

2

0.0

23

clust

ere

d

3

62

1

145.7

9

0.1

95

ns

132

10k

m b

uff

er

All

9

113

6

313.2

1

0.0

08

clust

ere

d

11

110

7

332.2

2

0.0

01

clust

ere

d

256

Ves

per

tili

onid

ae

6

100

6

219.9

0

0.5

65

ns

7

94

11

241.1

7

0.2

05

ns

224

Myo

tis

5

51

3

152.3

5

0.0

18

clust

ere

d

1

57

1

119.1

5

0.4

53

ns

118

10k

m g

rid

All

13

199

8

514.2

8

0.0

08

clust

ere

d

18

189

13

523.2

5

0.0

04

clust

ere

d

440

Ves

per

tili

onid

ae

8

156

10

349.1

9

0.4

72

ns

11

151

12

362.7

0

0.2

83

ns

348

Myo

tis

3

56

1

148.2

4

0.0

41

clust

ere

d

1

58

1

126.4

4

0.3

26

ns

120

50k

m g

rid

All

14

151

6

413.7

8

0.0

05

clust

ere

d

16

144

11

436.4

8

0.0

00

clust

ere

d

342

Ves

per

tili

onid

ae

10

133

6

296.1

1

0.5

20

ns

12

127

10

328.6

6

0.1

07

ns

298

Myo

tis

2

65

8

169.7

2

0.1

29

ns

2

146

2

152.3

7

0.4

31

ns

150

50k

m c

ircle

All

11

93

8

295.8

1

0.0

01

clust

ere

d

10

91

11

297.3

7

0.0

01

clust

ere

d

224

Ves

per

tili

onid

ae

5

92

4

202.0

4

0.4

86

ns

4

89

8

220.9

8

0.1

71

ns

202

Myo

tis

3

56

8

149.1

1

0.1

76

ns

6

60

1

157.5

8

0.0

80

ns

134

100k

m

circ

le

All

9

61

7

204.8

5

0.0

04

clust

ere

d

9

61

7

222.1

3

0.0

00

clust

ere

d

154

Ves

per

tili

onid

ae

4

68

4

150.1

1

0.5

28

ns

6

66

4

178.0

8

0.0

73

ns

152

Myo

tis

4

50

3

137.3

7

0.0

67

ns

6

48

3

138.6

7

0.0

58

ns

114

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

162

Tab

le S

2:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

dat

a fo

r al

l des

erts

com

bin

ed f

or

each

co

mm

unit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“all

tax

a”, “a

ll v

esper

tili

onid

s”, an

d “

all M

yotis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

clust

ere

d

com

munit

ies

ran

do

m

com

munit

ies

overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

resu

lts

clust

ere

d

com

munit

ies

rando

m

com

munit

ies

overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

resu

lts

df

5k

m b

uff

er

All

11

153

8

403.5

1

0.0

15

clust

ere

d

14

150

8

435.2

7

0.0

01

clust

ere

d

344

Ves

per

tili

onid

ae

8

128

8

285.6

6

0.5

28

ns

8

124

12

305.1

4

0.2

33

ns

288

Myo

tis

4

55

7

167.9

4

0.0

19

clust

ere

d

3

62

1

156.6

8

0.0

70

ns

132

10k

m b

uff

er

All

10

112

6

315.1

9

0.0

07

clust

ere

d

11

110

7

341.7

2

0.0

00

clust

ere

d

256

Ves

per

tili

onid

ae

6

99

7

224.0

6

0.4

86

ns

8

92

12

243.6

5

0.1

75

ns

224

Myo

tis

5

51

3

151.8

7

0.0

19

clust

ere

d

1

57

1

123.1

9

0.3

54

ns

118

10k

m g

rid

All

16

197

7

510.6

0

0.0

11

clust

ere

d

17

196

7

520.2

1

0.0

05

clust

ere

d

440

Ves

per

tili

onid

ae

9

154

11

354.7

5

0.3

90

ns

13

149

12

366.6

0

0.2

36

ns

348

Myo

tis

1

58

1

151.8

2

0.0

26

clust

ere

d

1

58

1

134.4

5

0.1

74

ns

120

50k

m g

rid

All

14

149

8

415.4

4

0.0

04

clust

ere

d

15

149

7

444.4

6

0.0

00

clust

ere

d

342

Ves

per

tili

onid

ae

13

127

9

304.2

5

0.3

89

ns

12

123

14

334.5

3

0.0

71

ns

298

Myo

tis

2

64

9

171.1

5

0.1

14

ns

1

72

2

156.5

9

0.3

40

ns

150

50k

m c

ircle

All

7

96

9

289.0

2

0.0

02

clust

ere

d

8

34

10

300.6

0

0.0

01

clust

ere

d

224

Ves

per

tili

onid

ae

5

90

6

204.7

1

0.4

34

ns

6

89

6

216.4

6

0.2

31

ns

202

Myo

tis

3

56

8

149.5

2

0.1

70

clust

ere

d

4

62

1

154.6

7

0.1

07

ns

134

100k

m c

ircle

All

8

62

7

204.9

5

0.0

04

clust

ere

d

12

58

7

226.1

1

0.0

00

clust

ere

d

154

Ves

per

tili

onid

ae

5

68

3

155.5

7

0.4

05

ns

6

66

4

176.7

9

0.0

82

ns

152

Myo

tis

4

49

4

138.1

6

0.0

61

ns

3

51

3

138.7

5

0.0

57

ns

114

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

163

Tab

le S

3:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r w

ing d

ata

for

all

des

erts

com

bin

ed f

or

each

co

mm

unit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“all

tax

a”, “a

ll v

esper

tili

onid

s”, an

d “

all M

yotis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

clust

ere

d

com

munit

ies

ran

do

m

com

munit

ies

overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

resu

lts

clust

ere

d

com

munit

ies

rando

m

com

munit

ies

overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

resu

lts

df

5k

m b

uff

er

All

13

148

11

407.2

7

0.0

11

clust

ere

d

18

142

12

417.7

4

0.0

04

clust

ere

d

344

Ves

per

tili

onid

ae

10

128

6

275.3

2

0.6

94

ns

9

131

4

298.9

5

0.3

16

ns

288

Myo

tis

6

57

3

146.6

8

0.1

81

ns

2

62

2

135.6

1

0.3

97

ns

132

10k

m b

uff

er

All

8

111

9

305.1

4

0.0

19

clust

ere

d

11

110

7

311.8

2

0.0

10

clust

ere

d

256

Ves

per

tili

onid

ae

7

100

5

219.8

5

0.5

66

ns

7

101

4

238.6

3

0.2

39

ns

224

Myo

tis

4

53

2

130.3

4

0.2

06

ns

3

55

1

112.2

5

0.6

32

ns

118

10k

m g

rid

All

19

189

12

525.9

7

0.0

03

clust

ere

d

19

189

12

533.2

3

0.0

02

clust

ere

d

440

Ves

per

tili

onid

ae

10

156

8

340.1

9

0.6

08

ns

9

157

8

358.9

1

0.3

32

ns

348

Myo

tis

5

54

1

131.8

1

0.2

17

ns

3

54

3

123.2

2

0.4

02

ns

120

50k

m g

rid

All

13

148

10

400.3

4

0.0

16

clust

ere

d

16

144

11

422.6

0

0.0

