Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

American pika population genetic structure, demographic history, and behavior in an atypical environment Robson, Kelsey 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2013_fall_Robson_Kelsey.pdf [ 2.45MB ]
Metadata
JSON: 24-1.0074070.json
JSON-LD: 24-1.0074070-ld.json
RDF/XML (Pretty): 24-1.0074070-rdf.xml
RDF/JSON: 24-1.0074070-rdf.json
Turtle: 24-1.0074070-turtle.txt
N-Triples: 24-1.0074070-rdf-ntriples.txt
Original Record: 24-1.0074070-source.json
Full Text
24-1.0074070-fulltext.txt
Citation
24-1.0074070.ris

Full Text

AMERICAN PIKA POPULATION GENETIC STRUCTURE, DEMOGRAPHIC HISTORY, AND BEHAVIOR IN AN ATYPICAL ENVIRONMENT  by  Kelsey Michelle Robson  B.Sc., The University of Memphis, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  THE COLLEGE OF GRADUATE STUDIES  (Biology)   THE UNIVERSITY OF BRITISH COLUMBIA   (Okanagan)  August 2013   ? Kelsey Michelle Robson, 2013  ii Abstract  Anthropogenic impacts on biodiversity are large and varied, from habitat destruction and fragmentation to climate change.  In response to these threats, wildlife species must rapidly adapt within their geographic range, or disperse to different areas that have become environmentally suitable. If not, population decline, extirpation, and eventual species extinction will result. There is a current need for research into the ability of organisms to persist at the tolerance limits of their bioclimatic envelope, as this information will help assess potential responses to changing environments. The American pika (Ochotona princeps) is an appropriate model species for studies of adaptability and persistence in atypical environments. The geographic range of these climate-sensitive mammals extends across a large, variable landscape. Pikas typically inhabit alpine talus that is patchily distributed; as such, they are also a model species for studies of metapopulation dynamics in a fragmented landscape. This study used microsatellite genotypic data to investigate a) population genetic variation and demographic history, b) relatedness and inbreeding, and c) population structure and connectivity of American pika inhabiting an atypical environment in the Columbia River Gorge, Oregon. A total of 316 hair samples were non-invasively collected from 11 sites across an elevation gradient ranging from 46-1260 m. There were 155 pikas identified in the sample. For this system, high inbreeding and low genetic variation best characterized pikas within a site. A high degree of structure was detected among sites, and differentiation increased where topographical features potentially served as dispersal barriers. Although pikas inhabiting geographically proximate sites tended to cluster at similar elevations, there was little evidence of statistically significant migration. Indirect measures, however, such as within-site relatedness and inbreeding, strongly suggested a pattern of male-biased dispersal. This work addressed a knowledge gap in the pika literature by reporting on the population genetics and behavior of populations inhabiting an atypical environment. In order to properly evaluate the conservation status of the American pika, and inform sound management policies, it is necessary to consider the entire species distribution and compare populations from different parts of the range. iii Preface  The following parts of this thesis have been published in or submitted to peer reviewed journals:  Information from Chapter 2 contributed to a recent publication. Lemay, M.A., P. Henry, C.T. Lamb, K.M. Robson, and M.A. Russello. (2013) Novel genomic resources for a climate-change sensitive mammal: characterization of the American pika transcriptome. BMC Genomics 14: 311. I was responsible for collecting and processing pika hair samples from the Columbia River Gorge, OR, USA, and contributing to single-nucleotide polymorphism discovery and characterization.  All sex-specific analyses and interpretations are based on work conducted in UBC Okanagan?s Ecological and Conservation Genomics Laboratory by Dr. Michael A. Russello and Clayton T. Lamb. The methodology has been submitted for publication. Lamb, C.T., K.M. Robson, and M.A. Russello. (submitted) Development and application of a molecular sexing protocol in the climate-change sensitive American pika. Conservation Genetics Resources. I contributed to data collection and analysis.   iv Table of Contents  Abstract.................................................................................................................................... ii Preface..................................................................................................................................... iii Table of Contents ................................................................................................................... iv List of Tables .......................................................................................................................... vi List of Figures........................................................................................................................ vii Acknowledgements ................................................................................................................ ix Chapter 1 Introduction......................................................................................................... 10 1.1 Global Biodiversity................................................................................................................. 10 1.1.1 Climate Change ............................................................................................................... 10 1.1.2 Conservation Genetics..................................................................................................... 13 1.2 The American pika ................................................................................................................. 14 Chapter 2 Population genetics and behavior of O. princeps in an atypical environment................................................................................................................................................. 19 2.1 Background............................................................................................................................. 19 2.2 Methods .................................................................................................................................. 26 2.2.1 Sample collection ............................................................................................................ 26 2.2.2 Laboratory methods......................................................................................................... 28 2.2.2.1 DNA extraction ........................................................................................................ 28 2.2.2.2 Primer optimization.................................................................................................. 28 2.2.2.3 Multiplex Polymerase Chain Reaction (PCR) ......................................................... 29 2.2.2.4 Molecular sexing...................................................................................................... 30 2.2.3 Data analysis.................................................................................................................... 32 2.2.3.1 Quality assessment................................................................................................... 32 2.2.3.2 Population genetic variation and demographic history ............................................ 32 2.2.3.3 Relatedness and inbreeding...................................................................................... 33 2.2.3.4 Population structure and connectivity...................................................................... 33 2.3 Results .................................................................................................................................... 35 2.3.1 Quality assessment .......................................................................................................... 35  v 2.3.2 Population genetic variation and demographic history ................................................... 36 2.3.3 Relatedness and inbreeding ............................................................................................. 36 2.3.4 Population structure and connectivity ............................................................................. 36 2.4 Discussion............................................................................................................................... 38 2.4.1 Population genetic variation and demographic history ................................................... 39 2.4.2 Relatedness and inbreeding ............................................................................................. 40 2.4.3 Population structure and connectivity ............................................................................. 43 2.4.4 Quality assessment .......................................................................................................... 45 2.4.5 Summary ......................................................................................................................... 46 2.5 Figures .................................................................................................................................... 49 2.6 Tables...................................................................................................................................... 60 Chapter 3 Conclusion ........................................................................................................... 69 3.1 Limitations.............................................................................................................................. 70 3.2 Future Work............................................................................................................................ 71 References .............................................................................................................................. 75  vi List of Tables  Table 1    Site locations for O. princeps in the Columbia River Gorge.................................. 60 Table 2    Information on the microsatellite loci retained in the study of O. princeps samples from the Columbia River Gorge. The size range is measured in the number  of base pairs (bp). ................................................................................................... 61 Table 3    Summary of within site measures of genetic variation for a sample of O. princeps. n is number of individuals, Rxy is relatedness according to the Queller and Goodnight method (Queller & Goodnight 1989), FIS is the inbreeding  coefficient, Ho is observed heterozygosity, and He is expected heterozygosity..... 62 Table 4    Relatedness and inbreeding of O. princeps male and female individuals within sites. Rxy is relatedness according to the Queller and Goodnight method  (Queller & Goodnight 1989), and FIS is the inbreeding coefficient. ...................... 63 Table 5    Pairwise site comparisons of genetic differentiation (?) for a sample of  O. princeps. ............................................................................................................ 64 Table 6    Bi-directional migration rates between sites for O. princeps. Values organized by column represent immigration, and values organized by row represent emigration. Migration rates represent the proportion of the population consisting of  genetic migrant individuals per generation. ........................................................... 65 Table 7    Pairwise site comparisons of genetic differentiation (?) for O. princeps males and females. Values above the diagonal represent females and values below  the diagonal represent males. ................................................................................. 66 Table 8    Bi-directional migration rates for O. princeps males between sites. Values organized by column represent immigration, and values organized by row represent emigration. Migration rates represent the proportion of the  population consisting of genetic migrant individuals per generation. ................... 67 Table 9    Bi-directional migration rates for O. princeps females between sites. Values organized by column represent immigration, and values organized by row represent emigration. Migration rates represent the proportion of the  population consisting of genetic migrant individuals per generation. ................... 68   vii List of Figures  Figure 1    Distribution of Ochotona princeps across the Intermountain West. Grey areas are the approximate range, and black dots are known pika locations. The X indicates the sampling area for this study. The inset at the top corner shows the approximate distribution of the collared pika. Image modified from  Galbreath et al. (2010). ......................................................................................... 49 Figure 2    Distribution of sampling sites for O. princeps in the Columbia River Gorge, Oregon, USA. Sites H, T, Y, J, and M are low-elevation sites (below  300 m). Sites U, I, R, D, B, and Q are high-elevation sites (above 1000 m)........ 50 Figure 3    O. princeps occupy low-elevation talus slopes that are characterized by  small-medium sized boulders, moss cover, and abundant vegetation. ................. 51 Figure 4    O. princeps occupy high-elevation talus slopes that are characterized by medium-large sized boulders and surrounding shrubby vegetation. A baited snare is  shown in the image on the right. ........................................................................... 52 Figure 5    A pika latrine site. An indicator of the presence of O. princeps at a talus  slope. ..................................................................................................................... 53 Figure 6    Packing tape snares.  Tripped snares provided hair samples of O. princeps (indicated by the arrow). ....................................................................................... 54 Figure 7    Hair obtained from O. princeps. An ideal sample composed of numerous  guard hairs............................................................................................................. 55 Figure 8    Representative gel image used to identify sex in O. princeps, including males (?), females (?), 100 bp DNA ladder (NEB), and a negative control. The lower molecular weight product is from the sex-determining region on the Y chromosome (SRY) and is found in males.  An autosomal microsatellite  (Ocp10) was used as a PCR positive control, and should be present in  both sexes.............................................................................................................. 