University of Colorado, Boulder
CU Scholar
Ecology & Evolutionary Biology Graduate Theses &
Dissertations
Ecology & Evolutionary Biology
Spring 1-1-2014
Population Genetics of Ovis and Phylogenetics of
Caprinae: A Comparison of Different Genetic
Markers for Evaluation of Diversity at Multiple
Taxonomic Levels
Catherine Colleen Driscoll
University of Colorado Boulder, [email protected]
Follow this and additional works at: http://scholar.colorado.edu/ebio_gradetds
Part of the Genetics Commons, and the Population Biology Commons
Recommended Citation
Driscoll, Catherine Colleen, "Population Genetics of Ovis and Phylogenetics of Caprinae: A Comparison of Different Genetic
Markers for Evaluation of Diversity at Multiple Taxonomic Levels" (2014). Ecology & Evolutionary Biology Graduate Theses &
Dissertations. 49.
http://scholar.colorado.edu/ebio_gradetds/49
This Dissertation is brought to you for free and open access by Ecology & Evolutionary Biology at CU Scholar. It has been accepted for inclusion in
Ecology & Evolutionary Biology Graduate Theses & Dissertations by an authorized administrator of CU Scholar. For more information, please contact
[email protected].
POPULATION GENETICS OF OVIS AND PHYLOGENETICS OF CAPRINAE: A
COMPARISON OF DIFFERENT GENETIC MARKERS FOR EVALUATION OF
DIVERSITY AT MULTIPLE TAXONOMIC LEVELS
by
CATHERINE C. DRISCOLL
B.A., University of California Berkeley 2000
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirement for the degree of
Doctor of Philosophy
Department of Ecology and Evolutionary Biology
2014
This thesis entitled:
Population Genetics of Ovis and Phylogenetics of Caprinae: A Comparison of Different Genetic
Markers For Evaluation of Diversity at Multiple Taxonomic Levels
Written by Catherine C. Driscoll
Has been approved for the Department of Ecology and Evolutionary Biology
_______________________________________
Jeffry B. Mitton
_______________________________________
Robert P. Guralnick
_______________________________________
Rebecca J. Safran
_______________________________________
David W. Stock
_______________________________________
John D. Wehausen
Date_____________________
The final copy of this thesis has been examined by the signatories, and we find that both the
content and the form meet acceptable presentation standards of scholarly work in the above
mentioned discipline
Driscoll, Catherine Colleen (Ph.D., Ecology and Evolutionary Biology)
Population Genetics of Ovis and Phylogenetics of Caprinae: A Comparison of Different Genetic
Markers For Evaluation of Diversity at Multiple Taxonomic Levels
Thesis directed by Professor Jeffry B. Mitton
This research was initiated due to concern that one of the four Rocky Mountain National
Park (RMNP) bighorn sheep (Ovis canadensis) herds (the Mummy) could be suffering from
inbreeding depression following a documented pneumonia outbreak and die-off in 1994. Failure
of this herd to recover led to a comprehensive population genetic analysis of five bighorn sheep
herds (Big Thompson, Continental Divide, Mummy, Never Summer and St. Vrain) to address
four main questions:
1) Using data from 13 microsatellite loci: How much genetic variation exists in non-coding
DNA in each herd?
2) Using sequence data from two distinct regions of the maternally inherited mitochondrial
genome (Cytochrome Oxidase I gene (COI) and D-Loop control region (DL)): To what
extent are ewes a source of gene flow between herds?
3) Using sequence data from two regions of the Y-chromosome gene SRY (promoter (PRO)
and complete gene coding region (ORF)): to what extent are rams a source of gene flow
between herds?
4) Do these data support evidence of a recent and severe genetic bottleneck in any of these
five herds such that inbreeding depression could be a significant factor in long-term
viability?
iii
Our results support that all five herds have high levels of heterozygosity, low population
substructure maintained by on-going gene flow, and no evidence of a recent genetic bottleneck.
All data clearly refute the hypothesis that decreased genetic variation is a significant factor in the
Mummy herd’s failure to thrive and instead suggest that environmental factors should be
investigated.
Unfortunately, our analyses of Y-chromosome data showed the SRY sequences to be
invariant among RMNP bighorn herds; however we realized these data would still be useful in a
broader context and we therefore expanded analyses to include all genera within the subfamily
Caprinae (Artiodactyla, Bovidae) and incorporated comparative data from the mitochondrial
gene ND5. These data provide comprehensive and complementary phylogenetic analyses for
which the more conserved Y-chromosome sequences generally showed better resolution of the
most basal nodes, while the more variable mitochondrial sequences were usually better able to
resolve relationships among recently diverged genera as well as between species within
individual genera.
iv
DEDICATION
I dedicate this document to my father, James Glynn Driscoll III, without whom this research
would not have been possible, and this dissertation may never have been completed. And I am
certain that had you not been there to guide me, I would still be wandering the Never Summer
Mountains, trying to find my way home.
ACKNOWLEDGMENTS
This dissertation is truly the product of both time and energy contributed by so many
people and I am eternally grateful to each and every one of them. I am indebted to my advisor,
Dr. Jeff Mitton, and my CU committee: Drs. Rob Guralnick, Rebecca Safran and David Stock,
who worked tirelessly to help me reach this goal; providing support, feedback and countless edits
along the way. I would also like to especially thank my outside committee member, Dr. John
Wehausen, who not only taught me many of the laboratory and analytical techniques necessary
for this research but who also provided endless support throughout this process. Thank you
John, our long discussions have been invaluable.
Dr. David Armstrong, you are one of the most influential professors that I have ever
encountered and I aspire to be the educator you are. Thank you for setting such an impressive
example, I will endeavor to follow in your path. Dr. John Basey, thank you for your tireless
dedication to undergraduate education; I have admired your ability to inspire both teaching
assistants and students alike since I came to CU. I hope you keep up your wonderful work. Dr.
Erin Bissell, thank you for the many semesters of teaching we accomplished together, we
weathered the education equivalent of “Apocalypse Now” and survived, for several years. I
learned so much along the way.
Graduate school has been described as a war of attrition, which requires unwavering
resolve in the face of seemingly insurmountable obstacles; even when you cannot find the
courage to carry on, even when you believe you cannot succeed. That resolve was only possible
for me because I had the support of such wonderful people, and I cannot express the depth of my
gratitude for the love and support of so many family and friends
vi
I want to thank my father, James Driscoll III, who contributed to this process in too many
ways to name. Beginning with our field work: I will never forget the cold, early mornings
searching for bighorn, and inevitably finding them on the absolute top of the ridge of the highest
peak around and steadfastly making the arduous climb straight up the side of some incredibly
steep mountainside. Day after day we labored to find and collect bighorn sheep samples for
these genetic analyses found herein. We saw beauty that is unrivaled by any experience in my
life. I will never forget those weeks in the backcountry of Rocky Mountain National Park.
I want to thank my mother, Patricia Martin, and my step father, Dan Martin, for
unending, unwavering love. There is no way to express the value of your support. Dan, you
have always been proud of me and I hope you know how much that has, and still does, mean.
Thank you. Mom, we weathered lean times while I was growing up and your incredible
optimism made my childhood rich in too many ways to name. I want to thank you especially for
coming to Colorado for my defense and sitting with me, waiting, for one of the most frightening
moments of my life. You are a rock, mom, and I love you.
Of course there are so many friends and colleagues to thank that I just have to simply list
them and hope they know how much they mattered. Dr. Robert Baker, we met the first day of
graduate school, and you have been a friend ever since. I feel so lucky to have gotten to know
you so much better in the last year. Tim Singleton, you have been a wonderful friend and
roommate almost from the moment I arrived in Colorado, thank you so much. Dr. Mike
Robeson, you have been a great friend and often advisor on the finer points of phylogenetics and
I hope we continue our video chats for years to come. Dr. Se Jin Song, I can’t imagine anyone
putting up with sharing a tiny office with me for so many years and managing to like me
anyway, I treasure your friendship. Mathew Arnold, thank you for being my intrepid explorer, I
vii
will always remember how much fun we had weathering the rainstorms during field collection
and I am so proud of you. Dr. Sarah Wise, thank you for, quite simply, being you. Your star
shines so brightly. Joey Hubbard, you are a great friend and roommate and I am so glad to know
you. Dr. Loren Sackett, I will always remember your smiling face. Ryan Lynch, thank you for
pushing me when I needed it and never letting me back down, you helped me find the strength to
see this through. Dr. Rob Roy Ramey III, thank you for unwavering support and for helping me
find my voice. Corey Hazekamp, you were the best undergraduate I ever had the pleasure of
working with, I can’t wait to see what great things you accomplish.
I would especially like to thank the EBIO staff, specifically Jill Skarstad and Linda
Bowden, to whom I went for innumerable questions and who always had the answers. Thank
you for your time and help and seemingly unending knowledge.
Finally. I would like to thank the National Park Service and the National Science
Foundation for providing the majority of the funding for my research. This work would not have
been possible without these resources, including the contributions of Judy Visty, Sherri Huwer,
and Janet George.
viii
CONTENTS
CHAPTER I
Introduction and Background…………………………………………..…………1
Development of Molecular Techniques, Markers, and Their Uses….….1
Population Genetics………………………………………….………...… 3
Phylogenetics………………………………..………….…………………6
CHAPTER II
Population Genetics of Rocky Mountain Bighorn Sheep Estimated From
Microsatellite and Mitochondrial Data ………………………………………….10
Introduction………………………………………….…………...……....10
Study Area…………………………...……...……………………...…....13
Materials and Methods…………………………….………...…………...18
Results……………………………………………………...…………….24
Discussion……………………………...…..………………...……......…41
CHAPTER III
Evaluating the Relative Resolving Power of Y-Chromosome and Mitochondrial
Markers for Inferring Phylogenic relationships among genera within the
Subfamily Caprinae (Artiodactyla, Bovidae)……………………………………47
Introduction…….……………………………………………….…….…47
Materials and Methods………………………….…….……………...….49
Results……………………………………………………………………54
Discussion………….……………….………………………..…………..61
ix
CHAPTER IV
General Conclusions……………………………………………………………..70
Population Genetics of Rocky Mountain Bighorn Sheep………………..70
Phylogenetics of Caprinae Genera………………………………...……..73
LITERATURE CITED……………………………………………………….………………….76
x
TABLES
TABLE
1.
Historical data for sheep translocations involving the five RMNP herds
(BT, CD, MM, NS and SV) movements of sheep into, between, and out
of these native herds.………………………………………………….............….17
2.
Microsatellite loci general information by locus including: location in
genome (Chr.), range of allele sizes (in base pairs) identified in this study
(Range), total number of different alleles detected (Alleles), PCR annealing
temperature in degrees Celsius (Anneal), fluorescent label for each primer
pair (Tag), source reference for locus information (Cited source) and
NCBI accession number (Accession)………………………...….....................…20
3.
Data collection summary by marker type and individual herd. Number
of sample collection sites (N), total microsatellite genotypes
(Microsatellite), total mitochondrial Cytochrome Oxidase I amplifications
(mtDNA COI) and total mitochondrial D-Loop control region
amplifications (mtDNA DL) for each of the five bighorn sheep herds
(BT: Big Thompson, CD: Continental Divide, MM: Mummy, NS: Never
Summer, SV: St. Vrain) in north-central Colorado, USA………………..….......25
4.
Genetic variation estimates of microsatellite genotype data, including
range and average number of alleles per locus (Na), Heterozygosity
excess (He), Heterozygosity expected at equilibrium (Heq), probability of
heterozygosity excess (One-tailed Wilcoxon test, P(e)), probability of
heterozygosity excess or deficiency (Two-tailed Wilcoxon Test, P(w)) and
probability of heterozygosity excess under TPM (Sign Test, P(s)) ……….….…26
5.
Pairwise fixation indices (Fst; left side) and effective migration rates
(Nm; right side) estimated from microsatellite genotypes (above diagonal)
and mitochondrial haplotypes (below diagonal) among all five bighorn
sheep herd pairs in north-central Colorado, USA……………………………..…26
xi
6.
Coefficient of inbreeding (Fis) estimates for three microsatellite data
subsets including: genotypes for only the four potentially “non-neutral”
loci: OLA, KERA, SOMA and TCRBV (Four loci), genotypes for the
nine putatively “neutral” loci: ADC, AE16, FCB266, FCB304, HH62,
MAF33, MAF36, MAF48 and MAF209 (Nine loci) and genotypes from
the complete data set of all 13 Loci (All loci) for five bighorn sheep herds
in north-central Colorado, USA.………………………….………………..….....30
7.
Pairwise isolation-by-distance (IBD) estimates from mitochondrial and
microsatellite data for among all five bighorn sheep herd……………………….32
8.
Distribution and abundance of each of the twelve mitochondrial haplotypes
identified in this study among all five herds of bighorn sheep in
north-central Colorado, USA.…………………………………………..…..……37
9.
Estimates of genetic diversity for the mitochondrial DNA
haplotypes comprising 1,937 bp of Cytochrome Oxidase I (COI) and D-Loop
(DL) sequences for each of the five bighorn sheep herds from north-central
Colorado, USA including total sample size (N), segregating sites (S),
number of haplotypes (h), haplotype diversity (Hd), average number of
nucleotide differences (K), nucleotide diversity (π).…..……………………......38
10.
Sample information by individual Caprinae genus for Y-chromosome
(SRY ORF and SRY PRO) and mitochondrial (ND5) data sets. ……….….…...50
11.
SRY and ND5 primer sequences used for all sequence amplifications…….……51
12.
Summary DNA sequence statistics by data set including sequence length,
number of haplotypes (h), nucleotide diversity (π), nucleotide diversity
with Jukes-Cantor correction (πJC), number of segregating sites (S),
average number of nucleotide differences (K), theta per nucleotide site as
estimated from S (ӨW), and theta per sequence as estimated from S (Ө).…….…53
13.
Support of individual nodes for each data set as estimated with posterior
probabilities (PP), ML Bootstrap percentages (BPML) and MP bootstrap
percentages (BPMP). Nodes not found or with less than 50% bootstrap
support or 0.5 posterior probability are indicated with an asterisk (*).
Nodes for which support could not be evaluated because not all genera
were available are indicated with a dash (-)………………………………...……59
xii
FIGURES
FIGURE
1.
Relief map of north-central Colorado, USA showing 26 DNA
collection sampling sites for the five bighorn sheep (Ovis canadensis)
herds in and around Rocky Mountain National Park included in this study.
Polygons each represent collection sites for a single herd: BT (circles), CD
(diamonds), MM (squares), NS (hexagons) and SV (triangles).………….......…15
2.
Map of the study area showing general geographic ranges for each of the
five bighorn sheep herds. Overall fixation index (Fst) estimates for both
microsatellite and mitochondrial are indicated in the top left box. Pairwise
fixation indices are indicated in boxes between herd pairs the highest and
lowest values for each marker type in bold. Herd abbreviations: BT = Big
Thompson, CD = Continental Divide, MM = Mummy, NS = Never
Summer, SV = St. Vrain……………………………………………………...….28
3.
Map of study area showing general geographic ranges for each of the five
bighorn sheep herds. Overall effective migration (Nm) estimates for both
microsatellite and mitochondrial data are indicated in top left box.
Pairwise effective migration estimates are indicated in boxes between herd
pairs with the highest and lowest values for each marker type in bold. Herd
abbreviations: BT = Big Thompson, CD = Continental Divide, MM =
Mummy, NS = Never Summer, SV = St. Vrain.……...…………………...….....29
4.
Isolation-by-distance and reduced major axis regression analysis of
microsatellite genotype data. Plot of log genetic similarity (M) vs. log
geographic distance (km) for all ten bighorn sheep population pairs. Herd
abbreviations: BT = Big Thompson, CD = Continental Divide, MM =
Mummy, NS = Never Summer and SV = St. Vrain. Line of best fit for
these date follows the equation y = -1.28x + 2.55(R2 = 0.167, p = 0.13).…...…..33
5.
Scatter plots showing individual microsatellite deviations from expected
heterozygosity across seven different mutation models of microsatellite
evolution where IAM = Infinite Allele Model, SMM = Strict Stepwise
Mutation Model and TPM = Two-Phase Mutation Model as implemented
in the program BOTTLENECK.…..…………………………………………......35
xiii
6.
BOTTLENECK bar chart of allele distributions by frequency classes from
least common (1) to most common (10) (left); Scatter plot distributions of
individual microsatellite deviations from excepted heterozygosity at
equilibrium (right). Herd abbreviations: Big Thompson (BT), Continental
Divide (CD), Mummy (MM), Never Summer (NS), St. Vrain (SV).…………....36
7.
Isolation-by-distance and reduced major axis regression analysis of
mitochondrial sequence data as implemented in Isolation by Distance Web
Service. Plot of log genetic similarity (M) vs. log geographic distance
(km) for all ten bighorn sheep population pairs. Herd abbreviations: BT =
Big Thompson, CD = Continental Divide, MM = Mummy, NS = Never
Summer and SV = St. Vrain. Line of best fit for these date follows the
equation y = -1.54x + 0.56 (R2 = 0.69, p = 0.02).…………….……………….....40
8.
Comparative cladograms of Caprinae genera based on two regions of the
Y-chromosome gene SRY: the SRY promoter (PRO, left) and the SRY
open reading frame (ORF, right). Node supports are given as posterior
probabilities (PP, Bayesian, above branch) and bootstrap values (ML/MP,
below branch). Nodes with support of less than 50 were collapsed.…………....56
9.
Comparative cladograms of the subfamily Caprinae based on ND5 gene
sequence (left) and combined SRY ORF/PRO gene sequence (right).
Node supports are shown on branches with posterior probabilities
(Bayesian) above the line and bootstrap values (ML/MP) below the line.
Nodes with branch supports all below 50 were collapsed and are not
shown.…………………………………………..………………………….…….57
10.
Cladogram of Caprinae genera based from combined ND5/PRO/ORF
sequence data. Node supports are shown on branches with posterior
probabilities (Bayesian) above the line and bootstrap values (ML/MP)
below the line. Branches with supports all below 50 were collapsed and
are not shown……………………………………………………………………58
xiv
CHAPTER I
The study of evolution has been a central focus of biology and was greatly advanced by
the work of both Charles Darwin (1899-1882) and Alfred Russel Wallace (1823-1913) who each
independently articulated the theory of natural selection as a mechanism by which the
composition of heritable characteristics in a given population changed through time (Darwin
1859, Wallace 1870). At that time, however, the hereditary mechanism of such characteristics
was still unknown. Advancements in the field of genetics, many of which stemmed from the
work of Gregor Mendel (1822-1884), ultimately provided a theory of heredity to explain the
genetic mechanism of inheritance and its relation to the theory of natural selection. The
integration of Mendelian genetics and evolution by natural selection was one of the central
results of the Modern Synthesis (Huxley 1942) which brought together ideas and data from
multiple fields of study.
DEVELOPMENT OF MOLECULAR TECHNIQUES, MARKERS, AND THEIR USES
The development of molecular techniques for quantifying genetic variation has
revolutionized evolutionary studies (Lewontin and Hubby 1966). Before the development of
these methods the primary focus of scientific analyses included morphological, physiological,
and behavioral data to investigate evolutionary processes. Following the discovery of the
structure of DNA (Watson and Crick 1953), protein electrophoresis provided some of the first
data describing allozyme variation in populations based on differences in the number and
composition of amino acids (Manos 1987, Leberg 1992). However, this method had limited
ability to identify variation as it only discerned differences that affected amino-acid identity
1
within protein-coding regions of the genome. This was a significant drawback because not only
is a very small percentage of genomic DNA expressed as proteins, but these regions are often
subject to different selective pressures relative to neutral genetic markers. Subsequent
developments allowing analyses of DNA polymorphisms throughout the genome, including
single nucleotide polymorphisms (SNP), restriction fragment length polymorphisms (RFLP),
variable number of tandem repeats (VNTR), minisatellites and microsatellites, improved both the
breadth and depth of genomic variation detection. The invention of polymerase chain reaction
(PCR) by Kary Mullis in 1983 greatly increased the efficiency of genetic data collection.
Concurrent advances in DNA sequencing, from Sanger sequencing (Sanger et al. 1977), to fully
automated sequencing (Applied Biosystems, 1987), to more recent advances in next generation
sequencing, have resulted in virtually unlimited availability of information for diverse genetic
markers.