02

clust

ere

d

342

Ves

per

tili

onid

ae

11

134

4

284.8

4

0.6

98

ns

11

136

2

314.7

6

0.2

42

ns

298

Myo

tis

3

69

3

146.7

3

0.5

60

ns

1

69

4

151.2

7

0.4

56

ns

150

50k

m c

ircle

All

10

95

7

295.8

9

0.0

01

clust

ere

d

8

98

6

282.1

7

0.0

05

clust

ere

d

224

Ves

per

tili

onid

ae

6

92

3

201.1

5

0.5

04

ns

5

93

3

208.3

0

0.3

66

ns

202

Myo

tis

4

60

3

134.2

5

0.4

78

ns

3

63

1

156.9

6

0.0

85

ns

134

100k

m c

ircle

All

8

141

5

197.8

5

0.0

10

clust

ere

d

9

63

5

200.6

5

0.0

07

clust

ere

d

154

Ves

per

tili

onid

ae

2

69

5

139.0

6

0.7

66

ns

5

68

3

159.3

3

0.3

26

ns

152

Myo

tis

1

53

3

108.4

4

0.6

29

ns

3

51

3

135.9

3

0.0

79

ns

114

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

164

Tab

le S

4:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

and w

ing d

ata

for

the

Gre

at B

asin

Des

ert

for

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“GB

tax

a”, “G

B v

esper

tili

onid

s”, an

d “

GB

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

5

50

4

118.5

0

0.4

70

ns

5

48

6

130.6

7

0.2

00

ns

118

Ves

per

tili

onid

ae

3

48

5

106.9

9

0.6

16

ns

3

49

4

115.7

9

0.3

84

ns

112

Myo

tis

2

26

1

56.9

3

0.5

15

ns

4

25

0

58.0

2

0.4

75

ns

58

10k

m b

uff

er

All

4

42

3

101.3

1

0.3

89

ns

5

39

5

119.0

3

0.0

73

ns

98

Ves

per

tili

onid

ae

4

41

2

91.9

2

0.5

41

ns

5

40

2

111.3

7

0.1

07

ns

94

Myo

tis

2

28

0

65.2

8

0.2

99

ns

2

28

0

62.3

3

0.3

93

ns

60

10k

m g

rid

All

1

58

3

110.6

1

0.0

00

clust

ere

d

3

54

5

126.2

1

0.0

00

clust

ere

d

124

Ves

per

tili

onid

ae

1

58

2

108.5

8

0.8

02

ns

4

53

4

120.0

5

0.5

33

ns

122

Myo

tis

2

25

0

47.9

6

0.7

05

ns

3

24

0

53.2

3

0.5

04

ns

54

50k

m g

rid

All

3

53

3

118.0

0

0.4

83

ns

5

48

6

138.0

2

0.1

00

ns

118

Ves

per

tili

onid

ae

2

54

3

112.8

1

0.6

18

ns

3

67

4

131.0

4

0.1

94

ns

118

Myo

tis

1

29

0

56.9

5

0.5

88

ns

1

29

0

56.9

0

0.5

90

ns

60

50k

m c

ircle

All

4

30

5

92.4

3

0.1

26

ns

3

32

4

102.4

3

0.0

33

ns

78

Ves

per

tili

onid

ae

2

33

3

80.7

3

0.3

34

ns

3

32

3

85.6

8

0.2

10

ns

76

Myo

tis

1

27

1

57.0

1

0.5

12

ns

1

26

2

48.0

3

0.8

22

ns

58

100k

m

circ

le

All

2

27

3

62.0

8

0.5

45

ns

2

27

3

79.1

4

0.0

96

ns

64

Ves

per

tili

onid

ae

1

29

1

52.6

7

0.7

95

ns

2

28

1

65.0

1

0.3

72

ns

62

Myo

tis

1

24

0

48.8

9

0.5

18

ns

1

24

0

43.9

1

0.7

15

ns

50

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

165

Tab

le S

5:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

dat

a fo

r th

e G

reat

Bas

in D

eser

t fo

r ea

ch c

om

munit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“GB

tax

a”, “G

B v

esp

erti

lionid

s”, an

d “

GB

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

4

52

3

115.7

4

0.5

42

ns

4

51

4

124.4

9

0.3

23

ns

118

Ves

per

tili

onid

ae

2

50

4

104.0

6

0.6

91

ns

4

49

3

110.7

0

0.5

17

ns

112

Myo

tis

0

28

1

53.4

9

0.6

44

ns

0

28

1

53.4

4

0.6

45

ns

58

10k

m b

uff

er

All

3

44

2

99.8

2

0.4

30

ns

5

40

4

117.6

1

0.0

86

ns

98

Ves

per

tili

onid

ae

3

42

2

89.7

0

0.6

06

ns

5

39

3

106.2

7

0.1

82

ns

94

Myo

tis

2

28

0

63.5

8

0.3

52

ns

1

28

1

60.2

6

0.4

66

ns

60

10k

m g

rid

All

1

58

3

108.1

9

0.8

43

ns

3

56

3

117.8

4

0.6

39

ns

124

Ves

per

tili

onid

ae

1

58

2

106.5

2

0.8

40

ns

4

54

3

112.0

2

0.7

30

ns

122

Myo

tis

0

27

0

45.5

8

0.7

86

ns

0

26

1

47.2

4

0.7

31

ns

54

50k

m g

rid

All

2

54

3

116.4

2

0.5

24

ns

5

49

5

135.9

9

0.1

23

ns

118

Ves

per

tili

onid

ae

2

54

3

112.3

9

0.6

28

ns

3

51

5

129.2

2

0.2

26

ns

118

Myo

tis

0

30

0

55.4

2

0.6

44

ns

0

29

1

53.3

4

0.7

16

ns

60

50k

m c

ircle

All

3

31

5

88.0

4

0.2

05

ns

3

33

3

95.1

4

0.0

91

ns

78

Ves

per

tili

onid

ae

3

30

5

78.4

5

0.4

01

ns

3

32

3

82.5

4

0.2

85

ns

76

Myo

tis

2

25

2

54.7

2

0.5

98

ns

1

26

2

53.1

0

0.6

58

ns

58

100k

m

circ

le

All

2

27

3

61.0

8

0.5

80

ns

3

25

4

74.6

7

0.1

70

ns

64

Ves

per

tili

onid

ae

2

59

1

53.4

4

0.7

72

ns

3

26

2

63.5

7

0.4

21

ns

62

Myo

tis

1

22

2

47.1

3

0.5

89

ns

0

25

1

46.4

7

0.6

16

ns

50

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

166

Tab

le S

6:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r w

ing d

ata

for

the

Gre

at B

asin

Des

ert

for

each

com

munit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“GB

tax

a”, “G

B v

esp

erti

lionid

s”, an

d “

GB

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

7

51

1

130.0

3

0.2

12

ns

7

48

4

146.6

9

0.0

38

clust

ere

d

118

Ves

per

tili

onid

ae

5

47

4

114.7

9

0.4

09

ns

6

49

1

127.2

3

0.1

54

ns

112

Myo

tis

3

25

1

68.6

5

0.1

60

ns

3

26

0

74.3

7

0.0

73

ns

58

10k

m b

uff

er

All

4

43

2

112.5

9

0.1

49

ns

6

40

3

125.9

3

0.0

30

clust

ere

d

98

Ves

per

tili

onid

ae

4

40

3

98.0

4

0.3

67

ns

6

40

1

109.6

2

0.1

29

ns

94

Myo

tis

2

27

1

61.6

2

0.4

18

ns

1

28

1

56.4

1

0.6

08

ns

60

10k

m g

rid

All

5

54

3

121.0

5

0.2

63

ns

5

53

4

144.9

8

0.0

96

ns

124

Ves

per

tili

