UNLV Retrospective Theses & Dissertations

1-1-1994

Systematics and biogeography of the Great Basin pocket mouse, Perognathus parvus

Carolyn Sue Ferrell University of Nevada, Las Vegas

Follow this and additional works at: https://digitalscholarship.unlv.edu/rtds

Repository Citation Ferrell, Carolyn Sue, "Systematics and biogeography of the Great Basin pocket mouse, Perognathus parvus" (1994). UNLV Retrospective Theses & Dissertations. 446. http://dx.doi.org/10.25669/9m3y-u5z4

This Thesis is protected by copyright and/or related rights. It has been brought to you by Digital Scholarship@UNLV with permission from the rights-holder(s). You are free to use this Thesis in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s) directly, unless additional rights are indicated by a Creative Commons license in the record and/ or on the work itself.

This Thesis has been accepted for inclusion in UNLV Retrospective Theses & Dissertations by an authorized administrator of Digital Scholarship@UNLV. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the qualify of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely afreet reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

A Bell & Howell Information Com pany 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600

Systematics and Biogeography of the Great Basin Pocket Mouse, Perognathus parvus

by

Carolyn Sue Ferrell

A thesis submitted in partial fulfillment of the requirements for the degree of

Master of Science

in

Biology

Department of Biological Sciences University of Nevada, Las Vegas May 1995 DMI Number: 1374875

UMI Microform 1374875 Copyright 1995, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103 The thesis of Carolyn S. Ferrell for the degree of Master of Science in Biology is approved.

Chairperson, Brett R. Riddle, Ph.D.

Examining Committee Member, Peter Starkweather, Ph.D. //

Examining Committee Member, Stanley Smith, Ph.D.

Graduate faculty Representative, Rodney Metcalf, Ph.D.

Interim Dean of the Graduate College, Cheryl Bowles, Ed . D.

University of Nevada, Las Vegas May 1995 ABSTRACT

The Great Basin pocket mouse, Perognathusparvus, has a distribution that extends across three geological regions: Columbia Plateau, Great Basin and Colorado

Plateau. Limited ability of P. parvus to disperse over large areas and across areas of unsuitable habitat provide a foundation for studying biogeographic vicariance hypotheses related to its distribution. Vicariance hypotheses involving late Tertiary geotectonic events and dispersal hypotheses of Pleistocene glacial-interglacial climatic cycles that may be responsible for phylogeographic structure in P. parvus were tested using mitochondrial

DNA (mtDNA) sequences and Restriction Fragment Length Polymorphisms (RFLPs).

High levels of sequence divergence support the hypothesis of late Tertiary isolation of

Columbia Plateau from Great Basin populations of P. parvus due to late Miocene uplift of the Blue Mountains. Other species exhibit distributions similar to P. parvus and may show the same type of phylogeographic disjunction between populations from the Columbia

Plateau and populations from the Great Basin. Table of Contents

ABSTRACT ...... iii

LIST OF TABLES ...... v

LIST OF FIGURES...... vi

ACKNOWLEDGEMENTS...... vii

CHAPTER 1 INTRODUCTION ...... 1 Perognathus parvus...... 5 Fossil Record of Heteromyids and Geologic History ...... 9 Molecular Data ...... 14

CHAPTER 2 MATERIALS AND METHODS ...... 17 Specimen Collection ...... 17 DNA Extraction ...... 17 Polymerase Chain Reaction ...... 19 DNA Sequences ...... 20 Restriction Fragment Length Polymorphisms ...... 21

CHAPTER 3 RESULTS ...... 26 Sequence Results ...... 26 RFLP Results ...... 31

CHAPTER 4 DISCUSSION ...... 46 Future Research...... 52 Conclusions ...... 53

CHAPTER 5 Perognathus alticola and P. xanthonotus ...... 55

APPENDIX I ...... 60

BIBLIOGRAPHY...... 61 LIST OF TABLES

Table 1 Karyotype characteristics of P. p a r v u...... s 10 Table 2 Summary of nucleotide base substitutions ...... 27 Table 3 Genetic distance matrix from mtDNA sequences ...... 30 Table 4 Composite haplotypes of specimens from the Great Basin/Colorado Plateau . . 32 Table 5 Composite haplotypes of specimens from the Columbia Plateau ...... 35 Table 6 Matrix of genetic distances and standard errors: Great Basin/Colo. Plat 37 Table 7 Matrix of genedc distances and standard errors: Columbia Plateau ...... 41 Table 8 Haplotype and nucleotide diversity values ...... 44 Table 9 Monte Carlo simulation X2 values and P-values ...... 45 Table 10 Karyotype characteristics of P. alticola and P. xanthonotus...... 56

v LIST OF FIGURES

Figure 1 Distribution of sagebrush-steppe and grassland ...... 6 Figure 2 Distribution of Perognathus parvus ...... 7 Figure 3 Phylogenetic tree from Riddle (1995) ...... 28 Figure 4 Estimated phylogeny of exemplars from Riddle (1995) ...... 29 Figure 5 Distribution of haplotypes ...... 36 Figure 6 UPGMA dendrogram of Great Basin/Colorado Plateau haplotypes ...... 38 Figure 7 UPGMA dendrogram of Columbia Plateau haplotypes ...... 40 Figure 8 Distribution of haplotypes and composite populations ...... 43 Figure 9 Distribution of Perognathus alticola and P. xanthonotus...... 57 ACKNOWLEDGEMENTS

This study was conducted from January 1992 to January 1995.1 would like to thank the members of my graduate committee, Dr. Brett Riddle, Dr. Peter Starkweather,

Dr. Stan Smith and Dr. Rodney Metcalf, for providing input on this thesis and for taking the time to be on my committee. I would also like to thank Jeff Drumm, Bruce Ferrell and Tyler Ferrell for helping me in the field, David Nickle and David Orange for laboratory assistance as well as useful discussions about the research, and I would especially like to thank Jeff Drumm for moral support throughout the progress of this research. Thanks to CMBS and BBMilan for their hospitality and good food.

This research was funded by NSF grant BSR-9107520 awarded to Dr. Brett

Riddle and other grants awarded to Carolyn Ferrell by the Graduate Student Association of the University of Nevada, Las Vegas; a Theodore Roosevelt Memorial Fund Grant from the American Museum of Natural History; a Summer Research Enhancement Grant and a Travel Grant both from the National Science Foundation, EPSCoR Women in

Science.

Approval for experimentation on vertebrate animals was obtained from the

Institutional Animal Care and Use Committee of the University of Nevada, Las Vegas

(Protocol # R701-1290-051). Chapter 1

Introduction

Systematics is defined as "the scientific study of kinds of diversity of organisms and of any and all relationships among them" (Simpson, 1961). A major task of systematics is to classify organisms in a logical fashion and determine through comparison what the unique properties of taxa are and also what properties taxa have in common. Included in this task is the attempt to recover phylogenetic relationships among organisms and to reconstruct a phylogeny based on these unique and common properties. Phylogenetic systematics (cladistics-Hennig, 1966) uses branching points on a phylogenetic tree to explain the evolutionary process of speciation (cladogenesis) while ignoring anagenesis (Mayr and Ashlock, 1991). Cladistics has become the major approach to recovering phylogenetic relationships among organisms (Brooks, 1981;

Brooks, 1990; Brooks and McLennan, 1991; Brooks and Wiley, 1985; Crisci and

Stuessy, 1980; Farris, 1983; Felsenstein, 1983; Miyamoto and Cracraft, 1991; Nelson and Platnick, 1981; Wiley, 1981; Wiley et al., 1991).

Systematics has uses that are important not only in the recovery of phylogenies, but in other fields of biology as well by providing a classification scheme for populations, species and higher taxa used by all other scientists in biology. Without this classification scheme, it would be impossible to explain interactions that-are studied in

1 ecology and behavior (Brooks and McLennan, 1991).

The nature of species has been a controversial subject for many years. Early

definitions of species used similarity of types as the criterion. Later, Buffon developed

the idea of sterility as a basis for species (Mayr and Ashlock, 1991). His idea then

became the principle on which the Biological Species Concept (BSC) was developed.

The BSC requires that species be made up of reproductive communities that are isolated

from one another. A major problem with this concept is that it does not allow for a test

for reproductive barriers in allopatric populations. A more fundamental problem is that

reproductive isolation is not necessarily concordant with phylogenetic relationships

(McKitrick and Zink, 1988).

Other concepts that have been developed have not used a biological criterion but

have been based on types. The Typological Species Concept (TSC) states that there are

several universal types. This concept does not reflect relationships among organisms,

and it does not take into account phenotypic plasticity or sexual dimorphism within a

species. The Nominalistic Species Concept (NSC) states that only individuals exist and

that species are abstractions created by people. The NSC precludes any type of

classification, and is therefore useless to biologists trying to determine patterns of

behavior or distribution (Mayr and Ashlock, 1991).

The species concepts that have become most useful to systematists are the

Evolutionary Species Concept (ESC) and the Phylogenetic Species Concept (PSC)

(Avise, 1994; Avise and Ball, 1990; Cracraft, 1983). The ESC states that a species is a lineage that is evolving separately from others and has it's own unitary evolutionary role 3

and tendencies. This concept does not use reproductive criteria and so can include

allopatric populations. However, it does not take into account genetic variation among

populations. Such populations would be considered separate lineages and therefore

separate species. The PSC defines a species as a monophyletic group composed of "the

smallest diagnosable cluster of individual organisms within which there is a parental

pattern of ancestry and descent" (Cracraft, 1983). A monophyletic group is one in which

all members can be traced back to a single most recent common ancestor. This definition

seems to be the one most accepted by phylogenetic systematists. It too has problems, but

the requirement of monophyly is a fundamental component of cladistic principles. The

issue of monophyly involves the problem of determining what type of evidence is

required to diagnose a monophyletic group. Shared, derived characters, or

synapomoiphies, are the only type of characters that can be used to define a

monophyletic group. Shared ancestral characters can not be used in the construction of a

phylogeny because the common ancestor of the clade may not truly be the most recent

common ancestor.

Biogeography is the study of distributions of organisms both past and present

(Brown and Gibson, 1983). Biogeographers attempt to understand and describe patterns of distribution, limits to distribution, and the roles of climate and topography in determining the distribution of an organism (Brown and Gibson, 1983; Vrba, 1992).

Early biogeographers noticed several basic distributional patterns, including: the observation that all taxa are endemic or restricted to a given area; that certain kinds of organisms tend to occur together; and that similar kinds of organisms sometimes occur in widely separated areas (Brown and Gibson, 1983). The underlying goal of modem

biogeography is to explain the causes of these patterns. Several scientists have made

contributions to the task of explaining distributions, most notably Charles Darwin.