This diversity of potential genetic data requires a careful evaluation of which genetic
marker or markers are appropriate for an individual study. The variation among genetic markers
includes differences in ploidy, inheritance pattern, variation, and evolutionary rate.
Mitochondrial markers are among the more common employed, because they are haploid, and
thus easier to amplify and sequence, they are generally highly variable, and they exhibit maternal
inheritance. Mitochondrial DNA analyses include: use as a molecular clock for dating origins
and divergences among taxa (Hasegawa et al. 1985, Cann et al. 1987), for tracking femalemediated gene flow (Richards et al. 2003, Meadows et al. 2006), and for tracing female lineages
and introgression within and between groups (Avise and Ellis 1986, Lehman et al. 1991).
Nuclear autosomal markers exhibit a vast array of variation. Microsatellites are one of the more
popular markers, as they are both highly variable and relatively easy to amplify and score.
2
Because microsatellites are non-coding, they are generally considered neutral and can be used to
accurately evaluate extant genetic variation uninfluenced by selection. Microsatellites have been
used in a wide variety of analyses including evaluation of variation within populations for
analyses of inbreeding depression (Taylor et al. 1994, Luikart et al. 1998), population
substructure and connectivity within metapopulations (Johnson et al. 2003. Palo et al. 2004),
medical evaluation of disease risk (Hearne et al. 1992, Duyao et al. 1993), DNA fingerprinting
for identification of kinship (Queller et al. 1993, Blouin 2003) and forensic investigations
(Norrgard 2008, Katsanis et al. 2013). As a result, Y-chromosome represents a unique genetic
marker in that it is haploid, is generally highly conserved, is paternally inherited, and only exists
in males. Thus, Y-chromosome genetic data can be useful for tracking male-mediated gene flow
and introgression (Kikkawa et al. 2003, Rootsi et al. 2004) and for tracing patrilineal lineages
and origins of species (Hammer et al. 1995, Crusiani 2011). Moreover, precisely because the Ychromosome is unique, Y-chromosome genetic data can be used to compliment or to contrast
with analyses of other markers (Tosi et al. 2000, Wood et al. 2005, Pidancier et al. 2006).
POPULATION GENETICS
One of the fields concerned with evolutionary processes is population genetics, which has
more recently emerged and is mainly a product of the work of three scientists: John Haldane,
Ronald Fisher and Sewall Wright in the 1920s and 1930s. Haldane contributed to the
quantification of population genetic approaches, to the understanding of the effects of mutation
and migration on natural selection, and was the first to define the concept of mutational load to
explain costs associated with natural selection (Haldane 1932). R. A. Fisher was instrumental in
3
advancing statistical approaches for quantifying genetics, including analysis of variance and
maximum likelihood, and his work was the basis for the field of quantitative genetics. He was
also the first to define heterozygote advantage as a result of his work which involved
appreciating that many natural polymorphisms are not selectively neutral and thus selection was
often a significant force in natural populations (Fisher 1930). Wright is credited with defining
the mathematics of genetic drift to explain stochastic processes causing genetic change in
populations across generations as well as F-statistics to explain genetic variation within and
among individuals and populations of the same species.
Thanks in part to advances in molecular genetics, F-statistics are more commonly used
today as a method of comparing estimates of heterozygosity (H) at different levels within a
population in order to evaluate parameters such as inbreeding and the degree of structure within
a metapopulation, defined as a group of populations of the same species that are spatially
separated but interact to some degree, in other words, “A population of populations” (Levins
1969). The term heterozygosity refers to the existence of two different alleles at a particular
genetic locus. Heterozygosity can be used to measure the total proportion of heterozygous loci
within a single individual, and can be extended for use as a population parameter, to measure the
proportion of individuals who are heterozygous for a particular locus in a population. Fis, more
commonly referred to as the inbreeding coefficient, is the F-statistic concerned with comparing
the heterozygosity of a single individual to the heterozygosity of its population. Fis ranges from
negative one (all individuals are heterozygous, no observed homogygotes) to positive one (all
individuals are homozygous, no observed heterozygotes). Fis values near zero are typically
accepted as indicating a reasonably variable, “healthy” population while an Fis approaching
positive one is deemed increasingly inbred and potentially at risk for the effects of inbreeding
4
depression. Inbreeding measures the extent to which two alleles at a particular locus among
individuals are identical by descent. When the number of alleles or the number of individuals in
a population decrease, the degree of homozygosity and inbreeding generally increase. This can
have a significant effect on fitness by increasing the union of recessive deleterious alleles and
thus the occurrence of phenotypic traits less well suited for a particular environment. Several
factors can dramatically impact this relationship, however, including: population size, mutation
rate and levels of current genetic variation, mating structure and departures from non-random
mating (and resultantly effective population size), the type and extent of selection, and stochastic
forces such as genetic drift. Although real world population genetic analyses are complicated by
these factors, increased heterozygosity has been repeatedly linked to increased fecundity
(Cothran et al. 1983, McAlpine 1993, Gemmell and Slate 2006), resistance to infection and
disease (Allison 1954, Gulland et al.1993), life-span (Mitton and Jeffers 1989, Hartl et al. 1992)
and population persistence (Soulé 1976, Brook et al. 2002, Frankham 2005, O’Grady et al.
2006). Thus, the inbreeding coefficient is a commonly estimated population genetic parameter
when evaluating genetic variation in populations.
The F-statistic concerned with the degree of subdivision within a population or
metapopulation, Fst, compares the heterozygosity of a single population to that of the larger
population. Fst ranges from zero (a panmictic population with identical identities and frequencies
of all alleles between all subpopulations) to one (complete genetic division of subpopulations
with no shared alleles). The degree of substructure, or isolation of subpopulations, within a
metapopulation can affect overall heterozygosity, because uneven distributions of alleles will
influence the incidence of different heterozygous combinations and may also cause some alleles
to exist only in certain subpopulations and not in others (sometime referred to as “private
5
alleles”). The result of increasing substructure within a population is that certain heterozygous
combinations never actually occur, decreasing overall heterozygosity, known as the Wahlund
effect, even when the population is in Hardy-Weinberg equilibrium. This can be problematic for
evaluation of genetic variation in natural populations, because, with increased population
substructure (As Fst gets closer to 1), the absolute number and frequencies of alleles in the whole
population will belie the actual heterozygosities of each subpopulation which are, in reality,
much lower than expected by estimates made from data for the whole population.
PHYLOGENETICS
Another field concerned with quantifying the effects of evolutionary processes through
time is phylogenetics. The field of phylogenetics is primarily focused on resolving evolutionary
relationships between groups, or clades, which may comprise any taxonomic unit from a single
species through to entire domains of life. This often involves using cladistics to group organisms
based on their synapomorphies, or shared derived characteristics, in order to infer evolutionary
history from a most recent common ancestor.
One of the earliest phylogenetic theories was proposed by Ernst Haeckle which is
succinctly summarized as “ontongeny recapitulates phylogeny”. This idea centers around the
thought that the way in which organisms develop during their lifetime reflects the evolutionary
history of their species, and thus shared ontogenic traits can be used to infer phylogenetic
relationships among taxa. This theory has since been generally disproved but reflects one of the
main goals of phylogenetics which is to identify diagnostic data that can accurately explain the
evolutionary relationships among groups of organisms at all taxonomic levels.
6
The ability to statistically evaluate support for observed phylogenetic relationships has
been provided by three main methods: Maximum Parsimony (MP), Maximum Likelihood (ML)
and Bayesian Inference (Bayes). MP analyses, as first described by Walter Fitch (1971),
identifies the “most parsimonious” tree which is the tree requiring the fewest state changes to
explain the evolutionary relationships among taxa given observed data. This method of tree
estimation was the first to be widely implemented, mainly due to its relative simplicity and lower
computational intensity, though this does vary depending on the search method (Camin and
Sokal 1965, Fitch 1971). As computational capabilities allowed for more intensive analyses, the
popularity of both ML and Bayesian methods increased. Maximum Likelihood is a method of
tree estimation that seeks to identify the “most likely” phylogenetic tree (or evolutionary
hypothesis) given the observed data using a uniform probability distribution and various
substitution models. Popularized by R. A. Fisher in the early 20th century, ML analyses have
been used to quantify evolutionary relationships for many decades (Edwards and Cavalli-Sforza
1963, Neyman 1974). Bayesian tree estimation methods are based on the theorems developed by
reverend Thomas Bayes (1701-1761) and rediscovered posthumously in 1812 by Pierre-Simon
Laplace (Truscott and Emory 1902). Bayesian inference is closely related to ML methods with
the additional ability to set a priori assumptions to modify probability distributions from the
simple uniform distribution implemented in ML and, like ML, identify of the most likely tree
given the data, though Bayesian analyses also allow assignment of a prior.
Accurately identifying phylogenetic relationships can be complicated by the limited
ability to accurately place clades in the context of larger phylogenies and can be exacerbated by a
number of factors. Time since divergence of related taxa can be an important consideration
when selecting diagnostic data because less variable markers can make it difficult to discern
7
relationships among recently diverged taxa while more variable data can obscure relationships
among taxa that diverged in the more distant past. Horizontal gene transfer, the transmission of
genetic information between unrelated organisms within a single generation via interspecific
hybridization, can confuse evolutionary signals and result in underestimation of time since
divergence between taxa. Although horizontal gene transfer is less common in eukaryotes, it has
been repeatedly documented, particularly between mitochondrial DNA and nuclear DNA (Lopez
et al. 1994, Turner et al. 2003). Loss of information from incomplete fossil records, also known
as taxon sampling error, can also make phylogenetic inferences difficult, and even when the
fossil record is complete, there is often a bias for characters that preserve well, such as teeth or
bones, over those that don’t, such as soft tissues and ontological traits (Kidwell and Holland
2002, Crepet et al. 2004). Consideration of the effects of these potential complications is
important, but, despite this, phylogenetics has provided important data regarding past and present
evolutionary patterns at all taxonomic levels.
Obfuscation of evolutionary signals can also arise due to homoplasy, wherein the same
phenotypic character is independently derived in two separate lineages, often due to convergent
evolution, and not as a result of a derived character from recent common ancestor (Chen et al.
1997, Emery and Clayton 2004). Thus, homoplastic traits do not accurately reflect a shared
evolutionary history and can confuse taxonomies, especially those based on limited data, such as
analyses of just a single character. Discordant phylogenetic signals can also result from analyses
of different marker types, such as morphological data, behavioral data, and genetic data,
reflecting differences in the evolutionary histories of these data. More recently, genetic data has
been predominantly used in phylogenetic studies, and the disparity between analyses of diverse
genetic markers is often due to differences in variation, mutation rate, and inheritance pattern.
8
Phylogenetics is further complicated in that interpretations of analyses are relative to the scope of
inquiry. For example, comparisons between two taxa may appear monophlyletic, paraphyletic,
or polyphyletic depending on scope of analyses and the taxa included. Consideration of the
effects of these potential complications is important, but, despite this, there have been important
advances in the field of phylogenetics regarding past and present evolutionary patterns at all
taxonomic levels.
9
CHAPTER II
INTRODUCTION
Caughley (1994) distilled conservation biology into two topics of which one, the small
population paradigm, concerns the extinction vulnerability of small populations. Among the
potential factors involved in this vulnerability are loss of genetic diversity from genetic drift and
the potential for resulting inbreeding depression (Soulé 1980). The ability of immigration to
counteract the effects of genetic drift in largely isolated populations has long been recognized
(Mills and Allendorf 1996) and has led in part to the important inclusion of landscape-level
metapopulation considerations in conservation biology (Hanski and Gilpin 1997). Potential
negative effects of inbreeding, however, remain a conservation challenge for animal species
living in populations that are both small and more isolated than populations had been historically
(IUCN: http://www.iucn.org, Baillie et al. 2004, Reed 2005, Vilas et al. 2006). Inbreeding
problems associated with decreased heterozygosity can reflect increased expression of
deleterious recessive alleles, exacerbating the causes of inbreeding depression attributed to both
the overdominance and dominance hypotheses (Hedrick and Kallinowski 2000, Reed et al.
2002). Decreased heterozygosity is one factor which has been empirically demonstrated to have
negative effects on individual fitness and long-term persistence of natural populations (Koehn et
al. 1973, Raikkonen et al. 2006, Luikart et al. 2008a). While inbreeding depression is probably
most common and acute in captive breeding populations, such as those in zoos (Ralls and Ballou
1986, Ralls et al. 1988, Frankham 2005, Gilligan et al. 2003, Boaks et al. 2007), it can also occur
in parks and wildlife areas (Keller and Waller 2002, Crnokrak and Barrett 2002, O’Grady et al.
10
2006). If detected, genetic rescue through the influx of genes from nearby populations may be a
possible management strategy to alleviate the costs associated with inbreeding (Hedrick 2001,
Hogg et al. 2006, Bergl et al. 2008).
It has been repeatedly shown that levels of fitness and inbreeding depression vary
between laboratory and wild populations of the same species (Dobzhansky et al. 1963). Thus, it
is important to evaluate the factors relevant to maintenance of genetic variation in natural
settings when considering the risks associate with inbreeding depression. The degree to which
gene flow ameliorates the risk of inbreeding depression in natural populations of plants and
animals is dependent on several demographic factors, including: Life history and size of
individual populations (Barrett et al. 1991, Manel et al. 2003, Angeloni et al. 2011), the distance
between populations, the degree and pattern of individual dispersal among populations,
individual life span, mating and reproductive structure (Hanksi 1998, Saccheri et al. 1998, Manel
et al. 2003). Thus, species that are more continuously distributed generally have higher rates of
gene flow and are less likely to experience decreased heterozygosity.
Bighorn sheep (Ovis canadensis) are a North American species that, because of narrow
habitat requirements, occur in a naturally fragmented pattern of often small, disjunct populations.
Where small, fragmented populations are geographically isolated, genetic drift and inbreeding
will steadily erode genetic diversity (Schwartz et al.1986). Reduction of a population’s genetic
diversity has been documented to result in problems that affect individual and population fitness
(Hogg et al. 2006, Johnson et al. 2011, Rioux-Paquette et al. 2011, Miller et al. 2012). Gene
flow can alleviate some of these problems, however, and bighorn sheep fit a metapopulation
model well, with male biased interdemic movements commonly documented by radio telemetry
(Bleich et al. 1990, 1996). Resulting gene flow among subpopulations has been measured
11
genetically (Epps et al. 2005, 2006, 2007) and is recognized as a critical biological process for
this species that counteracts the effects of genetic drift in functional metapopulations (Schwartz
et al. 1986, Bleich et al. 1990, 1996).
Bighorn sheep in and around Rocky Mountain National Park (RMNP) potentially
represent a similar metapopulation structure as those documented in previous studies. This study
was initiated, in part, as a request for genetic data to test the hypothesis that one of the four
RMNP herds might be suffering from inbreeding depression. The Mummy herd declined in size
following a pneumonia outbreak in the mid 1990s with continued poor recruitment thereafter.
The failure of this herd to rebound to historic numbers could be due to a number of genetic or
environmental causes, of which the genetic hypothesis was most testable.
Testing the hypothesis that the Mummy herd is suffering from inbreeding depression
involves measuring appropriate genetic variables in the larger metapopulation that includes
RMNP and the surrounding area. A complete description of a metapopulation involves genetic
descriptions of the populations and estimates of gene migration among them. Nearby herds
within this metapopulation provide convenient comparisons for population estimates of
heterozygosity (He) and the inbreeding coefficient (Fis) to determine if the population of interest
is genetically depauperate. Estimates of migration among populations will allow managers to
assess whether particular herds are virtually isolated from the others or are regularly receiving
genes from nearby populations.
To test the hypothesis that inbreeding depression was the cause of dwindling herd size
and continued low recruitment in the Mummy herd, we examined data from both mitochondrial
and microsatellite markers to estimate several population genetic parameters including: effective
migration (Nm), heterozygosity, isolation-by-distance (IBD), population substructure (Fst), and
12
inbreeding (Fis). These analyses provide a comprehensive evaluation of both within and among
population genetic variation, the extent to which these individual herds are part of a larger
metapopulation structure, and whether the Mummy herd (or any of the other four herds) is
suffering from the effects of a recent and severe genetic bottleneck.
Like many hypotheses in population biology, an asymmetrical experimental design exists
in this research. A finding that genetic diversity is not compromised, gene flow is common, and
inbreeding coefficients are not remarkable will cleanly disprove the inbreeding hypothesis.
However, an alternative finding consistent with inbreeding does not necessarily imply that
inbreeding is the primary factor driving population dynamics, given numerous other potential
factors. Under such situations, heterozygosity-fitness correlations and genetic rescue
experiments would be needed to further evaluate the role of genetics in population dynamics.
STUDY AREA
Our study was conducted in the Rocky Mountains of north-central Colorado in an area
including RMNP, Indian Peaks Wilderness, Roosevelt National Forest and Comanche Peak
Wilderness (Fig. 1). Five Rocky Mountain bighorn sheep (Ovis canadensis) populations were
included in our analyses based on their presence in and relative proximity to RMNP and
assessment of potential for gene flow within the larger metapopulation. Four of these herds
(Continental Divide (CD), Mummy (MM), Never Summer (NS) and St. Vrain (SV)) maintain
some or all of their annual range within the 415 square miles (1,075 square kilometers) of RMNP
while the fifth (Big Thompson (BT)) maintains an annual range east of RMNP within Roosevelt
National Forest. Historic data identify a single East Side (ES) herd as the combination of two
13
distinct herds, the Mummy and the Cow Creek (CC); however, current management recognizes a
single herd (MM) in this range. In addition, the BT and SV herds were collectively referred to as
the BT herd by wildlife managers prior to 1998 after which the two populations were recognized
as separate herds. Most recent population estimates for the RMNP herds (CD: 120, MM: 64-78,
NS: 180, and SV: 50-100) and the BT herd (85) were based on McClintock and White (2007),
the Colorado Division of Wildlife (CDOW) (George et al. 2009) and RMNP data (Judy Visty
and Cherie Yost, personal communication).
14
Figure 1. Relief map of north-central Colorado, USA showing 26 DNA collection sampling sites
for the five bighorn sheep (Ovis canadensis) herds in and around Rocky Mountain National Park
included in this study. Polygons each represent collection sites for a single herd: BT (circles),
CD (diamonds), MM (squares), NS (hexagons) and SV (triangles).
15
CDOW and Park data show that these five herds have served as either the source or
destination for sheep transplantations 11 times in the past (Table 1). These include six
transplantations from study herds to herds outside our study area (NS: 1979, 1982, and 1984;
CC: 1983, 1988, and 1989), three transplantations between study herds (MM to BT: 1983 and
1985; CC to BT: 1987) and two transplantations into study herds from outside sources (Tarryall
to CC: 1977; Georgetown to BT: 2000). The three transplants of East Side (CC & MM) sheep to
the BT herd represent major induced “migration” events within this metapopulation. The
transplantations of outside sheep into study herds represented about 21% the subsequent CC herd
census (126 sheep) and 31% of the subsequent BT herd census (77 sheep) and thus were
substantial introductions of potentially novel genetic variation into this metapopulation. Based
solely on this historic transplantation data, the MM and BT herds should share a higher
percentage of genetic variation relative to the other three herds. This shared variation could be
predominantly due to mitochondrial haplotypes rather than microsatellite genotypes given that
most of the transplanted sheep were female, bighorn sheep exhibit a highly polygynous mating
structure, and bighorn ewes tend to remain philopatric to natal herds.
16
Table 1. Historical data for sheep translocations involving the five RMNP herds (BT, CD, MM,
NS and SV) movements of sheep into, between, and out of these native herds.