onid

ae

3

56

2

111.5

4

0.4

95

ns

5

53

3

129.9

4

0.2

95

ns

122

Myo

tis

3

21

3

60.1

1

0.2

64

ns

3

24

0

68.8

9

0.0

84

ns

54

50k

m g

rid

All

7

50

2

124.1

0

0.3

32

ns

6

48

5

142.2

9

0.0

63

ns

118

Ves

per

tili

onid

ae

5

51

3

116.6

8

0.5

17

ns

7

47

5

125.2

6

0.3

06

ns

118

Myo

tis

2

27

1

59.5

9

0.4

91

ns

1

28

1

62.6

4

0.3

83

ns

60

50k

m c

ircle

All

4

30

5

99.4

2

0.0

51

ns

5

29

5

108.3

4

0.0

13

clust

ere

d

78

Ves

per

tili

onid

ae

3

32

3

82.3

6

0.2

89

ns

4

31

3

84.6

7

0.2

32

ns

76

Myo

tis

2

26

1

50.7

6

0.7

39

ns

1

27

1

46.5

6

0.8

60

ns

58

100k

m c

ircle

All

2

28

2

65.9

1

0.4

11

ns

5

26

2

74.4

8

0.1

74

ns

64

Ves

per

tili

onid

ae

1

30

0

52.8

5

0.7

90

ns

3

28

1

60.0

8

0.5

45

ns

62

Myo

tis

1

23

1

41.4

6

0.8

00

ns

1

23

1

36.6

0

0.9

21

ns

50

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

167

Tab

le S

7:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

and w

ing d

ata

for

the

Moja

ve

Des

ert

for

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“MJ

tax

a”, “M

J ves

per

tili

onid

s”, an

d “

MJ M

yotis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

1

17

1

35.1

0

0.6

05

ns

2

16

1

43.0

4

0.2

64

ns

38

Ves

per

tili

onid

ae

1

18

0

36.6

3

0.5

33

ns

1

18

0

41.9

9

0.3

02

ns

38

Myo

tis

0

10

1

17.3

3

0.7

45

ns

0

11

0

17.5

5

0.7

32

ns

22

10k

m b

uff

er

All

0

10

0

18.5

8

0.5

50

ns

1

9

0

21.6

4

0.3

60

ns

20

Ves

per

tili

onid

ae

0

9

1

18.7

5

0.5

38

ns

1

7

2

24.3

3

0.2

28

ns

20

Myo

tis

0

7

0

8.7

9

0.8

44

ns

0

7

0

15.4

2

0.3

50

ns

14

10k

m g

rid

All

2

24

1

49.8

5

0.6

35

ns

3

22

2

60.0

1

0.2

67

ns

54

Ves

per

tili

onid

ae

2

21

2

46.5

3

0.6

13

ns

1

22

2

55.1

8

0.2

85

ns

50

Myo

tis

0

9

1

12.7

1

0.8

08

ns

0

9

0

15.8

3

0.6

05

ns

18

50k

m g

rid

All

1

17

0

43.1

4

0.1

92

ns

2

16

0

40.4

0

0.2

82

ns

36

Ves

per

tili

onid

ae

1

16

1

43.3

2

0.1

88

ns

1

16

1

43.2

6

0.1

89

ns

36

Myo

tis

0

8

2

15.9

2

0.7

22

ns

0

10

0

19.6

7

0.4

79

ns

20

50k

m c

ircle

All

0

16

0

28.1

2

0.6

63

ns

1

15

0

29.1

8

0.6

10

ns

32

Ves

per

tili

onid

ae

1

14

1

30.4

9

0.5

43

ns

2

13

1

44.3

2

0.0

72

ns

32

Myo

tis

0

11

1

17.8

3

0.8

11

ns

1

11

0

26.4

6

0.3

31

ns

24

100k

m

circ

le

All

1

8

1

23.5

3

0.2

64

ns

0

10

0

19.3

6

0.4

99

ns

20

Ves

per

tili

onid

ae

0

9

1

21.8

5

0.3

49

ns

1

9

0

27.4

9

0.1

22

ns

20

Myo

tis

1

5

1

13.8

9

0.4

58

ns

0

6

1

16.5

4

0.2

82

ns

14

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

168

Tab

le S

8:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

dat

a fo

r th

e M

oja

ve

Des

ert

for

each

com

munit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“MJ

tax

a”, “M

J ves

per

tili

onid

s”, an

d “

MJ M

yotis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

1

17

1

36.2

3

0.5

51

ns

2

16

1

45.5

6

0.1

87

ns

38

Ves

per

tili

onid

ae

1

17

1

36.5

3

0.5

37

ns

2

17

0

44.9

5

0.2

04

ns

38

Myo

tis

0

10

1

17.5

4

0.7

33

ns

0

11

0

18.1

8

0.6

95

ns

22

10k

m b

uff

er

All

0

10

0

19.6

7

0.4

79

ns

1

9

0

22.3

1

0.3

24

ns

20

Ves

per

tili

onid

ae

0

9

1

19.4

2

0.4

95

ns

1

7

2

23.0

7

0.2

85

ns

20

Myo

tis

0

7

0

7.6

1

0.9

09

ns

0

7

0

13.7

1

0.4

71

ns

14

10k

m g

rid

All

2

24

1

55.1

4

0.4

31

ns

3

23

1

63.7

2

0.1

72

ns

54

Ves

per

tili

onid

ae

2

22

3

53.5

2

0.3

41

ns

2

23

2

63.7

9

0.0

91

ns

50

Myo

tis

0

8

1

14.1

3

0.7

20

ns

0

9

0

17.4

8

0.4

91

ns

18

50k

m g

rid

All

1

17

0

45.2

0

0.1

40

ns

1

17

0

38.8

7

0.3

42

ns

36

Ves

per

tili

onid

ae

1

15

2

45.1

3

0.1

42

ns

2

15

1

41.5

8

0.2

41

ns

36

Myo

tis

0

9

1

13.1

2

0.8

72

ns

0

10

0

18.0

1

0.5

87

ns

20

50k

m c

ircle

All

0

16

0

29.4

3

0.5

97

ns

1

15

0

29.3

4

0.6

02

ns

32

Ves

per

tili

onid

ae

1

16

2

33.3

4

0.4

02

ns

2

12

2

42.4

8

0.1

02

ns

32

Myo

tis

0

11

1

14.9

8

0.9

21

ns

1

11

0

24.6

8

0.4

23

ns

24

100k

m

circ

le

All

2

7

1

24.8

5

0.2

07

ns

1

9

0

20.6

2

0.4

20

ns

20

Ves

per

tili

onid

ae

1

8

1

22.9

5

0.2

91

ns

1

9

0

26.6

8

0.1

45

ns

20

Myo

tis

0

7

0

12.6

8

0.5

52

ns

0

7

0

15.7

2

0.3

31

ns

14

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

169

Tab

le S

9:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r w

ing d

ata

for

the

Moja

ve

Des

ert

for

each

com

munit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“MJ

tax

a”, “M

J ves

per

tili

onid

s”, an

d “

MJ M

yotis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

1

17

1

34.8

4

0.6

17

ns

2

16

1

39.8

8

0.3

86

ns

38

Ves

per

tili

onid

ae

1

18

0

34.3

8

0.6

38

ns

2

17

0

36.6

3

0.5

33

ns

38

Myo

tis

1

10

0

20.0

1

0.5

82

ns

0

10

1

18.0

1

0.7

05

ns

22

10k

m b

uff

er

All

0

10

0

18.2

8

0.5

69

ns

0

10

0

18.8

5

0.5

31

ns

20

Ves

per

tili

onid

ae

0

9

1

19.1

7

0.5

11

ns

1

8

1

21.2

8

0.3

81

ns

20

Myo

tis

1

6

0

14.0

5

0.4

46

ns

0

7

0

16.7

3

0.2

71

ns

14

10k

m g

rid

All

1

24

2

50.3

1

0.6

18

ns

2

23

2

56.1

7

0.3