56 Figure 9    Isolation by distance relationship. Genetic distance between sites increases as the geographic distance increases. Genetic distance is measured in theta (?),  and geographic distance is measured in kilometers. ............................................. 57  viii Figure 10    Genetic friction across sites of O. princeps occupancy in the Columbia River Gorge, Oregon, USA. As the color gradient moves from light yellow to dark red, the genetic differentiation per unit of geographic distance increases. Longitude and latitude are recorded in decimal degrees format. Genetic friction is a  relative measure based on a modified version of theta (?).................................. 58 Figure 11    Clustering of talus sites based on the program STRUCTURE. Each bar of the graph represents a unique individual, and the proportion of the bar that is  blue or red indicates assignment to a particular genetic cluster. ......................... 59   ix Acknowledgements  I offer my gratitude to the faculty, staff and fellow students at the UBCO, who have inspired me to complete my work in this field. I owe special thanks to my supervisor Dr. M.A. Russello, who has mentored me through this experience and contributed immensely to its success, and also my committee members Dr. K. Hodges, and C. Ray, for their continual feedback and support.  The members of the Ecological and Conservation Genomics Lab were instrumental towards the completion of this thesis. In particular: Clayton Lamb for his contributions as an undergraduate researcher, including long days in the field and hours in the laboratory; Philippe Henry for his introduction to specific laboratory and field methods; and Evelyn Jensen and Matt Lemay for their incredible support and friendship.  I thank Johanna Varner for her enthusiasm to collaborate with fieldwork and for providing important hair samples. I would also like to extend gratitude to Meghan Perra as a volunteer field assistant. Finally, special thanks are owed to my parents, who have supported me throughout my years of education, both morally and financially.  My work was funded by a variety of sources, including a NSERC Discovery grant (#341711-07) to Dr. M.A. Russello and a Genome BC Strategic Opportunities Fund grant (#130) to Dr. M.A. Russello. I was also supported in part by a Graduate Entrance Scholarship from the UBCO, and a University Graduate Fellowship from the UBCO.    10Chapter 1 Introduction  1.1 Global Biodiversity It is estimated that about nine million types of plants, animals, protists, and fungi inhabit the Earth (Cardinale et al. 2012). Only recently have we truly started to recognize the importance of this biodiversity, its value and contribution to human culture, and its necessity for our economic and social well-being. A few decades ago at the first Earth Summit, the majority of the world?s nations denounced human activity as the cause of ecosystem damage and the loss of biological diversity (Cardinale et al. 2012). For many species, the major determinants of extinction risk are anthropogenic environmental impacts; specifically, human population density, agricultural and urban land-use, species exploitation, introduced species and disease, and anthropogenic climate change (Soul? 1991). One of the greatest threats to species biodiversity and ecosystem function may result from the high density and rapid growth of the human population (Kerr & Currie 1995; McKinney 2001; Ceballos & Ehrlich 2002). In fact, the number of threatened species in the average nation is expected to increase 14% by 2050, as forecast by human population growth alone (McKee et al. 2003). Anthropogenic-induced climate change is an even more ubiquitous threat, particularly because of its magnitude and the rate at which it is occurring.  1.1.1 Climate Change A meta-analysis of 334 species and a global analysis of 1,570 species showed highly significant, nonrandom patterns of change correlated with observed climate warming in the twentieth century (Parmesan & Yohe 2003). Some of these patterns of change included range shifts averaging 6.1 km per decade towards the poles (Parmesan & Yohe 2003), community disaggregation and representation by more warm-adapted species (Parmesan  112005), and significant mean advancement of spring events by 2.3 days per decade (Parmesan & Yohe 2003). Dobrowski et al. (2013) presented a geographic assessment of the climate forcings that organims experienced during the 20th century. Specifically, climate velocity (both climate displacement rate and direction) in the contiguous United States was evaluated. The analysis demonstrated that velocity vectors are different across regions, show variable and opposing directions among the variables considered, and shift direction through time (Dobrowski et al. 2013). Biotic responses to climate change can be expected to show a similar level of variation. As well, climate change may yield different responses from populations in different parts of a species? geographic range. There is evidence that suggests peripheral populations have distinct behavior, physiological tolerances, population dynamics, and genetic structure (Gaston 2003). Models of species declines predict that compared to core populations, peripheral populations may even be more likely to persist (Channell & Lomolino 2000). Typically, peripheral populations live under variable and unstable conditions, a situation that may be more common range-wide if climate change persists (Safriel et al. 1994). According to the Fisher (1930) hypothesis, core populations are subject to stabilizing selection that decreases within-population genetic variability, while peripheral populations undergo fluctuating selection and have high within-population genetic variability. Although the overall genetic diversity of peripheral populations may be low compared to the core, there is greater stress adaptation (Hardie & Hutchings 2010). The conservation value of peripheral populations should be considered high as they are a significant component of intraspecific biodiversity and a source of adaptive potential and persistence in the face of climate change (Gibson et al. 2009).   12In addition, alpine ecosystems, which are particularly sensitive to environmental conditions, represent an ideal model for studying the impacts of climate change. Patterns and processes vary along altitudinal gradients, and can be investigated without the spatially extensive sampling required by latitudinal/ longitudinal gradients. On mountains, variables such as temperature and precipitation can change quite significantly over a short distance. There is evidence that increasing temperatures are shifting communities toward higher elevations (le Roux & McGeoch 2008; Lenoir et al. 2008); however, upward dispersal is limited by the summit, and movement between mountains is discouraged by unfavorable conditions in the surrounding matrix. Of the literature documenting biotic responses to climate change along elevation gradients, a significant proportion is dedicated to plant communities. With an experiment using plots along elevation gradients in a grassland system, de Sassi & Tylianakis (2012) tested how biomass at the plant, herbivore, and parasitoid level responds to the effects of warming and nitrogen deposition. They concluded that reduced top-down regulation is likely to coincide with increased herbivory, initiating a cascade effect upon other ecosystem processes (de Sassi & Tylianakis 2012). In another study of plant and animal interactions along elevation transects, Colwell et al. (2008) found that tropical lowland biotas would experience higher biotic attrition (i.e. the loss of diversity as species emigrate from an area) than highland biotas. From these few examples, it is clear that climate change will affect plant community composition and interactions; these alterations will in turn impact other trophic levels, and the overall ecosystem dynamics. To mitigate the adverse effects of climate change, species may either disperse, adapt in situ, or undergo local extinction (Hewitt & Nichols 2005). Given the anticipated  13consequences, there is considerable interest in obtaining a more detailed understanding of species? responses to climate change.  1.1.2 Conservation Genetics An increasingly affordable and informative method of investigating ecological and evolutionary questions involves the application of genetic tools. This sort of application defines conservation genetics, a field which has made substantial progress since its foundation in the late 1970s (Frankham et al. 2009). To date, the most substantial effort has been directed towards resolving taxonomic relationships and defining species, while the application of genetics in the management of threatened species in the wild has yet to be fully developed (Frankham 2010). In particular, there is a need to investigate the extinction risks of small and isolated populations, in which the impacts of genetic drift and inbreeding are amplified (Angeloni et al. 2011). To this purpose neutral genetic markers (e.g. amplified fragment length polymorphisms and microsatellites) can be employed for elucidating demographic history and patterns of dispersal and gene flow (Angeloni et al. 2011). Low genetic diversity within a population may reduce the potential of the population to adapt to environmental change and thereby increase the probability of extinction (Hardie & Hutchings 2010). With recent technological advancements, the field of conservation genetics is making use of new genomic tools. It is now possible to study within and between population genetic variation at a level representative of the entire genome, and investigate gene activity as it relates to population and habitat characteristics, and ultimately, adaptation (Primmer 2009; Allendorf et al. 2010; Angeloni et al. 2011). However, there is still an important and necessary place for traditional genetic approaches, and many unresolved issues to address in  14conservation biology. The most important of these issues is the use of genetics in the management of fragmented wild populations of threatened species (Frankham 2010). 1.2 The American pika In developing a framework for the management of fragmented populations, it is useful to study the population genetics of species that inhabit naturally fragmented landscapes. The American pika (Ochotona princeps) is one such example. These small lagomorphs are found in talus habitat throughout the Intermountain West (Smith & Weston 1990). Pikas use the interstices of rocky slopes as travel routes, protection from predators and environmental stress, and as den sites for overwintering (Smith 1974b). Typical behavior for these diurnal mammals includes defending territories (using scent markings, aggression, and vocal calls), watching for predators, and foraging (Smith & Weston 1990). Food and material for haypiles is obtained from surrounding meadows and forested areas (Smith & Weston 1990). Pikas preferentially establish territories close to the talus-vegetation interface as it facilitates foraging and limits exposure to predators when feeding (Brandt 1983). Typically, individuals of opposite sex are found in adjacent home ranges and the population density of a habitat patch is low (Brown et al. 1989). Females produce an average of two offspring per year; the juveniles exhibit philopatry and quickly settle in the nearest available territory (Smith 1978). For small mammals, pikas are relatively long-lived, with a maximum lifespan of about seven years (Smith & Weston 1990). The origins and distribution of the American pika have been elucidated through the use of neutral markers. Two species of pika occur in North America: the American pika, and the collared pika (Ochotona collaris) (Smith & Weston 1990). Both likely evolved from an Asian species that dispersed across the Beringian Land Bridge during the Pliocene era, and extended the range southward through western North America (Niu et al. 2004). The  15geographic separation and consequent speciation of O. princeps and O. collaris is thought to result from climatic warming following the Wisconsinan glaciation: glacial retreat isolated pikas on high mountain refuges such that gene flow between northern populations of collared pika and southern populations of American pika was extremely limited (Galbreath & Hoberg 2012) (Figure 1). These conditions, and the influence of pika life history characteristics (e.g. habitat specificity and limited dispersal ability) also produced different lineages across mountain systems (Galbreath et al. 2009). Phylogenetic analyses of O. princeps mitochondrial DNA haplotypes identified five separate lineages: the Southern Rocky Mountains, the Northern Rocky Mountains, Central Utah, the Sierra Nevada, and Cascade Range (Galbreath et al. 2009). The two northern pika lineages (the Northern Rocky Mountains and Cascade Range) exhibit genetic signatures of range expansion in the last glacial maximum, followed by decline, whereas the three southern lineages (the Southern Rocky Mountains, Central Utah, and the Sierra Nevada) show a contrasting pattern of stability followed by contraction (Galbreath et al. 2009). It is apparent that the origin of pikas in North America was heavily influenced by climate-driven divergence and subsequent speciation; for the American pika there is also historic evidence for regional and population-level variation across the range, likely driven by local environmental factors (Galbreath et al. 2009). This pattern is reflected in the current distribution of the American pika, which extends from central British Columbia in the north, to California in the south, and New Mexico in the east (Figure 1). Considering such a varied landscape, it is probable that populations in different parts of the range will possess unique signatures of local adaptation and may not be equally impacted by climate change. The results of repeated surveys of  16Great Basin populations would suggest that the American pika is at risk of extinction: a nearly five-fold increase in the local extinction rate was documented during the last ten years (Beever et al. 2011); an upward range displacement of 145 m has been recorded (Beever et al. 2011); and over 25% of surveyed sites have been extirpated in the 20th century (Beever et al. 2003). Rising temperatures are not the only stressor for these populations, but also anthropogenic factors such as livestock grazing and transportation routes (Beever et al. 2008). However, populations in Wyoming, Colorado, and New Mexico have not exhibited comparable vulnerability to these factors or signs of decline (Erb et al. 2011).  