Source Herd Census
Destination Census Year Translocated sheep
Ram Ewe
Yrlng Lamb Total
Sheep movement out of study area
Bristol
ES (CC)
Head
-
1983
3
11
0
5
19
ES (CC)
200(ES)
Glenwood
Canyon
-
1988
-
-
-
-
27
ES (CC)
200(ES)
Lower
Poudre
50
1989
2
9
2
5
18
NS
-
AZ & NV
-
1979
1
11
0
8
20
NS
-
Purgatoire
Canyon
-
1982
2
10
0
5
17
NS
-
Dinosaur
South
-
1984
1
13
0
5
19
Sheep movement within study area
ES (MM)
43
BT
-
1983
2
9
0
8
19
ES (MM)
70
BT
100
1985
5
14
0
7
26
ES (CC)
200(ES)
BT
100
1987
0
9
8
9
26
Sheep movement into study area
Tarryall
CC
-
1977
2
14
0
4
20
Georgetown
50
2000
5
13
0
4
22
450
BT
Previously identified factors influencing bighorn sheep migration and gene flow between
these populations include both natural features (distance, landscape topography, forage and water
availability) as well as artificial features (roads, buildings, cultivated areas and fences). Three
distinct mountain ranges are included within this area: the Mummy Range crossing the
northeastern corner of RMNP; the Continental Divide transecting the Park from North to South;
17
and the Never Summer Range along the western border separating RMNP and Routt National
Forest. In addition, three U.S. highways (7, 34 and 36) and numerous hiking trails transect
annual sheep migration routes at multiple locations throughout the study area. This area also
includes the town of Estes Park, several small communities, and many farms and ranches.
MATERIALS AND METHODS
Bighorn sheep DNA samples were obtained non-invasively by repeated sampling across
36 locations within the study area between 2005 and 2010 (Fig. 1). Fresh fecal pellets were
collected from a single, distinct pile and allowed to dry immediately. The outermost layer of
each pellet was carefully scraped in an attempt to obtain intestinal cells, and total genomic DNA
was extracted from that material with the QIAamp DNA Stool MinikitTM (Qiagen) incorporating
modifications following Wehausen et al. (2004). We found this DNA collection approach,
described in studies of bighorn sheep and other ruminants (Maudet et al. 2004, Luikart et al.
2008b, Poole et al. 2011, Godinho et al 2012), to be an efficient and effective non-invasive
method for isolating usable DNA.
Microsatellite data development - Thirteen known and unlinked dinucleotide
microsatellite loci were selected to develop data on nuclear genetic population structure, from
which gene flow and other population genetic measures could be estimated (Table 2). Four of
these are considered candidate adaptive loci due to locations within known gene regions (ADC,
KERA, OLA and SOMA) of which two (ADC and OLA) are within genes with known diseaserelated function (Schwaiger et al. 1993, Wood and Phua 1994, Crawford et al 1995, Lucy et al.
18
1998; Table 2). The other nine loci are not associated with genes and have consistently met
expectations for neutral markers (Crawford et al. 2006, Luikart and Cornuet 2008b). All loci
were amplified using one of three kits: AmpliTaq Gold® PCR Kit (Applied Biosystems), Type-it
Microsatellite PCR Kit (Qiagen) or Multiplex PCR Kit (Qiagen). Amplifications were
performed with fluorescently labeled forward primers (FAM, HEX or TET) in a final volume of
25µl containing 50ng of template DNA, 2 mM MgCl2, 0.05 mM dNTPs, 0.4µM of each primer
and 0.25 - 2 units of polymerase. General PCR cycling steps were: initial denature at 93οC for
7:30’ followed by 30 cycles of: 95οC for 30”, 55 to 63οC (depending on locus) for 40”, 72οC for
30” and a final extension at 72οC for 2’. Annealing temperatures were optimized for individual
microsatellite loci to maximize primer binding specificity and peak signal strength. PCR
reactions were multiplexed where possible.
19
Table 2. Microsatellite loci general information by locus including: location in genome (Chr.),
range of allele sizes (in base pairs) identified in this study (Range), total number of different
alleles detected (Alleles), PCR annealing temperature in degrees Celsius (Anneal), fluorescent
label for each primer pair (Tag), source reference for locus information (Cited source) and NCBI
accession number (Accession).
Locus
Chr. Range
Alleles Anneal Tag
Cited source
Accession
AE16
13
82-106
11
62
HEX
ADC***
-
85-113
8
62
FAM
FCB266
25
88-102
6
62
FAM
MAF36
22
88-104
8
56
TET
SOMA**
16
94-118
7
59
TET
MAF209
17
107-119
6
62
TET
HH62
16
108-122
9
56
HEX
MAF33
9
123-131
6
56
HEX
MAF48*
-
123-133
7
62
FAM
FCB304
19
134-150
9
62
FAM
TCRBV62**
4
167-175
4
62
FAM
KERA**
-
175-179
3
62
HEX
OLADRBps***
_
278-302
8
62
HEX
Penty et al. 1993
Crawford et al.
1995
Buchanan and
Crawford, 1993
Swarbrick et al.
1991
Lucy et al. 1998
Buchanan and
Crawford, 1992a
Ede et al. 1994
Buchanan and
Crawford, 1992b
Buchanan et al.
1991
Buchanan and
Crawford, 1993
Buitkamp et al.
1993
Tasheva et al.
1998
Beraldi et al.
2006
L11047
N/A
L01535
M80519
U15731
M80358
N/A
M77200
M62645
L01535
L18957
AF036962
AH003856
*Locus location unknown; **Loci within a gene region; ***Loci within disease-related
gene region.
20
PCR products were electrophoresed with TAMRA size standards (350 or 500) on an ABI
Prism 377 automated sequencer and chromatograms were analyzed using GENESCAN software
(Applied Biosystems). Samples were genotyped independently two and five (average four)
times to minimize errors due to allelic drop-out and/or false alleles. Duplicate samples with the
same genotypes were identified and removed from analyses.
Microsatellite data analyses - Microsatellite data were used to estimate heterozygosity
(He) and deviations from Hardy-Weinberg equilibrium as implemented in GENEPOP Web
version 4.2 (Raymond and Rousset 1995, Rousset 2008). Independent analyses were performed
on three different data sets: A) all 13 microsatellites, B) the nine neutral loci, and C) the four
candidate adaptive loci, in order to evaluate whether these microsatellites are all selectively
neutral or if there is evidence of selection acting on any individual locus. These data were also
used to estimate overall and pairwise genetic distances (Fst) and overall and per locus inbreeding
coefficients (Fis). Both overall and pairwise effective migration rates (Nm) were estimated using
the private alleles method (those alleles present in just a single herd) following Barton and
Slatkin (1986).
The relationship between genetic similarity and geographic distance was analyzed in the
program Isolation by Distance Web Service (IBDWS) version 3.23 (Jensen et al. 2005).
Approximate geographic coordinates (latitude, longitude) for all sample collection sites were
determined using Google Earth. Pairwise distances among herds were measured from the
estimated center of each herd’s annual range using the spherical distance (in kilometers) between
coordinate pairs. Genetic relatedness was estimated following Slatkin’s measure of genetic
similarity, defined as M = (1/Fst – 1)/4 (Slatkin 1993). The correlation between geographic
distance and genetic distance was evaluated statistically with a Partial Mantel test using 10,000
21
iterations as implemented in IBDWS. The significance of this relationship was evaluated by
regressing the log10 transformed pairwise geographic distances against the corresponding log10
transformed genetic distances using the Reduced Major Axis (RMA) regression.
Genetic structure was also examined using a model-based Bayesian clustering method as
implemented in STRUCTURE 2. 2 (Pritchard et al. 2000). This method clusters samples into K
populations based on their microsatellite genotypes and the allele frequencies at each locus in a
way that minimizes both Hardy-Weinberg disequilibrium and linkage disequilibrium (LD). K
was set to 1-8 and for each K the program was run with 750,000 MCMC steps after a burn-in of
250,000 iterations using the admixture model with correlated allele frequencies and no prior
sample information. The Evanno method was used to correct for overestimation of K and the
most likely number of populations was estimated from the rate of change in log probabilities
between each K evaluated (Evanno et al. 2005). Our a priori expectation was significant support
for K= 5 herds if genetic differentiation (Fst) was high and effective migration (Nm) was low
among herds.
We evaluated whether a recent (2Ne - 4Ne generations) and severe reduction in effective
population size (Ne) had occurred which would provide evidence for a genetic bottleneck as
implemented in the program BOTTLENECK (Cornuet and Luikart 1996, Luikart and Cornuet
1998, Luikart et al. 1998). This analysis operates on the principle that allele number (Na) is
reduced faster than observed heterozygosity immediately after a bottleneck. Evidence of a
bottleneck is evaluated by comparison of observed heterozygosity (He), calculated from allele
frequencies, with heterozygosity expected at mutation-drift equilibrium (Heq), calculated from
allele number. Analyses of data were run under three mutation models: Infinite allele model
(IAM), the stepwise mutation model (SMM), and the two phase model (TPM) that combines
22
IAM and SMM. We evaluated data under seven different model parameter combinations: TPM
with 5% IAM and variance of 12, TPM with 5% IAM and variance of 30, TPM with 10% IAM
and variance of 12, TPM with 10% IAM and variance of 30, TPM with 20% IAM and variance
of 12, TPM with 20% IAM and variance of 30 and a strict SMM model. These parameter
combinations were selected based on author recommendations (Piry et al. 1999) and previous
work in bighorn sheep (Ramey et al. 2000, Luikart et al. 2011) and all were run with 10,000
iterations. BOTTLENECK also evaluates allele frequency distributions graphically via a
qualitative “mode-shift” indicator of bottlenecked populations identified by a disproportionate
loss of rare alleles. The Wilcoxon test was used to identify significant deviations from expected
heterozygosities at mutation-drift equilibrium.
Mitochondrial DNA data development - In order to increase the likelihood of obtaining
data with optimal variation for high population genetic resolution, we selected two regions of the
mtDNA genome for inclusion in these analyses. Both Cytochrome Oxidase I (COI) and the DLoop control region (DL) had been previously analyzed in multiple Ovis species, providing
potentially useful comparative data for future analyses. Sequence amplification included 1,053
bp of COI and 963 bp of the DL using the AmpliTaq Gold® PCR Kit with individual reaction
composition as previously described for microsatellite PCR. COI was amplified with primers
COIF (5’-GCAGAGTTTGAAGCTGCT-3’) and COIR (5’-AGCTGACGTGAAGTAAGC-3’)
(Hiendleder et al. 1998) using a PCR profile which included an initial denature at 94οC for
7.5min followed by 35 cycles of: 94οC for 1min, 60οC for 70sec, 72οC for 90sec and a final
extension at 72οC for 5min. DL was amplified with primers P1F (5’CAACACCCAAAGCTGAAGTTC-3’) and P4R (5’-CTAGGCATTTTCAGTGCCTTGC-3’)
(Zhao et al. 2001) using a PCR profile which included an initial denature at 94ο for 7:30’
23
followed by 30 cycles of: 94οC for 1min, 62οC for 70sec, 72οC for 70sec and a final extension at
72οC for 5min. All sequences were aligned and analyzed in SequencherTM version 4.6 (Gene
Codes Corp.). COI and DL sequences were each analyzed separately and then concatenated into
a single haplotype per individual for all subsequent analyses.
Mitochondrial DNA data analyses - Molecular diversity for each population was
evaluated for several parameters including: the number of haplotypes (h), haplotype diversity
(Hd), nucleotide diversity (π), effective migration (Nm) and population differentiation (Fst) in the
program DnaSP (Librado and Rozas 2009). The correlation between log genetic similarity (M)
and log geographic distance (km) was analyzed in the program Isolation by Distance Web
Service (IBDWS) version 3.23 (Jensen et al. 2005) as described for microsatellites above with
the addition that mtDNA data was analyzed using both traditional Fst as well as Φst to allow
incorporation of sequence distance information.
RESULTS
Microsatellite data analyses - Microsatellite data yielded genotypes for 137 individuals
after removal of poor quality amplifications and duplicate samples (Table 3). Observed
heterozygosity, averaged across all thirteen loci, ranged from 0.59 to 0.72 and no estimates
deviated significantly from expected under Hardy-Weinberg equilibrium (Table 4). Overall and
pairwise population subdivision, as estimated by Fst, was low and comparable among all three
data sets (13, nine, and four loci respectively), and we therefore present results for analyses of
the complete 13 loci genotypes. Overall and pairwise Fst estimates were all low (0.02-0.09 for
24
pairwise; Table 5, Fig. 2). Likewise, pairwise and overall Fst estimates from individual loci were
all near zero (Fst = -0.021 to 0.158 & Fst = 0.023 to 0.070, respectively).
Table 3. Data collection summary by marker type and individual herd. Number of sample
collection sites (N), total microsatellite genotypes (Microsatellite), total mitochondrial
Cytochrome Oxidase I amplifications (mtDNA COI) and total mitochondrial D-Loop control
region amplifications (mtDNA DL) for each of the five bighorn sheep herds (BT: Big
Thompson, CD: Continental Divide, MM: Mummy, NS: Never Summer, SV: St. Vrain) in northcentral Colorado, USA.
Herd
N
Microsatellite
mtDNA COI
mtDNA DL
BT
3
15
22
22
CD
12
48
70
76
MM
10
28
36
42
NS
6
29
50
52
SV
5
17
31
34
Total
36
137
209
226
25
Table 4. Genetic variation estimates of microsatellite genotype data, including range and
average number of alleles per locus (Na), Heterozygosity excess (He), Heterozygosity expected
at equilibrium (Heq), probability of heterozygosity excess (One-tailed Wilcoxon test, P(e)),
probability of heterozygosity excess or deficiency (Two-tailed Wilcoxon Test, P(w)) and
probability of heterozygosity excess under TPM (Sign Test, P(s))
Herd
Na
He
Heq
Fis
P(e)
P(w)
P(s)
BT
2-7 (4.5)
0.66
0.61
0.09
0.08
0.17
0.55
CD
3-14 (7.4)
0.72
0.70
0.09
0.17
0.34
0.33
MM
4-7 (5)
0.59
0.60
-0.01
0.68
0.68
0.48
NS
3-8 (5)
0.66
0.63
-0.01
0.37
0.73
0.54
SV
3-7 (4.23)
0.63
0.59
0.04
0.17
0.34
0.55
Table 5. Pairwise fixation indices (Fst; left side) and effective migration rates (Nm; right side)
estimated from microsatellite genotypes (above diagonal) and mitochondrial haplotypes (below
diagonal) among all five bighorn sheep herd pairs in north-central Colorado, USA.
Fst
BT
CD
MM
NS
SV
Nm
BT
CD
MM
NS
SV
BT
–
0.02
0.03
0.09
0.02
BT
–
1.91
2.66
1.11
1.67
CD
0.26
–
0.03
0.03
0.04
CD
2.25
–
3.40
4.28
1.70
MM
0.25
0.02
–
0.09
0.05
MM
2.01
2.01
–
2.16
1.15
NS
0.44
0.10
0.18
–
0.09
NS
1.74
13.14
1.80
–
1.04
SV
0.25
0.16
0.19
0.19
–
SV
2.57
2.62
1.27
3.09
–
Estimates of both overall and pairwise effective migration (Nm) were comparable for
analyses of the four, nine and 13 loci data sets (Nm = 8. 96, range: 0.79 to 6.06, Nm = 8.82,
range: 0.96 to 4.41 and Nm = 8.82, range: 1.04 to 4.28 respectively), and we therefore present
26
data for all 13 loci. Effective migration was moderate to high for both overall and pairwise herd
comparisons (Table 5, Fig. 3). These estimates suggest that gene flow is likely contributing to
low population subdivision among the five herds.
27
Figure 2. Map of the study area showing general geographic ranges for each of the five bighorn
sheep herds. Overall fixation index (Fst) estimates for both microsatellite and mitochondrial are
indicated in the top left box. Pairwise fixation indices are indicated in boxes between herd pairs
the highest and lowest values for each marker type in bold. Herd abbreviations: BT = Big
Thompson, CD = Continental Divide, MM = Mummy, NS = Never Summer, SV = St. Vrain.
28
Figure 3. Map of study area showing general geographic ranges for each of the five bighorn
sheep herds. Overall effective migration (Nm) estimates for both microsatellite and
mitochondrial data are indicated in top left box. Pairwise effective migration estimates are
indicated in boxes between herd pairs with the highest and lowest values for each marker type in
bold. Herd abbreviations: BT = Big Thompson, CD = Continental Divide, MM = Mummy, NS =
Never Summer, SV = St. Vrain.
29
Inbreeding, as measured with the inbreeding coefficient (Fis), was estimated for each herd
using three data sets consisting of four, nine and 13 loci (Table 6). Results between analyses
were comparable for each herd and no large deviations were observed between these three
estimates. All Fis values were near zero as expected for non-bottlenecked herds with high gene
flow; these estimates show no evidence of inbreeding.
Table 6. Coefficient of inbreeding (Fis) estimates for three microsatellite data subsets including:
genotypes for only the four potentially “non-neutral” loci: OLA, KERA, SOMA and TCRBV
(Four loci), genotypes for the nine putatively “neutral” loci: ADC, AE16, FCB266, FCB304,
HH62, MAF33, MAF36, MAF48 and MAF209 (Nine loci) and genotypes from the complete
data set of all 13 Loci (All loci) for five bighorn sheep herds in north-central Colorado, USA.
Coefficients of inbreeding (Fis)
Herd
Four loci
Nine loci
All loci
Big Thompson
–0.104
0.164
0.090
Continental Divide
0.117
0.067
0.090
Mummy
–0.068
0.019
–0.010
Never Summer
–0.047
0.019
–0.010
St. Vrain
0.118
0.005
0.040
Isolation-by-distance analyses of the 10 herd pairs indicated a trend towards a negative
relationship between log genetic relatedness (M) and log geographic distance (R2 = 0.167, p =
0.130, y = -1.28x + 2.55), but no statistically significant relationship was revealed (Table 7). We
find that the herd pair with the lowest M is also the farthest geographically (NS/BT; 47.54 km),
30
while the pair with the highest M is the nearest geographically (CD/NS; 11.58 km); however,
this trend is not consistent across all herd pairs (Fig. 4).
31
Table 7. Pairwise isolation-by-distance (IBD) estimates from mitochondrial and microsatellite
data for among all five bighorn sheep herds.
Log
Log
Log
Geographic
Genetic
Genetic
geographic
genetic
genetic
Herd pair distance
similarity
similarity
distance
similarity
similarity
(km)
(M)
(M)
(km)
(M)
(M)
Mitochondrial data
Microsatellite data
BT/CD
37.33
1.57
0.47
-0.33
10.4
1.017
BT/MM
25.26
1.40
2.08
0.32
8.46
0.927
BT/NS
47.54
1.70
0.39
-0.42
2.58
0.412
BT/SV
24.37
1.39
0.75
-0.13
11.63
1.066
CD/MM
13.835
1.141
0.746
-0.127
8.1
0.908
CD/NS
11.577
1.064
4.151
0.618
6.92
0.840
CD/SV
31.828
1.503
0.834
-0.079
6.92
0.840
MM/NS
22.501
1.352
0.67
-0.174
2.56
0.669
MM/SV
28.568
1.456
0.714
-0.146
4.67
0.669
NS/SV
43.318
1.637
0.691
-0.161
2.66
0.425
32
Figure 4. Isolation-by-distance and reduced major axis regression analysis of microsatellite
genotype data. Plot of log genetic similarity (M) vs. log geographic distance (km) for all ten
bighorn sheep population pairs. Herd abbreviations: BT = Big Thompson, CD = Continental
Divide, MM = Mummy, NS = Never Summer and SV = St. Vrain. Line of best fit for these date
follows the equation y = -1.28x + 2.55(R2 = 0.167, p = 0.13).
33
Qualitative evaluation of overall metapopulation substructure (from Q plots), as
evaluated in STRUCTURE, shows no clear genetic division between herds for any of the three
microsatellite data sets (four, nine and 13 loci). Maximum likelihood (ML) values for predicted
herd number (K) were not significantly different between K = four, five or six herds, indicating
low confidence in assignment of individuals to specific herds based on microsatellite genotypes.
These data further support Fst and Nm estimates, suggesting low metapopulation substructure
among these five herds.
Of the 455 microsatellite heterozygosity (He) estimates obtained from BOTTLENECK
across all herd/locus/model combinations, only 63 deviated significantly from expected
heterozygosity at mutation-drift equilibrium (Fig. 5). Within these deviations we did not find a
consistent pattern for any specific locus among herds or across multiple loci for any single herd.