94

ns

54

Ves

per

tili

onid

ae

1

24

0

45.5

4

0.6

53

ns

1

24

0

47.9

9

0.5

55

ns

50

Myo

tis

0

9

0

17.2

5

0.5

06

ns

0

9

0

18.5

1

0.4

23

ns

18

50k

m g

rid

All

1

17

0

38.6

4

0.3

51

ns

2

16

0

43.4

7

0.1

83

ns

36

Ves

per

tili

onid

ae

1

17

0

38.1

6

0.3

72

ns

2

16

0

46.9

6

0.1

05

ns

36

Myo

tis

0

10

0

19.0

7

0.5

17

ns

0

10

0

23.0

0

0.2

89

ns

20

50k

m c

ircle

All

1

15

0

27.2

4

0.7

06

ns

0

16

0

26.0

2

0.7

63

ns

32

Ves

per

tili

onid

ae

1

15

0

33.6

0

0.3

90

ns

1

14

1

39.5

2

0.1

69

ns

32

Myo

tis

0

12

2

18.2

6

0.7

90

ns

1

10

1

28.3

4

0.2

46

ns

24

100k

m

circ

le

All

1

8

1

22.1

0

0.3

35

ns

0

9

1

19.9

6

0.4

61

ns

20

Ves

per

tili

onid

ae

0

10

0

21.2

8

0.3

81

ns

1

9

0

23.7

7

0.2

53

ns

20

Myo

tis

0

7

0

12.4

6

0.5

69

ns

1

6

0

17.6

6

0.2

23

ns

14

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

170

Tab

le S

10:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

and w

ing d

ata

for

the

Sonora

n D

eser

t fo

r ea

ch

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“SN

tax

a”, “S

N v

esper

tili

onid

s”, an

d “

SN

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

0

25

1

65.7

2

0.0

96

ns

3

21

2

67.2

1

0.0

76

ns

52

Ves

per

tili

onid

ae

1

14

2

33.8

7

0.4

74

ns

1

15

1

41.0

1

0.1

90

ns

34

Myo

tis

1

4

0

15.5

4

0.1

14

ns

1

2

1

16.3

5

0.0

90

ns

10

10k

m b

uff

er

All

2

25

1

69.6

4

0.1

04

ns

2

24

2

69.2

5

0.1

10

ns

56

Ves

per

tili

onid

ae

1

20

1

44.7

7

0.4

39

ns

0

21

1

43.5

7

0.4

90

ns

44

Myo

tis

2

5

0

22.4

6

0.0

70

ns

2

3

1

20.7

5

0.1

08

ns

14

10k

m g

rid

All

1

37

5

111.3

8

0.0

34

clust

ere

d

4

36

3

127.2

4

0.0

03

clust

ere

d

86

Ves

per

tili

onid

ae

2

20

3

56.5

3

0.2

44

ns

2

20

3

65.3

7

0.0

71

ns

50

Myo

tis

1

5

0

17.6

6

0.1

27

ns

1

4

1

13.5

8

0.3

29

ns

12

50k

m g

rid

All

6

26

1

96.2

7

0.0

09

clust

ere

d

7

25

1

97.0

9

0.0

08

clust

ere

d

66

Ves

per

tili

onid

ae

2

21

2

54.5

1

0.3

07

ns

2

22

1

70.1

4

0.0

32

clust

ere

d

50

Myo

tis

1

10

1

25.4

0

0.3

84

ns

1

9

2

26.5

1

0.3

28

ns

24

50k

m c

ircle

All

4

21

2

71.3

1

0.0

57

ns

3

23

1

69.7

8

0.0

73

ns

54

Ves

per

tili

onid

ae

0

17

1

32.9

1

0.6

16

ns

0

18

0

43.6

6

0.1

78

ns

36

Myo

tis

1

7

0

17.7

7

0.3

38

ns

2

5

1

28.7

7

0.0

26

clust

ere

d

16

100k

m

circ

le

All

1

10

1

30.6

6

0.1

64

ns

1

11

0

32.0

6

0.1

25

ns

24

Ves

per

tili

onid

ae

0

11

1

24.5

5

0.4

31

ns

0

12

0

29.1

4

0.2

15

ns

24

Myo

tis

0

7

1

13.3

0

0.6

50

ns

1

6

1

21.9

1

0.1

46

ns

16

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

171

Tab

le S

11:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

dat

a fo

r th

e S

on

ora

n D

eser

t fo

r ea

ch c

om

munit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“SN

tax

a”, “S

N v

esper

tili

onid

s”, an

d “

SN

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

0

25

0

64.0

4

0.1

22

ns

3

23

0

71.0

3

0.0

409

clust

ere

d

52

Ves

per

tili

onid

ae

0

16

1

34.3

5

0.4

51

ns

1

15

1

42.4

5

0.1

516

ns

34

Myo

tis

2

3

0

16.7

0

0.0

81

ns

2

2

1

17.6

2

0.0

618

ns

10

10k

m b

uff

er

All

2

25

1

68.4

7

0.1

23

ns

2

25

1

72.7

3

0.0

658

ns

56

Ves

per

tili

onid

ae

1

20

1

43.8

6

0.4

78

ns

1

20

1

44.3

3

0.4

577

ns

44

Myo

tis

2

5

0

21.9

1

0.0

80

ns

2

4

1

22.2

0

0.0

746

ns

14

10k

m g

rid

All

0

40

3

100.5

1

0.1

36

ns

5

37

1

121.8

2

0.0

067

clust

ere

d

86

Ves

per

tili

onid

ae

2

19

4

52.5

3

0.3

76

ns

3

19

3

67.1

4

0.0

532

ns

50

Myo

tis

1

5

0

17.1

2

0.1

45

ns

1

4

1

13.2

0

0.3

548

ns

12

50k

m g

rid

All

4

28

1

96.4

1

0.0

09

clust

ere

d

7

26

0

101.6

0

0.0

032

clust

ere

d

66

Ves

per

tili

onid

ae

2

21

2

53.4

6

0.3

43

ns

4

19

2

75.7

6

0.0

108

clust

ere

d

50

Myo

tis

2

9

1

25.1

0

0.4

00

ns

1

9

2

24.4

9

0.4

339

ns

24

50k

m c

ircle

All

4

20

3

69.4

9

0.0

76

ns

3

24

1

71.9

9

0.0

514

ns

54

Ves

per

tili

onid

ae

0

17

1

33.1

8

0.6

03

ns

2

16

0

45.6

5

0.1

3

ns

36

Myo

tis

1

7

0

17.4

6

0.3

56

ns

1

6

1

22.9

1

0.1

162

ns

16

100k

m

circ

le

All

2

9

1

30.9

3

0.1

56

ns

1

11

0

33.9

9

0.0

849

ns

24

Ves

per

tili

onid

ae

0

12

0

24.7

4

0.4

20

ns

0

12

0

28.3

5

0.2

456

ns

24

Myo

tis

0

7

1

13.1

6

0.6

61

ns

1

6

1

22.0

2

0.1

425

ns

16

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

172

Tab

le S

12:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r w

ing d

ata

for

the

Son

ora

n D

eser

t fo

r ea

ch c

om

munit

y

del

imit

atio

n m

ethod. S

pec

ies

pools

use

d w

ere

“SN

tax

a”, “S

N v

esper

tili

onid

s”, an

d “

SN

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

3

20

3

65.6

23

0.0

971

ns

4

20

2

61.8

3566519

0.1

65

ns

52

Ves

per

tili

onid

ae

2

13

2

37.2

48

0.3

219

ns

1

15

1

44.4

017728

0.1

09

ns

34

Myo

tis

1

4

0

16.0

10

0.0

994

ns

2

2

1

17.8

2728737

0.0

58

ns

10

10k

m b

uff

er

All

0

26

2

68.0

11

0.1

30

ns

1

25

2

65.5

291802

0.1

80

ns

56

Ves

per

tili

onid

ae

1

20

1

46.1

07