Regional differences across the range of the American pika likely originated as a result of climatic fluctuations throughout the Pleistocene. Population expansion and increased gene flow occurred during glacials, followed by population contraction and decreased gene flow during interglacials. Within mountain ranges, the dominant pattern of movement was along elevation gradients, in and out of high-elevation refugia as the climate warmed and cooled. Low-elevation populations that established gene flow between mountain ranges were the most likely to be extirpated due to warming temperatures during interglacials. The isolating effects of past interglacials are similar to the impacts of current global warming. The limited dispersal capacity, heat intolerance, and philopatric behavior exhibited by pikas only serve to increase population isolation (Smith 1974a; Peacock & Smith 1997a; Henry et al. 2012b). Local adaptation may be occurring in populations persisting under unique environmental conditions as a result of differential selection pressures across gradients (longitude, latitude, and elevation) (Henry et al. 2012a; Henry et al. 2012b). A comparative study of Ochotona princeps and Pantholops hodgsonii (an antelope that inhabits the Qinghai-Tibetan Plateau along with several subspecies of  17Ochotona, including O. curzoniae and O. daurica) revealed signals of positive selection in the American pika for genes involved in DNA repair and the production of ATPase (Ge et al. 2013). Ge et al. (2013) hypothesized that convergent evolution was acting upon genes associated with hypoxia, and that common genetic mechanisms are responsible for high-altitude adaptation. The American pika is uniquely adapted, both behaviorally and physiologically, to alpine environments. For example, overwintering in situ is used as an alternative strategy to hibernation; survival is dependent upon haypiles that are constructed from forage collected throughout the summer (Smith & Weston 1990). Under changing climatic conditions though, this behavior may compromise individual fitness. If global warming decreases the average snowpack insulation, pikas may be exposed to increasing amounts of freezing rain and cold, thereby reducing survival rates due to the resource and energy demands of thermoregulation (Smith 1978). During the summer, pikas use behavioral thermoregulation (i.e. inactivity during the hottest hours of the day) to buffer their narrow range of thermal tolerance (Smith & Weston 1990). Exposure to acute and long-term temperature increases, however, may increase the overall extinction risk. In early 2010 the U.S. Fish and Wildlife Service declined a bid to extend endangered species status to the American pika, largely due to a lack of support for range-wide species decline and extirpations (USFWS 2010). It is apparent that there is still much to be learned about populations in different parts of the species? range, and about pika ecology and behavior in general. For example, just recently a study on herbivore interactions documented patch selection by collared pika for vegetation previously grazed by caterpillars (Barrio et al. 2013). As interest in the American pika is increased due to its potential as an  18indicator species for climate change, populations in new and unique environments are being targeted for study; specifically, populations at the geographic range periphery and bioclimatic periphery. However, the study of populations in unique environments is still underrepresented in the literature and the thesis project described here is designed to fill that knowledge gap. Once more is understood about the complex relationship between the American pika and its environment, it may be possible to better predict the species? responses to, and future distribution under, climate change.   19Chapter 2 Population genetics and behavior of O. princeps in an atypical environment   2.1  Background In highly anthropogenic environments, the impacts of human activities are both conspicuous and severe. Landscapes are transformed by buildings and developments, fragmented by travel routes, and mined or harvested for resources. While these factors alone pose a serious threat to biodiversity, there are also the synergistic and ubiquitous effects from climate change. According to the Intergovernmental Panel on Climate Change (2007) global temperatures are expected to increase 2-6?C by the end of the century. Concurrent with this dramatic shift is the exponential growth of the human population, projected by the United Nations (2011) to reach ten billion by 2100. The impacts on Earth?s biodiversity, in the form of species declines and extinctions, are predicted to be both acute and permanent (Gibson et al. 2009). These anthropogenic changes are occurring at a rate poorly matched by evolutionary forces such as natural selection and adaptation (Sgro et al. 2011). Given the anticipated consequences, there is considerable interest in obtaining a more detailed understanding of species? responses to this transforming world (Peck 2011). The effects of climate change on biological systems have been documented at multiple levels. There is evidence for altered life history strategies and adjustments to the phenology and physiology of organisms: for example, earlier spring flowering for plants, and earlier breeding times for birds (Wuethrich 2000; Walther et al. 2002). Plant community composition is changing, and with that, the interactions that drive ecosystem dynamics (McCarty 2002; Root et al. 2003). Geographic range shifts have been mapped for organisms such as alpine plants (Grabherr et al. 1994), butterflies (Parmesan 1996), and birds (Thomas  20& Lennon 1999), with movement patterns directed mostly poleward in latitude or upward in elevation. There have also been documented population declines and extirpations (Thomas et al. 2004); in fact, an estimated 25% of known species are currently facing extinction in the face of climate change (Schipper et al. 2008). While the aforementioned case studies and predictions appear serious enough, climate change is only one of a number of anthropogenic impacts on biodiversity. Habitat destruction and fragmentation are also major causes for conservation concern. Habitat loss has a large, negative impact on biodiversity (Haila 2002); however, the effects of fragmentation are more difficult to characterize. Across numerous studies, a positive relationship was identified between movement and corridors, and species richness and connectivity (Debinski & Holt 2000). Yet it is difficult to make broad generalizations based on responses to fragmentation. Insights from habitat fragmentation studies vary depending upon the scale (from highly localized to continental analyses), the measured response variable (abundance, distribution, movement, etc), and the taxon of interest (Fahrig 2003). Consequently, it is necessary to investigate impacts and responses to habitat fragmentation on a case-by-case basis. In particular, studying organisms that have evolved to exist in metapopulations within inherently patchy habitats, and characterizing life history traits such as mating system and dispersal strategies, will help predict the extent to which a species may be affected by habitat fragmentation. In response to various anthropogenic threats, wildlife species must either rapidly adapt to new conditions within their geographic range, or disperse to different areas that have become environmentally suitable. If neither option is possible, population decline, extirpation, and eventual species extinction will result. There is a pressing need for research  21into the ability of organisms to persist at the tolerance limits of their bioclimatic envelope, as this information will help assess the likelihood of adaptation or dispersal in the face of changing environmental conditions.  The American pika (Ochotona princeps) is an appropriate model species for studies of metapopulation dynamics in a fragmented landscape, as well as studies of species adaptability and persistence in atypical environments. These small lagomorphs inhabit talus that is distributed patchily across the landscape (Smith & Weston 1990). Forest matrix and other dispersal barriers separate suitable habitat patches; as such, a natural metapopulation structure is established. The American pika is considered the best-known mammalian example of a classic metapopulation with significant population turnover, primarily influenced by habitat characteristics (Moilanen & Smith 1998). Vegetation and meadows adjacent to talus slopes provide important forage for pikas; typically, they must rely on constructed haypiles as a food source during the winter (Smith & Weston 1990). The talus slopes themselves also have thermoregulatory properties (Beever et al. 2010; Millar & Westfall 2010; Henry et al. 2012a). Pikas have a limited thermal tolerance and will use the rocks and underlying passageways to thermoregulate behaviorally (Macarthur & Wang 1974). Thermoregulation is particularly important in environments subject to conditions at the tolerance limit of the species? bioclimatic envelope. Due to some of its unique characteristics, the American pika has been previously described as a habitat-specialist species (Smith & Weston 1990). However, there is significant climatic and topographic variation across the species? range, and considering evidence of local adaptive behavior (Simpson 2009; Varner, personal communication), the American pika may be more of a habitat generalist than previously thought.  22The geographic range of the American pika extends from the southern Sierra Nevada and Rocky Mountains up to the northern Cascade, Coast, and Rocky Mountains (Figure 1). Particularly in the Great Basin area, there have been documented population extirpations and upward range retraction in the past few decades (Beever et al. 2003; Beever et al. 2011). Climate metrics have been identified as good predictors of persistence patterns; chronic heat stress and acute cold stress are some of the best predictors (Beever et al. 2010). Dynamic models of climate-mediated extinction predict that pikas may be lost from as much as 80% of their current range by the end of the century (USFWS 2010). However, more work may be required to improve current models as the relationship between predictor variables and pika persistence is highly correlational. As well, it remains poorly understood whether losses will be highest in populations experiencing the greatest climate change, or those currently living in sites at the edge of the species? bioclimatic envelope (Beever et al. 2010). Using a species-distribution model to predict distribution of suitable habitat under range-wide temperature increases, Calkins et al. (2012) predicted that range collapses would proceed until only populations in island-biogeographic ?mainlands? remained. These mainlands were not located in the geographic range center, but reflect instead the general topography of the distributional area of pikas, wherein large, high-elevation mountain ranges occur in peripheral areas (Calkins et al. 2012). Most previously published research has focused on sites at the range core, and populations inhabiting alpine talus, without explicitly investigating more complex habitat associations that may exist at bioclimatic and range peripheries, nor the potential of American pikas to expand into new areas as traditional habitat is lost. However, there are many examples of pikas inhabiting atypical sites. As far back as 1953, pikas were  23documented in quarries and roadcuts located in the Oregon Cascade Mountains (Roest 1953). Lutton (1975) found pika populations inhabiting the tailings from abandoned mines, and establishing territories in lumber piles in the Rocky Mountains of Colorado. Some of the most extensive, spatially contiguous potential habitat available to the American pika is found in the relatively low-elevation lava flows of Craters of the Moon National Monument and Preserve in southern Idaho (Rodhouse et al. 2010). Recently, Millar and Westfall (2010) conducted a broad-scale study across the central-eastern Sierra Nevada and adjacent Great Basin mountain ranges, documenting pika occurrences in stone walls, mine tailings, rockwork dams, and the foundations of historic buildings. Similarly, a study by Manning and Hagar (2011) in Western Oregon found pikas in 42 sites of anthropogenic origin, including roadcuts, quarries, and railroad riprap. Many of these sites were below 1200 m, and covered a range of slope aspects (Manning & Hagar 2011). Recurrent use of such atypical habitats suggests that pikas actually have a broader range of suitable habitat than the talus model suggests, and that the definition of ?typical? habitat needs to be revisited.  Regardless of the specific habitat type that pikas are colonizing, a common feature is the structure of patches surrounded by unsuitable matrix.  In the resulting metapopulations, isolated populations can experience restricted gene flow, inbreeding, lower effective population size, and reduced genetic variation (Peacock & Smith 1997a). Moreover, juveniles show a high tendency to remain on their natal or adjoining home range (philopatry), and in general, dispersal events are discouraged by high predation risk and unsuitable conditions across non-talus habitat (Smith & Weston 1990). Depending on the elevation and latitude, maximum dispersal distances range from 300 m to 10 km, with relatively few long-distance dispersals compared to short movements between neighboring  24patches (Smith 1974a; Hafner & Sullivan 1995; Beever et al. 2010). In a system with limited gene flow, inbreeding is an inevitable consequence. Yet, a population level study by Peacock and Smith (1997b) observed specific mate selection for individuals of intermediate relatedness. These results suggest that in potentially isolating talus habitat, behavioral mechanisms have developed to avoid close inbreeding and the associated genetic consequences. In order to understand more about pika persistence and adaptability, it is necessary to study populations that occur in different habitat types, and under a variety of environmental conditions, across the entirety of the pikas? geographic range. One area of interest is the Columbia River Gorge, located in the Pacific Northwest of the United States. The canyon follows the Columbia River for approximately 130 km and cuts a path through the Cascade Mountains. Travelling the Columbia River Gorge from east to west, the landscape changes from dry, rocky hillsides and open farmland to one dominated by mountains, rivers, and lush forests; it is in the western part of the Columbia River Gorge where pika populations have been documented (Simpson 2009). Here, pikas inhabit low-elevation talus slopes previously considered outside the species? bioclimatic range (Simpson 2009). This unique habitat may have selected for adaptations not present in typical high-elevation, montane populations. For example, the abundant vegetation and mild winter climate of the area allow for year-round foraging; as such, pikas are rarely observed using haypiles to cache food (Simpson 2009). Understanding how pikas use atypical habitats will assist in determining the function and quality of the habitats, and thus serve to direct management and conservation strategies.  