The largest number of deviations was found in the Mummy herd (19 of 91) and the fewest in the
Big Thompson herd (7 of 91). For each individual herd the scatter plots patterns were consistent
across all seven microsatellite mutation models (Fig. 6). We identified a clear positive
correlation between increasing percentage of stepwise mutations (SMM) in the two-phase
mutation model (TPM) and number of significant deviations from expected heterozygosity. This
is consistent with the general understanding that the strict SMM is a poor model for describing
microsatellite evolution. In addition, the Sign Test revealed no significant heterozygosity excess,
and the Wilcoxon test found no significant deviations from expected heterozygosity for any of
the five herds under any of the seven mutation models tested (Table 4). The “mode-shift” bar
graphs showed normal, L-shaped distributions of microsatellite variation by frequency class
across all five herds, with the possible exception of the Big Thompson herd, but this is most
likely due to small sample size.
34
Figure 5. Scatter plots showing individual microsatellite deviations from expected
heterozygosity across seven different mutation models of microsatellite evolution where IAM =
Infinite Allele Model, SMM = Strict Stepwise Mutation Model and TPM = Two-Phase Mutation
Model as implemented in the program BOTTLENECK.
35
Figure 6. BOTTLENECK bar chart of allele distributions by frequency classes from least
common (1) to most common (10) (left); Scatter plot distributions of individual microsatellite
deviations from excepted heterozygosity at equilibrium (right). Herd abbreviations: Big
Thompson (BT), Continental Divide (CD), Mummy (MM), Never Summer (NS), St. Vrain (SV).
36
Mitochondrial sequence variation and genetic diversity - High quality sequence data for
both COI and DL were obtained from 211 DNA samples and, after removal of duplicates, our
final mitochondrial data set included complete haplotypes for 190 individuals from all five
populations (Table 3). The full length 1,937 bp sequence contained 105 polymorphic sites, five
within COI and 100 within DL. A single invariant 79 bp indel (insertion or deletion) was
identified within the DL sequence and found to be polymorphic in the Big Thompson, Mummy
and St. Vrain herds while only the deletion was detected in the Continental Divide and Never
Summer herds. A total of twelve unique haplotypes were identified from among these herds
(Table 8).
Table 8. Distribution and abundance of each of the twelve mitochondrial haplotypes identified
in this study among all five herds of bighorn sheep in north-central Colorado, USA.
Haplotype Big Thompson Continental Divide Mummy Never Summer St. Vrain Total
1
13
2
21
0
1
37
2
7
0
0
0
16
23
3
0
11
7
10
0
28
4
0
36
0
13
0
49
5
0
9
1
16
9
35
6
0
10
0
0
0
10
7
0
1
0
0
0
1
8
0
1
0
0
2
3
9
0
0
1
0
0
1
10
0
0
0
1
0
1
11
0
0
0
1
0
1
12
0
0
0
0
1
1
Total
20
70
30
41
29
190
37
No single haplotype was shared among all five herds. Six of the 12 total haplotypes were
found in only a single herd. Overall haplotype diversity (Hd) and nucleotide diversity (π) were
0.826 and 0.129 respectively. The highest haplotype and nucleotide diversities were found in the
Never Summer herd while the lowest were found in the Big Thompson herd (Table 9).
Table 9. Estimates of genetic diversity for the mitochondrial DNA haplotypes comprising 1,937
bp of Cytochrome Oxidase I (COI) and D-Loop (DL) sequences for each of the five bighorn
sheep herds from north-central Colorado, USA including total sample size (N), segregating sites
(S), number of haplotypes (h), haplotype diversity (Hd), average number of nucleotide
differences (K), nucleotide diversity (π).
Herd
N
S
h
Hd
K
π
Big Thompson
20
3
2
0.479
1.44
0.014
Continental Divide
70
42
7
0.682
12.07
0.115
Mummy
30
38
4
0.469
13.18
0.130
Never Summer
41
40
5
0.703
16.56
0.159
St. Vrain
29
23
6
0.633
9.88
0.093
Total
190
45
12
0.826
13.03
0.129
Overall population subdivision was much higher for mitochondrial haplotypes (Fst =
0.25) than for microsatellite genotypes. The range of pairwise Fst estimates was likewise larger
(Fst = 0.02 to 0.26) with a comparable low value but a much larger high value than for
microsatellite data. The lowest pairwise estimate was between CD and MM and the largest was
between BT and NS (Table 5, Fig. 2). The overall effective migration among herds estimated
from mitochondrial sequences (Nm = 8.82) was lower than from microsatellite genotypes (Fig.
3). The range of pairwise effective migration estimates between herd pairs was likewise larger
than for microsatellite data with a similar low but a much larger high value (Table 5, Fig 3).
38
Isolation-by-distance analyses of the 10 herd pairs indicated a significant negative
correlation between genetic relatedness (M) and log geographic distance for all four analysis
combinations of geographic distance or log geographic distance vs. genetic relatedness or log
genetic relatedness (Table 7). These data were also consistent for analyses using both Fst and Φst.
Consistent with our microsatellite data, we find that the herd pair with the lowest M is also the
farthest geographically (NS/BT) while the pair with the highest M is the closest geographically
(CD/NS), however this trend is not consistent across all herd pairs and these results may reflect
influences of translocations to BT and MM from outside herds (Fig. 7).
39
Figure 7. Isolation-by-distance and reduced major axis regression analysis of mitochondrial
sequence data as implemented in Isolation by Distance Web Service. Plot of log genetic
similarity (M) vs. log geographic distance (km) for all ten bighorn sheep population pairs. Herd
abbreviations: BT = Big Thompson, CD = Continental Divide, MM = Mummy, NS = Never
Summer and SV = St. Vrain. Line of best fit for these date follows the equation y = -1.54x +
0.56 (R2 = 0.69, p = 0.02).
40
DISCUSSION
The distribution of bighorn sheep is typically naturally fragmented due to specific habitat
requirements, and population sizes are typically small, making bighorn sheep susceptible to
genetic drift, which can result in loss of allelic variation and decreased heterozygosity (Schwartz
et al.1986, Bleich et al. 1990, Gutiérrez-Espeleta et al. 2000, Berger 1990). However, migration
and resulting gene flow among neighboring herds can ameliorate the effects of genetic drift by
introducing new alleles and contributing to the maintenance of genetic variation in individual
herds. We found no evidence of a recent genetic bottleneck or severe reduction in genetic
diversity for any of these five Rocky Mountain bighorn sheep herds in and near RMNP. Our
data do not support the hypothesis that inbreeding depression was a factor in the failure of the
Mummy herd to recover to historic numbers following the pneumonia outbreak in the mid 1990s.
Our data also suggest that heterozygosity within these five herds is maintained by gene flow
within this metapopulation.
Heterozygosity is an important component of population genetic analyses, because it is
often positively correlated with increased fitness, resistance to disease and long-term persistence
of natural populations (Keller and Waller 2002, Jannikke et al. 2006, Xu et al. 2007, O’Grady et
al. 2006). Positive correlations between heterozygosity and both fitness (FitzSimmons et al.
1995) and resistance to disease have also been documented in bighorn sheep (Coltman et al.
1999, Luikart et al. 2008a). Disease is often a significant factor in long-term persistence, of
bighorn populations because they are particularly susceptible to pneumonia outbreaks which can
result in large die-offs or local extinctions (Bailey 1986, Shackleton et al. 1999, Cassirer and
Sinclair, 2007, Wehasausen et al. 2011). We found expected heterozygosities in RMNP (He =
41
0.59-0.72) to be comparable to the highest values reported in other studies of native bighorn
sheep populations (Boyce et al. 1997, Forbes and Hogg 1999, Epps et al. 2010, Johnson et al
2011, Luikart et al. 2011) and higher than those reported for restored bighorn sheep populations
(Whitaker et al. 2004).
Our inbreeding coefficient estimates support these heterozygosity analyses in that no
significant loss of heterozygosity, as indicated by a positive Fis, was detected. All Fis estimates
were near zero (-0.01 – 0.09) suggesting that no herd has experienced a recent genetic bottleneck
which would be indicated by a disproportionate loss of rare alleles. It is therefore unlikely that
any of these herds are suffering from the effects of inbreeding depression. Factors other than
genetic diversity are therefore likely responsible for the Mummy herd’s failure to thrive.
Metapopulation analyses suggest that high heterozygosity is being maintained by gene
flow among these five herds. Estimates of population substructure from microsatellite data were
low (Fst = 0.02 - 0.09, Fig. 2), effective migration rates were reasonably high (Nm = 1.04 to 4.28,
Fig. 3), and no significant correlation was found between allelic distribution and geographic
distance (Fig. 4). In contrast, mitochondrial data showed higher genetic isolation among herds
(Fst = 0.02 to 0.44, Fig. 2) and a significant negative correlation between haplotype distribution
and geographic distance (Fig. 6). The ranges of both effective migration rates (Nm = 1.04 to
13.14) and population substructure were larger for mitochondrial data than for microsatellite
data, suggesting that ewes preferentially move between certain herds more than between others.
Additionally, the pattern of population substructure is asymmetrically distributed across this
metapopulation. The two lowest Fst estimates involved the CD herd (CD/MM = 0.02, CD/NS =
0.10) which could be attributed increased migration between these herds due to shared lambing
areas where herds overlap, such as at Specimen Mountain on the west side of RMNP, which
42
could also explain the extremely high effective migration estimate between the NS and CD herds
(Nm = 13.14). The four highest Fst estimates were between the BT herd and each of the Park
herds (BT/NS = 0.44, BT/CD = 0.26, BT/MM = 0.25, BT/SV = 0.25) which could indicate that
ewe-specific movement is inhibited to some extent by the town of Estes Park (population 6,017),
which lies directly between RMNP and the BT herd.
Recent research has shown that both slope and ruggedness of terrain are significant
factors in bighorn habitat selection (Sappington et al. 2007) and that highways block gene flow
in this species (Epps et al. 2005). This could help explain the limited dispersal of ewes across
the relatively flat landscape between RMNP and the BT range which includes highways 36, 14
and 7. In contrast, our microsatellite data do not show this same disparity between movement
among Park herds and movement between RMNP herds and the BT herd, suggesting that ram
dispersal across populated areas is not inhibited to the same extent as ewe dispersal or that
motivation to migrate is higher in rams because of the potential for increased mating success in
non-natal herds.
The disparity between Fst estimates from microsatellite and mitochondrial data could also
be explained to some extent by differences in the evolution and inheritance patterns of these two
markers. Genes located on the mitochondrial chromosome are vital to metabolic processes
involved in aerobic respiration and are thus more conserved than microsatellite alleles which are
neutral and rapidly mutate via replication slippage and are thus generally highly variable in
populations. Mitochondria are also haploid and maternally inherited while microsatellites are
diploid and biparentally inherited. Thus, when the sex ratio is equal, mitochondrial markers have
an effective population one quarter that of microsatellite markers (Birky et al. 1983). However,
this disparity is offset, to some extent, by the highly polygynous mating structure of bighorn
43
sheep in which most adult ewes reproduce each year while generally only the most dominant
rams successfully mate (Geist 1966, Geist 1971, Coltman et al. 2005).
A closer look at mitochondrial haplotype distributions also reveals an unexpected pattern
among these herds in that haplotypes one and two are the only haplotypes detected in the BT
herd while the NS herd, at the opposite end of this metapopulation, has five haplotypes, but is
lacking haplotypes one and two (Table 8). Thus, our data show no haplotypes in common
between the BT and NS herds, which suggests that females may not have moved between these
populations and bred successfully in recent history. The mtDNA differentiation can be seen
more clearly if haplotypes one and two are pooled into "haplotype E" and all other haplotypes
are pooled into "haplotype W", so that the BT herd is fixed (frequency is 100%) for the E
haplotype, and the NS herd is fixed for the W haplotype. If we use Sewall Wright's calculation of
Fst to compare these populations for the pooled data, then Fst = 1.0. But when we use Nei's
formulation (Nei et al. 1975) for multiallelic data, the value is Fst = 0.25. This is because Nei's
method, used most commonly for multiallelic data, considers only the average decrement in
heterozygosity within sub populations in comparison to heterozygosity in the full population, but
it does not measure which alleles are present in the various subpopulations, or the degree of
allelic differentiation among sub populations. This complete mitochondrial differentiation
between the NS and BT herds raises questions concerning whether the intervening territory
marks an historic boundary to ewe-mediated gene flow or whether this differentiation results
from the genetic signature of novel mitochondrial haplotypes introduced via sheep translocation
from the Georgetown herd and that gene flow has yet to occur to any great extent since then.
Interpretation of these metapopulation substructure data must be tempered, however, by
the history of sheep translocations involving study herds (Table 1). Starting in 1977, 20 sheep
44
from the Tarryall population (South Platte River Canyon, CO) were introduced into the MM
herd. During the next decade, a total of 71 sheep were translocated between the MM herd to the
BT herd on three separate occasions; in 1983 (19 sheep), in 1985 (26 sheep) and in 1987 (26
sheep). The most recent translocation occurred in 2000 when 22 sheep from the Georgetown
population (Georgetown, CO) were introduced into the BT herd. These translocations were
primarily composed of ewes, between 59 and 99 of the total 113 sheep moved, depending on the
unknown sexes of both yearlings and lambs (George et al. 2009).
This, and the highly polygynous mating structure of bighorn sheep, leads to two
predictions. First The MM and BT herds should each have a higher proportion of specific
mitochondrial haplotypes and microsatellite alleles compared with other herds due to the induced
migration from the Tarryall and Georgetown herds. If incorporated into local gene pools, this
induced migration should result in higher pairwise Fst estimates between each of these two herds
and the other three (CD, NS and SV) herds. Second, due to the repeated translocation of sheep
(primarily ewes) from the MM herd to the BT herd, these herds should exhibit a higher
proportion of shared genetic variation, primarily mitochondrial haplotypes, resulting in lower Fst
estimates relative to the other herds. Our data, however, do not match either of these predictions.
Although BT mitochondrial pairwise Fst estimates are the highest in this metapopulation, the
BT/MM estimate is also high. The corresponding BT microsatellite pairwise Fst estimates are
all low and comparable with other pairwise values (Figure 3). MM mitochondrial pairwise Fst
estimates are variable, ranging from 0.02 to 0.26 and the Mm microsatellite pairwise Fst
estimates are also low and comparable with others in this metapopulation. This could be
attributed to low transplantation success rates which have been repeatedly documented in
bighorn, to the loss of transplanted sheep subsequent loss during the pneumonia induced die-off.
45
Taken together, these data indicate that the degree to which translocations of sheep have
influenced our results is unclear.
Quantification of population genetics for natural populations is hindered by a number of
factors, including incomplete historical data, uneven sampling, potentially unrealistic
assumptions for parameter estimations, and potential species-specific complications. Thus,
interpretation of population genetic data can be complex and, while comprehensive evaluation of
different genetic markers can help elucidate important patterns, data limitations need to be
acknowledged. Although the transplantation history makes some interpretations difficult, our
analyses of both mitochondrial and microsatellite data nevertheless found that the five Rocky
Mountain bighorn sheep herds have high heterozygosity, show no evidence of inbreeding, and
exhibit low metapopulation substructure due to on-going gene flow, a finding consistent with
other analyses of bighorn sheep metapopulations with unimpaired migration. While our data are
clearly not consistent, they do clearly disprove the hypothesis that decreased genetic variation is
responsible for the Mummy herd’s failure to thrive and suggest that other factors should be
considered.
46
CHAPTER III
INTRODUCTION
The subfamily Caprinae (Artiodactyla, Bovidae) arose around 14.1 MYA in Asia with
subsequent radiation around the world (Hernández-Fernández and Vrba 2005). The Caprinae
includes sheep, goats and closely related horned species that are often characterized by
adaptations to northern latitudes and high altitudes. Unlike other, more homogeneous taxonomic
groups, genera of Caprinae are characterized by a diverse set of morphological traits and do not
share any unambiguous synapomorphies (Gentry 1992), which may explain why many
taxonomies have been proposed for this group and a consensus has not been reached. In the last
fifty years many different genera have been proposed within this group. Of these, 10 to 12 are
currently generally accepted: Ammotragus (aoudad), Budorcas (takin), Capra (goat, ibex,
markhor, tur), Capricornis (serow), Naemorhedus (goral), Oreamnos (mountain goat), Ovibos
(musk ox), Ovis (true sheep), Pseudois (bharal, blue sheep) and Rupicapra (chamois). The
additional genus Hemitragus (tahr) was recently found to be polyphyletic and two additional
genera, Nilgiritragus (Nilgiri tahr) and Arabitragus (Arabian tahr), have been added as a result
(Ropiquet and Hassanin 2005b). Both of these newer genera have been subsequently supported
by additional data (Shafer and Hall 2010, Bibi et al. 2012). Inclusion of two additional genera,
Saiga (saiga) and Pantholops (chiru), has been historically debated (Simpson 1945, Nowak
1991, Gentry 1992, Groves and Shields 1996, Gatesy et al. 1997, Hassanin et al. 1997, Hassanin
and Douzery 1999), yet more recent molecular analyses have placed Saiga with Gazella in the
subfamily Antilopinae (Ropiquet and Hassanin 2005a), while Pantholops has been supported as
47
sister to all other Caprinae genera (Hassanin et al. 2009, Shafer and Hall 2010, Bibi et al. 2012,
but see Arif et al. 2012).
Not only are the absolute number of genera of the Caprinae debated, but also the
evolutionary relationships among them, for which many different taxonomies have been
proposed based on a diverse set of morphological, behavioral, ecological, and genetic data.
These analyses have led to taxonomies which range from including all genera within a single
tribe (Caprini) (Hassanin and Douzery 1999) to dividing genera among as many as four different
tribes (Caprina, Ovibovini, Panthalopina, and/or Rupicaprini) (Simpson 1945, Groves and
Shields 1996, Gatesy 1997, Hassanin et al. 2009, Arif et al. 2012, Bibi et al. 2012). One reason
for the incongruence of findings among studies is that very few have included most or all of the
genera in this subfamily, instead focusing on the placement of only one or a few taxa. In
addition, most genetic analyses to date have focused solely on mitochondrial markers, with the
majority of these focused on a single gene (Groves and Shields 1996, Gatesy et al. 1997,
Hassanin et al. 1998, Hassanin and Douzery 1999, Hassanin et al. 2009, An et al. 2010, Bibi et
al. 2012), while only a few included multiple mitochondrial markers (Hassanin and Douzery,
2003, Ropiquet and Hassanin 2005a) or whole mitogenome analyses (Hassanin et al. 2009, Arif
et al. 2012). At the same time, relatively few nuclear markers have been developed for these
genera (Matthee and Davis 2001, Ropiquet and Hassanin 2003, Ropiquet and Hassain 2005b,
Ropiquet and Hassanin, 2006, Pérez et al. 2011), and studies comparing nuclear and
mitochondrial phylogenies report that results are often not concordant between data sets
(Ropiquet and Hassanin 2006, Schaefer and Hall 2010).
To address the previous lack of comprehensive phylogenetic analyses of Caprinae, we
provide data in this study which include most or all Caprinae genera. Our phylogenetic analyses
48
include twelve of the thirteen extant genera of Caprinae (excepting Nilgritragus). To address the
prior paucity of comparative analyses incorporating diverse markers, our data include three
distinct sequences from two different markers; the mitochondrial gene ND5 and two regions of
the nuclear gene sex-determining region Y (SRY). Our data provide the first SRY open reading
frame (ORF) sequence data for 15 species in nine genera and the first SRY promoter (PRO)
sequence data for 13 species in 11 genera. With these data we compare the relative resolving
power of the highly variable ND5 gene with the highly conserved SRY gene to determine if and
when each marker is better at clarifying the evolutionary relationships among Caprinae genera.
MATERIALS AND METHODS
Data sets - Three distinct data sets were developed for this study; 1) the full-length SRY
ORF sequence (ORF); 2) the partial SRY promoter sequence (PRO); and the partial ND5
sequencer (ND5). We combined all available non-redundant published sequences (ORF: 2,
PRO: 11, ND5: 5) with our own sequence data (ORF: 31, PRO: 26, ND5: 27) to ultimately
include data for 12 genera and 24 species within the subfamily Caprinae. We analyzed each of
the three data sets independently as well as a fourth combined Y-chromosome data set
(PRO/ORF) and a fifth complete dataset (ND5/PRO/ORF). For combined analyses we only
included samples with both PRO and ORF data and only samples with at least two of the three
markers were included in the complete ND5/PRO/ORF data set (Table 10).
49
Table 10. Sample information by individual Caprinae genus for Y-chromosome (SRY ORF and
SRY
PRO) and mitochondrial (ND5) data sets.