0.3

851

ns

0

21

1

49.5

1911329

0.2

62

ns

44

Myo

tis

2

5

0

21.8

36

0.0

82

ns

2

4

1

20.3

7903709

0.1

19

ns

14

10k

m g

rid

All

6

37

6

127.2

72

0.0

026

clust

ere

d

7

33

3

122.4

911208

0.0

06

clust

ere

d

86

Ves

per

tili

onid

ae

5

17

3

65.1

93

0.0

731

ns

3

21

1

68.5

3128163

0.0

42

clust

ere

d

50

Myo

tis

2

6

0

20.5

77

0.0

569

ns

1

4

1

17.0

8661856

0.1

46

ns

12

50k

m g

rid

All

3

28

2

89.8

15

0.0

273

clust

ere

d

5

25

3

84.9

3644255

0.0

58

ns

66

Ves

per

tili

onid

ae

1

22

2

53.3

39

0.3

471

ns

3

22

0

63.7

098937

0.0

92

ns

50

Myo

tis

1

10

1

24.4

97

0.4

335

ns

2

8

2

29.8

1617523

0.1

91

ns

24

50k

m c

ircle

All

2

23

2

65.6

68

0.1

327

ns

2

25

0

60.7

393254

0.2

46

ns

54

Ves

per

tili

onid

ae

0

17

1

31.8

82

0.6

648

ns

0

18

0

38.8

141313

0.3

44

ns

36

Myo

tis

1

7

0

21.4

01

0.1

636

ns

3

4

1

31.0

3610761

0.0

13

clust

ere

d

16

100k

m

circ

le

All

1

11

0

26.3

41

0.3

361

ns

0

12

0

27.1

2961775

0.2

98

ns

24

Ves

per

tili

onid

ae

0

11

1

24.7

22

0.4

21

ns

0

12

0

25.9

2598422

0.3

57

ns

24

Myo

tis

0

7

1

11.6

35

0.7

687

ns

0

7

1

16.4

3540991

0.4

23

ns

16

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

173

Tab

le S

13:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r both

skull

and w

ing d

ata

for

the

Chih

uah

uan

Des

ert

for

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“CH

tax

a”,

“CH

ves

per

tili

onid

s”, an

d “

CH

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t

stat

isti

c

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

6

56

6

164.7

0

0.0

47

clust

ere

d

4

57

7

162.7

9

0.0

58

ns

136

Ves

per

tili

onid

ae

4

46

2

101.1

5

0.5

61

ns

3

43

6

104.9

4

0.4

56

ns

104

Myo

tis

1

20

0

45.5

3

0.3

27

ns

2

18

1

50.0

4

0.1

84

ns

42

10k

m b

uff

er

All

4

33

2

84.5

2

0.2

87

ns

2

34

3

85.9

2

0.2

53

ns

78

Ves

per

tili

onid

ae

0

30

1

50.9

2

0.8

42

ns

0

31

0

57.3

7

0.6

43

ns

62

Myo

tis

0

13

0

22.5

3

0.6

59

ns

0

13

0

23.6

9

0.5

94

ns

26

10k

m g

rid

All

7

77

2

189.9

8

0.1

65

ns

4

79

3

175.2

8

0.4

16

ns

172

Ves

per

tili

onid

ae

3

56

2

116.8

0

0.6

16

ns

4

53

4

112.7

6

0.7

14

ns

122

Myo

tis

0

17

0

35.0

3

0.4

19

ns

0

16

1

26.7

3

0.8

08

ns

34

50k

m g

rid

All

3

48

2

111.2

7

0.3

44

ns

1

48

4

116.3

8

0.2

31

ns

106

Ves

per

tili

onid

ae

1

41

0

77.0

3

0.6

92

ns

2

38

2

76.6

6

0.7

03

ns

84

Myo

tis

3

18

0

47.3

3

0.2

64

ns

1

19

1

42.0

9

0.4

67

ns

42

50k

m c

ircle

All

2

24

4

69.7

5

0.1

82

ns

1

24

5

70.0

0

0.1

77

ns

60

Ves

per

tili

onid

ae

0

26

3

46.5

3

0.8

60

ns

0

26

3

49.2

9

0.7

85

ns

58

Myo

tis

0

17

1

35.3

1

0.5

01

ns

1

17

0

46.4

2

0.1

14

ns

36

100k

m c

ircle

All

2

20

1

51.5

0

0.2

67

ns

2

19

2

61.5

4

0.0

62

ns

46

Ves

per

tili

onid

ae

1

21

1

42.9

7

0.6

00

ns

2

18

3

55.6

4

0.1

56

ns

46

Myo

tis

1

15

1

36.8

4

0.3

39

ns

1

15

1

38.6

7

0.2

67

ns

34

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

174

Tab

le S

14:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r sk

ull

dat

a fo

r th

e C

hih

uah

uan

Des

ert

for

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“CH

tax

a”, “C

H v

esper

tili

onid

s”, an

d “

CH

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

7

55

6

164.0

5

0.0

51

ns

6

58

4

171.9

2

0.0

202

clust

ere

d

136

Ves

per

tili

onid

ae

4

48

4

105.4

3

0.4

43

ns

2

45

5

108.5

8

0.3

597

ns

104

Myo

tis

1

20

0

48.2

8

0.2

34

ns

2

18

1

54.7

8

0.0

893

ns

42

10k

m b

uff

er

All

4

33

2

89.4

3

0.1

77

ns

4

33

2

92.6

7

0.1

228

ns

78

Ves

per

tili

onid

ae

2

28

1

53.6

2

0.7

67

ns

0

31

0

58.7

8

0.5

925

ns

62

Myo

tis

0

13

0

23.2

9

0.6

16

ns

0

13

0

21.4

0

0.7

208

ns

26

10k

m g

rid

All

6

78

2

185.8

1

0.2

23

ns

5

77

4

177.1

6

0.3

778

ns

172

Ves

per

tili

onid

ae

3

55

3

120.6

2

0.5

18

ns

4

53

4

115.5

1

0.6

481

ns

122

Myo

tis

0

17

0

38.9

9

0.2

55

ns

0

16

1

34.0

7

0.4

644

ns

34

50k

m g

rid

All

2

48

3

114.7

7

0.2

64

ns

2

48

3

119.0

8

0.1

817

ns

106

Ves

per

tili

onid

ae

2

38

2

80.9

9

0.5

73

ns

1

39

2

79.2

0

0.6

277

ns

84

Myo

tis

3

18

0

48.4

5

0.2

29

ns

2

18

1

43.1

2

0.4

233

ns

42

50k

m c

ircle

All

2

24

4

67.6

6

0.2

32

ns

1

25

4

71.7

2

0.1

429

ns

60

Ves

per

tili

onid

ae

2

26

1

50.6

8

0.7

415

ns

0

27

2

52.6

5

0.6

739

ns

58

Myo

tis

0

17

1

37.3

0

0.4

09

ns

1

17

0

40.9

7

0.2

615

ns

36

100k

m

circ

le

All

2

19

2

50.9

7

0.2

84

ns

2

19

2

63.3

2

0.0

459

clust

ere

d

46

Ves

per

tili

onid

ae

1

21

1

44.8

1

0.5

22

ns

2

18

3

55.4

3

0.1

607

ns

46

Myo

tis

2

14

1

37.4

9

0.3

12

ns

1

17

1

39.5

1

0.2

374

ns

34

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

175

Tab

le S

15:

Res

ult

s of

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

on M

CS

anal

yse

s fo

r w

ing d

ata

for

the

Chih

uah

uan

Des

ert

for

each

com

munit

y d

elim

itat

ion m

ethod. S

pec

ies

pools

use

d w

ere

“CH

tax

a”, “C

H v

esper

tili

onid

s”, an

d “

CH

Myo

tis”

.