Darwin (1859) made observations of biological variation between organisms from

separate islands in the Galapagos. He explained that this variation was due to natural

selection on certain traits that were advantageous to individual populations on different islands. A. R. Wallace observed gaps in the distributions of animals in Southeast Asia.

As he got further south, the fauna changed and became more like the fauna in Australia

(Brown and Gibson, 1983). Both Darwin's and Wallace's observations involve a dispersal explanation to account for the distributional patterns.

Explanations of biogeographic patterns became more robust in the 1960s and

1970s due to the acceptance of the theory of plate tectonics and continental drift (Brown and Gibson, 1983). Acceptance of this theory made it possible for distributions to be explained by vicariance events rather than dispersal. Vicariance events (i.e. movements of tectonic plates, orogeny, river formation, desertification, glaciation) isolate once continuous populations from each other, thus preventing gene flow between the two new populations. Prolonged isolation will eventually lead to a unique evolutionary trajectory for each of these populations and may result in speciation via an allopatric or peripatric mode.

Phylogeography involves the study of the distribution of geographically localized genealogical lineages and the processes governing those distributions (Avise, 1994).

Nuclear genotypes and mtDNA haplotypes are geographically localized in several 5

species. In general, differences in the ability of organisms to disperse and fragmentation

of suitable habitat exert strong influences on patterns of mtDNA phylogeographic

structure, and localization of mtDNA haplotypes often reflects limited gene flow among populations (Avise, 1994). Intrinsic factors influencing distributional patterns of mtDNA haplotypes also include rates of evolution. Rates of evolution can significantly affect the amount of diversity and divergence in a taxon. MtDNA evolves at a rate faster than most other single copy DNA (Vawter and Brown, 1986), but within the mtDNA genome, different genes are evolving at different rates (Riddle, 1995). Therefore, sampling of different gene regions or different genomes may reveal different patterns of diversity and divergence within a single taxon.

Perognathus parvus

The genus Perognathus (family Heteromyidae, subfamily Perognathinae) consists of arid-adapted rodents that are endemic to North America and whose range extends throughout the desert shrub-steppe and grassland-steppe communities of the west (Figure

1) (Hall, 1981; Schmidley et al., 1993). Perognathus parvus is a geographically widespread species that is distributed throughout the Great Basin, Columbia Plateau and onto the Colorado Plateau (Figure 2).

The Perognathus parvus species group was originally thought to comprise the species Perognathus parvus, P. alticola, and P. xanthonotus (Osgood, 1900; Williams,

1978) However, it is now postulated that P. xanthonotus is a subspecies of P. parvus based on biochemical data (Sulentich, 1983; Williams, 1993).

The Great Basin pocket mouse was originally described by Peale in 1848 as 6

CO O CO Figure Figure 1. Distribution of sagebrush-steppe and grassland in western North America. 7 Great Great Basin, and onto the Colorado Plateau. 8

Cricetodipus parvus, based on a holotype from Oregon (Merriam, 1889). It has

subsequently had a complex nomenclatural history. Baird distinguished Perognathus

from Cricetodipus based on body size and whether or not there was fur on the soles of

the feet. Osgood (1900) assigned the name Perognathus parvus based on Peale's

description and the fact that the Wilkes expedition, of which Peale was a member, and

the location where Baird caught his specimen were determined to be the same.

Most research conducted on Perognathus parvus has consisted of ecological and

physiological studies (Cramer and Chapman, 1990; French, 1993; Kenagy and Banes,

1984; O'Farrell et al., 1975; Price and Brown, 1983; Small and Verts, 1983; Verts and

Carraway, 1986). Patton (1967) and Williams (1978) studied the karyological affinities of

P. parvus; but did not formally propose inter- or intraspecific relationships. No studies

have addressed the historical biogeography of this widespread species.

Patton (1967) first described the karyotype of P. parvus without suggesting any phylogenetic hypotheses about it's evolutionary history relative to other members of the genus Perognathus or intraspecific relationships among subspecies. Williams (1978) examined the karyotypes of several subspecies of P. parvus as well as P. alticola and P. xanthonotus, which were at the time considered closely related sister taxa and members of the Parvus species group. He found that P. parvus columbianus from the Columbia

Basin had a karyotype that was extremely divergent from the karyotypes of P. parvus subspecies that had been collected in the Great Basin (Table 1). The high fundamental number (FN) and a submetacentric rather than an acrocentric Y chromosome in P. p_arvus columbianus but not other subspecies of P. parvus represent derived traits, and indicate 9

Table 1. Karyotype characteristics of Perognathus parvus spp. from the Columbia Plateau and Great Basin/Colorado Plateau (Williams, 1978). Haplotype of P. p. columbianus is divergent from other subspecies of P. parvus based on high fundamental number and X and Y chromosome. 2N=diploid number, FN=fundamental number, BA=biarmed chromosome, A=uniarmed chromosome, ST=subtelocentric, SM=submetacentric, A=acrocentric.

Autosomal, pairs Subspecies m EM BA AX Y. P. parvus trumbullensis 54 70 8 18 ST A

P. parvus olivaceous 54 74 11 15 ST A

P. parvus clarus 54 76 12 14 STA

P. parvus columbianus 54 106 0 26 SiM SM 10

independent evolution and divergence of P. parvus columbianus from other subspecies of

P. parvus (Hafner and Hafner, 1983). He also found that P. alticola and P. xanthonotus

had karyotypes that were identical to those found in P. parvus from the Great Basin.

Fossil Record of Heteromyids and Geologic History

Members of the family Heteromyidae first appear in the mid Oligocene fossil record during the Orellan North American Land Mammal Age (NALMA) (Wahlert,

1993). Thus, the systematic and biogeographic history of members of the family should reflect various aspects of geological, climatic, and vegetational history of western North

America during the past 30 million years.

The genus Perognathus also has a long history in North America. Fossil records of Perognathus occur from the Miocene to the Pleistocene in western North America

(Wahlert, 1993). The first appearance in the fossil record is of a specimen from the early

Hemingfordian (mid Miocene) of the John Day formation of northern Oregon (James,

1963). However, it is thought that the subfamilies Heteromyinae and Perognathinae were not yet clearly separated in the fossil record by the early Miocene (Wilson, 1960). A specimen of P. maldei, now extinct, from the late Pliocene of Idaho is believed to most closely resemble modem P. parvus, and it is thought to be ancestral to modem P. parvus, which first appears in the fossil record during the Rancholabrean NALMA (mid

Pleistocene) of Idaho. There are slight differences between the fossils found in the John

Day formation of northern Oregon and the Idaho fossils, but they could be phylogenetically related (Zakrzewski, 1969). 11

The region in which the oldest fossils of Perognathus occur is the Columbia

Plateau (James, 1963; Zakrzewski, 1969). The Columbia Plateau formed as a series of basaltic volcanic plains during the mid to late Miocene, approximately the time corresponding to the fossil record. As basalt accumulated in the Pasco Basin of central

Washington, tectonic deformation was taking place, resulting in a complex folding pattern throughout the Columbia Plateau. In the late Miocene, the longest structure formed during the height of volcanic activity was the Blue Mountains. This mountain chain borders the Columbia Plateau along most of its southern boundary and separates it from the Oregon Plateau. The High Lava Plains arose during the late Miocene and form a geological transition between the Columbia Plateau and Great Basin. It separates an area of declining volcanic activity on the Columbia Plateau from an area of tectonic extension in the Great Basin (Christiansen and McKee, 1977). The Oregon Plateau is part of the

High Lava Plains and forms the northernmost region of the Basin and Range Province. It is bordered on the north by the Blue Mountains and on the south by the Orevada rift

(Carlson and Hart, 1987). Fossil records of Perognathus from the Oregon Plateau are reported from the Barstovian NALMA (mid Miocene). Basaltic volcanism in the Steens

Mountain area of southern Oregon was at its peak during the mid Miocene (Carlson and

Hart, 1987). Volcanism in the Steens Mountain area continued sporadically until approximately 11 mya, and on the southeastern edge of the Oregon Plateau, volcanism has continued until the Holocene. This southeastern edge forms the border between the

Oregon Plateau and the Great Basin. Late Miocene basalts from the Steens Mountain area of southern Oregon are similar in composition to those found in the northern Great Basin, 12

and it is thought that they are related (Carlson and Hart, 1987).

Geologically the Great Basin is relatively young. The Great Basin had

considerable tectonic relief due to late Cretaceous and Tertiary orogeny, but the Basin and

Range topography did not exist until the late Miocene (Christiansen and McKee, 1977).

Volcanism appears to have begun along the northern axis of the Basin and Range province. By 17 to 14 mya the Great Basin already had considerable tectonic relief.

Much of the surface of the region was covered by 14 to 12 mya. Numerous ashflow tuffs of early Miocene age in Nevada indicate that there was little topographic relief comparable to present-day Basin and Range topography (Christiansen and McKee, 1977). Although much of the volcanic activity occurred prior to this time, the Great Basin has remained tectonically active to present day. Areas of seismicity in this region are relegated primarily to the eastern and southwestern edges, the Wasatch and Sierra Nevada ranges, respectively, which form the east and west borders of the Great Basin.

Ancestral members of the lineage giving rise to extant Perognathus parvus were distributed throughout most of the Columbia and Oregon Plateaus by tire mid Miocene

(Wahlert, 1993). Perognathus parvus fossils did not appear in the Great Basin fossil record until the Rancholabrean NALMA (mid Pleistocene). An ancestor to P. parvus may have evolved on the Columbia Plateau and subsequently migrated southward into the

Great Basin prior to the uplift of the Blue Mountains. One would predict that, based on the geological and fossil records, the Blue Mountains uplift presented a significant barrier to dispersal. This barrier would have isolated northern from southern populations as long ago as the late Miocene. The loss of gene flow between populations from these areas 13

would have resulted in separate evolutionary trajectories for northern versus southern populations of P. parvus. Williams' (1978) karyotype data support the hypothesis of a

separation of northern from southern populations.

P. parvus has long been postulated to be a single, widespread species. However, several alternative vicariance hypotheses can be constructed relative to the response of P. parvus to geological and climatic history. If P. parvus was split into two or more geographically isolated populations by geological or climatic perturbations, I would predict phylogeographic disjunction among populations and geographic variation corresponding to geologic formations within the range of P. parvus.

Three alternative biogeographic models may be used to suggest possible geographic mtDNA haplotype variation within and among populations of P. parvus on the

Columbia Plateau and in the Great Basin. First, late Tertiary geotectonic vicariance may be responsible for isolating Columbia Plateau from Great Basin populations of P. parvus.