Table 1. Sample information by genus for each marker
Genus
Ammotragus
SRY ORF
no. of
Species samples
lervia
Reference
ND5
no. of
samples
EU938019
Meadows & Kijas, 2009
1
1
EU938020
Meadows & Kijas, 2009
1
EU938021
Meadows & Kijas, 2009
――
――
――
present study
1
――
――
present study
1
Reference
――
――
――
1
――
――
――
――
――
――
――
Arabitragus
jayakai
――
Budorcas
taxicolor
1
――
Capra
SRY PRO
no. of GenBank Accession
sample
no.
GenBank
Accession no.
――
――
Reference
present study
――
1
present study
1
present study
――
1
present study
aegagrus
1
hircus
1
ibex
1
present study
1
present study
1
falconeri
D82963
GenBank
Accession
FJ207523
Hassanin, 2009
present study
1
D82963
1
AF53341
1
present study
1
present study
1
Capricornis sumatrensis
1
present study
1
present study
1
present study
Hemitragus jemehicus
1
present study
1
present study
1
present study
1
present study
1
present study
1
present study
1
present study
1
present study
1
present study
Oreamnos americanus
Ovibos
Moschatus
1
Ovis
ammon
1
JN992654
present study
1
JN992655
present study
1
JN992693
present study
1
present study
1
JN992656
present study
1
JN992692
present study
1
present study
1
JN992657
present study
1
JN992691
present study
1
present study
1
EU938024
Meadows & Kijas, 2009
EU938028
Meadows & Kijas, 2009
present study
1
present study
present study
2
present study
1
2
JN992664, JN992665
present study
2
JN400247 JN400247
JN400253
2
JN992658, JN992659
present study
1
JF680994
1
JN992660
present study
――
――
――
1
present study
1
JN992669
present study
1
JN992680
present study
1
present study
1
JN992668
present study
1
JN992688
present study
1
present study
present study
1
JN992689
present study
canadensis many JN992662, JN992663
dalli dalli
1
1
present study
many
present study
present study
1
JN992681
present study
dalli stonei
1
JN992670
present study
1
JN992679
present study
1
nivicola
1
JN992671
present study
――
――
――
――
1
JN992672
present study
1
JN992682
present study
1
present study
present study
present study
――
――
1
JN992673
present study
1
JN992683
present study
1
o. orientalis
1
Z30265
Payen & Cotinot, 1994
1
EU938022
Meadows & Kijas, 2009
1
1
EU938023
Meadows & Kijas, 2009
o. musimon
1
present study
1
present study
1
present study
1
present study
――
――
――
1
present study
1
present study
――
――
――
1
present study
1
present study
1
present study
1
present study
1
present study
1
present study
1
present study
vignei
1
EU938031
Meadows & Kijas, 2009
――
――
――
――
――
――
1
present study
Pantholops
hodgonii
――
――
――
――
――
――
1
present study
Pseudois
nayaur
1
present study
1
present study
1
scaeferi
――
――
――
――
――
1
Naemordehdus crispus
――
present study
JQ040802
Zou et al. 2011
Saiga
tartarica
1
present study
1
present study
1
Rupicapra
rupicapra
――
――
――
1
JN547786
Perez et al. 2011
1
FJ207539 Hassanin et al. 2009
pyrenaica
――
――
――
1
JN547785
Perez et al. 2011
1
FJ207537 Hassanin et al. 2009
50
present study
Sequence analyses - The complete 727 bp SRY ORF was amplified with primers
previously designed from mouflon sheep (O. o. musimon) genomic sequence (Payen and Cotinot
1994). A 481 bp segment of the SRY promoter located 1,238 bp upstream from the ORF start
codon was amplified with primers designed for this study (Table 11). All Y-chromosome
amplifications were conducted in 25µl reactions containing 50ng of template DNA, 2mM
MgCl2, 0.05mM dNTPs, 0.4µM of each primer and 0.25 units of HotStart-IT Taq DNA
Polymerase (Affymetrix Inc. Santa Clara, CA). The PCR profile included: One cycle of 5min at
94◦C followed by 40 cycles of: 1min at 94◦C, 1min at 61◦C (PRO) or 52◦C (ORF), 1min at 72◦C
and one final cycle of 5min at 72◦C. Sequencing was conducted by Functional Biosciences, Inc.
(Madison, WI). All alignments, including both sequences from single samples as well as
complete alignments including all Caprinae samples, were performed in Sequencher™ version
4.6 (Gene Codes Corporation; Ann Arbor, MI). All sequences were checked by hand for
accuracy and polymorphisms were confirmed manually.
Table 11. SRY and ND5 primer sequences used for all sequence amplifications
Locus Primer
Primer sequence (5’-3’)
bp
Tm Reference
name
(οC)
ORF
SRY1-F
TGCCAGGAGGTATTGAGGGG
723 52
Payen and
SRY1-R
CAGAGGAGCAGTTATTTTGG
Cotinot, 1994
PRO
SRY3A-F
SRY3A-R
AACAATTGCATGTAGCTCCAGA
CAATGACAGTAGAAACTTCAAAGG
481
61
present study
ND5
ND5A-F
ND5A-R
AGTAGTTATCCATTCGGTCTTAGG
AAGATTTGTTGGAGATCTCAGGTG
665
58
ND5
ND5B-F
ND5B-R
CCAACACAGCAGCCTTACAAGC
GGTCTTTGGAGTAGAATCCGGTGAG
690
61
(Wehausen,
unpublished
data)
(Wehausen,
unpublished
data
51
We sequenced 1,110 bp of the mtDNA gene ND5, the largest and one of the faster
evolving mitochondrial protein coding genes (Lopez-Ramirez and Saavedra-Molina, 1997) via
two overlapping segments amplified with separate primer pairs and annealing temperatures
(Table 11). For 22uL reactions, PCRs included 1X PCR buffer, 1.9mM MgCl2 , 0.16mM each
dNTP, 5X BSA (New England Biolabs), 0.4uM each primer, and 8.8 units Amplitaq Gold
(Applied Biosystems). The PCR profile included one cycle of 7 min at 94◦C followed by 35
cycles of: 1 min at 94◦C, 70 sec at 58/61◦C, and 1.5 min at 72◦C. After examining 5uL of PCR
products on an agarose gel, cleanup was: 12uL PCR product + 2.4uL shrimp alkaline
phosphatase (SAP) + 2.4uL exonuclease I (Exo1) at 37◦C for 30min then 80◦C for 10min. Cycle
sequencing reactions combined 7.5uL of cleaned PCR product, 3uL BigDye3.1 (Applied
Biosystems), 0.255uL of 10uM primer, and 4.295uL water, and used the following termpeature
profile: 25 cycles of 96◦C 10sec, 50◦C 5sec, and 60◦C 4min. This was cleaned with DyeEx
columns (Qiagen) and dried down. Rehydration used 0.4uL dextran blue dye and 5uL deionized
formamide. All samples were sequenced in both directions and electrophoresed on an ABI 373
or 377 automated DNA sequencer using associated Sequence Analysis software (ABI). Forward
and reverse sequences were aligned and edited using Sequencher software. Calculations are
summarized in Table 12, including: haplotype number (h), number of segregating sites (S),
genetic distance (K), and estimates of nucleotide diversity (π, ө) were performed in DnaSP
version 4.0 (Rozas et al. 2003).
52
Table 12: Summary DNA sequence statistics by data set including sequence length, number of
haplotypes (h), nucleotide diversity (π), nucleotide diversity with Jukes-Cantor correction (πJC),
number of segregating sites (S), average number of nucleotide differences (K), theta per
nucleotide site as estimated from S (ӨW), and theta per sequence as estimated from S (Ө).
Length
Data Set
(bp)
h
π
π jc
S
K
ӨW
Ө
ORF
727
21
0.024
0.025
116
17.059
0.038
26.667
PRO
481
21
0.026
0.026
61
11.909
0.032
15.030
ND5
1110
39
0.108
0.118
471
119.520
0.101
111.403
PRO/ORF
1208
18
0.025
0.024
142
28.100
0.034
40.661
ND5/PRO/ORF
2318
27
0.072
0.121
442
122.248
0.104
114.674
Phylogenetic analyses - Each of the three individual data sets (ORF, PRO and ND5)
were analyzed separately as well as together in a concatenated fourth (PRO/ORF) and fifth
(ND5/PRO/ORF) data set to maximize molecular information and total number of genera
included. All PRO sequences were assigned a simple nucleotide model while the codon model
with partitioned base sites was used for all ORF and ND5 sequences. The combined Ychromosome data set was partitioned as PRO (1-481 bp ), ORF (482-1,208 bp) and the full
concatenated data set was partitioned as: ND5 (1-1,110 bp), PRO (1,111-1,591 bp), ORF (1,5922,318 bp). Maximum parsimony (MP), maximum likelihood (ML) and Bayesian methods were
used to reconstruct phylogenies for each data set. All trees were rooted with the outgroup Saiga
tartarica and were analyzed and edited in FigTree 1.3.1 (Rambaut 2006).
Appropriate models of nucleotide substitution for ML and Bayesian analyses were
identified using both the –lnL and the Akaike Information Criterion (AIC) scores (Akaike 1974,
53
Hurvich and Tsai 1989) as implemented in jModelTest 0.1.1 (Posada and Crandall 1998; Posada,
2008). Selected models were: GTR + G for ORF, PRO and PRO/ORF data sets and GTR+ I + G
for ND5 and ND5/PRO/ORF data sets.
MP analyses were performed in PAUP 4.0 (Swofford 2003). Unweighted MP trees
(Hasegawa 1985) were estimated using heuristic searches with indels designated as a 5th state
change and no limit on the number of trees retained. Branch support was assessed with 10,000
bootstrap replicates using tree bisection-reconnection (TBR) branch swapping. ML analyses
were performed in RAxML 1.3 (Silvestro and Michalak 2011). Branch supports were calculated
with 10,000 thorough bootstrap replicates on the extended majority rule (MRE) consensus tree
for each data set. Bayesian analyses were performed in MrBayes version 3.2.1 (Huelsenbeck
and Ronquist 2001) using the GTR + G model for ORF, PRO and PRO/ORF data sets and the
GTR + I + G for the ND5 and ND5/PRO/ORF data sets as identified in JModelTest from both –
Lnl and Aikakie Information Criteria (AIC) scores. All analyses included 1,000,000 runs
sampling every 100 trees with a burn-in of 25% (2,500 trees).
RESULTS
We sequenced 727 bp of the Y-chromosome SRY ORF which included the entire gene
coding region. Our final data set included 44 unique sequences representing eight Caprinae
genera (not including the outgroup S. tartarica). The SRY ORF contained 116 variable sites and
55 of these were parsimony informative. Two indels were detected, both of which occurred in C.
sumatrensis: a three base insertion of GGT for at position 72 and a single base insertion of A for
at position 679 which results in a frameshift and a protein that is 15 amino acids shorter than that
54
of the other genera examined. Whether these indels are characteristic of C. sumatrensis remains
to be independently confirmed, as our sequence was determined from two reads of a single
individual and no other SRY ORF sequence was not publicly available we were unable to
confirm these indels with independent data. The 483 bp SRY PRO data set contained 21 unique
sequences representing 10 genera for which 61 sites were polymorphic, and 33 were parsimony
informative. The 1,110 bp ND5 data set contained 26 unique sequences representing 13 genera
for which 471 sites were polymorphic, and 361 were parsimony informative. The 1,208 bp
PRO/ORF data set contained 18 unique sequences representing seven genera for which 136 sites
were polymorphic, and 46 were parsimony informative. The 2,318 bp ND5/PRO/ORF data set
contained 27 unique sequences representing 10 genera for which 442 sites were polymorphic,
and 324 were parsimony informative.
Phylogenetic Analyses – DNA polymorphism and statistics are summarized by data set
in Table 12. Three cladograms of Caprinae genera based on different combinations of data are
presented in Figures 7, 8 and 9. Support of specific nodes for each data set, as estimated with
posterior probabilities (PP), maximum likelihood (BPML), and maximum parsimony (BPMP)
are shown in Table 13.
55
Figure 8. Comparative cladograms of Caprinae genera based on two regions of the Ychromosome gene SRY: the SRY promoter (PRO, left) and the SRY open reading frame (ORF,
right). Node supports are given as posterior probabilities (PP, Bayesian, above branch) and
bootstrap values (ML/MP, below branch). Nodes with support of less than 50 were collapsed.
56
Figure 9. Comparative cladograms of the subfamily Caprinae based on ND5 gene sequence
(left) and combined SRY ORF/PRO gene sequence (right). Node supports are shown on
branches with posterior probabilities (Bayesian) above the line and bootstrap values (ML/MP)
below the line. Nodes with branch supports all below 50 were collapsed and are not shown.
57
Figure 10. Cladogram of Caprinae genera based from combined ND5/PRO/ORF sequence data.
Node supports are shown on branches with posterior probabilities (Bayesian) above the line and
bootstrap values (ML/MP) below the line. Branches with supports all below 50 were collapsed
and are not shown.
58
Table 13. Support of individual nodes for each data set as estimated with posterior probabilities
(PP), ML Bootstrap percentages (BPML) and MP bootstrap percentages (BPMP). Nodes not
found or with less than 50% bootstrap support or 0.5 posterior probability are indicated with an
asterisk (*). Nodes for which support could not be evaluated because not all genera were
available are indicated with a dash (-).
Species included in node
PRO
BPML BPMP PP
PP
Capra
ORF
PRO / ORF
BPML BPMP PP
BPML BPMP PP
ND5
ND5 / PRO / ORF
BPML BPMP PP
BPML BPMP
1.00
90
86
1.00
98
93
1.00
99
99
1.00
82
69
1.00
100
100
*
*
*
1.00
99
98
1.00
81
86
1.00
100
100
1.00
99
100
Ovis - New World
(canadensis / dalli / nivicola )
0.96
85
64
*
61
*
0.90
94
67
0.95
99
79
1.00
97
100
Ovis - Old World
(ammon / orientalis / vignei )
*
82
*
1.00
99
98
1.00
100
100
1.00
99
100
0.98
100
100
Ovis / Oreamnos
*
*
*
0.99
92
77
*
*
*
*
*
*
*
*
*
Oreamnos / Budorcas
*
*
*
*
*
*
*
*
*
0.96
73
52
*
77
53
1.00
73
97
1.00
99
93
1.00
81
86
1.00
87
*
1.00
90
91
Capra / Hemitragus
*
*
*
0.96
86
62
0.97
83
63
1.00
82
69
1.00
95
81
Pseudois
-
-
-
-
-
-
-
-
-
1.00
100
100
_
_
_
1.00
90
82
1.00
97
90
1.00
100
98
*
*
*
1.00
100
96
Ovibos / Capricornis / Naemorhedus
-
-
-
-
-
-
-
-
-
1.00
84
88
-
-
-
Capricornis / Naemorhedus
-
-
-
-
-
-
-
-
-
0.75
79
79
-
-
-
Rupicapra
1.00
100
100
-
-
-
-
-
-
1.00
100
100
1.00
100
100
Budorcas / Pseudois / Hemitragus / Capra
/ Oreamnos
1.00
85
79
*
*
*
1.00
93
87
*
*
*
*
*
*
Budorcas / Pseudois / Hemitragus / Capra
0.92
59
*
*
*
*
0.96
96
81
-
-
-
*
*
*
*
*
*
0.52
92
77
*
*
*
*
*
*
0.8
*
*
Budorcas / Capricornis / Ovibos / Capra /
Hemitragus / Pseudois / Oreamnos
1.00
84
79
*
*
*
1.00
93
87
*
*
*
0.95
72
66
Budorcas / Capricornis / Ovibos / Capra /
Hemitragus / Pseudois
0.92
70
53
*
*
*
0.96
96
81
*
*
*
0.56
*
*
Budorcas
-
-
-
-
-
-
-
-
-
1.00
100
100
1.00
100
100
Budorcas / Capra
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Ammotragus / Arabitragus
-
-
-
-
-
-
-
-
-
1.00
97
87
-
-
-
Pantholops
-
-
-
-
-
-
-
-
-
1.00
99
89
-
-
-
Ovis
Capra / Hemitragus / Psuedois
Ovibos / Capricornis
Budorcas / Capricornis / Ovibos
59
MP analyses consistently provided lower support and less resolution relative to both ML
and Bayesian methods for individual nodes in each of the four trees. Bayesian posterior
probabilities tended to be higher than either MP or ML bootstrapping at individual nodes. In
general the Y-chromosome data provided better resolution, as judged by posterior probabilities
and bootstrap values, than the mitochondrial data for the earliest divergences within Caprinae
(Table 13). Within the Y-chromosome data, SRY PRO was more variable than SRY ORF and
generally provided better resolution, though these data sets were individually less informative
than the combined PRO/ORF data set (Fig. 9) and exhibit some topological incongruities, with
several nodes showing low to no support across analyses. These incongruent or poorly
supported nodes include the placement of Budorcas and Oreamnos as well as the organization of
species within both Capra and Ovis, which likely reflects low variation among these relatively
recent divergences within these clades (Fig. 8).
The combined SRY data provided better resolution at more basal nodes than did ND5,
which was unable to resolve relationships among five (PP and ML) or seven (MP) of the major
Caprinae clades, resulting in a large basal polytomy. However, SRY variation was insufficient
to resolve relationships among species within more recently diverged groups, while ND5, in
contrast, provided better resolution at these more distal nodes across data sets and tree estimation
methods. Table 13 summarizes support for individual nodes across all three analyses for each of
the five data sets.
Mitochondrial data strongly supported the placement of Pantholops as sister to all other
genera of Caprinae as well as grouping of Ammotragus and Arabitragus (Fig. 9). These data also
support the division of tahrs (originally the single genus Hemitragus) into the Himalayan tahr
(genus Hemitragus) and the Arabian tahr (genus Arabitragus). Four of the five data sets show
60
high support for Hemitragus as sister to Capra while the fifth, PRO, lacked sufficient resolution
and instead included it in a polytomy with Capra and Pseudois. All ND5 analyses group the
Arabian tahr with the aoudad (Ammotragus lervia) but were unable to resolve placement of this
clade within the Caprinae, instead including it in the large basal polytomy.
DISCUSSION
Clades supported by both Y-chromosome and mitochondrial phylogenies - Several
nodes within our Caprinae phylogenetic trees are supported independently by all available data,
including Capra, Capra/Hemitragus/Pseudois, Ovibos/Capricornis, and Rupicapra. Additional
nodes that are supported by at least two of three markers or four of five data sets included Ovis
(excluding PRO), Ovis New World clade (Excluding ORF), Ovis Old World clade, and
Capra/Hemitragus, Ovibos/Capricornis (Table 13).
The monophyly of true sheep (genus Ovis) was strongly supported for four of the five
data sets (Table 13) and three of these, ND5, PRO/ORF and ND5/PRO/ORF, also supported the
separation of Ovis into two distinct clades: Old World species (O. ammon, O. orientalis and O.
vignei) and New World species (O. dalli, O. canadensis and O. nivicola) (Figs. 8 and 9) while
the fourth, ORF, also supported the old world clade (PP = 1.0, ML = 99, MP = 98) but found low
(ML = 61) to no (MP and Bayes) support for the new world clade (Fig. 8). In contrast, PRO
suggested paraphyly of Ovis with strong support for the grouping of new world sheep (PP = 1,
ML = 94, MP = 98) with all other Caprinae genera separate from the other true sheep (Old
World) species (Fig. 8). It is likely that low Y-chromosome variation is contributing to poor
resolution within Ovis due to the relatively recent divergence and radiation of this genus (Ramey
61
1993, Bunch et al. 2006, Loehr et al. 2006, Rezaei et al. 2010, Meadows et al. 2011), and that
this is responsible for the observed discordance between our topologies. Our data thus best
supports the hypothesis that Ovis is a monophyletic group in accordance with the majority of our
data and previous findings (Hassanin and Douzery 1999, Hernandez-Fernandez and Vrba 2005,
Ropiquet and Hassanin 2005a, 2005b).
Relationships among species within each of the Ovis clades also differed among data sets.
Both ND5 (PP = 1, ML = 99, MP = 100) and the combined ND5/PRO/ORF (PP = 1, ML = 100,
MP = 100) data sets show high support for the grouping of O. o. musimon with O. o. orientalis.