Currently, the American pika is a species of interest for conservation. In late 2010, the USFWS did not accept a petition to list the American pika as an endangered species  25(USFWS 2010). Evidence was presented for population declines and extirpations linked to climate change in parts of the species range; however, the ruling was justified based on a lack of consistency for this pattern across the entire distribution (USFWS 2010). As previously described, there are incidents of population persistence in parts of the range where environmental conditions were previously considered unsuitable for pikas. It remains to be determined whether the American pika faces a genuine extinction risk, warranting increased conservation protection, or whether the current status of ?least concern? should be maintained.  The purpose of this study was to investigate the population genetic structure, demographic history and behavior of American pika inhabiting an atypical environment. The Columbia River Gorge was selected as the study area because pika habitat is found there at low elevations (Manning & Hagar 2011), remains free of snow cover during the winter (Simpson 2009), and experiences conditions at the limit of the pikas? bioclimatic envelope (Simpson 2009). These habitat characteristics are largely considered atypical for pika, which are more broadly found under the following conditions: talus located at elevations ranging from sea level to 3000 m in the northern part of the range, and elevations above 2500 m near the southern limits of the distributional range (Smith & Weston 1990); areas where deep snow cover provides winter insulation (Smith 1978); and habitat that experiences, on average, <30 days/year with temperatures above 35?C (short summers), and >150 days/year with temperatures below 0?C and a continuous freeze-free period <120 days (long winters) (Hafner 1993). I used microsatellite genotypic data to quantify levels of genetic variation within and among sampled sites in the Columbia River Gorge. These data were used to reconstruct patterns of within-site relatedness, inbreeding, and demographic  26history, as well as among-site population structure and connectivity. Paired with sex identification using a novel molecular test, these genetic data provided inferences into mating behavior and sex-specific dispersal strategies. The findings from the Columbia River Gorge were subsequently compared with patterns that have emerged from previous studies at sites more central to the pika?s range, as well as on the northern periphery of its range, to expand our understanding of pika habitat associations and explore the implications for conservation and management.  2.2 Methods Research was carried out in the Columbia River Gorge, Oregon, USA (Figure 2). In order to select study sites along the Columbia River Gorge, Google Earth satellite imagery was compared against topographical trail maps. Talus slopes were identified by their color and defined rock formations; patches that were accessible by road or trail were visited to determine if the habitat supported pikas.  A total of 11 sites were established: five low-elevation sites between 0-300 m, and six high-elevation sites between 1000-1325 m (Figure 2, Table 1). An individual site consisted of either one large talus patch enclosed by forest, or a few smaller patches connected by talus corridors.  2.2.1 Sample collection Once a potential site was accessed, the presence of pikas was confirmed by: a) sighting a pika, b) hearing an alarm call, or c) finding a latrine site. The latter of these indicators proved the most important, as in many cases pikas were not seen or heard at a site until several days after the first visit. In order to sample individuals, noninvasive hair snares were adapted from Henry and Russello (2012) and set into crevices in the talus.  27At low-elevation sites, tape snares were most effective when set deeper into the rock (decreasing exposure to moisture that was found to dissolve sticky adhesive and increase visibility of the clear tape), and when set around pika ?travel routes? clearly worn into the moss. These modifications suited the cool, wet conditions typical of low-elevation talus, where a carpet of moss covered the rock and vegetation encroached into the talus (Figure 3). Pikas were more vocal at the low-elevation sites (personal observation), and thus areas where an alarm call was heard were the primary targets for snaring. A different approach was used to set snares at the high-elevation sites. Rock size was larger, creating a deep network of elaborate pathways and openings. As a result, a ?baited? snare was used: fake haypiles made from grasses and flowers served to attract pikas to an area where tape webs had been set (Figure 4). This approach worked well on dry, sun-exposed talus that was free of moss (Figure 4). Because pikas were relatively less vocal at the high-elevation sites, the location of latrine sites helped guide where baited snares were set (Figure 5). Approximately one week was dedicated to each site. An initial observation period was carried out to: 1) determine prospective locations for snare setting (based on where pikas were observed and heard, and where latrine sites were found); 2) mark locations with neon duct tape; and 3) take digital images. An intensive, all-day effort was required to set tape webs across the entirety of a talus slope (Figure 6). The site was then revisited two days later for sample collection: using sterile technique, sections of hair-covered tape were cut from the snare and maneuvered into a sample tube (Figure 7). New snares were then set, and the sample collection procedure repeated two days later. Snare success decreased dramatically when wet, and sites had to be visited more frequently to reset snares during  28rainy periods. The sampling period spanned June 15 to July 15, 2012, resulting in the acquisition of 316 samples from 11 sites. 2.2.2 Laboratory methods  2.2.2.1 DNA extraction The first step in the DNA extraction required the removal of pika hair from the tape adhesive. After experimenting with treatments of acetone, baby oil, hairspray, and olive oil, I found that Goo Gone? cleaning solution effectively loosened the adhesive without deteriorating the quality of the extract and PCR amplification product.  Samples were submerged in Goo Gone? for several seconds, and forceps were used to scrape the hair off the tape and rinse the hair through a warm water bath. The DNA extraction was carried out using a Promega Tissue and Hair Extraction Kit (for use with DNA IQ?), following a modified protocol: 1) each sample was incubated in 100 ?l of Incubation Buffer/ Proteinase K solution at 56?C for 3 hours under light agitation; 2) a volume of 200 ?l of lysis buffer was added following the incubation; and 3) depending on the amount of resin left in the tube after subsequent washes, a volume of 60-100 ?l of elution buffer was used.  2.2.2.2 Primer optimization  Twenty-eight microsatellite loci have been previously characterized in the American pika (Peacock et al. 2002; Peacock & Kirchoff 2009). Previous optimization of these microsatellites by Henry et al. (2012b) identified a subset of 10 loci that successfully amplified in individuals from pika populations at the northern periphery of the range (in British Columbia). However, varied success was encountered using these microsatellites with individuals from the study populations in the Columbia River Gorge, and only six loci from the subset of 10 were retained (Ocp 02, Ocp 06, Ocp 07, Ocp 11, Ocp 13, and Ocp 15)  29(Peacock et al. 2002). Nine additional microsatellite loci from the original list were therefore tested using the Columbia River Gorge samples, three of which (Ocp 10, Ocp 19, and Ocp 25) (Peacock & Kirchoff 2009) were retained after optimization.  In total, 9 microsatellite loci were used to provide genotypic data for all downstream analyses (Table 2). 2.2.2.3 Multiplex Polymerase Chain Reaction (PCR)  Due to low DNA quality and quantity typical of hair as a starting material, a two-step, multiplex approach was implemented that enhances the amplification of microsatellite loci (Piggott et al. 2004; Henry et al. 2009). Specifically, the first step involved a multiplex PCR that simultaneously amplified multiple loci in a single reaction. Combinations of forward and reverse primer pairs were as follows: a) Ocp 02, Ocp 07, Ocp 11, Ocp 13, and Ocp 25; and b) Ocp 06, Ocp 10, Ocp 15, Ocp 19, and Ocp 22. The multiplex PCRs were carried out on an ABI Veriti thermal cycler in 25 ?l reactions containing 1-25 ng of DNA, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.08 ?M of each primer (10 primers in total), 0.2 mM dNTPs, 10 ?g bovine serum albumin, and 1 U of AmpliTaq Gold DNA polymerase (Applied Biosystems). Cycling parameters consisted of 95?C for 10 minutes, 25 cycles of 95?C for 30 seconds, 50?C for 90 seconds, 72?C for 45 seconds, and a final extension at 72?C for 10 minutes.  The product of the multiplex PCR was then used as the genetic material for subsequent individual PCRs with each primer pair. All forward primers were 5?-tailed with an M13 sequence [5?-TCCCAGTCACGA-CGT-3?] to facilitate automated genotyping. Specifically, the M13-tailed forward primer was used in combination with an M13 primer of the same sequence, 5?-labeled with one of four fluorescent dyes (6-FAM, VIC, NED, PET), effectively incorporating the fluorescent label into the resulting PCR amplicon (Schuelke  302000). In addition, reverse primers were modified following Brownstein et al. (1996) to improve genotyping. The 13 ?l reactions for the second step contained: 1.5 ?l of PCR product from the first step, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.08 ?M of forward primer, 0.8 ?M each of the reverse and M13 primer, 0.2 mM dNTPs, 10 ?g bovine serum albumin, and 0.5 U AmpliTaq Gold DNA polymerase (Applied Biosystems). A touchdown cycling program was used for all loci except Ocp 06, with parameters set at 95?C for 10 minutes, 35 cycles of 95?C for 30 seconds, 30 seconds of annealing, 72?C for 45 seconds, and a final extension at 72?C for 10 minutes. The annealing temperature decreased by 1?C per cycle from 60-55?C until reaching the sixth cycle, at which point the 29 remaining cycles continued at 55?C. For locus Ocp 06, cycling parameters were maintained except for the annealing temperature, which was held constant at 50?C. Finally, PCR products were run alongside a 100 bp ladder (New England Biolabs, MA, USA) on a 1.5% agarose gel containing 2.5% SYBR Safe DNA gel stain (Invitrogen, Carlsbad, CA, USA) and visualized using a RED personal gel imaging system (Alpha Innotech, San Leandro, CA, USA).  PCR products were co-loaded and run on an ABI 3130XL DNA automated sequencer. Two independent investigators manually scored all alleles in GENEMAPPER 4.0 (Applied Biosystems).   2.2.2.4 Molecular sexing A novel molecular sexing protocol was used to classify each individual as male or female (Lamb et al. 2013). As no O. princeps sex-determining region (SRY) sequences had  31been previously generated, the rabbit (Oryctolagus cuniculus) SRY gene DNA sequence (GENBANK accession # HM230423) was downloaded from GENBANK. This sequence was then used in BLAST searches within the ENSEMBL genomic browser of unannotated scaffolds deposited for O. princeps. Recovered scaffolds of high sequence similarity were then aligned with the O. cuniculus SRY gene fragment in SEQUENCHER 5.0 (Gene Codes Corporation). Primers were designed to target a specific region (~117 base pairs) of the O. princeps SRY gene using PRIMER3 (Rozen & Skaletsky 2000). The SRY fragment was PCR co-amplified with a higher molecular weight nuclear microsatellite (~200 base pairs; Ocp 10). Due to the nature of the assay, males exhibited both SRY and Ocp 10 products, while females only possessed the Ocp 10 PCR product (Figure 8). The accuracy of the protocol was validated using samples of known sex from different starting materials (hair and liver) (Lamb et al. 2013). For the individuals sampled in the Columbia River Gorge, PCRs were carried out in 12.5 ?l reactions containing 1-25 ng of DNA, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.8 ?M of each primer (forward and reverse), 0.2 mM dNTPs, 10 ?g bovine serum albumin, and 1 U ABI AmpliTaq Gold DNA polymerase. Cycling parameters were set at 95?C for 10 minutes, 40 cycles of 95?C for 30 seconds, 48?C for 30 seconds, and 72?C for 60 seconds, and a final extension at 72?C for 7 minutes. PCR products were run alongside a 100 bp ladder (New England Biolabs, MA, USA) on a 1.5 % agarose gel containing 2.5% SYBR Safe DNA gel stain (Invitrogen, Carlsbad, CA, USA) and visualized using a RED personal gel imaging system (Alpha Innotech, San Leandro, CA, USA). Gel images were analyzed for the presence of clear, bright bands, and samples coded as male or  32female (Figure 8). The sexing protocol was used on each sample twice, and any samples that had incongruences between the two trials (approx. 5%) were repeated a third time. 2.2.3 Data analysis 2.2.3.1 Quality assessment  The genotypic data set was examined for the presence of null alleles using MICROCHECKER (Van Oosterhout et al. 2004). Given the non-invasive sampling approach, there is a probability that the same individual may be represented by more than one hair sample. In order to identify repeat individuals, a multilocus match analysis was performed in GENALEX (Peakall & Smouse 2006). From a pair of samples that matched at all loci, or all loci but one (excluding missing data), the sample with the most missing data was discarded. Deviations from Hardy-Weinberg equilibrium (HWE) and linkage equilibrium were assessed using exact tests, as implemented in GENEPOP 3.3 (Raymond & Rousset 1995; Rousset 2008). Type I error rates for tests of linkage disequilibrium and departure from HWE were corrected for multiple comparisons using the sequential Bonferroni method (Rice 1989). The single locus and multilocus probability of identity (i.e. the probability that two individuals drawn at random from a population will have the same genotype) was calculated in GENALEX (Peakall & Smouse 2006).  2.2.3.2 Population genetic variation and demographic history  Allelic diversity, observed heterozygosity (Ho), and expected heterozygosity (He) were calculated for each site using ARLEQUIN (Excoffier & Lischer 2010).   Genetic signatures of recent demographic contraction, that is within the past few generations, were assessed within each site using: 1) the heterozygote excess test; and 2) the mode-shift test, both implemented in BOTTLENECK 1.2.02 (Cornuet & Luikart 1997). These  33tests are based on a comparison between observed allele frequencies, and what would be expected in a population at mutation-drift equilibrium. For the heterozygote excess test, significance was assessed using 100,000 iterations with the Wilcoxon sign-rank test and under a Two Phase Model with 90% stepwise mutations.  