MP

D

MN

TD

Del

imit

atio

n

met

hod

T

axo

n

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s

Clu

stere

d

com

munit

ies

Ran

do

m

com

munit

ies

Overd

isper

sed

com

munit

ies

Tes

t st

atis

tic

p-

valu

e

Res

ult

s d

f

5k

m b

uff

er

All

5

57

6

155.3

2

0.1

23

ns

5

57

6

153.8

1

0.1

41

ns

136

Ves

per

tili

onid

ae

1

47

4

87.3

1

0.8

81

ns

1

46

5

95.8

7

0.7

03

ns

104

Myo

tis

2

18

1

39.7

5

0.5

70

ns

1

17

3

45.2

9

0.3

36

ns

42

10k

m b

uff

er

All

0

37

2

70.4

4

0.7

16

ns

0

37

2

71.2

5

0.6

93

ns

78

Ves

per

tili

onid

ae

0

29

2

47.0

4

0.9

21

ns

1

29

1

52.3

9

0.8

03

ns

62

Myo

tis

1

12

0

24.3

3

0.5

57

ns

1

12

0

28.8

9

0.3

16

ns

26

10k

m g

rid

All

8

70

8

194.4

9

0.1

15

ns

8

73

5

187.6

0

0.1

97

ns

172

Ves

per

tili

onid

ae

2

56

3

110.8

5

0.7

56

ns

2

55

4

110.0

0

0.7

74

ns

122

Myo

tis

1

15

1

25.4

5

0.8

55

ns

1

13

3

29.2

0

0.7

02

ns

34

50k

m g

rid

All

2

48

3

105.4

1

0.4

98

ns

3

48

2

114.1

4

0.2

77

ns

106

Ves

per

tili

onid

ae

0

42

0

69.1

9

0.8

78

ns

1

39

2

76.5

2

0.7

07

ns

84

Myo

tis

2

18

1

38.2

9

0.6

35

ns

0

20

1

43.0

9

0.4

25

ns

42

50k

m c

ircle

All

2

25

3

70.7

9

0.1

61

ns

1

27

2

63.0

2

0.3

70

ns

60

Ves

per

tili

onid

ae

0

27

2

46.2

5

0.8

67

ns

0

27

3

49.3

0

0.7

85

ns

58

Myo

tis

1

17

0

35.1

3

0.5

10

ns

2

15

1

51.7

0

0.0

44

ns

36

100k

m

circ

le

All

1

20

2

51.5

9

0.2

64

ns

2

21

0

54.7

0

0.1

78

ns

46

Ves

per

tili

onid

ae

1

21

1

40.0

2

0.7

20

ns

2

19

2

52.0

3

0.2

51

ns

46

Myo

tis

0

16

1

28.9

1

0.7

15

ns

0

16

1

45.7

7

0.0

86

ns

34

df=

2*(n

um

ber

of

com

munit

ies)

Tes

t st

atis

tic=

χ2

ns=

not

signif

ican

tly d

iffe

rent

from

ran

dom

ly a

ssem

ble

d c

om

munit

ies

176

Tab

le S

16. P

ears

on p

rod

uct

-mom

ent

corr

elat

ion c

oef

fici

ents

fo

r P

CS

and s

kull

and w

ing M

CS

(a)

SE

S-M

PD

and (

b)

SE

S-M

NT

D.

Gra

y c

ells

indic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5

a)

Tax

on

D

elim

itat

ion

m

eth

od

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

ua

All

5k

m b

uff

er

0.680

0.634

0.707

0.589

0.751

10k

m b

uff

er

0.716

0.631

0.494

0.666

0.649

10k

m g

rid

0.607

0.597

0.613

0.068

0.682

50k

m g

rid

0.780

0.575

0.704

0.649

0.787

50k

m c

ircl

e 0.779

0.778

0.314

0.626

0.750

100k

m c

ircl

e 0.860

0.753

0.663

0.536

0.864

Ves

per

tili

on

idae

5k

m b

uff

er

0.675

0.673

0.747

0.689

0.544

10k

m b

uff

er

0.701

0.658

0.816

0.673

0.533

10k

m g

rid

0.658

0.641

0.749

0.700

0.489

50k

m g

rid

0.704

0.613

0.753

0.676

0.562

50k

m c

ircl

e 0.695

0.802

0.780

0.533

0.246

100k

m c

ircl

e 0.701

0.735

0.725

0.625

0.423

Myo

tis

5k

m b

uff

er

0.503

0.481

-0.024

0.772

0.605

10k

m b

uff

er

0.678

0.612

0.245

0.814

0.762

10k

m g

rid

0.505

0.011

0.281

0.740

0.822

50k

m g

rid

0.611

0.506

0.303

0.720

0.602

50k

m c

ircl

e 0.630

0.425

0.023

0.211

0.646

100k

m c

ircl

e 0.785

0.620

0.444

0.837

0.818

b) Tax

on

D

elim

itat

ion

m

eth

od

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.677

0.790

0.535

0.319

0.640

10k

m b

uff

er

0.682

0.778

0.611

0.503

0.567

10k

m g

rid

0.632

0.811

0.650

0.183

0.578

50k

m g

rid

0.676

0.778

0.761

0.355

0.495

50k

m c

ircl

e 0.639

0.793

0.566

0.587

0.491

100k

m c

ircl

e 0.567

0.677

0.400

0.328

0.441

Ves

per

tili

on

idae

5k

m b

uff

er

0.641

0.777

0.552

0.543

0.594

10k

m b

uff

er

0.667

0.746

0.839

0.450

0.677

10k

m g

rid

0.639

0.812

0.667

0.679

0.546

50k

m g

rid

0.557

0.737

0.621

0.387

0.334

50k

m c

ircl

e 0.653

0.771

0.834

0.295

0.726

100k

m c

ircl

e 0.541

0.589

0.848

0.254

0.694

Myo

tis

5k

m b

uff

er

0.087

0.312

-0.469

0.565

-0.125

10k

m b

uff

er

0.321

0.429

0.044

0.825

0.183

10k

m g

rid

0.133

0.056

-0.025

0.836

-0.011

50k

m g

rid

0.168

0.319

-0.049

0.645

0.057

50k

m c

ircl

e -0.049

0.107

-0.364

0.550

0.220

100k

m c

ircl

e 0.165

0.271

-0.055

0.708

0.332

177

Tab

le S

17. P

ears

on p

rod

uct

-mom

ent

corr

elat

ion c

oef

fici

ents

fo

r P

CS

and s

kull

MC

S (

a) S

ES

-MP

D a

nd (

b)

SE

S-M

NT

D. G

ray c

ells

indic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5

a)