Second, late Quaternary glacial-interglacial climatic cycles causing expansion and contraction of areas of suitable habitat may be responsible for isolating Columbia Plateau and Great Basin populations. Finally, P. parvus populations on the Columbia Plateau may have shifted their range southward during each Pleistocene glaciation in response to decreased winter temperatures and the southward shift of shrub-steppe habitat followed by northward expansion during interglacials. This model would predict no detectable "deep" phylogenetic splits between Columbia Plateau and Great Basin populations.

In this study, I have addressed the following questions: 1) How is mtDNA variation apportioned among populations of Perognathus parvusl 2) To what extent does 14

geographic variation in mtDNA distribution and divergence differentiate among the three

models presented above?

Molecular Data

The hypotheses stated above were tested using mitochondrial DNA (mtDNA)

sequences and Restriction Fragment Length Polymorphisms (RFLPs). The mitochondrial

genome contains non-recombining, maternally inherited DNA that evolves faster than

other single copy DNA (Vawter and Brown, 1986). Gene arrangement in mtDNA

appears generally stable, and nearly the entire genome is a coding region so that

hypervariable regions such as introns are not part of the data set (Avise, 1994). The

amount of variability in the mtDNA genome is largely due to its high rate of mutation,

thus it provides substantial numbers of characters for comparisons between closely related

species and populations. Many studies have empirically demonstrated geographic variation within and among populations using mtDNA (Avise et al., 1987; Avise et al.,

1988; Avise, 1992; Moritz et al., 1987; Riddle et al., 1993).

Molecular data have several attributes that provide an advantage over morphological data when reconstructing phylogenetic hypotheses. Morphological data represent a small subset of genetic information that can be environmentally plastic and may not reflect the true evolutionary relationships of organisms. Morphological characters in some taxa may not be sufficient for inferring a robust phylogeny, and certain characters can be misleading because they represent special environmentally induced adaptations that may not have a genetic basis. Molecular data, on the other hand, have a distinct 15

advantage over morphological data in that the data are sampled directly from the genome of the organism, thus ensuring that characters are hereditary (Avise, 1994; Hillis, 1987).

Genomes provide an enormous amount of information. The genome codes for proteins and other cellular material, as well as retaining a record of evolutionary change in its nucleotides. Molecular characters can provide an extensive, detailed data set, limited only by the number of nucleotide pairs in the DNA, that permits the reconstruction of a well- resolved phylogeny that can either refute or support a phylogeny that was constructed using morphological characters (Hillis, 1987; Mayr and Ashlock, 1991). Morphological and behavioral characters may often be convergent. For instance, long tails in kangaroo rats (Dipodomys sp.) and jumping mice (Zapiis sp.) are believed to act as a rudder for balance. Because the tail has a similar function in both of these mice, this character might be incorrectly considered homologous. In reality these two taxa are only distantly related.

Because similar molecular characters are unlikely to be convergent, molecular data often allow one to separate homologous characters from convergent ones by separating morphologically similar organisms from one another based on genetic information rather than shared morphological characteristics that may have evolved convergently.

Comparisons of distantly related taxa using morphological characters can be confounded because of the lack of comparable characters between taxa. Molecular data allow for the direct comparison of levels of genetic divergence among nearly any group of organisms. Molecular data can also provide a means for estimating times of divergence based on a molecular clock that is calibrated using fossil material or biogeographic hypotheses, thus molecular characters can add an explicit temporal dimension by 16

calculating the distance between taxa and translating it into a time of divergence between

the taxa (Mayr and Ashlock, 1991; Moritz and Hillis, 1990).

The three models presented above can be related to phylogeographic hypotheses that have derived from studies of mtDNA gene genealogies (Avise, 1994). Large phylogenetic gaps between monophyletic groups usually arise from long-term barriers to dispersal. A late Tertiary geotectonic vicariance would most likely result in a "deep" level of divergence between northern and southern populations of P. parvus. Reciprocal monophyly of northern and southern populations would have occured over time through stochastic lineage sorting. The time of divergence calculated from a molecular clock could be related to a geotectonic vicariance event such as the uplift of the Blue Mountains.

One would expect shallow levels of genetic divergence if barriers to dispersal had not been present for long periods of time. Pleistocene glacial-interglacial expansion and contraction of suitable habitat would result in shallower levels of divergence than would a long-term isolation event from the late Tertiary. Phylogeographic population structure showing shallow levels of divergence would be expected in species with life histories conducive to dispersal or in populations that have few barriers to dispersal. If populations of P. parvus migrated southward during periods of glacial maxima, low levels of divergence would be found corresponding to a Pleistocene time of divergence. Reciprocal monophyly from a

Pleistocene divergence may not have occured in such a short time. Chapter 2

Material and Methods

Specimen Collection

A total of 45 specimens were collected from throughout the range of Perognathus

parvus (Appendix 1). Specimens were caught in Sherman live traps using chicken feed

as bait. Specimens were euthanized using isohalane according to the University of

Nevada, Las Vegas Animal Care and Use protocol. Heart, liver, kidney, and brain tissue

were removed and stored in cryotubes in liquid nitrogen, or they were preserved at

ambient temperature in "Lysis Buffer" (Longmire and Baker, 1994). Tissues are stored

at the University of Nevada, Las Vegas , whereas skins, skeletons and skulls were

prepared as museum specimens and are housed at the Museum of Southwestern Biology,

University of New Mexico.

DNA Extraction

Total genomic DNA was extracted from heart or brain tissue using either of two

DNA isolation procedures. The first DNA isolation protocol, a slightly modified version

of Hillis et al. (1990), involved macerating approximately lOOmg of heart, liver, or brain tissue and adding 500pl of STE buffer (0.1 M NaCl, 0.05 M Tris-HCl, pH 7.5, 0.001 M

EDTA) with 25pl of (10 ug/pl) Proteinase K to a 1.5 ml centrifuge tube. Twenty-five

17 18

(25)pl of 20% SDS (sodium dodecyl sulfate) was then added to lyse cell membranes. The mixture was incubated at 55°C for two hours with occasional mixing. After incubation, an equal volume of phenol:chloroform:isoamyl alcohol (PCI) with a ratio of 25:24:1 was added. The mixture was incubated at room temperature for 5 minutes with regular mixing to prevent phases from separating. Tubes were then microcentrifuged for 5 minutes, and the aqueous layer was removed and saved. This procedure was performed again and followed by twice washing the removed aqueous layer with an equal volume of chloroformrisoamyl alcohol (Cl, 24:1), incubating for 5 minutes at room temperature, microcentrifuging for 3 minutes, and finally removing and saving the aqueous layer. The extracted genomic DNA was then incubated overnight at -20°C in one ml of cold absolute ethanol and 1/10 volume of 2M NaCl to precipitate the DNA. The next day the tubes were centrifuged for five minutes, and the liquid was decanted. The remaining pellet in the bottom of the tube was washed with 500 p. 1 of 80% ethanol by centrifuging for 5 minutes and decanting the ethanol. Once washed, the DNA pellet was vacuum centrifuged for 20 minutes. When the pellet had dried, it was resuspended in 150 to

200 p. 1 of IX TE buffer (0.001 M Tris-HCl, pH 7.5, 0.0001 M EDTA) and stored at 4°C.

The second DNA isolation protocol of Longmire and Baker (1994) was much easier to use in the field and was used for specimens captured during July and August of

1994. Approximately 0.4g of tissue (one entire heart or brain)was macerated in a petri dish and placed in a 15ml polypropylene Coming tube with 5 ml of "Lysis Buffer." The macerated tissue could then be stored indefinitely out of sunlight at ambient temperature. 19

"Lysis Buffer" consists of 0.5 M EDTA, 20% SDS, and Tris-HCl. For extraction, 250|il of (lOOpg/ml) Proteinase K was added to the tubes and rotated overnight at 37°C. After overnight rotation, an equal volume of phenol buffered with TE was added and rotated for another 30 minutes at 37°C. Tubes were then centrifuged at 2000 rpm for five minutes to separate the phases. The aqueous layer was removed and placed in dialysis tubing.

Samples were dialyzed for 24 to 36 hours against three changes of IX TE buffer (0.001 M

Tris-HCl, pH 7.5, 0.0001 M EDTA). The genomic DNA remaining in the dialysis tubing was placed in 15 ml tubes and stored at 4°C.

Polymerase Chain Reaction

The Polymerase Chain Reaction (PCR) was used to amplify target gene regions.

PCR uses oligonucleotide primers to amplify specific sequences of DNA in a cyclic fashion. Template DNA is placed in a tube containing a 10X solution of reaction buffer

(buffers are different for each type of polymerase), lOmM of each oligonucleotide primer,

8 mM of each of four deoxynucleotides (dNTPs) (dATP, dCTP, dGTP, dTTP), and 4u/pl of Vent® DNA polymerase (Hillis et al., 1990). Vent® DNA polymerase has a much lower rate of error in incorporating wrong nucleotides during the primer extension phase of PCR than Taq DNA polymerase does (Catalog reference). The mixture is then overlaid with mineral oil to prevent evaporation. Template DNA is denatured at 95°C for one minute in the first step. In the second step, the oligonucleotide primers anneal to the 5' ends of the denatured DNA at 50 - 60°C for one minute. In the final step, extension of the primers occurs by addition of the individual dNTPs to the 3' end of the primer at 72°C for 2 20

minutes. Each of these steps is repeated for 25 cycles, creating millions of copies of the

original double-stranded target sequence. Repeating this procedure using only one

oligonucleotide primer allows the generation of single-stranded DNA that is suitable for

DNA sequencing.

DNA Sequences

Most of the 784 base pair mt DNA Cytochrome oxidase HI (COIQ) gene was

amplified using PCR. Primers designed by Riddle (unpublished) and identified according

to the 3' base number in the published sequence of Miis (Bibb et al., 1981) are as follows:

L 8618 5' CAT GAT AAC ACA TAA TGA CCC ACC AA -3', H 9323 5'- ACT ACG

TCT ACG AAA TGT CAG TAT CA -3’. A section of this region containing 326 base

pairs was sequenced from three specimens, one from the lower Columbia Plateau and two

from the Great Basin, using the Sanger dideoxy sequencing reaction with Sequenase 2.0

DNA Sequencing kit (Allard et al., 1991; Sanger et al., 1977). Single-stranded DNA is

annealed to one primer (the sequencing primer). This mixture is separated into 4 sub­

samples containing all four deoxynucleotides and a single dideoxynucleotide (ddNTP) that lack a 3' hydroxyl (3'-OH) group. The DNA strand is made radioactive using 35‘S that has

been added to the ddNTPs. Primer extension is catalyzed by DNA polymerase, and sequence extension occurs through the attachment of nucleotides to a free 3'-OH.