SRY PRO also shows moderate support for this association (PP = .62, ML = 68, MP < 50), and
all of these data agree with previous findings that mouflon sheep are the wild progenitor of
domestic sheep (Rezaei et al. 2010, Meadows et al. 2011). However, the two Y-chromosome
data sets showed low (ORF/PRO; PP = .75) to no (ORF) support for this grouping.
The monophyly of goats and related species (genus Capra) was well supported across all
analyses and data sets. Likewise, all analyses supported a Capra/Pseudois/Hemitragus clade
except ND5 MP which excluded Pseudois. For four of the five data sets Hemitragus was sister
to Capra while Pseudois was more distantly related to both genera. The fifth data set, PRO, was
unable to resolve these relationships, showing a polytomy at this node. Resolution of species
within Capra differed between data sets. ORF showed division of all four species with high
support for the earliest divergence of the ibex (C. ibex), followed by the markhor (C. falconeri)
with wild goat (C. aegagrus) and domestic goat (C. hircus) most recently diverged (Fig. 8). The
PRO data also supported the early divergence of ibex but was unable to resolve the relationships
among the other three species (Fig. 8). The combined PRO/ORF data set showed high support
for all divisions within Capra in agreement with the ORF data set (Fig. 9). In contrast, our ND5
62
data found wild goat and markhor to be most closely related and were unable to further resolve
the Capra relationships, showing a three-way polytomy at this node (Fig 8). The combined
ND5/PRO/ORF data set gave high support to the early divergence of ibex, followed by
divergence of markhor but showed low confidence and conflicting placements for wild goat and
domestic goat among tree estimation methods (Fig. 10).
While several studies support the monophyly of Capra and its association with Pseudois
and Hemitragus (Groves and Shields 1996, Hassanin et al. 1998, Shafer and Hall 2010), many
others have yielded conflicting results (Pidancier et al. 2006, Ropiquet and Hassanin 2006,
Hassanin et al. 2009, Bibi et al. 2012). Both Pidancier et al. (2006) and Ropiquet and Hassanin
(2006) found strong discordance for relationships among Capra species between mtDNA and Ychromosome markers which they attributed to lineage sorting of ancestral polymorphisms,
introgressive hybridization of mtDNA haplotypes, and differing selection pressures acting on
each marker type. Hassanin et al. (2009) found that evolutionary signals differed among genes
across the mtDNA genome and that results for ND5 were particularly discordant with the other
mitochondrial genes, placing C. hircus completely outside the “goat-like” clade
(Pseudois/Hemitragus/Capra). They did not support the paraphyly of Capra however, instead
attributing this discordance to either exogenous (via interspecies hybridization) or endogenous
(via amplification of nuclear copies of mtDNA, or Numts) contamination. Additional analyses
of all available C. hircus mtDNA data found many problems in the goat reference genome
including detection of several Numt sequences, one of which covered most of ND5 and another
which was in Hemitragus and Pseudois, suggesting a nuclear integration event occurred in the
common ancestor to this clade. Bibi et al. (2012) used both mtDNA and morphological data to
place an extinct Capra species (C. wodaramoya) and found several areas of conflict between
63
phylogenies. Both placement of the genus, as either sister to or polyphyletic with Hemitragus,
and relationships among Capra species differed between datasets. For example, morphological
data supports monophyly of ibex species (C. ibex, C. nubiana, C. pyrenaica) while mtDNA
shows a paraphyletic grouping of these species. A common problematic area was placement of
the basal relationships among Caprinae genera which they suggest is due to rapid adaptive
radiation of these genera in the Miocene.
The monophyly of the Chamois (R. rupicapra and R. pyrenaica) was also supported by
both ND5 and ND5/PRO/ORF data sets. The placement of this genus within Caprinae did vary
among data sets, however, with our PRO data placing it sister to the Capra/Hemitragus/Pseudois
clade (PP: 1, ML BP: 95, MP BP: 93) while ND5 was unable to resolve this placement and
ND5/PRO/ORF analyses were incongruent in that Bayesian (PP: 1) and ML (BP: 100) analyses
agree with the PRO analyses in placing Rupicapra sister to Capra/Hemitragus/Pseudois while
MP (BP: 100) analyses group Rupicapra with Ammotragus all with high support. Perez et al.
(2011) also found phylogenetic discordance between Y-chromosome and mtDNA markers.
Their SRY data supported a weak association between Rupicapra, Capra and Ammotragus with
all Rupicapra belonging to a single unique clade while mtDNA data formed three, well
differentiated, clades within Rupicapra. They also find that the distance between species pair for
mtDNA was about two to three times that of SRY and suggest that the phylogenetic discordance
between mtDNA and Y-chromosome data is due to disparity in time of divergence of haplotypes
for each and that all modern Rupicapra are descended from a single very young male lineage.
The Ovibos/Capricornis clade (which also included Naemorhedus where data were
available) was well supported across all five data sets; placement of this clade within Caprinae
differed among data sets, however. ORF lacked adequate resolution to place this clade,
64
including it in a basal polytomy, while both PRO and PRO/ORF data sets placed this clade with
good statistical support as the second divergence within the larger
Capra/Hemitragus/Pseudois/Budorcas/(Rupicapra/Ammotragus) group after the divergence of
Oreamnos (Figs. 8 and 9). ND5 ML and Bayesian analyses both highly supported placement of
this clade as the earliest divergence within the Caprinae while MP was unable to resolve this
branch and instead includes it in a large basal polytomy (Fig. 9). ND5/PRO/ORF analyses are
also incongruent in that there was low support for grouping with Budorcas (Bayesian), while
both ML and MP were unable to resolve the placement of Ovibos/Capricornis and instead
included it in a three-way polytomy within the larger clade including all genera except Ovis (Fig.
10).
Clades with discordance between Y-chromosome and mitochondrial phylogenies - We
found conflicting results for the placement of the takin (genus Budorcas). Our ND5 data shows
an association of Budorcas with Oreamnos (PP: .96, ML BP: 73, MP BP: 52), while our
ND5/PRO/ORF data shows weak (PP: .52, ML BP: 54) to no (MP) association with the
Capricornis/Ovibos clade, and both the ND5 and ORF data sets lack sufficient resolution to
reliably place these clades within the larger Caprinae phylogeny. While our PRO/ORF analyses
place Budorcas sister to the Capra/Hemitragus/Pseudois clade (PP: 1, ML BP: 100, MP BP: 98)
in agreement with Hassanin et al., (2009), our PRO data place it sister to the
Capra/Hemitragus/Pseudois/Rupicapra clade but with low (ML BP: 53) to no (PP, MP) support.
Placement of the mountain goat (genus Oreamnos) was inconsistent among data sets and
analyses. Three of our data sets (PRO, PRO/ORF & ND5/PRO/ORF) place Oreamnos sister to a
large clade including all Caprinae genera except Ovis and Ammotragus. However, these
individual data sets differ in the relative relationship of Oreamnos to specific genera within this
65
large clade. The PRO and PRO/ORF data both show a closer association with the Capricornis/
Ovibos clade, while the full ND5/PRO/ORF data set associates Oreamnos with a
Budorcas/Capricornis/ Ovibos clade. In contrast, the ORF data include Oreamnos in a clade
with Ovis, while the mitochondrial data group Oreamnos with Budorcas as part of a large basal
polytomy.
The Oreamnos/Budorcas clade was strongly supported for ND5 ML and Bayesian (but
not MP) analyses and this relationship is in agreement with previous findings (Ramey 2000)
based on half the ND5 data. This is not consistent, however, with the findings of Hassanin et al.
(2009), which supports our ORF data in associating Oreamnos with Ovis based on analyses of
multiple mitochondrial markers, though interestingly their ND5 data, like ours, did not support
this association. Other studies have grouped Oreamnos with Rupicapra (Fernandez and Vrba
2005), with a Capricornis/Ovibos/Neamorhedus clade using cytochrome b mitochondrial data
(Hassanin et al. 1998), or have found conflicting results among data sets or analyses (Groves and
Shields 1996, Shafer and Hall 2010, Bibi et al. 2012). The genus Oreamnos has been labeled a
“rogue taxon” because its placement within the Caprinae is sensitive to both the types of markers
used and the analytical methods employed. A total evidence approach employed by Shafer and
Hall (2010) found the most support for a relatively basal position for Oreamnos sister to all
Caprinae genera except Naemorhedus and Ovibos. This is similar to findings by Ropiquet and
Hassanin (2005), but a definitive consensus has yet to be reached. It is possible that placement
of Oreamnos is hindered by the lack of variation or speciation within the genus. This could be
due to their adaptation to extreme environments, slow growth rates, and long generation times,
which may have decreased the evolutionary rate in mountain goats.
66
Comparison of taxonomic resolution for ND5 and SRY - Several evolutionary and
population dynamic factors contribute to diversity among specific genetic markers including
evolutionary rate, selection, inheritance pattern, ploidy, and recombination. Population dynamic
factors include generation time, effective population size, sex-biased dispersal patterns, and
mating structure. Both Y-chromosome and mitochondrial markers are haploid and uniparentally
inherited, which predict an effective population (Ne) one-quarter that of diploid nuclear markers.
However, differing mutation rates and species-specific population dynamics also contribute to
nucleotide diversity. Most Caprinae genera exhibit dispersal patterns which would predict
higher Y-chromosome diversity. However, this is potentially offset, to some extent, by highly
polygynous mating structures where a small number of males contribute a disproportionately
large amount of genetic variation to subsequent generations, ultimately decreasing the Ychromosome Ne in Caprinae genera.
While available data are limited, both theoretical and empirical data suggest that Ychromosome markers could provide a unique and informative phylogenetic perspective for
understanding Caprinae taxonomy. For example, it has been shown that increased germ-cell
divisions in spermatogenesis is responsible for higher male mutation rates in humans (Haldane
1947), and this has been subsequently supported in ungulates with data from five Caprinae
genera where the observed mutation rate was three to four times higher in males than females
(Lawson and Hewitt 2002). In contrast, observed Y-chromosome nucleotide diversity (πy = 0.90
x 10-4) in Ovis was approximately 10% of expected (πy = 8.36 x 10-4), even after correction for a
lower male Ne and a higher male-to-female mutation rate. This Ovis Y-chromosome nucleotide
diversity estimate corresponded to about 1.4% of the mitochondrial observed nucleotide diversity
(πmt = 6.15 ± 0.87 x 10-3) (Meadows et al. 2004). This can be at least partly explained by
67
purifying selection acting to conserve the SRY coding sequence which is necessary for normal
development of male anatomy. Thus, there is reason to expect that Caprinae Y-chromosome
markers will be more conserved than mitochondrial markers, which would prove more useful for
identifying basal divergences within this subfamily due to a lack of sequence saturation over
long periods of evolutionary time (Hartl and Clark 1997, Barbosa and Carranza 2010, Perez et al.
2011). Our data supports this hypothesis in that the substitutions per site (K) for ND5 was seven
(ORF) to ten (PRO) times higher than SRY (Table 12). Furthermore, all three tree estimation
methods find a large basal polytomy for ND5 (Fig. 9), while all tree estimation methods find
better resolution for all three SRY data sets with strong support for basal divergences (Figs. 8
and 9).
One reason that has been proposed for the observed discordance in evolutionary patterns
between Y-chromosome and mitochondrial markers in the Caprinae data is that all 13
mammalian mitochondrial genes are involved in the production of energy used to make ATP and
maintain body temperature and are therefore especially important for adaptations necessary for
living in colder environments at higher latitudes and altitudes. Because many genera of the
Caprinae utilize these environments, it is hypothesized that selection acting on mitochondrial
genes has contributed to the evolution of this group. For example, positive selection for 75 bp
D-loop repeats as been documented and attributed to a response to the more severe oxidative
stress and higher metabolic rates resulting from high-altitude life (Hassanin et al. 2009). In
addition, it has been suggested that phylogenetic placement of problematic genera such as Capra
and Oreamnos has been difficult because of past interspecific hybridization and subsequent
selection for mitochondrial genotypes more fit for these environments, ultimately leading to
68
discordant evolutionary signals between nuclear and mitochondrial markers (Ropiquet and
Hassanin 2006).
Caprinae phylogenetics - The taxonomic structure of the subfamily Caprinae has been
problematic since Simpson’s original classification (1945) which recognized 11 genera divided
into four tribes: 1) Caprini (goats, sheep, and related species), 2) Rupicaprini (chamois, goral,
and serow), 3) Ovibovini (muskox and takin) and 4) Saigini (antelope and chiru). Subsequent
molecular studies have found all four of these tribes to be either poly- or paraphyletic (Gentry
1992, Gatesy 1997, Hassanin et al. 1998, Ropiquet and Hassanin 2005, Hassanin et al. 2009,
Shafer and Hall 2010, Arif et al. 2012) leading to the proposal of many additional taxonomies.
This taxonomic discordance has been attributed to a number of factors, including the lack of any
diagnostic synapomorphies uniting this group. Rather, the subfamily Caprinae includes genera
with diverse adaptations which have contributed to rapid radiation throughout high altitude
habitats across Africa, Asia, Europe, and North America. Additionally, individual genera vary
both in times since divergence as well as diversity and species composition within clades.
Additional factors, such as interspecific hybridization, variable evolutionary rates between
markers, and differing life history and demographic variables, all contribute to conflicting results
among the many phylogenetic analyses including genera within this subfamily. This discordance
is exacerbated by a paucity of comprehensive comparative analyses incorporating diverse
markers, because the overwhelming majority of available data is from mitochondrial markers.
Thus continued development of additional nuclear markers, including more diverse Ychromosome markers, should complement existing mitochondrial data and prove useful for
increasing resolution of the evolutionary relationships within this subfamily.
69
CHAPTER IV
POPULATION GENETICS OF ROCKY MOUNTAIN BIGHORN SHEEP
Population genetic analyses have proved incredibly useful for evaluation of genetic
variation, gene flow, metapopulation structure, inbreeding and ultimately informed management
of populations. Comparative analyses of different markers can significantly improve the value of
genetic data, however correct interpretation requires consideration of the differences inherent in
the markers used. Additionally, interpretation of these data is often context sensitive in that
individual populations or species differ in myriad ways. Quantification of population genetics
for natural populations, specifically, is often hindered by a number of factors, including
incomplete historical data, uneven sampling, unrealistic assumptions for parameter estimations,
and complications due to species-specific demography and dynamics. Thus, interpretation of
population genetic data can be complex and comprehensive evaluation of different genetic
markers may help discern patterns but conflicts and discrepancies should also be acknowledged.
Bighorn sheep often exist in small, isolated populations due to specific habitat
requirements and naturally isolated ranges that are increasingly fragmented due to anthropogenic
factors. Thus, bighorn herds are often susceptible to inbreeding depression due to factors
including disease and genetic drift, which can result in loss of allelic variation and decreased
heterozygosity. Migration and gene flow among neighboring herds can ameliorate the effects of
these factors by introducing new alleles that contribute to the maintenance of genetic variation in
individual herds. However, bighorn herds are often subject to the pressures associated with
habitat fragmentation, recent research has shown that both slope and ruggedness of terrain are
70
significant factors in bighorn habitat selection (Sappington et al. 2007) and that highways block
gene flow in bighorn (Epps et al. 2005), which can limit dispersal of sheep among adjacent
herds.
Five Rocky Mountain bighorn sheep (Ovis canadensis) populations were included in our
analyses based on their presence in and relative proximity to RMNP and assessment of potential
for gene flow within the larger metapopulation. Four of these herds (Continental Divide (CD),
Mummy (MM), Never Summer (NS) and St. Vrain (SV)) maintain some or all of their annual
range within the 415 square miles of RMNP while the fifth (Big Thompson (BT)) maintains an
annual range east of RMNP within Roosevelt National Forest. This dissertation research was
initiated due to concern regarding long-term persistence of the Mummy herd following a
pneumonia outbreak in the mid 1990s with a significant die-off and continued poor recruitment
thereafter. This could be due to a number of genetic or environmental causes, and we therefore
gathered data from both mitochondrial haplotypes and microsatellite genotypes to determine if
decreased genetic variation could explain the Mummy herd’s failure to rebound to historic
numbers.
We find no evidence of a severe reduction in genetic diversity or recent genetic
bottleneck for any of these five Rocky Mountain bighorn sheep herds. Observed
heterozygosities were comparable to (Forbes and Hogg 1999, Epps et al. 2010, Johnson et al
2011, Luikart et al. 2011) or higher than (Boyce et al. 1997) those reported for other native
bighorn populations and higher than those reported for restored bighorn populations (Whitaker et
al. 2004). Metapopulation analyses suggest that high heterozygosity is being maintained, to some
extent, by gene flow among these five herds. Estimates of population subdivision from
microsatellite data were low and, likewise, effective migration rates were reasonably high, and
71
no significant correlation was found between allelic distribution and geographic distance. In
contrast, mitochondrial data show higher genetic isolation among herds and a significant
negative correlation between haplotype distribution and geographic distance. The ranges of both
effective migration rates and population substructure were larger for mitochondrial data than for
microsatellite data, suggesting that ewes preferentially move between some herds more than
between others.
The disparity between metapopulation structure as estimated from microsatellite and
mitochondrial data could also be explained to some extent by differences in the evolution and
inheritance patterns of these two markers. Genes located on the mitochondrial chromosome are
vital to metabolic processes involved in aerobic respiration and are thus more conserved than
microsatellite alleles which are neutral and rapidly mutate via replication slippage and are thus
generally highly variable in populations Ballard and Whitlock, 2004). Mitochondria are also
haploid and maternally inherited while microsatellites are diploid and biparentally inherited.
Thus, when the sex ratio is equal, mitochondrial markers have an effective population one
quarter that of microsatellite markers (Birky et al. 1983). However, this disparity is offset, to
some extent, by the highly polygynous mating structure of bighorn sheep in which most adult
ewes reproduce each year while generally only the most dominant rams successfully mate (Geist
1966, Geist 1971, Coltman et al. 2005).
Interpretation of these metapopulation substructure data must be tempered, however, by
the history of sheep movement both into and between study herds. These translocations were
primarily composed of ewes (George et al. 2009) and this, together with the highly polygynous
mating structure of bighorn sheep, suggests that the MM and BT herds should each have a higher
proportion of private mitochondrial haplotypes and microsatellite alleles and higher pairwise Fst
72
estimates between each of these two herds and the other three (CD, NS and SV) herds, and the
BT and MM herds should exhibit a higher proportion of shared genetic variation, primarily
mitochondrial haplotypes, resulting in lower subdivision relative to the other herds. Our data do
not match either of these predictions, however, and suggest that these transplantations are not
appreciably altering interpretation of the metapopulation structure. This could be explained by
low transplantation success which has been documented in bighorn (Gross et al. 2000, Singer et
al. 2000) which may have resulted in few transplanted ewes surviving and reproducing in their
transplanted herds.
Despite mitigating factors inherent in evaluation of genetic variation in natural
populations our data clearly show that none of these five Rocky Mountain bighorn herds are
suffering from a genetic bottleneck and that this is likely due to continued gene flow within this
metapopulation. Our results also reiterate the utility of comparative mitochondrial and
microsatellite analyses which have proved useful for discerning relationships both within and
among many ungulate species (Sarno et al. 2004, Randi et al. 2005, Shafer et al. 2010) and
highlight the advantages of evaluating population genetic parameters from complementary
markers for other species (Brunner et al. 1998, Shaw et al. 1999).
PHYLOGENETICS OF CAPRINAE GENERA
Phylogenetics is a discipline that has greatly benefitted from relatively recent advances in
the ability to quantify individual variation at the molecular level, augmenting or complementing
previous analyses centered on morphological and behavioral data. This is especially true in
73
genera and species with low evolutionary rates or complicated evolutionary and demographic
histories such that divergence is difficult to assess and multiple lines of evidence are required
(Avise et al. 1992, Martin et al. 1992, Kuhner and Felsenstein 1994, Yang 1996, Wendel and
Doyle 1998). Taxonomic organization of genera within the subfamily Caprinae is particularly
difficult because no unique synapomorphies unite this diverse group and phylogenetic analyses
to date have focused primarily on mitochondrial data. This has resulted in continued
disagreement regarding resolution of Caprinae phylogenetics, and recent papers have repeatedly
called for analyses which incorporate more diverse markers to order to aid elucidation of the
relationships within this group and related ungulate genera (Zachos et al. 2003, HernandezFernandez and Vrba 2005, Lorenzen et al. 2008).