2.2.3.3 Relatedness and inbreeding Mating behavior was indirectly assessed through measures of within-site relatedness and inbreeding. Relatedness among individuals was estimated according to the method of Queller and Goodnight (1989), as implemented in GENALEX (Peakall & Smouse 2006). A measure of the degree of relatedness between individuals (r) was calculated, which ranges from 0 (no alleles in common across the assessed genetic markers), to 1 (all alleles in common). Inbreeding coefficients (FIS) were calculated in FSTAT (Goudet 1995). The above analyses were repeated for males and females separately. 2.2.3.4 Population structure and connectivity  Levels of genetic differentiation among sites were estimated by pairwise site comparisons of ?, an unbiased measure of FST (Weir & Cockerham 1984), as implemented in GENETIX using 1000 permutations (Belkhir et al. 2004). The directions and magnitudes of contemporary migration rates among sites were estimated using a non-equilibrium Bayesian method implemented in BAYESASS v. 3 (Wilson & Rannala 2003). The program run length was 10,000,000 MCMC replicates after a burn-in period of 1,000,000, sampling the chain every 100 iterations. Consistency of the results was assured by running the program five times using different random seeds and monitoring the output as visualized in TRACER (Rambaut & Drummond 2007).  34 To test for nonrandom associations between genetic differentiation and geographic distance, a Mantel test was performed between the geographic distance matrix and a second matrix of ? values, using the ISOLATION BY DISTANCE WEB SERVICE (Jensen et al. 2005). The geographic distance matrix was created using GOOGLE EARTH. First, straight-line paths were drawn to connect sites; second, the elevation profile of every path was accessed, which takes the topography into consideration and calculates vertical displacement from sea level; and third, the length of the hypotenuse was calculated. To further examine isolation by distance patterns, a non-stationary genetic friction map was generated, representing the relative rate at which genetic differentiation changes per unit of geographic distance. The analysis used a similarity matrix of 1-? values, a table of longitude and latitude coordinates for each site, four simulated neighbors at a distance of 0.035 km apart, and 100 posterior replicates, as implemented in LOCALDIFF (Duforet-Frebourg & Blum 2012). The output was then processed in Rstudio? and associated with a topographical map of the study area. To assess levels of population subdivision within the data set, a Bayesian clustering analysis was implemented in STRUCTURE (Pritchard et al. 2000). Run length was set to 500000 MCMC replicates after a burn-in period of 250000 using correlated allele frequencies under a straight admixture model, as well as using the LOCPRIOR option (which uses sampling locations as prior information to assist in clustering). The most likely number of clusters was determined by: 1) varying the number of clusters (K) from 1 to 15 with ten iterations per value of K; and 2) implementing the ?K approach (Evanno et al. 2005) in STRUCTURE HARVESTER (Earl & von Holdt 2011).  35Analyses of molecular variance (AMOVAs) were performed in ARELQUIN (Excoffier & Lischer 2010) to explore the distribution of genetic variation across the following grouping strategies: 1) by site; and 2) by inferred STRUCTURE clusters. For both male and female data sets, levels of genetic differentiation among sites were estimated by pairwise site comparisons of ?  (Weir & Cockerham 1984), performed in GENETIX using 1000 permutations (Belkhir et al. 2004). Sex-specific migration rates were calculated in BAYESASS v. 3 (Wilson & Rannala 2003) as above. 2.3 Results Of the 316 extracts tested, six did not PCR amplify at any loci and were presumed to be of non-pika origin. From the remaining samples, 202 were retained with information at six or more loci. Lastly, the conservative approach to detect repeat individuals removed 47 samples, producing a final count of 155 pikas. The overall dataset contained 19.4 % missing data, ranging from 0% (Ocp19) to 59% (Ocp25). 2.3.1 Quality assessment There was some evidence for null alleles, but no consistent patterns emerged that warranted the removal of a locus. Loci Ocp 07, Ocp 15, and Ocp 06 were flagged for potential null alleles in three, four, and five sites, respectively. Other loci (Ocp 02, Ocp 10, Ocp 13, and Ocp 25) were found to have null alleles in only a small proportion of sites (n=1-2).  Deviation from HWE was found at some loci at some sampling sites. Specifically, Ocp 06 deviated at three of the eleven sites, Ocp 15 at three sites, Ocp 25 at two sites, and Ocp 07 at two sites. There was no single site with more than two deviating loci. Subsequent analyses with and without the allele calls for those individuals at those loci produced highly similar results (data not shown). No instances of non-random associations between  36genotypes were reported when tested for linkage disequilibrium. The probability of identity varied across the nine microsatellites, ranging from 0.14 (Ocp 02 and Ocp 06) to 0.73 (Ocp 25). The multilocus probability of identity for this suite of nine microsatellite loci was 7.3x10-4, providing confidence in our ability to detect unique individuals. 2.3.2 Population genetic variation and demographic history Within site genetic variation was relatively low, averaging 3.14 alleles per locus, and mean observed and expected heterozygosities of 0.39 and 0.54, respectively (Table 3).  There was some evidence for recent demographic contraction, based on the heterozygote excess and mode-shift tests. Sites I, Q, and R showed a significant pattern of excess heterozygosity relative to the number of alleles expected at mutation-drift equilibrium (Table 3). In addition, over half the sites displayed shifted allele-frequency distributions (sites I, Q, R, B, D, and J) (Table 3). 2.3.3 Relatedness and inbreeding Pairwise relatedness at the site level ranged from -0.08 to 0.31, with an average of 0.14 (Table 4). Inbreeding coefficients (FIS) were always positive and differed significantly from zero. Values ranged from 0.21 to 0.40, with an average of 0.30 (Table 4).   Approximately 96% of sampled individuals (n=150) were successfully sexed using the newly developed molecular approach. A biased ratio of 93 males: 57 females was revealed. Mean male relatedness (0.153; range -0.043 to 0.317) was lower than that recovered in females (0.260; range -0.083 to 0.567) (Table 4). Likewise, mean inbreeding coefficients (FIS) for males (0.448; range 0.104 to 1.174) were lower than those in females (0.648; range 0.068 to 2.383) (Table 4). 2.3.4 Population structure and connectivity  37Significant genetic differentiation was detected for most (82%) of the pairwise site comparisons of ?  (P<0.05) (Table 5), with average ? values centered around 0.119. This high degree of differentiation was reflected in the low estimates of recent migration rates (m; the proportion of migrant individuals per population per generation), which ranged from 0.009 to 0.142, with an average of 0.023 (Table 6). These values were compared to the suggested migration rate at which two populations exchange sufficient migrants to influence each other?s dynamics (m=0.1). Significant migration was found between two high-elevation sites (from site D to site R).  Regarding sex-specific patterns, significant genetic differentiation was detected for 43% of the male pairwise site comparisons of ? (P<0.05) (Table 7), compared to 65% of the female comparisons (Table 7). While there were no instances of significant migration for males (average 0.028; range 0.011 to 0.099) (Table 8), for females, significant migration occurred from site H to site Y and from site Q to site Y (average 0.027; range 0.016 to 0.012) (Table 9).  An overall pattern of isolation by distance was detected (r2=0.170, P=0.006) based on Mantel tests, where among-site genetic variation was positively correlated with geographic distance (Figure 9). The genetic friction map showed a progressive decrease in genetic friction moving across the study area from the western-most low-elevation sites to the eastern-most high-elevation sites (Figure 10). In areas of elevated genetic friction, a relatively higher rate of genetic differentiation per unit of distance was occurring.  The clustering analysis performed in STRUCTURE revealed significant population subdivision within the dataset. There were two genetic clusters detected, with ?K modal at K=2 (Figure 11). The individuals and sampling sites that have relatively uniform  38membership to a single cluster can be described as follows: low-elevation sites (H, T, Y, and J), and high-elevation sites (U, I, Q, B, D, R). The low-elevation site M is the one exception, as it grouped with the cluster of high-elevation sites. However, the AMOVA revealed that the among-site component of genetic variation was maximized by the eleven groupings based on sampling location, as opposed to the two groupings based on STRUCTURE. Under the STRUCTURE groupings, 6.1% of the total variation was among groups, and 93.9% was within groups. Under the sampling site groupings, 12.5% of the total variation was among groups, and 87.5% was within groups. 2.4 Discussion Much of the current pika literature addresses populations at the core of the species? distribution, and populations inhabiting typical alpine talus. Although pika occupancy has been reported in atypical habitats, ranging from areas once thought to be outside their bioclimatic envelope (Simpson 2009), to artificial sites such as abandoned mines, lumber piles, and other human-made structures (Smith 1974a; Millar & Westfall 2010), little is known regarding population structure, demographic history, and behavior in such areas. In order to properly evaluate the status of the American pika (whether to be listed as least concern, threatened, or endangered), and create sound conservation and management policies, it is necessary to consider the entire species distribution and compare populations from different parts of the range. This work addresses the knowledge gap in the literature by reporting on the population genetics and inferred behavior of American pika inhabiting an atypical environment in the Columbia River Gorge.  Several patterns emerged regarding levels of population genetic variation and structure, dispersal and mating strategies, and population connectivity. For this system, high inbreeding and low genetic variation best characterized pikas within a site. A high degree of  39structure was detected among sites, and differentiation increased where topographical features potentially served as dispersal barriers. Although pikas inhabiting geographically proximate sites tended to cluster at similar elevations, there was little evidence of statistically significant migration. Indirect measures, however, such as within-site relatedness and inbreeding, strongly suggested a pattern of male-biased dispersal. 2.4.1 Population genetic variation and demographic history  Low levels of allelic variation and observed heterozygosity (0.39) characterized the sampled sites in the Columbia River Gorge (Table 3). These values are lower than those reported at more typical, high elevation sites in Nevada, California, and Montana (0.64, SD=0.15) (Merideth 2002; Henry 2011). In contrast, metrics of site-level genetic variation were more similar between the Columbia River Gorge and Bella Coola, the latter of which is located at the northern periphery of the American pika distribution in British Columbia (Henry et al. 2012b). Similar to the findings in BC, pikas from the Columbia River Gorge sites consistently exhibited a heterozygote deficit, characteristic of small populations in which equilibrium conditions are not met (i.e. genetic drift, non-random mating, and migration are occurring). These findings are in contrast to a microsatellite-based study of American pika in their range core, which found relatively higher levels of genetic variation (Ho= 0.64) despite small population sizes of less than 10 individuals per site (Merideth 2002). Consequently, it appears different levels of variation exist between populations at the range core, and those at geographic and climatic peripheries. This pattern has been previously demonstrated in an investigation of climate change impacts on two lagomorph species, Romerolagus diazi and Lepus timidus (Anderson et al. 2009). In this case, Anderson et al. (2009) developed models that predicted differential responses to change of the leading and trailing edge margins, and a greater sensitivity to climate change at the range limits,  40when compared to the metapopulation centroid. Peripheral populations, and those adapted to atypical environments, may exhibit distinctive characteristics and genetic reserves of biodiversity even when the level of heterozygosity within a population is low (Gibson et al. 2009).  These results are consistent with population genetic theory, especially given the evidence for demographic contraction. The high-elevation sites were over-represented in the heterozygote and mode-shift tests; observations from the field found these sites relatively hotter and drier, and also noted several nearby sites with old signs of pika occupancy (e.g. decaying and scattered haypiles and single fecal pellets) but no current activity. The high elevation sites may be more susceptible to stochastic events like environmental stress, and the underlying low levels of genetic variation impact the ability of a population to respond to environmental change. Adaptation to novel environments is dependent upon: 1) selection on new mutations; and 2) selection on pre-existing genetic variation (Barrett & Schluter 2007). Case studies of ecologically relevant genes in several different species, from beach mice (Peromyscus polionotus) to threespine stickleback fish (Gasterosteus aculeatus), support the hypothesis that standing variation plays an important role in rapid adaptation to novel environments (Colosimo et al. 2005; Barrett & Schluter 2007; Steiner et al. 2007). Determining the source and extent of genetic variation in different populations could help better predict responses to change. 