Wherever a ddNTP has been incorporated further strand extension is halted (Avise, 1994).

The reaction occurs such that DNA fragments of differing lengths are produced. Each of the four reactions is electrophoresed through a polyacrylimide gel for approximately 5 21

hours and visualized on an autoradiograph. The DNA sequences obtained from the COm

gene region were read into the MacVector program version 3.5 (International

Biotechnologies, Inc., New Haven, CT) for alignment and editing.

Transition substitutions in mtDNA accumulate at a stochastically-linear rate of

divergence initially but because of the high rate of substitution, multiple substitutions at a

site eventually produce a high level of phylogenetic noise. Even if transition substitutions are saturated, transversion parsimony, which uses only transversion substitutions in analysis, can still be employed because transversions accumulate in a linear fashion long after the sequences have diverged beyond the saturation of transitions (Brown, 1985).

Riddle (1995) provided evidence that divergence among most geographically and evolutionary distinct lineages of pocket mice exceeded the point of transition substitution saturation. Therefore transversion sequence divergence was calculated using the Jukes-

Cantor formula (DNADIST in PHYLEP v.3—Felsenstein, 1993). Phylogenetic trees were calculated using the branch and bound algorithm in PAUP version 3.0 (Swofford, 1991) and mixed parsimony analysis, which employs generalized parsimony (transition and transversion base substitutions weighted equally) for 1st and 2nd codon positions and transversion parsimony (transition substitutions excluded) for 3rd positions, with 100 bootstrap iterations. Chaetodipiis hispidiis and C. intermedins were used as outgroups.

Restriction Fragment Length Polymorphisms

Restriction endonucleases cleave DNA at specific sites. The resulting fragment pattern can then be used to infer restriction sites along the sequence of DNA being 22

analyzed (Avise, 1994). The fragments are separated by electrophoresis according to

molecular weight- Differences in fragment size may result from base substitutions within

cleavage sites, additions or deletions of DNA, or sequence rearrangements.

The NADH dehydrogenase 2 (ND2) gene region and a portion of the Cytochrome

oxidase I (COI) gene region containing 2150 base pairs was amplified using PCR with primers designed by Dr. J. C. Patton (Riddle et al., 1993). Primer sequences are: L3880,

5'-TAA GCT ATC GGG CCC ATA CC-3'; and H6033, 5'-ACT TCA GGG TGC CCA

AAG AAT CA-3'. Ten microliters of the amplified product were subjected to restriction endonuclease digestion according to manufacturer's protocols using the following 13 restriction enzymes: Bsp 12861. BslNI, BstUI, Haell. HaelU, Hindi, HindlH. Hinfl,

Mbol. MspI, Rsal. Taql. and Xbal. Digested samples were electrophoresed through a

1.5% agarose gel for approximately 3 hours. Banding patterns were then visualized after staining the gel with ethidium bromide and placing it on an ultra-violet transilluminator.

Polaroid photographs were taken of each gel.

After scoring fragments according to distance travelled on the gel, restriction sites were inferred under assumptions of single-site substitution patterns (Dowling et al., 1990).

Data were analyzed using RESTSITE (Miller, 1991) and REAP (McElroy et al., 1991).

Distance matrices were calculated in RESTSITE using the Jukes-Cantor formula, and were used to generate UPGMA dendrograms. Distance matrices plus standard errors of distances were calculated using the DSE program in REAP. DSE estimates genetic distance values using equations set out by Nei and Tajima (1981). Standard .errors are computed according to Nei and Tajima (1983). 23

Haplotype diversity and nucleotide diversity and divergence were calculated in the

DA program of REAP. Haplotype diversity within populations (h) was calculated using

equation 7 of Nei and Tajima (1981):

h = n(l - Sxj2) / (n - 1),

where x; is the population frequency of the ith haplotype and n is the number of individuals

sampled. In this case X; is the frequency of haplotype (i) occurring in the ND2/COI region

of the mitochondrial genome. Standard error of haplotype diversity (VS11/2) was calculated using equation 8.12 of Nei (1987):

VSI(h)= (2 / ( 2n - 1)) { 2 (2n - 2) [E X;3 - ( E x 3 )2] + 2 x3-(2 Xj2 )2},

where again n is the number of individuals sampled, and X; is the frequency of haplotype (i) occurring in the ND2/COI region of the mitochondrial gene region. Estimated nucleotide diversity within populations (n) was calculated using equation 10.19 of Nei (1987) which corrects for small sample sizes:

n=(nx/nx- 1 ) SgXiXjdg,

where nx is the number of sequences sampled, dy estimates the number of nucleotide 24

substitutions per restriction site between the ith and the jth haplotypes, and x{ and Xj are

the sample frequencies of the ith and jth haplotypes for the population being analyzed.

Nucleotide diversity among populations is calculated using Nei's equation:

^ x y = 2 , ^ yj d^,

where d^- estimates the nucleotide substitutions between the ith haplotype from population

X and the jth haplotype from population Y. Nucleotide divergence between populations

(dA) is calculated using Nei's equation 10.21:

= ^XY ' ( + dY ) / 2.

Geographic heterogeneity in haplotype distributions among populations was estimated using the Monte program of REAP. This program uses a Monte Carlo simulation to analyze the extent of geographic heterogeneity among populations (Roff and

Bentzen, 1989). The program generates, through a large number of randomizations of the data set, the distribution of x2 expected if the null hypothesis of homogeneity were true for the particular data set being studied. Because the frequency of some haplotypes is so low that the expected distribution generated by this statistic may not conform to that of y2. the

Monte Carlo statistic is designated as X2. The probability, P, of obtaining a value of X2 equal to or larger than that obtain from the actual observations is given by the equation: 25

P = n / N,

where n is the number of randomizations that generate the large value of X2, and N is the number of randomized sets. The standard error of P is:

s.e. = [P (1 - P ) / N]1/2. Chapter 3

Results

Sequence Results

Analysis of sequence divergence in the C01H gene indicated that there are at least two divergent mtDNA lineages within P. parvus. Average transversion sequence divergence between specimens from Crook County, Oregon (Columbia Plateau) and two specimens from the Great Basin (Cassia County, Idaho and Nye County, Nevada) was estimated to be 7.3%. A summary of base substitutions among these three specimens is given in Table 2.

Figure 3 depicts a phylogeny that was estimated using data from both the COm gene and the mtDNA cyt b gene (Riddle, 1995). Bootstrap support for the P. parvus node is 84% versus 54% (Figure 4) when exemplars from this data set included only sequence from the COIII gene. Although the P. parvus clade from the smaller data set had lower bootstrap support, P. parvus from the Columbia Plateau did not tend to cluster with any other clade at a similar frequency as it did with the P. parvus clade. Bootstrap support for Columbia Plateau P. parvus clustering with the P. apache clade or the P. longimembris clade was 5% and 25%, respectively. A genetic distance matrix for these sequences is shown in Table 3. Note that distances between Columbia Plateau and Great

26 27

Table 2. Pairwise comparison of nucleotide base substitutions from 326 base pairs of the mtDNA COm gene obtained from 3 specimens of Perognathus parvus. Collection localities, Las Vegas Tissue (LVT) numbers and numbers under which sequences were published in GenBank are indicated. Substitutions are recorded as number of transitions/transversions per codon position.

Codon position Locality LVT GenBank 1st 2nd 3rd

OR: Crook Co./ID: 1921/794 U2 1624 4/1 3/1 37/20 Cassia Co.

OR: Crook Co./NV: 1921/1806 U2 1622 5/1 3/0 36/21 Nye Co.

ED: Cassia Co./NV: 794/1806 U2 1621 1/0 0/0 0/2 Nye Co. 28

CO CO CO CO CO CO CO CO

CO CO

a

to 00

C\

CO C\ C\ evolution evolution by Riddle (1995). Reproduced with permission. 29

CO

CO

-c

§ CQ >>

•U 2?0h a § Table 3. Transversion distance matrix of 326 base pairs of the COIII gene from sequences of several pocket mouse species, calculated using Jukes-Cantor formula in the DNADIST algorithm of PHYLIP. These specimens were used to construct a phylogeny including specimens of Perognathus parvus from the Great Basin and the southern Columbia Plateau. GRB=Great Basin, CLP=Columbia Plateau, HISP=HISPIDUS, APA=APACHE, FASC=FASCIATUS, FLAV=FLAVUS, FLVS=FLAVESCENS, LONG=LONGIMEMBRIS, PARV1=PARVUS_GRB1, PARV2=PARVUS_GRB2, PARV3= PARVUS_CLP. frj > O 2 O < Hi < > Specimen -IISP. APA FASC FLVS 06 PARV2 PARV3

HISPIDUS vo o o APACHE CO CO o r xf o n o o cn m FASCIATUS —'4 1 o CN O T—

FLAVESCENS n o o N O in o O V o o o r”H CO T—< oo o o o r—< 1—4N* cs O C CO LONGIMEMBRIS 0.1296 1 O V o O V *—H o o O V o o O o r—( CN CO cs i—H rt O O r-- r-H T“H

PARVUS_GRB1 0.1068 n n 3 o o r*Hm o\ CN O oo O O O O C o o o r— cs CS CS o o o r- O CO o o r***OO n PARVUS_GRB2 1 0.0950 O V t O O oo oo o o CO o O o o m o o t O o vo o r- r—Hy—4in i—H r-* i—t -H —H n PARVUS_CLP n 0.1073 0.0712 xf cs i 00 O V O V r-» N" o 1 o 1—

Basin P. parvus (x = 7.3) is clearly less than average distances between either Columbia

Plateau or Great Basin P. parvus and the Flavus group and Longimembris group (x =

11.4, 10.3) or Fasciatus group (x = 10.6) representatives.

RFLP Results

Restriction Fragment Length Polymorphism (RFLP) analysis revealed eighteen mtDNA haplotypes (7 from the Columbia Plateau and 11 from the Great Basin/Colorado

Plateau). The COIII sequence divergence among samples of P. parvus support a high level of divergence between Columbia Plateau and Great Basin/Colorado Plateau haplotypes (Table 2). Restriction site analysis methods employed in the study were unable to discern specific site gains and losses between haplotypes from the Columbia Plateau relative to haplotypes from the Great Basin/Colorado Plateau because double enzyme digestions were not performed on restriction fragments. Additionally, sites that could be scored as shared between regions could not be assumed to be uniquely derived due to high probabilities of multiple losses and gains. Therefore, the Columbia Plateau and Great

Basin haplotypes were analyzed as separate restriction-site data sets (Tables 4 and 5).

Distribution of mtDNA RFLP haplotypes is shown in Figure 5.