To address the previous lack of comprehensive phylogenetic analyses of Caprinae, we
provide comparative analyses of two different markers; the mitochondrial gene ND5 and two
regions of the nuclear gene sex-determining region Y (SRY) for twelve of the thirteen extant
Caprinae genera. Our data provide the first SRY open reading frame (ORF) sequence data for15
species in nine genera and the first SRY promoter (PRO) sequence data for 13 species in 11
genera. With these data we compare the relative resolving power of the highly variable ND5
gene with the highly conserved SRY gene to determine if and when each marker is better at
clarifying the evolutionary relationships among Caprinae genera. Our data highlight the
complex evolutionary history of Caprinae genera and the importance of incorporating data from
diverse markers with disparate evolutionary signals by highlighting inherent incongruities
between them, resulting from differences including evolutionary rate, variation, inheritance
pattern and effective population size. These analyses suggest that both ND5 and SRY have
complementary phylogenetic value, in accordance with results from other species (Sasazaki et al.
74
2007, Schiffer et al. 2007, Yoon et al. 2008, Nussberger et al. 2013). Our more variable ND5
data were more useful for discerning recent divergences, which agrees with analyses in other
species (Horai et al. 1995, Questiau et al. 1998, Tserenbataa et al. 2004, Palsboll et al. 2007,
Smith et al. 2008). In contrast, our more conserved SRY data proved more useful for discerning
past divergences, which has been shown in other species (Jobling et al. 2003, Geraldes et al.
2005, Rangel-Villalobos et al. 2009, Lippold et al. 2011), although the low variation in the SRY
markers impeded our analyses of population genetics in bighorn sheep, in congruence with other
findings from other sheep species (Meadows and Kijas 2009, Oner et al. 2011, Shafter et al.
2011).
75
LITERATURE CITED
Akaike, H. 1974. A new look at statistical model identification. IEEE Transaction on Automatic
Control 19: 716-723.
Allison, A. C. 1954. Protection afforded by sickle-cell trait against subtertian malarial infection.
British Medical Journal 1: 290-294.
An, J., H. Okumura, Y.-S. Lee, K.-S. Kim, M.-S. Min, and H. Lee. 2010. Organization and
variation of the mitochondrial DNA control region in five Caprinae species. Genes and
Genomics 32: 335–344.
Angeloni, F., N. J. Ouborg, and R. Leimu. 2011. Meta-analysis on the association of population
size and life history with inbreeding depression in plants. Biological Conservation 144:
35-43.
Arif, I. A, M. A. Bakir, and H. A. Khan. 2012. Inferring the phylogeny of bovidae using
mitochondrial DNA sequences: resolving power of individual genes relative to complete
genomes. Evolutionary Bioinformatics Online 8: 139–150.
Avise, J. C., and D. Ellis. 1986. Mitochondrial DNA and evolutionary genetics of higher animals
[and discussion]. Philosophical Transactions of the Royal Society B: Biological Sciences
312: 325-342.
Avise, J. C. 1992. Molecular population structure and the biogeographic history of a regional
fauna: a case history with lessons for conservation biology. Oikos 63: 62-67.
76
Ballard J. W. O. and M. C. Whitock. 2004. The incomplete natural history of mitochondria.
Molecular Ecology 13: 729-744.
Bailey, J. A. 1986. The increase and die-off of Waterton Canyon bighorn sheep: biology,
management and dismanagement. Biennial Symposium of the North American Wild
Sheep and Goat Council 5: 325-340.
Baillie, J. E. M., C. Hilton-Taylor, and S. N. Stuart. 2004. 2004 IUCN Red List of Threatened
Species. A Global Species Assessment. IUCN, Gland, Switzerland and Cambridge, UK.
Barbosa, A. M., and J. Carranza. 2010. Lack of geographic variation in Y-chromsomal introns of
red deer (Cervus elaphus). Journal of Negative Results 7: 1-4.
Barrett, S. C. H., J. R. Kohn, D. A. Falk, and K. E. Holsinger. 1991. Genetic and evolutionary
consequences of small population size in plants: implications for conservation. Genetics
and Conservation of Rare Plants: 3-30.
Barton, N. H., and M. Slatkin. 1986. A quasi-equilibrium theory of the distribution of rare alleles
in a subdivided population. Heredity 56: 409–415.
Beraldi, D., A. McRae, J. Gratten, J. Slate, P. M. Visscher, and J. M. Pemberton. 2006.
Development of a linkage map and mapping of phenotypic polymorphisms in a freeliving population of Soay sheep (Ovis aries). Genetics 173: 1521-1537.
Berger, J. 1990. Persistence of different-sized populations: An empirical assessment of rapid
extinction in bighorn sheep. Conservation Biology 4: 91-98
77
Bergl, R. A., B. J. Bradley, A. Nsubuga, and L. Vigilant. 2008. Effects of habitat fragmentation,
population size and demographic history on genetic diversity: The Cross River gorilla in
a comparative context. American Journal of Primatology 70: 848-859.
Bibi, F., E. Vrba, and F. Fack. 2012. A new African fossil caprin and a combined molecular and
morphological Bayesian phylogenetic analysis of caprini (Mammalia: Bovidae). Journal
of Evolutionary Biology 25: 1843–1854.
Birky, C. W. Jr., T. Maruyama, and P. Fuerst. 1983. An approach to populatio nand evolutionary
genetic theory for genes in mitochondrial and chloroplasts, and some results. Genetics
103: 513-527.
Bleich V.C., J. D. Wehausen, and S. A. Holl. 1990. Desert-dwelling mountain sheep:
conservation implications of a naturally fragmented distribution. Conservation Biology 4:
383–390.
Bleich VC, J. D. Wehausen, R. R. Ramey, and J. L. Rechel. 1996. Metapopulation theory and
mountain sheep: implications for conservation. In: Metapopulations and Wildlife
Conservation (ed. McCullough DR). Island Press, Covelo, CA.
Boyce, W. M., P. W. Hedrick, N. E. Muggli-Cockett, S. Kalinowski, M. C. T. Penedo and R. R.
Ramey II. 1997. Genetic variation of Major Histocompatibility Complex and
microsatellite loci: A comparison in bighorn sheep. Genetics 145: 421-453.
78
Brook, B. W., D. W. Tonkyn, J. J. O'Grady and R. Frankham. 2002. Contribution of inbreeding
to extinction risk in threatened species. Conservation Ecology 6: 16.
Brunner, P. C., M. R. Douglas, and L. Bernatchez. 1998. Microsatellite ans mitochondrial DNA
assessment of population structure and stocking effects in Arctic charr Salvelinus alpinus
(Teleostei: Salmonidae) from central Alpine lakes. Molecular Ecology 7: 209-223.
Buchanan, F. C., P. A. Swarbrick, and A. M. Crawford. 1991. Ovine nucleotide repeat
polymorphism at the MAF48 locus. Animal Genetics 22: 379-380.
Buchanan, F. C., and A. M. Crawford. 1992a. Ovine dinucleotide repeat polymorphism at the
MAF209 locus. Animal Genetics 23: 183.
Buchanan, F. C., and A. M. Crawford. 1992b. Ovine dinucleotide repeat polymophism at the
MAF33 locus. Animal Genetics 23: 186.
Buchanan, F. C., and A. M. Crawford. 1993. Ovine microsatelittes at the OarFCB11,
OarFCB128, OarFCB193, OarFCB266 and OarFCB304 loci. Animal Genetics 24:,145..
Buitkamp, J., F.-W. Schwaiger, and J.-T. Epplen. 1993. Ovis aries T cell receptor gene V-region,
exons 1 (3’ end) and 2 (5' end). NCBI.
Cann, R. L., M. Stoneking, and A. C. Wilson. 1987. Mitochondrial DNA and human evolution.
Nature 325: 31-36.
Cassirer, E. F., and A. R. E. Sinclair. 2007. Dynamics of pneumonia in a bighorn sheep
metapopulation. Journal of Wildlife Management 71: 1080-1088.
79
Caughley, G. 1994. Directions in conservation biology. Journal of Animal Ecology 63: 215-244.
Chen, L., A. L. DeVries, and C-H, Cheng. 1997. Convergent evolution of antifreeze
glycoproteins in Antarctic notothenioid fish and Arctic cod. Proceedings of the National
Academy of Sciences 94: 3817-3822.
Cornuet, J. M., and G. Luikart. 1996. Description and power analysis of two tests for detecting
recent population bottlenecks from allele frequency data. Genetics 144: 2001–2014.
Coltman, D. W., J. G. Pilkington, J. A. Smith, and J. M. Pemberton. 1999. Parasite-mediated
selection against inbred Soay sheep in a free-living, island population. Evolution. 12591267.
Coltman, D. W., P. O'Donoghue, J. T. Hogg, and M. Festa-Bianchet. 2005. Selection and genetic
(co)variance in bighorn sheep. Evolution 59: 1372-1382.
Cothran, E. G., R. Chesser, M. H. Smith, and P. E. Johns. 1983. Influences of genetic variability
and maternal factors on fetal growth in white-tailed deer. Evolution 37: 282-291.
Crawford, A. M., K. G. Dodds, A. J. Ede, C. A. Pierson, G. W. Montgomery, H. G.
Garmonsway, A. E. Beattie, K. Davies, J. F. Maddox, S. W. Kappes, R. T. Stone, T. C.
Nguyen, J. M. Penty, E. A. Lord, J. E. Broom, J. Buitkamp, W. Schwaiger, J. T. Epplen,
P. Mathew, M. E. Matthew, D. J. Hulme, K. J. Beh, R. A. Mcgraw, and C. W. Beattie.
1995. Genetic linkage map of the sheep genome. Genetics 140: 703–724.
Crawford, A. M., K. A. Paterson, K. G. Dodds, C. D. Tascon, P. A. Williamson, M. R.
Thompson, S. A. Bisset et al. 2006. Discovery of quantitative trait loci for resistance to
80
parasite nematode infection in sheep: I. Analysis of outcross pedigrees. BMC Genomics
7: 178.
Crnokrak, P., and S. C. H. Barrett. 2002. Perspective: Purging the genetic load: A review of the
experimental evidence. Evolution 56: 2347-2358.
Darwin, C. 1859. On the Origin of Species by Means of Natural Selection, or the Preservation of
Favoured races in the Struggle for Life. D. Appleton, New York, NY.
Dobzhansky, T., B. Spassky, and T. Tidwell. 1963. Genetics of natural populations XXXII.
Inbreeding and the mutational and balanced loads in natural populations of Drosophila
obscura. Proceedings of the National Academy of Sciences U.S. 53: 482-486.
Duyao, M., C. Ambrose, and R. Meyers, A. Novelletto, F. Persichetti, M. Frontali, and M.
MacDonald. 1993. Trinucleotide repeat length instability and age of onset in
Huntington’s disease. Nature Genetics 4: 387-392.
Emery, N. J., and N. S. Clayton. 2004. The mentality of crows: Convergent evolution of
intelligence in corvids and apes. Science 306: 1903-1907.
Epps, C. W., P. J. Palsbøll, J. D. Wehausen, G. K. Roderick, R. R. Ramey, and D. R.
McCullough. 2005. Highways block gene flow and cause a rapid decline in genetic
diversity of desert bighorn sheep. Ecology Letters 8: 1029–1038.
Epps C. W., P. J. Palsbøll, J. D. Wehausen, G. K. Roderick, and D. R. McCullough. 2006.
Elevation and connectivity define refugia for mountain sheep as climate warms.
Molecular Ecology 15: 4295-4302.
81
Epps, C. W., J. D. Wehausen, V. C. Bleich, S. G. Torres, and J. S. Brashares. 2007. Optimizing
dispersal and corridor models using landscape genetics. Journal of Applied Ecology 44:
714-724.
Epps, C., J. D. Wehasuen, J. D. Palsboll and D. R. McCullough. 2010. Using genetic tools to
track desert bigorn sheep colonizations. The Journal of Wildlife Management 74: 522531.
Evanno, G., S. Regnaut, and J. Goudet. 2005. Detecting the number of clusters of individuals
using the software STRUCTURE: A simulation study. Molecular Ecology 14: 2611–
2620.
Fisher, R. A. 1930. The Genetical Theory of Natural Selection. Clarendon Associates, Oxford,
U.K.
FitzSimmons, N. N., S. W. Buskirk, and M. H. Smith. 1995. Population history, genetic
variability, and horn growth in bighorn sheep. Conservation Biology 9: 314-323.
Forbes, S. H., and J. T. Hogg. 1999. Assessing population structure at high levels of
differentiation: microsatellite comparisons of bighorn sheep and large carnviores. Aminal
Conservation 2: 223-233.
Frankham, R. 2005. Genetics and extinction. Biological Conservation 126: 131-140.
82
Gatesy, J., G. Amato, E. Vrba, G. Schaller, and R. DeSalle. 1997. A cladistic analysis of
mitochondrial ribosomal DNA from the Bovidae. Molecular Phylogenetics and Evolution
7: 303–319.
Geist, V. 1966. The evolutionary significant of mountain sheep horns. Evolution 20: 558-566.
Geist, V. 1971. Mountain Sheep. Chicago Press, Chicago, IL, USA.
Gemmell, N. J. and J. Slate. 2006. Heterozygote advantage for fecundity. PLos ONE 1: e125.
Gentry, A. W. 1992. The subfamilies of the family Bovidae. Mammal Review 22: 1-32.
George, J. L., R. Kahn, M. W. Miller, and B. Watkins. 2009. Colorado bighorn sheep
management plan 2009 − 2019. Special report no. 81.
Geraldes, A., C. Rogel-Gaillard, and N. Ferrand. 2005. High levels of nucleotide diversity in the
European rabbit (Oryctolagus cuniculus) SRY gene. Animal Genetics 36: 349-351.
Gilligan, D.M., D. A. Briscoe, and R. Frankham. 2003. Dynamics of genetic adaption to
captivity. Conservation Genetics 4: 189–197.
Godinho, R., T. Abáigar, S. Lopes, A. Essalhi, L. Ouragh, M. Cano, and N. Ferrand. 2012.
Conservation genetics of the endangered Dorcas gazelle (Gazella dorcas spp.) in
Northwestern Africa. Conservation Genetics 13: 1003–1015.
Gross, J. E., F. J. Singer, and M. E. Moses. 2000. Effects of disease, dispersal, and area on
bighorn sheep restoration. Restoration Ecology 8: 25-37.
83
Groves, P., and G. F. Shields. 1996. Phylogenetics of the Caprinae based on cytochrome b
sequence. Molecular Phylogenetics and Evolution 5: 467–476.
Gulland, F. M. D., S. D. Albon, J. M. Pemberton, P. R. Moorcroft and T. H. Clutton-Brock.
1993. Parasite-associated polymorphism in a cyclic ungulate population. Proceedings of
the Royal Society London Series B - Biological Sciences 254: 7-13.
Gutiérrez-Espeleta, G. A., S. T. Kallinowski, W. M. Boyce, and P. W. Hedrick. 2000. Genetic
variation and population structure in desert bighorn sheep: Implications for conservation.
Conservation Genetics 1: 3-15.
Haldane, J. B. S. 1932. The Causes of Evolution. Longmans, Green, New York.
Haldane, J. B. S. 1947. The mutation rate of the gene for haemophilia, and its segregation ratios
in males and females. Annuals of Eugenics 13: 262-271.
Hammer, M. F. 1995. A recent common ancestry for human Y chromosomes. Nature 378: 376–
378.
Hanski, I. A., and M. E. Gilpin (eds.).1997. Metapopulation Biology; Ecology, Genetics, and
Evolution. Academic Press, San Diego, CA.
Hartl, G. B., J. Markowski, A. Swiatecki, T. Janiszewski, and R. Willing. 1992. Studies of the
European hare .43. Genetic diversity in the Polish brown hare Lepus-europaeus pallas,
1778 – implications for conservation and management. Acta Theriologica 37: 15-25
84
Hartl, D. L., and A. G. Clark. 1997. Principles of Population Genetics. 3rd ed. Sinauer
Associates, Inc. Sunderland, MA, pp. 320-328.
Hasegawa, M, H. Kishino, and T. Yano. 1985. Dating of the human-ape splitting by a molecular
clock of mitochondrial DNA. Journal of Molecular Evolution. 22: 160-174.
Hassanin, A., E. Pasquet, and J. Vigne. 1998. Molecular systematics of the subfamily Caprinae
(Artiodactyla, Bovidae) as determined from cytochrome b sequences. Journal of
Mammalian Evolution 5: 217-236.
Hassanin, A., and E. J. Douzery. 1999. The tribal radiation of the family Bovidae (Artiodactyla)
and the evolution of the mitochondrial cytochrome b gene. Molecular Phylogenetics and
Evolution 13: 227–243.
Hassanin, A., and E. J. Douzery. 2003. Molecular and morphological phylogenies of Ruminantia
and the alternative position of the Moschidae. Systematic Biology 52: 206-228.
Hassanin, A., and A. Ropiquet. 2004. Molecular phylogeny of the tribe Bovini (Bovidae,
Bovinae) and the taxonomic status of the Kouprey, Bos sauveli, Urbain 1937. Molecular
Phylogenetics and Evolution 33: 896-907.
Hassanin, A., A. Ropiquet, A. Couloux, and C. Cruaud. 2009. Evolution of the mitochondrial
genome in mammals living at high altitude: new insights from a study of the tribe Caprini
(Bovidae, Antilopinae). Journal of Molecular Evolution 68: 293–310.
Hedrick P., and S. T. Kalinowski. 2000. Inbreeding depression in conservation biology. Annual
Review of Ecological Systems 31: 139–162.
85
Hedrick, P. 2001. Conservation genetics: Where are we now? Trends in Ecology and Evolution
16: 629-636.
Hernández-Fernández, M., and E. S. Vrba. 2005. A complete estimate of the phylogenetic
relationships in Ruminantia: a dated species-level supertree of the extant ruminants.
Biological Reviews of the Cambridge Philosophical Society 80: 269–302.
Hiendleder, S., K. Mainz, Y. Plante, and H. Lewalski. 1998. Analysis of mitochondrial DNA
indicates that domestic sheep are derived from two different ancestral maternal sources:
no evidence for contributions from urial and argali sheep. Journal of Heredity 89: 113–
120.
Hogg, J. T., S. H. Forbes, B. M. Steele, and G. Luikart. 2006. Genetic rescue of an insular
population of large mammals. Proceedings of the Royal Society London Series B Biological Sciences 273: 1491–1499.
Hurvich C. M., and C. L. Tsai. 1989. Regression and time series model selection in small
samples. Biometrika 76: 297-307.
Huelsenbeck, J. P., and F. Ronquist. 2001. MrBayes: Bayesian inference of phylogeny.
Bioinformatics 17: 754-755.
Huxley, J. 1942. Evolution: The Modern Synthesis. Alen and Unwin, London, UK.
International Union for Conservation of Nature ad Natural Resources. 2004. 2004 IUCN
Red List of Threatened Species, IUCN-SSC http://www.iucnredlist.org.
86
Jensen, J. L., A. J. Bohonak, and S. T. Kelley. 2005. Isolation by distance, web service. BMC
Genetics 6: 13.
Jobling, M. A., and C. Tyler-Smith. 2003. The human Y chromosome: an evolutionary marker
comes of age. Nature Reviews Genetics 4: 598-612.
Johnson, J. A., J. E. Toepfer, and P. O. Dunn. 2003. Contrasting patterns of mitochondrial and
microsatellite population structure in fragmented populations of greater prairie-chickens.
Molecular Ecology 12: 3335-3347.
Johnson H. E., L. S. Mills, J. D. Wehausen, T. R. Stephenson, and G. Luikart. 2011. Translating
effects of inbreeding depression on component vital rates to overall population growth in
endangered bighorn sheep. Conservation Biology 25: 1240-1249.
Katsanis, S. H., and J. K. Wagner. 2013. Characteriszation of the standard and recommended
CODIS markers. Journal of Forensic Sciences 58: S169-S172.
Keller L. F., and D. M. Waller. 2002. Inbreeding effects in wild populations. Trends in Ecology
and Evolutionary Biology 17: 230–241.
Kidwell, S. M. and S. M. Holland. 2002. The quality of the fossil record: Implications for
evolutionary analyses. Annual Review of Ecologic Systematics 33: 561-88.