2.4.2 Relatedness and inbreeding Averaging across all sites in the Columbia River Gorge, there was high relatedness and high levels of inbreeding. However, some curious patterns emerge when studying the results at the individual site level. Specifically, site I reported a negative relatedness, yet held the highest inbreeding value. High-elevation sites B and Q similarly reported very low  41relatedness values coupled with high inbreeding coefficients. In contrast to these findings, it is expected that inbreeding increases the genotypic correlation between relatives above the level it would be in a system of random mating (Pamilo 1985), and that both relatedness and inbreeding are increased in smaller populations. The low-elevation site H, for which there was a relatively large sample size, showed the highest relatedness out of all sites. It is possible that the conflicting results for some of the sites are a result of small sample sizes, and do not accurately reflect the biology and behavior of the system. As well, there can be inaccuracies when using microsatellite marker-based estimators of relatedness, often a result of genotyping errors and discrepancies between observed and true genotypes (Blouin et al. 1996; Russello & Amato 2004). If biologically meaningful, my results are comparable to those of Peacock and Smith (1997b), where mating between individuals of intermediate relatedness (e.g. mating between half-siblings, r=0.25) was documented. In the Columbia River Gorge system, pikas showed a mean relatedness of 0.140 across all sites; this value is within the range of what might be produced by mating between individuals of intermediate relatedness. Still, further work is required to determine whether this behavior is site-specific, regional, or prevalent across the entire range of the species.  The molecular sexing data enabled further exploration of patterns of relatedness and inbreeding. When the male and female datasets were compared, levels of relatedness and inbreeding were lower among males than females. Along with fewer instances of significant genetic differentiation between male pairwise site comparisons, this data provided preliminary evidence for male-biased dispersal was uncovered. However, the only instances of statistically significant migration rates were reported for females, not males. This finding  42should be interpreted with caution though, as it was generated from a limited sample size (Table 6), and is likely not representative of what is actually happening in the system. Excluding the few instances of significant female migration, the overall results fit well into the literature on mammalian dispersal patterns and the prevalence of male-biased dispersal (Greenwood 1980; Dobson 1982; Handley & Perrin 2007). However, according to Zgurski and Hik (2012), pikas in North America use a monogamous or polygynandrous mating system, which is characteristic of mammal species without sex-biased dispersal. Indeed, their study of collared pika (n=364) found evidence of equal dispersal rates for females and males over a period of nine years (Zgurski & Hik 2012). They also reported low levels of inbreeding and relatedness values that did not differ from that expected under a model of random choice. While these results may accurately characterize the mating system for O. collaris, it is not appropriate to assume that O. princeps exhibit similar mating behavior, considering that the species diverged during the Pleistocene and have evolved in different ecoregions (Galbreath & Hoberg 2012).  Smith and Ivins (1983) classified the mating system of O. princeps as facultatively monogamous, based on observations of paired configurations, persistence in time, and interplay of aggression and social tolerance. Brandt (1985) further defined the mating system as driven by female choice. For the study system here, there is support for female choice if the unbalanced sex ratio in the dataset is considered: there were ~60% more males than females represented by hair snares. However, it is possible that this result is not representative of the sex ratio for the talus slope, and is instead a product of sampling bias. Hair snares usually targeted areas where a pika was seen or heard; thus, if males were more actively issuing alarm calls and defending their territory, that area would have been  43preferentially snared. It is also possible that on average, males showed less caution and had a greater curiosity toward the tape snares, thereby increasing their sampling number. The proportion of 93 males to 57 females sampled from the Columbia River Gorge deviated significantly from the null hypothesis of an even sex ratio when a goodness of fit test was applied (G=8.70, P=0.003). There is little evidence in the literature supporting an uneven sex ratio for populations of American pika. In two related studies of pika mating behavior, Peacock and Smith (1997b) identified 23 mated pairs, comprised of 23 females and 18 males, and recorded female to male ratios from sample sites of 4:9, 9:14, 17:17, and 11:15 (Peacock & Smith 1997a). A goodness of fit test of these ratios did not find any significant deviations from the null hypothesis of an even sex ratio. Based on these statistics, and the targeted snaring method, it is very likely that my study was affected by sample bias in the field. 2.4.3 Population structure and connectivity  Considered as a whole, the sampled sites separated into two clusters that are well explained by elevation and topography. Sites H, T, Y and J (composing one cluster) are all low-elevation sites situated beneath large cliff faces that would serve as a barrier to upward dispersal. Sites H and T are 350 m apart, and sites Y and J are 690 m apart (Figure 2). The signature of genetic similarity shared by these sites suggests that gene flow is occurring; in fact, there were several instances when pikas calls were heard in the forest and small, overgrown talus patches were encountered when travelling between sites (personal observation). The other cluster is composed of high-elevation sites (U, I, Q, B, D, and R). The distances between these sites ranges from 0.87 km to 7 km, with an average distance of 3.6 km. Based on this information, and recorded maximum dispersal distances for pika ranging from 300 m to 10 km (Smith 1974b; Hafner & Sullivan 1995; Beever et al. 2010), it  44is reasonable to assume that gene flow occurs between these sites, particularly because there are a number of intermittent talus patches. Site M is the one anomaly here; this low-elevation site showed membership within the cluster of high-elevation sites. The nearest sampled high-elevation site is 5.9 km away; however, a close examination of the topography around site M does reveal a more gradual slope, and several stepping stone talus patches up to an elevation of ~1180 m. It is quite possible that these factors encouraged dispersal and gene flow along the altitudinal gradient. There was also evidence for substructure within the STRUCTURE clusters, as the results of the AMOVA make a case for assigning more groupings to the data set. Similarly, a heterozygote deficit and lack of allelic richness within all sites were detected. Such measures are characteristic of isolated populations with limited gene flow, and conform to what is expected of pikas living in fragmented talus habitat. As well, a weak signal of isolation by distance was detected (based on a low r2 value of 0.170, which indicates the strength of the relationship). Talus sites that were geographically farther apart showed a higher level of genetic differentiation. In such cases, there is limited gene flow between far-apart sites as dispersal events are less successful across greater distances of poor quality habitat, likely due to increased exposure to predation and unfavorable climatic conditions.  Spatially-explicit analyses provided additional insights, recovering a clear pattern of decreasing genetic friction from the western-most low-elevation sites to the eastern-most high-elevation sites. It appears that for sites located on a slope, where most dispersal events would require a significant change in elevation, there is greater genetic variation per unit of distance. The high-elevation sites R, D, B, and Q, all contained in a region of low genetic friction, are located on a plateau. Site M is once again an exception to the trend, as this low- 45elevation talus site is found in an area of relatively low genetic friction. Such a result may be a product of the small sample size at site M, or may implicate surrounding talus patches (leading up to the high-elevation plateau) as dispersal stepping-stones.  As a whole, dispersal appeared to be limited within the study area. Based on among site estimates of contemporary migration rates, only sites R and D were connected by substantial gene flow. The lack of significant migration between sampling sites is consistent with the characterization of pikas as philopatric mammals with a limited dispersal capacity. However, based on the geographic distance separating sites, there are some cases where significant migration might be expected but was not reported (e.g. between sites H and T, <400 m apart, and between sites Y and J, <700 m apart). The low-elevation sites H, T, Y and J are located in dense forest; it is possible that this habitat matrix discouraged dispersal. An alternative explanation is that habitat saturation, as a result of abundant vegetation resources, increases behavioral conflict. Agonistic behavior by resident adults toward juvenile immigrants may discourage dispersal. Therefore, rather than claim that pikas have a limited dispersal ability, it might be more appropriate to recognize that effective pika dispersal can be limited by many factors, such as landscape heterogeneity and behavioral exclusion.  2.4.4 Quality assessment  For many wildlife species, acquiring sufficient sample sizes to inform population genetic analyses can be challenging. Pika studies that involve genetic work have typically employed trapping methods, which are invasive and time-consuming (Peacock & Smith 1997a, b; Peacock et al. 2002). This study benefitted from the use of a non-invasive hair snare sampling approach (Henry & Russello 2012), which enabled the acquisition of samples from ~50% more individuals over a four week period than a comparable study in  46the Sierra Nevada that conducted trapping over four years (Peacock & Smith 1997a, b). However, the larger sample size per unit time comes at the cost of the quality of acquired material, as well as the lack of other individual metrics associated with the sample (mass, sex, age class, etc). Although the final sample size of 155 individuals is robust for the scale of the study, a much larger number of hair samples were collected but failed to amplify consistently using PCR methods. It is likely that the snares collected an abundance of shed hair that typically contains highly degraded DNA in comparison to freshly plucked hair or other more conventional tissues (e.g. blood, skin) that may be collected from trapped individuals. However, relative to other types of non-invasively collected samples (e.g. fecal pellets), hair typically provides DNA of greater quality and quantity (Waits & Paetkau 2005). The major consequence of using a noninvasive sampling approach was that a large number of samples were excluded from the analysis, and moderate to high levels of missing data were present in the final data set, limiting the explanatory power of some analyses.  2.4.5 Summary   Overall, results from this study have added depth to two known aspects of pika biology; namely, non-random mating and limited dispersal ability. At a larger scale, the connectivity between sites and talus distribution suggests that the study system may exhibit metapopulation dynamics. As no obvious source population was identified, a classic Levins metapopulation, describing a population of populations which go extinct locally and recolonize, is arguably the best model (Akcakaya et al. 2007). While it may be that populations of pika in the Columbia River Gorge are well adapted to their environment, based on past records (National Parks Service, personal communication) and continued persistence, the impacts of climatic factors have yet to be fully explored. It is also likely that the microclimate of the talus is a more important variable for pika persistence than the  47macroclimate of the region. Indeed, observations from the field revealed a notable difference in conditions between the high and low-elevation sites; these qualitative observations need to be followed up with quantitative monitoring and data. Currently, climatic variables (temperature and relative humidity) are being measured in the Columbia River Gorge at several of the sites sampled in this study (Varner, personal communication). Data loggers have been placed at and below the talus surface, and will therefore capture the microclimate at each habitat patch. To properly test the hypothesis that talus microclimate is a better predictor of habitat suitability and pika persistence than regional macroclimate, it will be necessary to conduct a long-term study with various measures of environmental and genetic variation.  In addition, there remains much that is still unknown about pika habitat associations. To fully explore this relationship, long-term monitoring of populations inhabiting unique patches (for example, mine tailings and quarries) and atypical environments will be required. Without this temporal aspect, and measures of reproductive success and fitness, it is not possible to determine the true value of the habitats: Are they temporary sites that facilitate dispersal? Can they support source or only sink populations? A more thorough understanding of how pikas use different types of habitat will provide important direction to conservation. Models currently used to predict future distribution patterns make projections based on available patches of ?suitable talus habitat? (Beever et al. 2010). In light of evidence that pikas may have more complex habitat associations, these models should be adjusted to consider non-talus sites and areas outside the species typical bioclimatic zone.  As well, because human-made sites can provide suitable habitat for pika (Moilanen & Smith  481998), land managers may have the opportunity to create dispersal corridors and possibly even develop long-term sites for pika occupancy.    49 2.5 Figures                   Figure 1. Distribution of Ochotona princeps across the Intermountain West. Grey areas are the approximate range, and black dots are known pika locations. The X indicates the sampling area for this study. The inset at the top corner shows the approximate distribution of the collared pika. Image modified from Galbreath et al. (2010).   50                 Figure 2. Distribution of sampling sites for O. princeps in the Columbia River Gorge, Oregon, USA. Sites H, T, Y, J, and M are low-elevation sites (below 300 m). Sites U, I, R, D, B, and Q are high-elevation sites (above 1000 m).  51  Figure 3. O. princeps occupy low-elevation talus slopes that are characterized by small-medium sized boulders, moss cover, and abundant vegetation.  52    Figure 4. O. princeps occupy high-elevation talus slopes that are characterized by medium-large sized boulders and surrounding shrubby vegetation. A baited snare is shown in the image on the right.   