A UPGMA dendrogram of haplotypes from the Great Basin/Colorado Plateau, based on the genetic distance matrix in Table 6, is shown in Figure 6. The most common haplotype found among Great Basin/Colorado Plateau samples was number 1 (Figure 5); this haplotype was found in 15 specimens from five localities in central to southern

Oregon and from one specimen located in north-central Nevada (55% of the total). Table 4. Composite haplotypes of specimens from the Great Basin/Colorado Plateau used in restriction site analysis. Enzyme class (Enz.cl), enzyme, and Las Vegas Tissue (LVT) are indicated. Composite haplotype (COMP) numbers are identified in the last column. d c f Tf Tf n Tf n Tf TT © C i s . 5.33 4.67 5.33 5.33 © >-4 l 1 i OO -4 > 1 >— < 1 t— »— ( a T—^ 3 s < > H COMP N O - ~ r m < < < < < PQ < pq t < c < < 4 “ O C o < < o u < < —4 r < < PQ < po < < < < < cs 0 0 o O N < < < < < < U < < < < < cn qpq pq < < < h O O o < pq < PQ —i i— - ~ C < < < < < < St < < o o QPQ PQ < < < < < < T—H O C < ■ < < < < < < NO < < O C —4 t— f N • < < < < < < r-H < < < < O QPQ PQ < < < i—H < < < < < < < O O O N < < < —t T— < < < < < < < WO N O O C < < < cs < < PQ QP PQ PQ PQ < “• 1“ PQ <

1935 n < < < < < < < < < < PQ < < < < < < m < < < 1937 < < < < < qpq pq < pq pq CQ 1938 < < < < pq pq pq 1939 < < < < W < < < < < < < < < pq < < < < 1940 »—H < < qpq pq < pq pq < 1941 < < < < < < < < < < < pq < 1942 < pq < < < < < < < < < pq pq < 1943 < < < < qpq pq pq pq < < < < 1944 < < < < < < < < < < qpq pq pq pq < < 1945 ' < < < < pq < 1946 < < 33 Table 4 Continued. Composite haplotypes of specimens from the Great Basin/Colorado Plateau used in restriction site analysis. Enzyme class (Enz.cl), enzyme, and Las Vegas Tissue (LVT) are indicated. Composite haplotype (COMP) numbers are identified in the last column. VO D V no f T > l f T T T f T rr f T f T Enz.cl 5.33 5.33 5.33 f T 1 1 z | t—< a 1 S f O O SO 3 > cl 1 H H i n c I I | COMP N O f T < < < < m m m < < < r-H - r < < m < < o o O f T < < < < < < < m < < < « n N O on f T < < < < < —( T— m < < m < < < s O -H 1- n i T—i < < < < < < < < < < m m m m < < m < O m s c < < < < < < m n N O n > < - t < < < < < < < < < < —• T— 34 Table 5. Composite haplotypes of specimens from the Columbia Plateau used in restriction site analysis. Enzyme class (Enz.cl), enzyme, and Las Vegas Tissue (LVT) are indicated. Composite haplotype (COMP) numbers are identified in the last column.. w f T f T f T v© f T f T vo f T f T c 5.33 4.57 5.33 5.33 i oo VO Z >-H l-H 1 1 P 1 LVT# ri Hnelll HincII HinrlHI Mbol XhnI COMP o o o z z z z Z Z z z z cl z z CS o o r-H

1915 , o z z o o CO O Z CL, Z Z Cu 1916 r-H o o z z z z z o z z o z A ~ r f T 1917 o o o o o o o o z z o z z z z Z wo z z z o o o o o o 1918 r-H z z z z z o z z z z z wo 1919 z t—H z O o o z cs r-H 1920 h z CL, z CL, VO z z P o i P o z z r-H 1921 a O CL, Z Z P P z VO Z Z P T— i P a a a 1950 H o z z z z z o z z z z o i P Z c~~ 1953 r-H o z o z z z z z pi z a z 1954 r-H r~ CTs vn i z o o O' o a a z z o o o OO z z y n *H

o a o o z o o z o o oo z z z z z 1956 r-H o a z z a z O OO z 1958 r-H 35 36

c

a CQ c3 o ~ - 3 2 «

C GO o S. S2 o CS rtCL <3 X i Cl 3 2 2 "S rt S E OCXO TD O L. rt ,5° Ot-. U_ o O u CO

*> r9 3 ^ »CO TDc Uh rt Table 6. Matrix of genetic distances and standard errors (Nei and Tajima, 1981) from southern populations of P. parvus. Genetic distances among eleven composite mtDNA haplotypes are shown below the major diagonal, and standard errors are shown above the major diagonal. d c - ffi / i V/“ m 1—1 o wr n s r H r H r T X s %

10 n o O o o O o o m p O O O o p H r H r o o o o o T—H f T T—H O t o cs o p o cs o m O T— O o o f T O C “H T“ ) o i ■ T— H - H H n n n 37 38 matrix matrix calculated using equations of Nei (1987) in RESTSITE (Miller, 1991). Figure Figure 6. UPGMA dendrogram of Great Basin/Colorado Plateau mtDNA haplotypes generated from genetic distani Haplotype number 10 was the next most closely related to haplotype 1 and was found

together with haplotype 1 in southeastern Oregon. Haplotype 7 clustered with haplotypes

1 and 10 but was found in a population that was located approximately 120 miles from the

nearest haplotype 1. Haplotype 7 was found in conjunction with one other haplotype,

number 8,with which it did not cluster and to which it was not found to be most closely

related. Haplotypes 6 and 9, from east-central and northeastern Nevada, respectively,

clustered together with approximately 0.25% sequence divergence between them.

Haplotypes 2 and 11 also clustered with 6 and 9 and were from the Nevada Test Site and central Nevada, respectively; haplotype 8, from the same locadon as haplotype 7, clustered outside this group. The most divergent haplotype was number 5. This was found in specimens collected from south-central Utah on the Colorado Plateau. It was found to be approximately 2.6% divergent from the nearest haplotypes, numbers 3 and 4, which were both found on the Nevada Test Site.

Eight composite haplotypes were found from six localities from the Columbia

Plateau specimens. The UPGMA dendrogram (Figure 7), constructed from the distance matrix in Table 7, showed a distinct demarcation between specimens collected in Oregon and those collected in Washington. Haplotype 18, which clustered with 16, was found in eastern Oregon. Both of the localities from which these specimens were collected were along the southernmost border of the Columbia Plateau. This cluster of haplotypes was found to be approximately 8.1% divergent from other haplotypes on the Columbia

Plateau. The five haplotypes found in Washington all clustered relatively closely with one another. Haplotypes 12 and 13 were both from the same locality, and they clustered most 40

cd CM 6 T—

ao CO in co o f g e x ’-1

< s z 2 co Q w o c W

ca cn B H d cn ^ a inCO lo .s o s c* o ■s So o , >w‘' 0 *53 s z o cd d— N. & ° O 2 g -c o -zs o O cd "O 3 < ST S ep O -S CO C L 2} o -o r - OJ d> cd j j Ll o Table 7. Matrix of genetic distances and standard errors (Nei and Tajima, 1981) from northern populations of P. parvus. Genetic distances among seven composite mtDNA haplotypes are shown below the major diagonal, and standard errors are shown above the major diagonal. r—i o v T—4 cn r r ^4 i—4 t" fH 0 . 0 2 3 1 o o o o o o d o o o 1—< o r—H n i o rH 0 0 1—< n i o o CS o o n i n i - r 41 42

closely with haplotypes 14 and 15 which were also from the same locality, which was

located approximately 60 miles from where 12 and 13 were found. Haplotype 17

clustered outside this group and was found to be about 2.3% divergent from it.

A total of six composite populations were compiled for use in a Monte Carlo

simulation; three populations within the Columbia Plateau and three populations within the

Great Basin/Colorado Plateau. Composite populations and distribution of haplotypes are

as shown in Figure 8. Columbia Plateau and Great Basin/Colorado Plateau populations were analyzed separately, and then the composite populations in each region were combined and analyzed as two populations. X2 values were calculated from the original data matrices and then compared to the X2 calculated after repeated random sampling of the data matrix. Haplotype diversity (h) and nucleotide diversity in the Columbia Plateau were much higher than in the Great Basin/Colorado Plateau (Table 8). Monte Carlo simulation results demonstrated significant geographic heterogeneity in both regions

(Table 9). 43

a *w o\ 8. . i c* m O'! i H O. d. ci >S3 od o o k s ^ « g CQ O. •5 73 cx ^ C coc: £> ^ *3 C4 a,

O 3 CO 3

S ° .2 ~3 X)3 u •c o 00 c Q 2 oo 3

CQ3 73O £ 3 Table 8 . Haplotype and nucleotide diversity values among composite mtDNA haplotypes in the Great Basin/Colorado Plateau and on the Columbia Plateau calculated in the DA program of REAP using equations from Nei (1987) and Nei and Tajima (1983).

Geographic location Haplotype Diversity Nucleotide diversity ( 7t) (h)

Great Basin/Colorado 0.3797 ± 0.08289 0.005087 ± 0.0000112 Plateau

Columbia Plateau 0.7556 ± 0.00605 0.02743 ± 0.0002384 45

Table 9. Monte Carlo simulation X 2 values and P-values among composite mtDNA haplotypes from the Great Basin/Colorado Plateau and the Columbia Plateau and between the Great Basin/Colorado Plateau and Columbia Plateau. X 2 values were calculated in the Monte program of REAP using the equations of Roff and Bentzen (1989).

Monte Carlo simulation X2 values P-values

Among Great Basin/Colorado 66.00 0.0001 Plateau populations

Among Columbia Plateau 22.75 0.003 populations

Between Great 46.00 0.0001 Basin/Colorado Plateau and Columbia Plateau populations Chapter 4

Discussion

Biogeographic hypotheses based upon Pleistocene glacial and interglacial cycles

have been used to explain distributions of rodents in western North America (Schmidley et

a l, 1994;) as well as diversity patterns of flora (Spaulding, 1990; Thompson, 1990).

Analyses reveal that there is considerable variation both within and among populations of

Perognathus parvus from the Columbia Plateau and the Great Basin. Prior studies hinted at the variation between populations from the Columbia Plateau versus the Great Basin

(Williams, 1978), but none exhibited this amount or pattern of variation. Variation within

P. parvus occurs in a well-defined hierarchy. Riddle (1995) demonstrated hierarchical structuring among other species of Perognathus, Chaetodipus, and Onychomys. DNA sequence divergence levels between Columbia Plateau and Great Basin/Colorado Plateau haplotypes are veiy high, while genetic distance values from RFLP analyses show much smaller levels of divergence among haplotypes within either region. Additional phylogeographic structure may be present among northern and southern Columbia Plateau populations.