Kikkawa, Y., T. Takada, K. Nomura, T. Namikawa, H. Yonekawa and T. Amano. 2003.
Phylogenies using mtDNA and SRY provide evidence for male-mediated intogression in
Asian domestic cattle. Animal Genetics 34: 96-101.
87
Lawson, L.-J., and G. M. Hewit. 2002. Comparison of substitution rates in ZFX and ZFY introns
of sheep and goat related species supports the hypothesis of male-biased mutation rates.
2002. Journal of Molecular Evolution 54: 54-61.
Lehman, N., A. Eisenhawer, K. Hansen, L. D. Mech, R. O. Peterson, P. J. P. Gogan, and R. K.
Wayne. 1991. Introgression of coyote mitochondrial DNA into sympatric North
American gray wolf populations. Evolution 45: 104-119.
Levins, R. 1969. Some demographic and genetic consequences of environmental heterogeneity
for biological control. Bulletin of the ESA 15: 237-240.
Lewontin, R. C., and J. L. Hubby. 1966. A molecular approach to the study of genic
heterozygosity in natural populations. II. Amount of variation and degree of
heterozygosity in natural populations of Drosophila pseudoobscura. Genetics 54: 595.
Loehr, J., K. Worley, A. Grapputo, J. Carey, A. Veitch, and D. W. Coltman. 2006. Evidence for
cryptic glacial refugis from North American mountain sheep mitochondrial DNA. Journal
of Evolutionary Biology 19: 419-430.
Librado, P., and J. Rozas. 2009. DnaSP v5: a software for comprehensive analysis of DNA
polymorphism data. Bioinformatics 25: 1451–1452.
Lopez, J. V., N. Yuhki, R. Masuda, W. Modi, S. J. O'Brien. 1994. Numt, a recent transfer of
tandem amplification of mitocondrial DNA to the nuclear genome of the domestic cat.
Journal of Molecular Evolution. 39: 174-190.
88
Lopez-Ramirez, A. and A. Saavedra-Molina. 1997. Several effects of D-amino acids in rat liver
mitochondria. Journal of the Federation of American Societies 11: A954.
Lorenzen, E. D., P. Arctander, amd H. R. Siegismund. 2008. High variation and very low
differentiation in wide ranging plains zebra (Equus quagga): insights from mtDNA and
microsatellites. Molecular Ecology 17: 2812-2824.
Lucy, M. C., G. S. Johnson, H. Shibuya, C. K. Boyd, and W. O. Herring. 1998. Rapid
communication: polymorphic (GT)n microsatellite in the bovine somatotropin receptor
gene promoter. Journal of Animal Science 76: 2209–2210.
Luikart, G., F. W. Allendorf, J. M. Cornuet, and W. B. Sherwin. 1998. Distortion of allele
frequency distributions provides a test for recent population bottlenecks. The Journal of
Heredity 89: 238–247.
Luikart, G., and J.-M. Cornuet. 1998. Empirical evaluation of a test for identifying recently
bottlenecked populations from allele frequency data. Conservation Biology 12: 228–237.
Luikart, G., K. Pilgrim, J. Visty, V. O. Ezenwa and M. K. Schwartz. 2008a. Candidate gene
microsatellite variation is associated with parasitism in wild bighorn sheep. Biology
Letters 4: 228-231.
Luikart, G., S. Zundel, R. Delphine, M. Christian, K. A. Keating, J. T. Hogg, B. Steele, K.
Foresman and P. Taberlet. 2008b. Low Genotyping rates and non-invasive sampling in
bighorn sheep. The Journal of Wildlife Management 7: 299-304.
89
Luikart, G., S. J. Amish, J. Winnie, A. Beja-Pereira, R. Godinho, F. W. Allendorf, and R. B.
Harris. 2011. High connectivity among argali sheep from Afghanistan and adjacent
countries: Inferences from neutral and candidate gene microsatellites. Conservation
Genetics 12: 921–931.
Manel, S., M. K. Schwartz, G. Luikart, and P. Taberlet. 2003. Landscape genetics: combining
landscape ecology and population genetics. Trends in Ecology and Evolution 18: 189197.
Manos, P. S. and D. E. Fairbrothers. 1987. Allozyme variation in populations of six Northeastern
American Red Oaks (Fagaceae: Quercus subg. Erythroblanus). Systematic Botany 12:
365-373.
Matthee, C. A., and S. K. Davis. 2001. Molecular insights into the evolution of the family
Bovidae: a nuclear DNA perspective. Molecular Biology and Evolution 18: 1220– 30.
Maudet, C., G. Luikart, D. Dubray, A. Von Hardenberg and P. Taberlet. 2004. Low genotyping
error rates in wild ungulate faeces sampled in winter. Molecular Ecology Notes 4: 72775.
McAlpine, S. Genetic heterozygosity and reproductive success in the green treefrog, Hyla cinera.
1993. Heredity 70:553-558.
McClintock, B. T. and G. C. White. 2007. Bighorn sheep abundance following a suspected
pneumonia epidemic in Rocky Mountain National Park. Journal of Wildlife Management
71:183–189.
90
Meadows, J. R. S., R. J. Hawken, and J. W. Kijas. 2004. Nucleotide diversity on the ovine Y
chromosome. Animal Genetics. 35: 379-385.
Meadows, J. R. S., O. Hanotte, C. Drogemuller, J. Calvo, R. Godfrey, D. Coltman, J. F. Maddox,
N. Marzanov, Juha Kantanen, and J. W. Kijas. 2006. Globally dispersed Y chromosomal
haplotypes in wild and domestic sheep. Animal Genetics 37: 444-453.
Meadows, J. R. S. and J. Kijas. 2009. Re-sequencing regions of the ovine Y chromosome in
domestic and wild sheep reveals novel paternal haplotypes. Animal Genetics 40:119-123.
Meadows, J. R. S., Hiendleder, S., Kijas, J. W. 2011. Haplogroup relationships between
domestic and wild sheep resolved using a mitogenome panel. Heredity 106, 700-706.
Miller, J. M., J. Poissant, J. T. Hogg, and D. W. Coltman. 2012. Genomic consequences of
genetic rescue in an insular population of bighorn sheep (Ovis canadensis). Molecular
Ecology 21:1583-1596.
Mills, L., and F. Allendorf. 1996. The one-migrant-per-generation rule in conservation and
management. Conservation Biology 10: 1509-1518.
Mitton, J. B. and R. M. Jeffers. 1989. The genetic consequences of mass selection for growth
rate in Engelmann spruce. Silvae Genetica 38:6-12.
Nakagome, S., Pecon-Slattery, J., Masuda, R. 2008. Unequal rates of Y chromosome gene
divergence during speciation of the family Ursidae. Molecular Biology and Evolution,
25:1344–56.
91
Nei, M., T. Maruyama, R. Chakraborty. 1975. The bottleneck effect and the genetic variability in
populations. Evolution 29:1-10.
Norrgard, K. 2008. Forensics, DNA fingerprinting and CODIS. Nature Education 1:1.
Nowak, R. M., Paradiso, J. L. 1991. Walker's Mammals of the World. Cambridge University
Press, Cambridge, MA.
Nussberger, B., M. P. Greminger, C. Grossen, L. F. Keller, and P. Wandeler. 2013. Development
of SNP markers identifying European wildcats, domestic cats, and their admixed
progeny. Molecular Ecology Resources 13: 447-460.
O'Grady, J. J., B. W. Brook, D. H. Reed, J. D. Ballou, D. W. Tonkyn and R. Frankham. 2006.
Realistic levels of inbreeding depression strongly effect extinction risk in wild
populations. Biological Conservation 133:42-51.
Oner, Y., J. H. Calvo, and C. Elmaci. 2011. Y chromosomal characterization of Turkish native
sheep breeds. Livestock Science. 136: 277-280.
Palo, J., S.-L. Varvio, I. Hansk, and R. VäinÖlä. 2004. Developing microsatellite markers for
insect population structure: Complex variation in a Checkerspot Butterfly. Hereditas 123:
295-300.
Palsbøll P. J., M. Berube, and F. W. Allendorf. 2007. Identification of management units using
population genetic data. Trends in Ecology and Evolution 22: 11-16.
Payen, E., Cotinot, C. 1994. Sequence evolution of SRY gene within Bovidae family.
Mammalian Genome 5: 723-725.
92
Penty, J., H. Henry, A. Ede, A. M. Crawford. 1993. Ovine microsatellites at the OarAE16.
OarAE54, OarAE57, OarAE119, and OarAE129 loci. Animal Genetics 24: 219.
Pérez, T., Hammer, S. E., Albornoz, J., and A. Domínguez. 2011. Y-chromosome phylogeny in
the evolutionary net of chamois (genus Rupicapra). BMC Evolutionary Biology 11: 272.
Pidancier, N., S. Jordan, G. Luikart, and P. Taberlet. 2006. Evolutionary history of the genus
Capra (Mammalia, Artiodactyla): discordance between mitochondrial DNA and Ychromosome phylogenies. Molecular Phylogenetics and Evolution 40: 739–749.
Piry, S., G. Luikart, and J.-M. Cornuet. 1999. BOTTLENECK: A computer program for
detecting recent reductions in the effective population size using allele frequency data.
The Journal of Heredity 90: 499–503.
Poole, K. G., D. M. Reynolds, M. Darryl, G. Mowat, and P. Paetkau. 2011. Estimating mountain
goat abundance using DNA from fecal pellets. The Journal of Wildlife Management 75:
1527-1534.
Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using
multilocus genotype data. Genetics 155: 945–959.
Questiau et al. 1998
Raikkonen, J., A. Bignert, P. Mortensen, and B. Fermholm. 2006. Congenital defects in a highly
inbred wolf population. Mammalian Biology 71: 65-73.
93
Ralls, K., and J. Ballou. 1986. Captive breeding programs for populations with a small number
of founders. Trends in Ecology and Evolution 1: 19-22.
Ralls, K., Ballou, J. D., Templeton, A. 1998. Estimates of lethal equivolents and the cost of
inbreeding in mammals. Conservation Biology 2: 185-193.
Ramey, R. R. II. 1993. Evolutionary genetics of North American mountain sheep. Ph. D.
Dissertation, Cornell University, Ithica, NY.
Ramey, R. R., G. Luikart, and F. J. Singer. 2000. Genetic bottlenecks resulting from restoration
efforts: the case of bighorn sheep in Badlands National Park. Restoration Ecology 8: 85–
90.
Randi, E. Management of wild ungulate populations in Italy: captive-breeding, hybridisation and
genetic consequenses of translocations. 2005. Veterinary Research Communications. 29:
71-75.
Raymond, M., and F. Rousset. 1995. GENEPOP (version 1.2): population genetics software for
exact tests and ecumenicism. The Journal of Heredity 86: 248–249.
Reed, D.H., D. A. Briscoe, and R. Frankham. 2002. Inbreeding and extinction: effects of
environmental stress and lineage. Conservation Genetics 3: 301-307.
Reed, D.H. 2005. Relationship between population size and fitness. Conservation Biology 19:
563-568.
94
Rezaei, H. R., S. Naderi, I. C. Chintauan-Marquier, P. Taberlet, A. T. Virk, H. R. Nagash, D.
Rioux, M. Kaboli, and F. Pompanon. 2010. Evolution and taxonomy of the wild species
of the genus Ovis (Mammalia, Artiodactyla, Bovidae). Molecular Phylogenetics and
Evolution 54: 315-326.
Richards, M., C. Rengo, F. Cruciani, F. Gratrix, J. F. Wilson, R. Scozzari, V. Macaulay, and A.
Torroni. 2003. Extensive female-mediated gene flow from Sub-Saharan Africa into Near
Eastern Arab Populations. American Journal of Human Genetics. 72:1058-1064.
Rioux-Paquette E., M. Festa-Bianchet, and D. W. Coltman. 2011. Sex-differential effects of
inbreeding on overwinter survival, birth date and mass of bighorn lambs. Journal of
Evolutionary Biology 24: 121-131.
Rootsi, S., T. Kivisild, G. Benuzzi, M. Bermisheva, I. Kutuev, L. Barac et al. 2004.
Phylogeography of Y-chromosome haplogroup I reveals distinct domains of prehistoric
gene flow in Europe. The American Journal of Human Genetics 75: 128-137.
Ropiquet, A., and A. Hassanin. 2005a. Molecular phylogeny of caprines (Bovidae, Antilopinae):
the question of their origin and diversification during the Miocene. Journal of Zoological
Systematics and Evolutionary Research 43: 49–60.
Ropiquet, A., and A. Hassanin. 2005b. Molecular evidence for the polyphyly of the genus
Hemitragus (Mammalia, Bovidae). Molecular Phylogenetics and Evolution 36: 154–68.
Ropiquet, A., and A. Hassanin. 2006. Hybrid origin of the Pliocene ancestor of wild goats.
Molecular Phylogenetics and Evolution 41: 395–404.
95
Rousset, F. 2008. GENEPOP ’007: a complete re-implementation of the GENEPOP software for
Windows and Linux. Molecular Ecology Resources 8: 103–106.
Sappington, J. M., K. M. Longshore and D. B. Thompson. 2007. Quantifying landscape
ruggedness for animal habitat analysis: A case study using bighorn sheep in the Mojave
Desert. Journal of Wildlife Management 71: 1419-1426.
Schwaiger, F.-W., J. Buitkamp, E. Weyers, and J.T. Epplen. 1993. Typing of artiodactyl
MHC-DRB genes with the help of intronic simple repeated DNA sequences. Molecular
Ecology 2: 55-59.
Schwartz, O. A., V. C. Bleich, and S. A. Holl. 1986. Genetics and conservation of mountain
sheep Ovis canadensis nelsoni. Biological Conservation 37: 179–190.
Shafer, A. B. A, and J. C. Hall. 2010. Placing the mountain goat: a total evidence approach to
testing alternative hypotheses. Molecular Phylogenetics and Evolution 55(1), 18–25.
Shackleton, D. M., C. C. Shank, and B. M. Wikeem. 1999. Natural history of Rocky Muontain
and California bighorn sheep. pp. 78-138 in R. Valdez and P. R. Krausman, editors.
Mountain Sheep of North America. University of Arizona Press, Tucson, USA.
Shaw, P. W., C. Turan, J. M. Wright, M. O'Commell and G. R. Carvalho. 1999. Micorsatellite
DNA analysis of population structure in Atlantic herring (Clupea harengus), with direct
comparison to allozyme and mtDNA RFLP analyses. Heredity 83: 490-499.
Silvestro, D., and I. Michalak. 2011. RaxmlGUI: a graphical front-end for RAxML. Organisms
Diversity and Evolution 12: 335-337.
96
Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bulletin
of the American Museum of Natural History 85: 151-156.
Singer, F. J., C. M. Papouchis, and K. K. Symonds. 2000. Transloactions as a tool for restoring
populations of bighorn sheep. Restoration Ecology 8: 6-13.
Slatkin, M. 1993. Isolation by distance in equilibrium and non-equilibrium populations.
Evolution 47: 264–279.
Soulé, M. E. 1976. Allozyme variation: its determinants in space and time. Molecular Evolution:
60-77.
Soulé, M. E. 1980. Thresholds for survival: maintaining fitness and evolutionary potential.
Pages 151-169 in M. E. Soulé and B. A. Wilcox, eds. Conservation Biology. An
Evolutionary-Ecological Perspective. Sinauer Associates, Sunderland, MA.
Swarbrick, P. A., F. C. Buchanan, and A. M. Crawford. 1991. Ovine dinucleotide repeat
polymorphism at the MAF36 locus. Animal Genetics 22: 377–378.
Swofford, D. L. 2003. "PAUP". Phylogenetic Analysis Using Parsimony, version 4.0. Sinauer
Associates, Sunderland, MA, USA.
Tasheva, E. S., J. L. Funderburgh, L. M. Corpuz, and G. W. Conrad. 1998. Cloning,
characterization and tissue-specific expression of the gene encoding bovine keratocan, a
corneal keratan sulfate proteoglycan. Gene 218: 63–68.
97
Taylor, A. C., W. B. Sherwin, and R. K. Wayne. 1994. Genetic variation of microsatellite loci in
bottlenecked species: the hairy-nosed wombat (Lasioshinus krefftii). Molecular Ecology
3: 277-290.
Tosi, J. A., J. C. Morales, and D. J. Melnick. 2000. Comparison of Y Chromosome and mtDNA
phylogenies leads to unique inferences of Macaque evolutionary history. Molecular
Phylogenetics and Evolution. 17: 133-144.
Truscott, F. W., and F. L. Emory. 1902. A Philophical Essay on Probabilities by Pierre Simon,
Marquis de Laplace. John Wiley & Sons, London, U.K.
Tserenbataa, T., R. R. Ramey, O. A. Ryder, T. W. Quinn, and R. P. Reading. 2004. A
population genetic comparison of argali sheep (Ovis ammon) in Mongolia using the ND5
gene of mitochondrial DNA; implications for conservation. Molecular Ecology 13: 13331339.
Turner, C., C. Killoran, N. S. T. Thomas, M. Rosenberg, N. A. Chuzhanova, J. Johnston, Y.
Kemel, D. N. Cooper and L. G. Biesecker. 2003. Human genetic disease caused by de
novo mitochondrial-nuclear DNA transfer. Human Genetics. 112: 303-309.
Vilas, C., E. S. Miguel, R. Amaro, and C. Garcia. 2006. Relative contribution of inbreeding
depression and eroded adaptive diversity to extinction risk in small populations of shore
campion. Conservation Biology 20: 229–238.
98
Wallace, A. R. 1858. On The Tendency of Varieties to Depart Indefinitely From the Original
Type. Proceedings of the Linnean Society of London, 3: 53-62.
Wallace, A. R. 1870. Contributions to the Theory of Natural Selection. Macmillan and Co.
London, UK.
Watson, J. D., and F. H. Crick. 1953. The structure of DNA. Cold Spring Harbor Symposia on
Quantitative Biology 18: 123-131.
Wehausen, J. D., R. R. Ramey, and C. W. Epps. 2004. Experiments in DNA extraction and PCR
amplification from bighorn sheep feces: the importance of DNA extraction method. The
Journal of Heredity 95: 503–509.
Whitaker, D. G., S. D. Osterman and W. M. Boyce. 2004. Genetic variability of reintroduced
California bighorn sheep in Oregon. Journal of Wildlife Management 68: 850-859.
Wood, N.J., and S.H. Phua. 1994. A dinucleotide repeat polymorphism at the adenylate
cyclase activating polypeptide locus in sheep. Animal Genetics 24: 329.
Wood. E. T., D. A. Stover, C. Ehret, G. Destro-Bisol, G. Spedini, H. McLeod, L. Louie, M.
Bamshad, B. I. Strassman, H. Soodyall, and M. F. Hammer. 2005. Contrasting patterns of
Y chromosome and mtDNA variation in Africa: evidence for sex-biased demographic
processes. European Journal of Human Genetics. 13: 867-876.
Xu, Y. C., S. G. Fang, and Z. K. Li. 2007. Sustainability of the South China tiger: implications of
inbreeding depression and introgression. Conservation Genetics 8: 1199-1207.
99
Yoon, D., Y. S. Kwon, K. Y. Lee, W. Y. Jung, S. Sasazaki, H. Mannen, J. T. Jeon, and J. H. Lee.
2008. Discrimination of Korean cattle (Hanwoo) using DNA markers derived from SNPs
in bovine mitochondrial and SRY genes. Asian Australian Journal of Animal Sciences
21:25.
Zeng, B., L. Xu, B. Yue, Z. Li, and F. Zou. 2008. Molecular phylogeography and genetic
differentiation of blue sheep Pseudois nayaurszechuanensis and Pseudois schaeferi in
China. Molecular Phylogenetics and Evolution 48: 387–395.
Zhao, X., M. Chu, N. Li, and C. Wu. 2001. Paternal inheritance of mitochondrial DNA in the
sheep (Ovine aries). Science in China Series C: Life Sciences 44: 321–326.
Zou, F., Q. Peng, L. Tang, and L. Tang. 2011. Mitogenome analysis of the dwarft blue shep
(Pseudois schaeferi): evidence of adaptive evolution of morphological variation in the
ATP synthase gene. Direct submission, NCBO accession JQ040802.1.
100