53    Figure 5. A pika latrine site. An indicator of the presence of O. princeps at a talus slope.   54    Figure 6. Packing tape snares.  Tripped snares provided hair samples of O. princeps (indicated by the arrow).    55    Figure 7. Hair obtained from O. princeps. An ideal sample composed of numerous guard hairs.   56             Figure 8. Representative gel image used to identify sex in O. princeps, including males (?), females (?), 100 bp DNA ladder (NEB), and a negative control. The lower molecular weight product is from the sex-determining region on the Y chromosome (SRY) and is found in males.  An autosomal microsatellite (Ocp10) was used as a PCR positive control, and should be present in both sexes.    57                     Figure 9. Isolation by distance relationship. Genetic distance between sites increases as the geographic distance increases. Genetic distance is measured in theta (?), and geographic distance is measured in kilometers.   58                   Figure 10. Genetic friction across sites of O. princeps occupancy in the Columbia River Gorge, Oregon, USA. As the color gradient moves from light yellow to dark red, the genetic differentiation per unit of geographic distance increases. Longitude and latitude are recorded in decimal degrees format. Genetic friction is a relative measure based on a modified version of theta (?). 59  Figure 11. Clustering of talus sites based on the program STRUCTURE. Each bar of the graph represents a unique individual, and the proportion of the bar that is blue or red indicates assignment to a particular genetic cluster.   60 2.6 Tables  Table 1. Site locations for O. princeps in the Columbia River Gorge. Site Latitude (o) Longitude (o) Elevation (m) B 45.637 -121.728 1247 D 45.636 -121.758 1254 H 45.672 -121.840 283 I 45.656 -121.778 1185 J 45.687 -121.788 194 M 45.690 -121.733 104 Q 45.633 -121.714 1174 R 45.631 -121.766 1292 T 45.673 -121.836 318 U 45.666 -121.787 1032 Y 45.688 -121.796 199  61 Table 2. Information on the microsatellite loci retained in the study of O. princeps samples from the Columbia River Gorge. The size range is measured in the number of base pairs (bp). Name Size range (bp) Dye Repeat motif Ocp 02 418-438 FAM (GATA)12 Ocp 06 346-366 FAM (TAGA)9 Ocp 07 300-314 NED (AC)15 Ocp 10 179-203 FAM (TAGA)8 Ocp 11 290-306 VIC (TAGA)9 Ocp 13 262-274 VIC (TAGA)10 Ocp 15 177-201 PET (TAGA)13 Ocp 19 214-229 NED (CAT)10 Ocp 25 230-244 FAM (CT)12    62 Table 3. Summary of within site measures of genetic variation for a sample of O. princeps. n is number of individuals, Rxy is relatedness according to the Queller and Goodnight method (Queller & Goodnight 1989), FIS is the inbreeding coefficient, Ho is observed heterozygosity, and He is expected heterozygosity.  Site n Rxy FIS Ho He Het excess Mode-shift Allelic diversity B 12 0.01 0.24 0.46 0.60 0.082 shifted* 3.56 D 17 0.20 0.26 0.38 0.51 0.248 shifted* 3.33 H 16 0.31 0.25 0.32 0.42 0.285 L-shaped 2.67 I 12 -0.08 0.40 0.40 0.62 0.014* shifted 3.44 J 10 0.18 0.35 0.38 0.53 0.125 shifted 2.63 M 6 0.23 0.24 0.42 0.53 n/a n/a 2.25 Q 11 0.00 0.21 0.47 0.59 0.010* shifted* 3.11 R 9 0.11 0.36 0.42 0.64 0.055* shifted* 3.14 T 21 0.23 0.23 0.38 0.49 0.674 L-shaped 3.44 U 13 0.14 0.38 0.36 0.56 0.367 L-shaped 3.56 Y 28 0.23 0.36 0.30 0.46 0.590 L-shaped 3.44 Average 14 0.14 0.30 0.39 0.54 0.245 n/a 3.14          *values in bold are statistically significant       63 Table 4. Relatedness and inbreeding of O. princeps male and female individuals within sites. Rxy is relatedness according to the Queller and Goodnight method (Queller & Goodnight 1989), and FIS is the inbreeding coefficient.  Site Males Females Rxy FIS       Males Females Males Females B 7 2 -0.04 -0.08 0.36 2.38 D 11 4 0.22 0.49 0.10 0.07 H 7 9 0.15 0.41 1.17 0.39 I 5 7 0.03 -0.02 1.17 0.35 J 6 4 0.12 0.60 0.19 0.10 M 4 2 0.24 0.57 0.26 2.38 Q 7 3 0.00 -0.03 0.32 0.11 R 5 4 0.03 0.07 0.31 0.13 T 14 7 0.23 0.27 0.31 n/a U 7 6 0.32 -0.07 0.28 0.19 Y 20 7 0.19 0.32 n/a 0.37 Average 9 5 0.15 0.26 0.45 0.65  64 Table 5. Pairwise site comparisons of genetic differentiation (?) for a sample of O. princeps.  Site D H I J M Q R T U Y B 0.097* 0.102* 0.064* 0.137* 0.048 0.052 0.020 0.118* 0.035 0.119* D  0.156* 0.132* 0.213* 0.098* 0.123* 0.028 0.152* 0.144* 0.181* H   0.115* 0.150* 0.143* 0.165* 0.165* 0.088* 0.089* 0.091* I    -0.006 0.091* 0.115* 0.081* 0.067* 0.068* 0.147* J     0.169* 0.219* 0.158* 0.115* 0.096* 0.129* M      0.072 0.008 0.152* 0.041 0.138* Q       0.106* 0.143* 0.138* 0.148* R        0.208* 0.006 0.158* T         0.174* 0.137* U                   0.130*            *values in bold are statistically significant (P<0.05)        65 Table 6. Bi-directional migration rates between sites for O. princeps. Values organized by column represent immigration, and values organized by row represent emigration. Migration rates represent the proportion of the population consisting of genetic migrant individuals per generation.  Site B D H I J M Q R T U Y B  0.033 0.026 0.015 0.031 0.018 0.019 0.015 0.017 0.016 0.015 D 0.014  0.049 0.012 0.015 0.020 0.016 0.012 0.017 0.012 0.016 H 0.013 0.013  0.013 0.014 0.013 0.032 0.012 0.017 0.013 0.017 I 0.033 0.030 0.016  0.083 0.018 0.056 0.015 0.020 0.016 0.018 J 0.018 0.018 0.022 0.017  0.017 0.018 0.016 0.019 0.016 0.040 M 0.058 0.047 0.023 0.020 0.025  0.040 0.019 0.030 0.021 0.023 Q 0.059 0.032 0.018 0.016 0.016 0.017  0.015 0.018 0.015 0.023 R 0.017 0.142* 0.021 0.017 0.020 0.018 0.025  0.019 0.019 0.019 T 0.012 0.012 0.025 0.011 0.015 0.011 0.032 0.010  0.010 0.048 U 0.081 0.058 0.036 0.017 0.022 0.017 0.020 0.014 0.017  0.020 Y 0.046 0.009 0.017 0.009 0.023 0.013 0.010 0.009 0.009 0.009               *the value in bold is statistically significant         66 Table 7. Pairwise site comparisons of genetic differentiation (?) for O. princeps males and females. Values above the diagonal represent females and values below the diagonal represent males.  Site B D H I J M Q R T U Y B  0.197* 0.203* 0.102 0.362* 0.185* 0.002 0.053 0.122* -0.092 0.140* D 0.051  0.355* 0.314* 0.363 0.298* 0.218* 0.011 0.355* 0.189* 0.273* H 0.057 0.096*  0.186* 0.263* 0.185* 0.202* 0.215* 0.096* 0.123* 0.130* I 0.123 0.111 0.158*  0.097 0.048 0.155* 0.147* 0.094 0.103 0.180* J 0.073 0.190 0.103 0.146*  0.228* 0.306* 0.260* 0.203* 0.187 0.297* M -0.047 -0.013 0.013 0.134 0.126  -0.011 -0.076 0.177 -0.042 0.216* Q 0.084 0.142* 0.157* 0.212* 0.254* 0.177  0.036 0.185* 0.078 0.166* R -0.002 0.085 0.117 0.050 0.111 0.045 0.168*  0.263* -0.091 0.092 T 0.114* 0.088* 0.068 0.160* 0.093 0.093 0.202* 0.214*  0.140* 0.157* U 0.072 0.200* 0.108 0.180* 0.037 0.037 0.239* 0.010 0.234*  0.133* Y 0.078 0.184* 0.060 0.273* 0.094 0.094 0.252* 0.179* 0.155* 0.170*               *values in bold are statistically significant (P<0.05)         67 Table 8. Bi-directional migration rates for O. princeps males between sites. Values organized by column represent immigration, and values organized by row represent emigration. Migration rates represent the proportion of the population consisting of genetic migrant individuals per generation.  Site B D H I J M Q R T U Y B  0.018 0.018 0.019 0.019 0.032 0.018 0.037 0.019 0.074 0.019 D 0.030  0.015 0.015 0.015 0.070 0.042 0.070 0.015 0.015 0.030 H 0.018 0.018  0.019 0.018 0.018 0.056 0.075 0.018 0.018 0.055 I 0.042 0.021 0.021  0.062 0.210 0.021 0.042 0.021 0.042 0.021 J 0.020 0.020 0.019 0.020  0.037 0.002 0.041 0.019 0.020 0.040 M 0.022 0.022 0.022 0.022 0.022  0.022 0.067 0.022 0.034 0.045 Q 0.018 0.022 0.018 0.019 0.019 0.018  0.034 0.018 0.056 0.019 R 0.021 0.021 0.021 0.021 0.021 0.200 0.021  0.021 0.021 0.021 T 0.027 0.027 0.013 0.013 0.019 0.015 0.099 0.027  0.040 0.040 U 0.018 0.019 0.019 0.018 0.019 0.018 0.019 0.930 0.055  0.018 Y 0.064 0.011 0.011 0.011 0.011 0.052 0.011 0.098 0.011 0.011    68 Table 9. Bi-directional migration rates for O. princeps females between sites. Values organized by column represent immigration, and values organized by row represent emigration. Migration rates represent the proportion of the population consisting of genetic migrant individuals per generation.  Site B D H I J M Q R T U Y B  0.025 0.030 0.026 0.026 0.026 0.026 0.026 0.003 0.026 0.072 D 0.023  0.022 0.022 0.220 0.022 0.022 0.044 0.022 0.022 0.022 H 0.017 0.017  0.017 0.017 0.017 0.017 0.016 0.018 0.017 0.115* I 0.018 0.018 0.074  0.037 0.018 0.019 0.019 0.022 0.018 0.018 J 0.022 0.022 0.045 0.022  0.022 0.022 0.022 0.026 0.220 0.022 M 0.026 0.026 0.038 0.025 0.025  0.026 0.026 0.021 0.026 0.065 Q 0.021 0.021 0.025 0.021 0.021 0.021  0.021 0.022 0.021 0.101* R 0.022 0.022 0.022 0.022 0.022 0.023 0.022  0.019 0.022 0.022 T 0.019 0.018 0.056 0.037 0.037 0.019 0.019 0.018  0.019 0.018 U 0.020 0.020 0.039 0.020 0.020 0.020 0.020 0.039 0.039  0.059 Y 0.018 0.018 0.019 0.019 0.190 0.018 0.018 0.019 0.019 0.019    *values in bold are statistically significant          69Chapter 3 Conclusion  Much of the current pika literature is biased towards populations at the core of the species? distribution, and populations inhabiting alpine talus habitat. This research fills that knowledge gap by assessing the population genetics and behavior of American pika at the tolerance limit of the species? bioclimatic envelope, in the atypical environment of the Columbia River Gorge, Oregon, USA. Patterns of connectivity emerged between sites, with increased differentiation in the presence of dispersal barriers. Some evidence for site isolation was found based on high levels of inbreeding and relatedness, as well as a low proportion of significant migration. The non-random mating behavior that produces some of these signatures is similar to that exhibited by populations at the core of the species? distribution, where inbreeding between individuals of intermediate relatedness was observed (Smith 1974a; Peacock & Smith 1997b). Preliminary evidence for male-biased dispersal was uncovered, but will require future investigation. Overall, levels of genetic variation within sites in the Columbia River Gorge were low in relation to what has been reported in core populations, and comparable to what has been found from locations at the northern range periphery (Henry et al. 2012b). A more thorough understanding of how pikas use different types of habitats will provide direction to conservation management. Under the philosophy of Gibson et al. (2009), pika habitat in the Columbia River Gorge should be considered a high priority area for conservation, as populations adapted to atypical environments may exhibit distinctive characteristics and genetic reserves of biodiversity. Conservation action in peripheral and atypical habitats will benefit from the study and identification of genes underlying local adaptation. To this end, a measure such as the population adaptive index (PAI) can be  70employed (Bonin et al. 2006). The population adaptive index is a parameter that describes the adaptive specificity of a population relative to all considered populations. If a consistent suite of variables is used to measure the PAI, it will be possible to compare populations across a species? distribution and identify high priority areas.  Still, there is an ongoing debate in the field of conservation genetics as to the relative importance of adaptive versus neutral variation for directing species management (Primmer 2009), and growing support for an integrative approach (Funk et al. 2012). For pika populations in the Columbia River Gorge it remains important to characterize patterns of neutral genetic variation, as my research has done here, in addition to investigating adaptive genetic variation.  3.1 Limitations The topography of the Columbia River Gorge prevented the investigation and sampling of all talus sites within the study area. Several locations were identified as potential pika habitat, but were left unexplored due to a lack of access roads or trails, steep and treacherous terrain, and large patches of poisonous plants (Toxicodendron diversilobum and Toxicodendron radicans). Consequently, the spatial extent of the collected data was inconsistent, and patterns of site connectivity were explained with less confidence than if every potential dispersal patch and inhabitable talus site had been sampled. Noninvasive genetic sampling has great potential as a research tool and for management applications in wildlife biology. However, the ability to obtain DNA from a variety of sources such as hair, feces, or saliva, without physically trapping or handling the study organism, often incurs the cost of poorer quality genetic material (Waits & Paetkau 2005). This problem was encountered when extracting and amplifying DNA from hair samples of American pika. Less than half of the samples collected in the field amplified at a sufficient number of microsatellite markers to be retained in the study, and the final dataset  71still had a high percentage of missing data. The low quality DNA also prevented the use of a larger suite of microsatellites, used successfully in previous studies of American pika (Peacock et al. 2002; Peacock & Kirchoff 2009; Henry et al. 2012b). These factors, combined with small sample sizes at some sites, affected the robustness of analyses and the interpretation of some results. In order to mitigate this problem it might be helpful to take extra measures to preserve DNA (e.g. more frequent sample collection to reduce exposure to the elements, immediate storage in liquid nitrogen, and/or expedited extraction perhaps even in a mobile field lab), or further optimize extraction techniques.  3.2 Future Work Neutral genetic markers, such as microsatellites, are widely employed to study contemporary processes (e.g. genetic drift, gene flow, and demographic effects) but can also be important tools for understanding evolutionary relationships and reconstructing phylogenetic trees (Angeloni et al. 2011). To this end it would be appropriate to broaden the scope of this research by including additional nuclear and mitochondrial DNA markers in the analysis. Specifically, mitochondrial DNA can provide insights on phylogeographic patterns that would facilitate range-wide comparisons of populations. For example, Lamb (2013) identified single nucleotide polymorphisms (SNPs) within the NADH dehydrogenase 5 (ND5) region of the mitochondrial genome of American pika across multiple transects in Bella Coola, BC. While this marker was not informative at the site level, it did reveal patterns of genetic variation among altitudinal transects (Lamb 2013). Future studies could explore patterns of ND5 variation in populations across the species? range, targeting sites exposed to different climatic conditions in order to better understand the historical forces and contemporary processes shaping patterns of variation in American pikas.  