Molecular clocks based on generation time, body size and metabolic rate allowed for the calibration of a time of divergence between populations of P. parvus (Li et al.,

46 47

Martin and Palumbi, 1993). The Mus - Rattus split about 10 mya provides a standard

against which rates of divergence can be calibrated in rodents of similar body size (Jaeger

et al., 1986). Riddle (1995) estimated a transversion substitution rate of about 1% per

million years in the COIU gene of Mus and Rattus. The transversion sequence divergence

of about 7.2% between northern and southern haplotypes of P. parvus therefore suggests

a late Miocene time of divergence. The demarcation between Columbia Plateau and

Great Basin populations of P. parvus occurs in central Oregon at the Blue Mountains.

The estimated time of divergence of approximately 7.2 million years ago roughly corresponds to the uplift of the Blue Mountains. The high level of bootstrap support for the P. parvus clade in Riddle (1995) supports the monophyletic nature of the P. parvus clade, and the level of transversion sequence divergence between Columbia Plateau and

Great Basin sequences supports the hypothesis thatP. parvus is actually made up of at least two distinct lineages. Genetic distances between Columbia Plateau and Great Basin sequences are greater than the distances between P. apache and P.fasciatus (5.9%), P. flavus and P.flavescens (4.9%), and between P. apache and P. flavescens (2.3%), which are all considered to be separate species (Nickle, 1994). This level of DNA sequence divergence is exceptionally high for an intraspecies comparison and indicates that what has traditionally been known as one species may actually represent two distinct lineages that diverged from one another following a large-scale vicariance event, thus warranting unique specific status for each. Further support for recognizing two distinct species comes from the high degree of karyotypic divergence that Williams (1978) found between northern and southern populations of P. parvus (Table 1). 48

The deep level of sequence divergence between the Columbia Plateau and Great

Basin populations of P. parvus and the estimated time of divergence at approximately 7.2 ma provide evidence for support of the hypothesis of a late Tertiary geotectonic vicariance event The time of the Blue Mountains orogeny (late Miocene) roughly corresponds to the time of isolation between Columbia Plateau and Great Basin populations of P. parvus.

Late Pleistocene vicariance as the result of habitat expansion and contraction would be reflected in lower levels of sequence divergence between northern and southern populations, as would Pleistocene dispersal due to climatic changes during glacial maxima.

Therefore, Pleistocene vicariance and dispersal hypotheses can be rejected in favor of a late Tertiary vicariance explanation for the distribution of northern and southern mtDNA haplotypes.

Nucleotide diversity among haplotypes from the Columbia Plateau is an order of magnitude higher than diversity found among the Great Basin haplotypes. Genetic distance values among haplotypes on the Columbia Plateau are also significantly higher than in the Great Basin. The high levels of diversity and divergence on the Columbia

Plateau are largely attributed to comparisons among Washington and Oregon haplotypes.

Haplotypes 16 and 18 from Oregon are more closely related to each other than they are to haplotypes from Washington; they also exhibit the largest amount of genetic divergence from the other haplotypes on the Columbia Plateau. Haplotypes 12 and 15 from one location are separated from haplotypes 13, 14, and 17, as well as another 12 and 15 by the

Columbia River. Gene flow probably occurred between these populations before the

Columbia River formed a possible barrier to dispersal. If the Columbia River does form a 49

barrier, it has not existed long enough to cause large amounts of divergence to occur between populations in Washington. Additional data are required to assess geographic distribution of haplotypes 16 and 18, which are most divergent from other Columbia

Plateau haplotypes. Although sequences from Washington haplotypes would allow for estimation of a more detailed phylogeny of the P. parvus clade, RFLP data indicate that additional phylogeographic structure is included in P. parvus.

Fossil pollen records from the Columbia Plateau indicate that xeric shrub-steppe habitat did not exist in that area as a dominant component until about 14 to 12 ma

(Leopold and Denton, 1987). Prior to this time the area contained mesic deciduous forest vegetation. The change from a mesic to xeric vegetation is believed to be due to the high level of active volcanism occurring during this time period. Volcanic activity can cause what appears to be climatic change, causing a moist habitat type to shift into a xeric one

(Cross and Taggart, 1982). Although xeric grassland and shrub-steppe habitat was not the dominant vegetation type in the area, it was present at lower elevations. Uplift of the

Cascades during the mid Miocene created a rain shadow that decreased the amount of precipitation east of the mountain range and facilitated the expansion of xeric habitat

(Leopold and Denton, 1987). P. parvus on the Columbia Plateau during the Miocene may have expanded it's range with the expansion of xeric shrub-steppe vegetation and decline of woodland habitat An hypothetical scenario that would explain the amount of diversity in the Columbia Plateau would be that an ancestral Perognathus parvus was present on the Columbia Plateau during middle Miocene. Beginning in the Miocene, the Columbia

Plateau was dominated by mesic woodland with patches of drier shrub-steppe type habitat 50

at lower elevations (Leopold and Denton, 1987). Advancement of woodland habitat

would have forced widespread P. parvus populations into increasingly smaller patches of

suitable habitat. Diversity that remained in those populations was a remnant of the earlier,

more widespread populations.

Another possible scenario to explain the amount of diversity on the Columbia

Plateau is that there was considerable basaltic volcanism occurring in the area during the

Miocene. Vertical tectonic rotation associated with volcanism caused a sharp folding pattern to the basaltic plains (Swanson et al., 1989). This folding could have caused isolation of populations and subsequent genetic differentiation from one another.

Readvancement of shrub-steppe habitat approximately 3 mya (Leopold and Denton, 1987) would have allowed populations to expand, and genetic diversity within and among populations would be left over from a time when populations were isolated from one another and evolving separately. Pleistocene climatic reconstructions show evidence for severe decreased winter temperatures on the Columbia Plateau. P. parvus most likely persisted on the Columbia Plateau during the height of Pleistocene glaciation instead of shifting its range southward off of the Columbia Plateau due to the fact that P. parvus is able to hibernate, slowing its metabolic rate so that it can withstand periods of unsuitable weather (French, 1993).

The amount of divergence occurring among haplotypes in the Great

Basin/Colorado Plateau is on average much lower than that on the Columbia Plateau.

This suggests that long-term effective population sizes of P. parvus on the Columbia

Plateau have been higher than in the Great Basin. The low levels of divergence among 51

Great Basin/Colorado Plateau haplotypes, and the wide distribution of haplotype 1

indicate few barriers to dispersal and gene flow in the northern part of the Great Basin, or

may indicate recent expansion of the range of P. parvus into the northern Great Basin

from southern refugia. Pleistocene glacial-interglacial cycles and expansion and

contraction of areas of suitable habitat in the Great Basin may be responsible for the

genetic divergence found among haplotypes from the Great Basin. During the Wisconsin

glaciation, shrub-steppe plant communities existed in the intermontane basins of the Great

Basin (Thompson and Mead, 1982). The distribution of this habitat may have provided

corridors for dispersal of organisms inhabiting it Subsequent climatic warming forced shrub-steppe communities upslope to mid elevations where they now exist. Populations of

P. parvus, which had a once more continuous distribution in the intermontane basins, became isolated from one another when suitable habitat became fragmented.

The correspondence of geological, fossil, and molecular data appears to be chronological, with the oldest basalt and fossils, and the greatest amount of genetic diversity occurring in the north on the Columbia Plateau. As one moves southward into the Great Basin, the relative age of basalt is younger, fossil records appear later, and the genetic diversity among haplotypes is lower.

There are other taxa that exhibit similar distribution patterns to P. parvus.

Spermophilus beldingi is distributed throughout the Great Basin much like Perognathus parvus, whileS. columbianus is distributed on the Columbia Plateau north of the Blue

Mountains (Hall, 1981). This is very similar to northern and southern populations of P. parvus. The Blue Mountains orogeny likely had extreme influence on the distributions of 52

animal and plant taxa and the genetic variation present in those taxa that underwent a

vicariance event of such magnitude. There is evidence for the presence of sciurid rodents in North America subsequent to the Barstovian NALMA (mid Miocene) (Savage and

Russell, 1983). If they were present on the Columbia Plateau as well as in the Great

Basin, populations would have been subjected to the same vicariance events as P. parvus.

I propose the hypothesis thatS. beldingi and S. columbianus, and possibly other species of small rodents exhibit a pattern of genetic variation very similar to that between northern and southern populations of P. parvus.

On another temporal scale, there are rather high levels of haplotype diversity in both northern and southern populations of P. parvus. Pleistocene glacial perturbations probably contributed to geographic variation among populations of P. parvus. Fossil pollen records show that the overall composition of vegetation in the Great Basin has changed very little since the Pleistocene, but there have been shifts in the distribution of the vegetation since the latest glacial episode (Thompson, 1990). Glacial-interglacial cycles are known to have caused expansion and contraction of biotic environments that in turn may be related to maintenance of haplotype diversity among populations of P. parvus.

Future Research

There are some concerns about interpreting molecular variation that comes from only one source (i.e. a single gene region or genome). It would be prudent to use nuclear

DNA to test conclusions in this study. Assuming that the northern and southern 53

populations have been separated for a long period of time, differences should also

accumulate in the nuclear genome (Avise et al., 1987). Use of allozymes, sequencing of

nuclear DNA, DNA-DNA hybridization, microsatellites, and microcomplement fixation

techniques would provide independent data sets.

The area of overlap between northern and southern populations of P. parvus in

central Oregon contains populations that have both northern and southern haplotypes

present. Because this study involved only the use of mtDNA, which is maternally

inherited, it is impossible to assess whether or not hybridization may be occurring between

northern and southern populations or if this area merely represents a zone of overlap in the distributions of the two reproductively-isolated populations of P. parvus. In order to test

the hypothesis of hybridization, it would be necessary to use analyses of nuclear DNA (i.e. allozymes, micro satellite DNA) to determine if there had been any gene flow between populations. Analyses of nuclear DNA would allow for the determination of heterozygosity in individuals as well as identification of origin of parental genotypes present in offspring.

Conclusions

The results of this study have clearly shown that althoughPerognathus parvus is monophyletic, it consists of at least two distinct lineages: one on the Columbia Plateau, north of the Blue Mountains, one in the Great Basin, south of the Blue Mountains, and possibly another in central Oregon along the southern edge of the Columbia Plateau.

Divergence values between the Columbia Plateau and Great Basin lineages suggest that northern and southern populations were isolated from one another during the mid to late

Miocene corresponding to the uplift of the Blue Mountains in northern Oregon.