Tax

on

D

elim

itat

ion

m

eth

od

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.649

0.529

0.647

0.567

0.761

10k

m b

uff

er

0.680

0.518

0.440

0.644

0.635

10k

m g

rid

0.594

0.468

0.545

0.076

0.673

50k

m g

rid

0.748

0.481

0.656

0.608

0.764

50k

m c

ircl

e 0.760

0.709

0.249

0.637

0.755

100k

m c

ircl

e 0.849

0.664

0.698

0.575

0.880

Ves

per

tili

on

idae

5k

m b

uff

er

0.634

0.594

0.667

0.626

0.511

10k

m b

uff

er

0.647

0.579

0.715

0.584

0.499

10k

m g

rid

0.621

0.533

0.674

0.651

0.471

50k

m g

rid

0.655

0.538

0.693

0.641

0.528

50k

m c

ircl

e 0.646

0.750

0.658

0.497

0.318

100k

m c

ircl

e 0.660

0.668

0.722

0.625

0.388

Myo

tis

5k

m b

uff

er

0.523

0.515

0.364

0.789

0.631

10k

m b

uff

er

0.697

0.648

0.331

-0.011

0.756

10k

m g

rid

0.539

0.165

0.634

0.694

0.843

50k

m g

rid

0.637

0.549

0.431

0.717

0.635

50k

m c

ircl

e 0.656

0.474

0.196

0.237

0.656

100k

m c

ircl

e 0.800

0.671

0.328

0.834

0.823

b) Tax

on

D

elim

itat

ion

m

eth

od

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.630

0.645

0.481

0.366

0.604

10k

m b

uff

er

0.615

0.628

0.555

0.518

0.483

10k

m g

rid

0.594

0.695

0.600

0.248

0.503

50k

m g

rid

0.629

0.698

0.732

0.350

0.386

50k

m c

ircl

e 0.638

0.747

0.497

0.616

0.517

100k

m c

ircl

e 0.579

0.678

0.320

0.371

0.465

Ves

per

tili

on

idae

5k

m b

uff

er

0.586

0.685

0.592

0.492

0.508

10k

m b

uff

er

0.597

0.665

0.807

0.266

0.587

10k

m g

rid

0.579

0.723

0.673

0.615

0.450

50k

m g

rid

0.485

0.678

0.639

0.283

0.178

50k

m c

ircl

e 0.600

0.720

0.805

0.192

0.606

100k

m c

ircl

e 0.494

0.571

0.821

0.151

0.621

Myo

tis

5k

m b

uff

er

0.158

0.413

-0.508

0.523

-0.179

10k

m b

uff

er

0.399

0.464

-0.075

0.794

0.386

10k

m g

rid

0.191

0.115

0.307

0.775

-0.059

50k

m g

rid

0.265

0.438

0.102

0.655

0.203

50k

m c

ircl

e 0.058

0.113

-0.095

0.539

0.149

100k

m c

ircl

e 0.258

0.094

-0.022

0.747

0.372

178

Tab

le S

18. P

ears

on p

rod

uct

-mom

ent

corr

elat

ion c

oef

fici

ents

fo

r P

CS

and w

ing M

CS

(a)

SE

S-M

PD

and (

b)

SE

S-M

NT

D. G

ray c

ells

indic

ate

signif

ican

t co

rrel

atio

n w

ith p

-val

ue

<0.0

5.

a)

Tax

on

D

elim

itat

ion

m

eth

od

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.642

0.778

0.798

0.517

0.578

10k

m b

uff

er

0.701

0.749

0.622

0.617

0.450

10k

m g

rid

0.555

0.762

0.729

0.036

0.553

50k

m g

rid

0.736

0.726

0.816

0.602

0.558

50k

m c

ircl

e 0.738

0.864

0.398

0.528

0.526

100k

m c

ircl

e 0.808

0.818

0.642

0.322

0.653

Ves

per

tili

on

idae

5k

m b

uff

er

0.696

0.771

0.795

0.674

0.562

10k

m b

uff

er

0.746

0.753

0.882

0.691

0.528

10k

m g

rid

0.673

0.784

0.785

0.655

0.463

50k

m g

rid

0.757

0.748

0.784

0.725

0.560

50k

m c

ircl

e 0.708

0.860

0.926

0.573

0.253

100k

m c

ircl

e 0.713

0.752

0.675

0.617

0.468

Myo

tis

5k

m b

uff

er

0.290

0.184

-0.476

0.862

0.262

10k

m b

uff

er

0.479

0.135

-0.423

0.576

0.374

10k

m g

rid

0.274

-0.257

-0.302

0.994

0.417

50k

m g

rid

0.357

-0.103

0.111

0.751

0.215

50k

m c

ircl

e 0.365

0.008

-0.375

0.095

0.515

100k

m c

ircl

e 0.646

0.307

0.128

0.704

0.701

b) Tax

on

D

elim

itat

ion

m

eth

od

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

ua

All

5k

m b

uff

er

0.676

0.822

0.594

0.348

0.605

10k

m b

uff

er

0.694

0.757

0.596

0.523

0.653

10k

m g

rid

0.630

0.803

0.666

0.063

0.635

50k

m g

rid

0.704

0.807

0.727

0.343

0.569

50k

m c

ircl

e 0.551

0.778

0.403

0.558

0.310

100k

m c

ircl

e 0.500

0.700

0.140

0.369

0.318

Ves

per

tili

on

idae

5k

m b

uff

er

0.666

0.745

0.619

0.613

0.676

10k

m b

uff

er

0.680

0.714

0.805

0.545

0.742

10k

m g

rid

0.666

0.796

0.667

0.665

0.644

50k

m g

rid

0.665

0.742

0.562

0.686

0.559

50k

m c

ircl

e 0.679

0.796

0.688

0.451

0.724

100k

m c

ircl

e 0.547

0.619

0.687

0.543

0.720

Myo

tis

5k

m b

uff

er

-0.021

0.012

-0.266

0.646

-0.040

10k

m b

uff

er

0.181

0.021

0.454

0.961

0.114

10k

m g

rid

0.120

-0.142

0.448

0.955

0.063

50k

m g

rid

0.031

-0.332

-0.002

0.818

-0.027

50k

m c

ircl

e -0.085

-0.172

-0.399

0.589

0.367

100k

m c

ircl

e 0.091

0.074

-0.020

0.714

0.241

179

a)

Tax

on

D

elim

itat

ion

met

hod

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.012

0.470

0.605

0.096

0.047

10k

m b

uff

er

0.008

0.389

0.550

0.104

0.287

10k

m g

rid

0.008

0.000

0.635

0.034

0.165

50k

m g

rid

0.005

0.483

0.192

0.009

0.344

50k

m c

ircl

e 0.001

0.126

0.663

0.057

0.182

100k

m

circ

le

0.004

0.545

0.264

0.164

0.267

Ves

per

tili

on

idae

5k

m b

uff

er

0.586

0.616

0.533

0.474

0.561

10k

m b

uff

er

0.565

0.541

0.538

0.439

0.842

10k

m g

rid

0.472

0.802

0.613

0.244

0.616

50k

m g

rid

0.520

0.618

0.188

0.307

0.692

50k

m c

ircl

e 0.486

0.334

0.543

0.616

0.860

100k

m

circ

le

0.528

0.795

0.349

0.431

0.600

Myo

tis

5k

m b

uff

er

0.023

0.515

0.745

0.114

0.327

10k

m b

uff

er

0.018

0.299

0.844

0.070

0.659

10k

m g

rid

0.041

0.705

0.808

0.127

0.419

50k

m g

rid

0.129

0.588

0.722

0.384

0.264

50k

m c

ircl

e 0.176

0.512

0.811

0.338

0.501

100k

m

circ

le

0.067

0.518

0.458

0.650

0.339

b) Tax

on

D

elim

itat

ion

met

hod

All

d

eser

ts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.002

0.200

0.264

0.076

0.058

10k

m b

uff

er

0.001

0.073

0.360

0.110

0.253

10k

m g

rid

0.004

0.000

0.267

0.003

0.416

50k

m g

rid

0.000

0.100

0.282

0.008

0.231

50k

m c

ircl

e 0.001

0.033

0.610

0.073

0.177

100k

m

circ

le

0.000

0.096

0.499

0.125

0.062

Ves

per

tili

on

idae

5k

m b

uff

er

0.293

0.384

0.302

0.190

0.456

10k

m b

uff

er

0.205

0.107

0.228

0.490

0.643

10k

m g

rid

0.283

0.533

0.285

0.071

0.714

50k

m g

rid

0.107

0.194

0.189

0.032

0.703

50k

m c

ircl

e 0.171

0.210

0.072

0.178

0.785

100k

m

circ

le

0.073

0.372

0.122

0.215

0.156

Myo

tis

5k

m b

uff

er

0.195

0.475

0.732

0.090

0.184

10k

m b

uff

er

0.453

0.393

0.350

0.108

0.594

10k

m g

rid

0.326

0.504

0.605

0.329

0.808

50k

m g

rid

0.431

0.590

0.479

0.328

0.467

50k

m c

ircl

e 0.080

0.822

0.331

0.026

0.114

100k

m

circ

le

0.058

0.715

0.282

0.146

0.267

Clustered

(sig.)

Clustered

(ns)

Not

significant

Overdispersed

(ns)

Overdispersed

(sig.)

Fig

ure

S1:

All

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

p-v

alues

for

all

spec

ies

pools

and d

elim

itat

ion m

ethods

for

skull

and w

ing d

ata

com

bin

ed, co

lor-

coded

by s

ignif

ican

ce. (

a) S

ES

-MP

D r

esult

s. (

b)

SE

S-M

NT

D r

esult

s.