Haplotype diversity on the Columbia Plateau is higher than in the Great Basin and suggests that effective population size has remained much higher over time. Chapter 5

Perognathus alticola and P. xanthonotus

Perognathus alticola and P. xanthonotus were originally hypothesized to be

members of the Perognathus parvus Species Group (Osgood, 1900). Williams' (1978)

study of karyotypes supported this hypothesis. Sulentich (1983) reviewed the systematics

of the Perognathus parvus Species Group in southern California and based on biochemical

data, he determined that P. xanthonotus was a subspecies of P. parvus and that P. alticola

alticola and P. alticola inexpectatus should be separated at the specific level. Williams

(1993) then accepted Sulentich's proposal that P. xanthonotus is a subspecies of P. parvus, but he did not accept the separation of P. alticola subspecies, thus disagreeing

with his own karyotype data showing that both P. alticola and P. xanthonotus had identical karyotypes to P. parvus from the Great Basin (Table 10).

The status of these two other members of the Perognathus parvus Species Group has not been satisfactorily resolved. These two species reside in the northern Mojave

Desert of southern California in isolated areas of habitat characterized by Joshua trees

(Yucca brevicauda), sagebrush (Artemisia sp.), juniper (Juniperus sp.) and pinyon pine

(Pinus sp) (Figure 9). Woodrat middens in the Mojave Desert suggest a cooler temperature regime for the area as evidenced by shadscale-juniper association during the

55

l 56

Table 10. Karyotype characteristics of species included in the Perognathus parvus species group showing similarity between P. p. olivaceous and P. alticola, and similarity between . p. clarus and P. xanthonotus. 2N=diploid number, FN=fundamental number, BA=biarmed chromosome, A=uniarmed chromosome, ST=subtelocentric, SM=submetacentric, A=acrocentric. Data are from Williams (1978).

Autosomal pairs SPECIES m EN BAAX X

Perognathus parvus olivaceous 54 74 11 15 STA

Perognathus parvus clarus 54 76 12 14 STA

Perognathus alticola 54 74 11 15 STA

Perognathus xanthonotus 54 76 12 14 ST A 57

rt .a a e 'Cr-! ^ Ic3 Q So U

•N*O) 2 o

Sc

oo o £ TD rtC O

OO O>*.

c o 3 X) •c Q

On K 3 OO £ 58

middle Wisconsin. Extirpation of shadscale around 19,000 ybp was followed by the

immigration of pinyon pine and shrubs (Spaulding, 1990), and desertscrub became the

dominant vegetation type during the late Pleistocene and early Holocene. These areas are

likely to be post-Pleistocene refugia, and P. xanthonotus and P. alticola may represent

isolated populations of previously more southern distribution of P. parvus. These areas

also support other taxa that may be Pleistocene relicts (Riddle, pers. comm.). Based on

the distribution, historical ecological evidence and Williams' (1978) karyotype data, I propose the hypothesis that bothP. alticola inexpectatus and P. xanthonotus are

subspecies of P. parvus. P. alticola alticola may truly represent a unique taxon based on its distribution in relation to the other two species and the possibility that the Sierra

Nevada acts as a barrier to dispersal. P. alticola inexpectatus and P. xanthonotus reside on the eastern side of the Sierra Nevada, while P. alticola alticola resides on the western side. Uplift of the Sierra Nevada during the early Miocene would have separated these populations long ago if they were present at that time.

It is not known whether or not P. alticola or P. xanthonotus are extant. Neither have been caught in several years (not for lack of effort), and P. alticola is a protected species in California. For this reason it will be difficult to test the hypotheses presented above using whole specimens. Recent techniques in isolating and analyzing DNA from museum skins and bones of extinct species have opened new areas of research (Higuchi et al., 1984, 1987; Hoss and Paabo, 1993; Paabo, 1985; Paabo, 1989; Paabo et al., 1989).

Isolation of DNA from museum specimens of P. alticola and P. xanthonotus, and use of primers designed for amplification of small diagnostic sequences of DNA would allow one 59

to compare sequences of these two species with those of P. parviis and determine levels of divergence among them. Taxonomic status of P. alticola and P. xanthonotus could then be assessed and biogeographic hypotheses evaluated. 60

APPENDIX I. List of specimens and localities used to construct phylogenetic relationships within Perognathus parvus and calculate genetic distances within and among mtDNA haplotypes in composite populations.

IDAHO: Cassia Co., (LVT 794, LVT 795)

NEVADA: Esmeralda Co., 11 mi. N, 7 mi. W Dyer, T1S, R34E, Sl/2, NE 1/4 sec. 19; (LVT 1931, LVT 1932, LVT 1933, LVT 1934). Nye Co., 36 mi. N, 24 mi. W Mercury, Nevada Test Site, (LVT 1805, LVT 1806); 30 mi. N, 17 mi. W Mercury, Nevada Test Site, (LVT 1807); 1 mi. W Manhattan, (LVT 1957). White Pine Co., (LVT 1929).

OREGON: Baker Co., 1 mi. S, 5 mi. E Baker, T9S, R40E, (LVT 1955, LVT 1956, LVT 1958). Crook Co., 10 mi. N, 5 mi. E Redmond; TBS, R14E sec. 34, (LVT 1921, LVT 1949, LVT 1950, LVT 1951, LVT 1952). Harney Co., 5 mi. S, 4 mi. W Hines; T24S, R30E sec. 8, (LVT 1923, LVT 1925, LVT 1927). Lake Co., 2 mi. S, 8 mi. E Hampton; T22S, R21E, (LVT1943, LVT 1944, LVT 1945, LVT 1946, LVT 1947, LVT 1948). Malheur Co., 1 mi. N, 1 mi. E Basque Station, (LVT 1937, LVT 1938, LVT 1939, LVT 1940, LVT 1941, LVT 1942).

WASHINGTON: Adams Co., 4 mi. S, 6 mi. E Ritzville; T19N, R36E sec. 3, (LVT 1919, LVT 1920). Douglas Co., 16 mi. N, 5 mi. E Wenatchee; T25N, R21W sec. 25, (LVT 1915, LVT 1916). Okanogan Co., 4 mi. S, 3 mi. W Oroville, Wannacut Lake; T39N, R27E, (LVT 1953, LVT 1954); 1 mi. S, 1.5 mi. W Okanogan; T33N, R26E sec. 19, (LVT 1917, LVT 1918). Bibliography

Allard, M. S., D. L. Ellsworth, and R. L. Honeycutt. 1991. The production of single- stranded DNA suitable for sequencing using the polymerase chain reaction. BioTechniques, 10: 24-26.

Avise, J. C. 1992. Molecular population structure and the biogeographic history of a regional fauna: A case history with lesson for conservation biology. Oikos, 63: 62-73.

Avise, J. C. 1994. Molecular markers, natural history and evolution. Chapman and Hall, Inc., New York, London.

Avise, J. C. and R. M. Ball. 1990. Principle of genealogical concordance in species concepts and biological taxonomy. Oxford Surveys of Evolutionary Biology, 7: 45-67.

Avise, J. C., J. Arnold, R. M. Ball, E. Bermingham, T. Lamb, J. E. Neigel, C. A. Reeb, and N. C. Saunders. 1987. Intraspecific phylogeography: The mitochondrial DNA bridge between population genetics and systematics. Annual Review of Ecology and Systematics, 18: 489-522.

Avise, J. C., R. M. Ball, and J. Arnold. 1988. Current versus historical populations sizes in vertebrate species with high gene flow: A comparison based on mitochondrial DNA lineages and inbreeding theory for neutral mutations. Molecular Biology and Evolution, 5(4): 331-344.

Bibb, M. J., R. A. Van Etten, and C. T. Wright. 1981. Sequence and gene organization of mouse mitochondrial DNA. Cell, 26:167-180.

Brooks, D. R. 1981. Hennig's parasitological method: A proposed solution. Systematic Zoology, 30: 229-249.

Brooks, D. R. 1990. Parsimony analysis in historical biogeography and coevolution: Methodological and theoretical update. Systematic Zoology: 14-30.

61 62

Brooks, D. R. and D. A. McLennan. 1991. Phylogeny, ecology and behavior: A research program in comparative biology. University of Chicago Press, Chicago, EL.

Brooks, D. R. and E. O. Wiley. 1985. Theories and methods in different approaches to phylogenetic systemadcs. Cladistics, 1: 1-11.

Brown, J. H. and A. C. Gibson. 1983. Biogeography. The C. V. Mosby Company, St. Louis, MO.

Brown, W. M. 1985. The mitochondrial genome of animals. In Molecular Evolutionary Genetics, R. J. MacIntyre, ed. Plenum Press, New York.

Carlson, R. W. and W. K. Hart, 1987. Crustal genesis on the Oregon Plateau, Journal of Geophysical Research 92(B7): 6191-6206.

Christiansen, R. L. and E. H. McKee, 1977. Late Cenozoic volcanic and tectonic evolution of the Great Basin and Columbia Intermontane Regions, Geologic Society of America Memoir No. 152: 283-311.

Cracraft, J. 1983. Species concepts and speciation analysis. Current Ornithology, 1: 159-187.

Cramer, K. L. and J. A. Chapman, 1990. Reproduction of three species of pocket mice (.Perognathus) in the Bonneville Basin, Utah. Great Basin Naturalist 50(4): 361- 365.

Crisci, J. V. and T. F. Stuessy. 1980. Determining primitive character states for phylogenetic reconstruction. Systematic Botany, 5: 112-135.

Cross, A. T. and R. E. Taggart, 1982. Causes of short-term sequential changes in fossil plant assemblages: Some considerations based on a Miocene flora of the Northwestern United States, Annals of the Missouri Botanical Garden 69: 676-734.

Darwin, C. 1859. On the origin of species by natural selection, or the preservation of favored races in the struggle for life. John Murray, London.

Dowling, T. E., C. Moritz, and J. D. Palmer. 1990. Nucleic Acids II: Restriction site analysis. In Molecular Systematics, D. M. Hillis and C. Moritz, eds. Sinauer Associates, Inc., Sunderland, MA.

Farris, J. S. 1983. The logical basis of phylogenetic analysis. In Advances in Cladistics, 2, N. I. Platnick and V.A. Funk, eds. Columbia University Press, New York. 63

Felsenstein, J. 1983. Parsimony in systematics: Biologican and statistical issues. Annual Review of Ecology and Systematics, 14: 313-333.

Felsenstein, J. 1993. PHYLIP (Phylogeny Inference Package, version 3.5) Distributed by the author. University of Washington, Seattle.

French, A. R. 1993. Physiological ecology of the Heteromyidae: Economics of energy and water utilization. In Biology of the Heteromyidae, H. H. Genoways and J. H. Brown, eds. The American Society of Mammalogists Special Publication No. 10: 1-719.

Hafner, J. C. and M. S. Hafner. 1983. Evolutionary relationships of Heteromyid rodents. In Biology of Desert Rodents, O. J. Reichman, ed. Great Basin Naturalist Memoirs No. 7: 3-29.

Hall, E. R., 1981. The Mammals of North America. John Wiley and Sons, Inc.

Hennig, W. 1966. Phylogenetic systematics. University of Illinois Press, Urbana.

Higuchi, R. G., B. Bowman, M. Freiberger, O. A. Ryder, and A. C. Wilson. 1984. DNA sequence from the quagga, an extinct member of the horse family. Nature, 312: 282-284.

Higuchi, R. G., L. A. Wrischnik, E. Oakes, M. George, B. Tong, and A. C. Wilson. 1987. Mitochondrial DNA of the extinct quagga: relatedness and extent of postmortem change. Journal of Molecular Evolution, 25: 283-287.

Hillis, D. M. 1987. Molecular versus morphological approaches to systematics. Annual Review of Ecology and Systematics, 18: 23-42.

Hillis, D. M., A. Larson, S. K. Davis, and E. A. Zimmer. 1990. Nucleic acids HI: Sequencing. In Molecular Systematics, D. M. Hillis and C. Moritz, eds. Sinauer Associates, Inc., Sunderland, MA.

Hoss, M. and S. Paabo. 1993. DNA extraction from Pleistocene bones by a silica-based purification method. Nucleic Acids Research, 21: 3913-3914.

Jaeger, J. -J., H. Tong, and C. Denys. 1986. The age of Miis-Rattus divergence: Paleontological data compared with the molecular clock. Comptes Rendus Academie des Sciences Paris (serie 2), 302: 917-922. 64

James, G. T. 1963. Paleontology and nonmarine strratigraphy of the Cuyama Valley Badlands, California. Part I. Geology, faunal interpretations, and systematic descriptions of Chiroptera, Insectivora, and Rodentia. University of California Publications in Geological Sciences, 45: 1-171.

Kenagy, G. J. and B. M. Banes, 1984. Environmental and endogenous control of reproductive function in the Great Basin pocket mouse Perognathus parvus. Biology of Reproduction 31: 637-645.

Leopold, E. B. and M. F. Denton, 1987. Comparative age of grassland and steppe east and west of the Rocky Mountains, Annals of the Missouri Botanical Garden 74: 841-867.

Li, W. -H., M Gouy, P. M. Sharp, C. O'Huigin, and Y. -W. Yang. 1990. Molecular phylogeny of Rodentia, Lagomorpha, Primates, Artiodactyla, and Carnivora and molecular clocks. Proceedings of the National Academy of Sciences, USA, 87: 6703-6707.

Longmire, J. L. and R. J. Baker. 1994 (In press). The use of "Lysis Buffer" in DNA isolation.

Martin, A. P. and S. R. Palumbi. 1993. Body size, metabolic rate, generation time, and the molecular clock. Proceedings of the National Academy of Sciences, USA, 90: 4087-4091.

Mayr, E. and P. D. Ashlock. 1991. Principles of systematic zoology, 2nd ed. McGraw- Hill, Inc., USA.

McElroy, D., P. Moran, E. Bermingham, and I. Komfield, REAP, Restriction Enzyme Analysis Package, Version 4.0.

McKitrick, M. C. and R. M. Zink. 1988. Species concepts in ornithology. The Condor, 90(1): 1-14.

Merriam, C. H., 1889. Revision of the North American Pocket Mice, North American Fauna No. 1: 1-29.

Miller, J. C., 1991. RESTSITE V 1.2, J. C. Miller

Miyamoto, M. M. and J. Cracraft. 1991. Phylogenetic analysis of DNA sequences. Oxford University Press, Inc., New York. 65

Moritz, C., T. E. Dowling, and W. M. Brown, 1987. Evolution of animal mitochondrial DNA: relevance for population ecology and systematics, Annual Review of Ecology and Systematics 18: 269-292.

Nei, M. 1987. Molecular evolutionary genetics. Columbia University Press, New York.

Nei, M. and F. Tajima, 1981. DNA Polymorphism detectable by restriction endonucleases, Genetics 97: 145-163.

Nei, M. and F. Tajima. 1983. Maximum likelihood estimation of the number of nucleotide substitutions from restriction site data. Genetics, 105: 207-217.

Nelson, G. and N. Platnick. 1981. Systematics and Biogeography: Cladistics and vicariance. Columbia University Press, New York.

Nickle, D. C. 1994. Molecular Phylogeographic structure of the Perognathus fasciatus group. M.S. Thesis, University of Nevada, Las Vegas.

O'Farrell, T. P., R. J. Olson, R. O. Gilbert, and J. D. Hedlund, 1975. A population of Great Basin pocket mice, Perognathus parvus, in the shrub-steppe of south-central Washington, Ecological Monographs 45: 1-28.

Osgood, W. H., 1900. Revision of the Pocket Nice of the Genus Perognathus, North American Fauna No. 18: 1-66.

Paabo, D. 1985. Molecular cloning of ancient mummy DNA. Nature, 314: 644-645.

Paabo, S. 1989. Ancient DNA: extraction, characterization, molecular cloning and enzymatic amplification. Proceedings of the National Academy of Sciences, USA, 86: 1939-1943.

Paabo, S., R. G. Higuchi, and A. C. Wilson. 1989. Ancient DNA and the polymerase chain reaction. Journal of Biological Chemistry, 264: 9709-9712.

Patton, J. L., 1967. Chromosomes and evolutionary trends in the pocket mouse subgenus Perognathus (Rodentia: Heteromyidae), The Southwestern Naturalist 12(4): 429- 438.

Price, M. V. and J. H. Brown. 1983. Patterns of morphology and resource use in North American desert rodent communities. In Biology of Desert Rodents, O. J. Reichman, ed. Great Basin Naturalist Memoirs No. 7: 117-134. 66

Riddle, B. R. 1995. Molecular biogeography in the pocket mice {Perognathus and Chaetodipus) and grasshopper mice ( Onychomys): The late Cenozoic development of a North American aridlands rodent guild. Journal of Mammalogy, 76(2): 283-301.

Riddle, B. R., R. L. Honeycutt, and P. L. Lee, 1993. Mitochondrial DNA phylogeography in northern grasshopper mice ( Onychomys leucogaster) - the influence of Quaternary climatic oscillations on population dispersion and divergence, Molecular Ecology 2: 183-193.

Roff, D. A. and P. Bentzen, 1989. The statistical analysis of mitochondrial DNA polymorphisms: %2 the problem of small samples.

Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain- terminating inhibitors. Proceedings of the National Academy of Sciences, USA, 74: 5463-5467.

Savage, D. E. and D. E. Russell. 1983. Mammalian paleofaunas of the world. Addison- Wesley, London.

Schmidley, D. J., K. T. Wilkins, and J. N. Derr. 1993. Biogeography, pp. 319-356. In Biology of the Heteromyidae, H. H. Genoways and J. H. Brown, eds. The American Society of Mammalogists Special Publication No. 10: 1-719.

Simpson, G. G. 1961. Principles of animal taxonomy. Columbia University Press, New York.

Small, R. J. and B. J. Verts, 1983. Responses of a population of Perognathus parvus to removal trapping. Journal of Mammalogy 64(1): 139-143.

Spaulding, W. G. 1990. Vegetational and climatic development of the Mojave Desert: The last glacial maximum to the present. In Packrat Middens, J. L. Betancourt, T. R. Van Devender, and P. S. Martin, eds., University of Arizona Press.

Sulentich, J. M., 1983. The systematics and evolution of the Perognathus parvus species group in southern California (Rodentia: Heteromyidae), M.S. thesis, University of California, Long Beach, 85 pp. 67

Swanson, D. A., K. A. Cameron, R. C. Evarts, P. T. Pringle, and J. A. Vance, 1989. IGC Field Trip T106: Cenozoic volcanism in the Cascade Range and Columbia Plateau, southern Washington and northernmost Oregon, American Geophysical Union, Washington, D. C.

Swofford, D. L. 1991. PAUP: Phylogenetic Analysis Using Parsimony, Version 3.0s. Computer program distributed by the Illinois Natural History Survey, Champaign, Illinois, 1985, 1991.

Thompson, R. S. 1990. Late Quaternary vegetation and climate in the Great Basin. In Packrat Middens, J. L. Betancourt, T. R. Van Devender, and P. S. Martin, eds., University of Arizona Press.

Thompson, R. S. and J. I. Mead. 1982. Late Quaternary environments and biogeography in the Great Basin. Quaternary Research, 17: 39-55.

Towner, J. W., 1965. The effect of radioactive fallout at the Nevada Test Site on the chromosomes of the pocket mouse, Health Physics 11: 1569-1571.

Upholt, W. B., 1977. Estimation of DNA sequence divergence from comparison of restriction endonuclease digests, Nucleic Acids Research 4: 1257-1265.

Vawter, L. and W. M. Brown, 1986. Nuclear and mitochondrial DNA comparisons reveal extreme rate variation in the molecular clock. Science 234: 194-196.

Verts, B. J. and L. N. Carraway, 1986. Replacement in a population of Perognathus parvus subjected to removal trapping. Journal of Mammalogy 67(1): 210-215.

Vrba, E. S. 1992. Mammals as a key to evolutionary theory. Journal of Mammalogy, 7: 31-28.

Wahlert, J. H., 1993. The Fossil Record, in The Biology of the Heteromyidae, H. H. Genoways and J. H. Brown, eds., The American Society of Mammalogists Special Publication No. 10: 1-37.

Wiley, E. O. 1981. Phylogenetics, the theory and practice of phylogenetic systematics. Wiley-Interscience, New York.

Wiley, E. O., D. Siegel-Causey, D. R. Brooks, and V. A. Funk. 1991. The Compleat Cladist: A primer of phylogenetic procedures. University of Kansas, Museum of Natural History Special Publication No. 19: 1-158. 68

Williams, D. F., 1978. Karyological affinities of the species groups of silky pocket mice (Rodentia, Heteromyidae), Journal of Mammalogy 59(3): 599-612.

Williams, D. F., H. H. Genoways, and J. K. Braun, 1993. Taxonomy, in The Biology of the Heteromyidae, H. H. Genoways and J. H. Brown, eds., The American Society of Mammalogists Special Publication No. 10: 38-196.

Wilson, R. W. 1960. Early Miocene rodents and insectivores from northeastern Colorado. The Univ. of Kansas, Paleontological Contributions, Vertebrata, 7: 1- 92. Zakrzewski, R. J., 1969. The Rodents from the Hagerman Local Fauna, Upper Pliocene of Idaho, Contributions to the Museum of Paleontology, The University of Michigan No. 23: 1-36.