180

a)

Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.015

0.542

0.551

0.122

0.051

10k

m b

uff

er

0.007

0.430

0.479

0.123

0.177

10k

m g

rid

0.011

0.843

0.431

0.136

0.223

50k

m g

rid

0.004

0.524

0.140

0.009

0.264

50k

m c

ircl

e 0.002

0.205

0.597

0.076

0.232

100k

m

circ

le

0.004

0.580

0.207

0.156

0.284

Ves

per

tili

on

idae

5k

m b

uff

er

0.528

0.691

0.537

0.451

0.443

10k

m b

uff

er

0.486

0.606

0.495

0.478

0.767

10k

m g

rid

0.390

0.840

0.341

0.376

0.518

50k

m g

rid

0.389

0.628

0.142

0.343

0.573

50k

m c

ircl

e 0.434

0.401

0.402

0.603

0.742

100k

m

circ

le

0.405

0.772

0.291

0.420

0.522

Myo

tis

5k

m b

uff

er

0.019

0.644

0.733

0.081

0.234

10k

m b

uff

er

0.019

0.352

0.909

0.080

0.616

10k

m g

rid

0.026

0.786

0.720

0.145

0.255

50k

m g

rid

0.114

0.644

0.872

0.400

0.229

50k

m c

ircl

e 0.170

0.598

0.921

0.356

0.409

100k

m

circ

le

0.061

0.589

0.552

0.661

0.312

b) Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.001

0.323

0.187

0.041

0.020

10k

m b

uff

er

0.000

0.086

0.324

0.066

0.123

10k

m g

rid

0.005

0.639

0.172

0.007

0.378

50k

m g

rid

0.000

0.123

0.342

0.003

0.182

50k

m c

ircl

e 0.001

0.091

0.602

0.051

0.143

100k

m

circ

le

0.000

0.170

0.420

0.085

0.046

Ves

per

tili

on

idae

5k

m b

uff

er

0.233

0.517

0.204

0.152

0.360

10k

m b

uff

er

0.175

0.182

0.285

0.458

0.593

10k

m g

rid

0.236

0.730

0.091

0.053

0.648

50k

m g

rid

0.071

0.226

0.241

0.011

0.628

50k

m c

ircl

e 0.231

0.285

0.102

0.130

0.674

100k

m

circ

le

0.082

0.421

0.145

0.246

0.161

Myo

tis

5k

m b

uff

er

0.070

0.645

0.695

0.062

0.089

10k

m b

uff

er

0.354

0.466

0.471

0.075

0.721

10k

m g

rid

0.174

0.731

0.491

0.355

0.464

50k

m g

rid

0.340

0.716

0.587

0.434

0.423

50k

m c

ircl

e 0.107

0.658

0.423

0.116

0.262

100k

m

circ

le

0.057

0.616

0.331

0.143

0.237

Clustered

(sig.)

Clustered

(ns)

Not

significant

Overdispersed

(ns)

Overdispersed

(sig.)

Fig

ure

S2:

All

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

p-v

alues

for

all

spec

ies

pools

and d

elim

itat

ion m

ethods

for

skull

dat

a, c

olo

r-co

ded

by

signif

ican

ce. (

a) S

ES

-MP

D r

esult

s. (

b)

SE

S-M

NT

D r

esult

s.

181

a)

Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.011

0.212

0.617

0.097

0.123

10k

m b

uff

er

0.019

0.149

0.569

0.130

0.716

10k

m g

rid

0.003

0.263

0.618

0.003

0.115

50k

m g

rid

0.016

0.332

0.351

0.027

0.498

50k

m c

ircl

e 0.001

0.051

0.706

0.133

0.161

100k

m

circ

le

0.010

0.411

0.335

0.336

0.264

Ves

per

tili

on

idae

5k

m b

uff

er

0.694

0.409

0.638

0.322

0.881

10k

m b

uff

er

0.566

0.367

0.511

0.385

0.921

10k

m g

rid

0.608

0.495

0.653

0.073

0.756

50k

m g

rid

0.698

0.517

0.372

0.347

0.878

50k

m c

ircl

e 0.504

0.289

0.390

0.665

0.867

100k

m

circ

le

0.766

0.790

0.381

0.421

0.720

Myo

tis

5k

m b

uff

er

0.181

0.160

0.582

0.099

0.570

10k

m b

uff

er

0.206

0.418

0.446

0.082

0.557

10k

m g

rid

0.217

0.264

0.506

0.057

0.855

50k

m g

rid

0.560

0.491

0.517

0.434

0.635

50k

m c

ircl

e 0.478

0.739

0.790

0.164

0.510

100k

m

circ

le

0.629

0.800

0.569

0.769

0.715

b) Tax

on

Del

imit

atio

n

met

hod

All

des

erts

Gre

at

Bas

in

Moja

ve

Son

ora

n

Chih

uah

uan

All

5k

m b

uff

er

0.004

0.038

0.386

0.165

0.141

10k

m b

uff

er

0.010

0.030

0.531

0.180

0.693

10k

m g

rid

0.002

0.096

0.394

0.006

0.197

50k

m g

rid

0.002

0.063

0.183

0.058

0.277

50k

m c

ircl

e 0.005

0.013

0.763

0.246

0.370

100k

m

circ

le

0.007

0.174

0.461

0.298

0.178

Ves

per

tili

on

idae

5k

m b

uff

er

0.316

0.154

0.533

0.109

0.703

10k

m b

uff

er

0.239

0.129

0.381

0.262

0.803

10k

m g

rid

0.332

0.295

0.555

0.042

0.774

50k

m g

rid

0.242

0.306

0.105

0.092

0.707

50k

m c

ircl

e 0.366

0.232

0.169

0.344

0.785

100k

m

circ

le

0.326

0.545

0.253

0.357

0.251

Myo

tis

5k

m b

uff

er

0.397

0.073

0.705

0.058

0.336

10k

m b

uff

er

0.632

0.608

0.271

0.119

0.316

10k

m g

rid

0.402

0.084

0.423

0.146

0.702

50k

m g

rid

0.456

0.383

0.289

0.191

0.425

50k

m c

ircl

e 0.085

0.860

0.246

0.013

0.044

100k

m

circ

le

0.079

0.921

0.223

0.423

0.086

Fig

ure

S3:

All

Fis

her

’s c

om

bin

ed p

robab

ilit

y t

est

p-v

alues

for

all

spec

ies

pools

and d

elim

itat

ion m

ethods

for

win

g d

ata,

colo

r-co

ded

by s

ignif

ican

ce. (

a) S

ES

-MP

D r

esult

s. (

b)

SE

S-M

NT

D r

esult

s.

Clustered

(sig.)

Clustered

(ns)

Not

significant

Overdispersed

(ns)

Overdispersed

(sig.)

182

VITA

Lorelei Patrick is originally from Lyle, Washington. She received her Associate of Arts degree

from Columbia Gorge Community College in The Dalles, Oregon in 2000 then transferred to

Portland State University in Portland, Oregon to earn her Bachelor of Science degree in Biology

in 2003. She continued at Portland State University earning a Master of Science degree in

Biology in 2007. While working on her MS, she was a research assistant on several projects

including small mammal trapping to record prevalence of Hanta virus, marine mammal

necropsies with the Stranding Network, bat surveys for the Forest Service, museum curatorial

assistant, and working in a conservation genetics lab. In 2008 she began her PhD program at

Louisiana State University where she was a research assistant trapping rodents in the Mojave

Desert and a teaching assistant for several semesters.

Documents

PHYLOGENETIC AND MORPHOLOGICAL COMMUNITY …myweb.ttu.edu/richstev/pubs/Dissertations/PatrickLoreleiDissertation... · Dr. Luis Ruedas at the Portland State University Museum of Vertebrate