72Genomic approaches, as described above, have the potential to improve studies in conservation biology by elucidating the relationship between climate change and species? responses. Current research on threespine stickleback (Gasterosteus aculeatus) is exemplary of how genetic patterns can be linked to environmental selective pressures (Hohenlohe et al. 2010; Jones et al. 2012). Jones et al. (2012) sequenced the genomes of threespine stickleback from different aquatic environments, and identified a genome-wide set of loci consistently associated with marine-freshwater divergence. Similarly, in a study of several species of marine fish, Nielsen (2009) discovered that adaptive loci show genetic divergence between populations that, when analyzed with neutral loci, had almost identical allelic frequencies. Recently, I was involved in a broad-scale, transcriptome-level investigation within the American pika (Lemay et al. 2013). Next-generation sequencing technology was harnessed in order to identify a large suite of SNPs that will enable future studies of local and range-wide patterns of neutral and adaptive variation. Targeted gene approaches may also be useful within this system. For example, genetic and molecular research on populations of plateau pika (Ochotona curzoniae) has identified several gene regions that code for proteins involved in cellular processes such as respiration and thermoregulation (Kitao et al. 2007; Yingzhong et al. 2007). Based on comparisons between humans and other vertebrates, these pika populations appear adapted to the cold climate and hypoxic conditions characteristic of the Qinghai-Tibet Plateau (Yingzhong et al. 2007). Still, more work needs to be done to locate these potentially adaptive gene regions in sequences from American pika, and develop molecular markers. Both targeted gene and genomic scan-based frameworks may enable future investigations  73into patterns of genetic divergence along environmental, elevation, and geographic gradients for further investigation into how pikas respond to changing environments.  It will be important to expand future studies to represent different parts of the North American range of pikas, as the effects of climate change are more likely to be revealed by analyses that consider entire landscapes, so that non-climatic influences that dominate at the local, short-term scale can be disentangled (Parmesan 1996). These lines of inquiry could prove particularly important for the American pika, as the current body of literature is deficient in broad scale investigations; the majority of studies are local-scale and site-specific, with exceptions by Galbreath et al. (2009) and Galbreath and Hoberg (2012). While important information has been documented with local and site-specific approaches, there are limitations for their utility in management and conservation. For example, over the last ten years Beever et al. (2011) documented a nearly five-fold increase in the local extinction rate in the Great Basin. While alarming, these extinction events may be isolated to the unique ecoregion. A recent decision by the USFWS not to list the American pika under threatened or endangered species status was justified based on the lack of evidence for population decline and extirpation across the entire species? range (USFWS 2010).  The distribution of the American pika extends across a large and variable landscape. Consequently, populations at the core and periphery experience different climatic and geographic conditions; these selection pressures may be driving local adaptation. Considering the current threat from climate change and anticipated range shifts, it is particularly important to continue to expand study efforts to populations in peripheral locations and atypical environments in order to inform future status assessments and management decisions. To date, my research represents the most comprehensive population  74genetics study ever conducted for pikas in an atypical habitat. It highlights differences between populations inhabiting different regions, and provides an important point of comparison for measures of population genetic variation, demographic history, and behavior, relative to studies conducted in different parts of the species? range.                       75References  Akcakaya HR, Mills G, Doncaster CP (2007) The role of metapopulations in conservation. Blackwell Publishing, Oxford, 64-84.  Allendorf FW, Hohenlohe PA, Luikart G (2010) Genomics and the future of conservation genetics. Nature Reviews Genetics, 11, 697-709.  Anderson BJ, Akcakaya HR, Araujo MB, Fordham DA, Martinez-Meyer E, Thuiller W, Brook BW (2009) Dynamics of range margins for metapopulations under climate change. Proceedings of the Royal Society B: Biological Sciences, 276, 1415-1420.  Angeloni F, Wagemaker N, Vergeer P, Ouborg J (2011) Genomic toolboxes for conservation biologists. Evolutionary Applications, 5, 130-143.  Barrett RDH, Schluter D (2007) Adaptation from standing genetic variation. TRENDS in Ecology and Evolution, 23, 38-44.  Barrio IC, Hik DS, Peck K, Bueno CG (2013) After the frass: foraging pikas select patches previously grazed by caterpillars. Biology Letters, 9.  Beever EA, Brussard PF, Berger J (2003) Patterns of apparent extirpation among isolated populations of pikas (Ochotona princeps) in the Great Basin. Journal of Mammalogy, 84, 37-54.  Beever EA, Wilkening JL, McIvor DE, Weber SS, Brussard PF (2008) American pikas (Ochotona princeps) in northwestern Nevada: a newly discovered population at a low-elevation site. Western North American Naturalist, 68, 8-14.  Beever EA, Ray C, Mote PW, Wilkening JL (2010) Testing alternative models of climate-mediated extirpations. Ecological Applications, 20, 164-178.  Beever EA, Ray C, Wilkening JL, Brussard PF, Mote PW (2011) Contemporary climate change alters the pace and drivers of extinction. Global Change Biology, 17, 2054-2070.  Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F (2004) GENETIX 4.05, logiciel sous Windows TM pour la g?n?tique des populations. Laboratoire G?nome, Populations, Interactions, CNRS UMR 5000, Universit? de Montpellier II, Montpelier.   76Blouin MS, Parsons M, Lacaille V, Lotz S (1996) Use of microsatellite loci to classify individuals by relatedness. Molecular Ecology, 5, 393-401.  Bonin A, Nicole F, Pompanon F, Miaud C, Taberlet P (2006) Population adaptive index: a new method to help measure intraspecific genetic diveristy and prioritize populations for conservation. Conservation Biology, 21, 697-708.  Brandt CA (1983) Den location by and reproductive success of pikas: a resource defence mating system. American Zoologist, 23, 932-936.  Brandt CA (1985) The evolution of sexual differences in natal dispersal: a test of Greenwood's hypothesis, in Migration: mechanisms and adaptive significance. University of Texas, 386-396.  Brown RN, Southwick CH, Golian SC (1989) Male-female spacing, territorial replacement, and the mating system of pikas (Ochotona princeps). Journal of Mammalogy, 70, 622-627.  Brownstein MJ, Carpten JD, Smith JR (1996) Modulation of non-templated nucleotide addition by taq DNA polymerase: primer modifications that facilitate genotyping. Biotechniques, 20, 1004-1010.  Calkins MT, Beever EA, Boykin KG, Frey JK, Andersen MC (2012) Not-so-splendid isolation: modeling climate-mediated range collapse of a montane mammal Ochotona princeps across numerous ecoregions. Ecography, 35, 780-791.  Cardinale BJ, Duffy EJ, Gonzalez A, Hooper DU, Perrings C, Venail P, Narwani A, Mace GM, Tilman D, Wardle DA, Kinzig AP, Daily GC, et al. (2012) Biodiversity loss and its impact on humanity. Nature, 486, 59-67.  Ceballos G, Ehrlich PR (2002) Mammal population losses and the extinction crisis. Science, 296, 904-907.  Channell R, Lomolino MV (2000) Trajectories toward extinction: dynamics of geographic range collapse. Journal of Biogeography, 27, 169-179.  Colosimo PF, Hosemann KE, Balabhadra S, et al. (2005) Widespread parallel evolution in sticklebacks by repeated fixation of ectodysplasin alleles. Science, 307, 1928-1933.   77Colwell RK, Brehm G, Cardelus CL, Gilman AC, Longino JT (2008) Global warming, elevational range shifts, and lowland biotic attrition in the wet tropics. Science, 10, 258-261.  Cornuet JM, Luikart G (1997) Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics, 144, 2001-2014.  de Sassi C, Tylianakis JM (2012) Climate change disproportionately increases herbivore over plant or parasitoid biomass. PLoS ONE, 7.  Debinski DM, Holt RD (2000) A survey and overview of habitat fragmentation experiments. Conservation Biology, 14, 342-355.  Dobrowski SZ, Abatzoglou J, Swanson AK, Greenberg JA, Mynsberge AR, Holden ZA, Schwartz MK (2013) The climate change velocity of the contiguous United States during the 20th century. Global Change Biology, 19, 241-251.  Dobson FS (1982) Competition for mates and predominant juvenile male dispersal in mammals. Animal Behaviour, 30, 1183-1192.  Duforet-Frebourg N, Blum MGB (2012) Non-stationary patterns of isolation-by-distance: inferring measures of genetic friction. Cornell University.  Earl DA, von Holdt BM (2011) STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Conservation Genetics Resources, 4, 359-361.  Erb LP, Ray C, Guralnick R (2011) On the generality of a climate-mediated shift in the distribution of the American pika (Ochotona princeps). Ecology, 92, 1730-1735.  Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology, 14, 2611-2620.  Excoffier L, Lischer HEL (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10, 564-567.  Fahrig L (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, Evolution, and Systematics, 34, 487-515.   78Fisher RA (1930) The General Theory of Natural Selection. Clarendon Press, Oxford.  Frankham R (2010) Where are we in conservation genetics and where do we need to go? Conservation Genetics, 11, 661-663.  Frankham R, Ballou JD, Briscoe DA (2009) Introduction to conservation genetics, 2nd edition. Cambridge University Press, Cambridge, UK.  Funk CW, McKay JK, Hohenlohe PA, Allendorf FW (2012) Harnessing genomics for delineating conservation units. Trends in Ecology and Evolution, 27, 489-496.  Galbreath KE, Hoberg EP (2012) Return to Beringia: parasites reveal cryptic biogeographic history of North American pikas. Proceedings of the Royal Society B: Biological Sciences, 279, 371-378.  Galbreath KE, Hafner DJ, Zamudio KR (2009) When cold is better: climate-driven elevation shifts yield complex patterns of diversification and demography in an alpine specialist (American pika, Ochotona princeps). Evolution, 63, 2848-2863.  Galbreath KE, Hafner DJ, Zamudio KR, Agnew K (2010) Isolation and introgression in the Intermountain West: contrasting gene genealogies reveal the complex biogeographic history of the American pika (Ochotona princeps). Journal of Biogeography, 37, 344-362.  Gaston K (2003) The structure and dynamics of geographic ranges. Oxford University Press.  Ge R-L, Cai Q, Shen Y-Y, San A, Ma L, Zhang Y, Yi X, Chen Y, Yang L, Huang Y, He R, Hui Y, Hao M, Li Y, Wang B, Ou X, Xu J, Zhang Y, Wu K, Geng C, et al. (2013) Draft genome sequence of the Tibetan antelope. Nature Communications, 4.  Gibson SY, Van der Marel RC, Starzomski BM (2009) Climate change and conservation of leading-edge peripheral populations. Conservation Biology, 23, 1369-1373.  Goudet J (1995) FSTAT (Version 1.2): A computer program to calculate F-statistics. The Journal of Heredity, 86, 485-486.  Grabherr G, Gottfried M, Pauli H (1994) Climate effects on mountain plants. Nature, 369, 448.   79Greenwood PJ (1980) Mating systems, philopatry and dispersal in birds and mammals. Animal Behaviour, 28, 1140-1162.  Hafner DJ (1993) North American pika (Ochotona princeps) as a Late Quaternary biogeographic indicator species. Quaternary Research 39, 373-380.  Hafner DJ, Sullivan RM (1995) Historical and ecological biogeography of Nearctic pikas (Lagomorpha, Ochotonidae). Journal of Mammalogy, 76, 302-321.  Haila Y (2002) A conceptual genealogy of fragmentation research: from island biogeography to landscape ecology. Ecological Applications, 12, 321-334.  Handley LJL, Perrin N (2007) Advances in our understanding of mammalian sex-biased dispersal. Molecular Ecology, 16, 1559-1578.  Hardie DC, Hutchings JA (2010) Evolutionary ecology at the extremes of species' ranges. Environmental Reviews, 18, 1-20.  Henry P (2011) Investigating the genetic basis of adaptation in a climate change sensitive species: the American pika. Unpublished Doctorate's thesis, The Univeristy of British Columbia Okanagan.  Henry P, Russello MA (2012) Obtaining high-quality DNA from elusive small mammals using low-tech hair snares. European Journal of Wildlife Research, 57, 429-435.  Henry P, Miquelle D, Sugimoto T, McCullough DR, Caccone A, Russello MA (2009) In situ population structure and ex situ representation of the endangered Amur tiger. Molecular Ecology, 18, 3173-3184.  Henry P, Henry A, Russello MA (2012a) Variation in habitat characteristics of American pikas along an elevation gradient at their northern range margin. Northwest Science, 86, 346-350.  Henry P, Sim Z, Russello MA (2012b) Genetic evidence for restricted dispersal along continuous altitudinal gradients in a climate change-sensitive mammal: the American pika. PLoS ONE, 7.  Hewitt GM, Nichols RA (2005) Genetic and evolutionary impacts of climate change. Yale University Press, Conneticut.   80Hohenlohe PA, Bassham S, Etter PD, Stiffler N, Johnson EA, Cresko WA (2010) Population genomics of parallel adaptation in threespine stickleback using sequenced RAD tags. PLoS Genetics, 6.  Intergovernmental Panel on Cliamte Change (2007) IPCC Fourth Assessment Report: Climate Change 2007.  (ed. Press CU), Cambridge, UK.  Jensen JL, Bohonak AJ, Kelley ST (2005) Isolation by distance, web service. BMC Genetics, 6.  Jones FC, Grabherr MG, Chan YF, Russell P, Mauceli E, Johnson J, Swofford R, Pirun M, Zody MC, White S, Birney E, Searle S, Schmutz J, et al. (2012) The genomic basis of adaptive evolution in threespine sticklebacks. Nature, 484, 55-61.  Kerr JT, Currie DJ (1995) Effects of human activity on global extinction risk. Conservation Biology, 9, 1528-1538.  Kitao N, Yahata T, Matsumoto T, Okamatsu-Ogura Y, Omachi A, Kimura K, Saito M (2007) Molecular cloning and tissue distribution of uncoupling protein 1 (UCP1) in plateau pika (Ochotona dauurica). Journal of Veterinary Medical Science, 69, 1065-1068.  Lamb CT (2013) SNP discovery and validation in the American pika (Ochotona princeps) reveals population structure among elevational transects at the northern range periphery. Unpublished Honor's Thesis.  Lamb CT, Robson KM, Russello MA (2013) Development and application of a molecular sexing protocol in the climate change-sensitive American pika. Conservation Genetics Resources, Submitted.  le Roux PC, McGeoch MA (2008) Rapid range expansion and community reorganization in response to warming. Global Change Biology, 14, 2950-2962.  Lemay MA, Henry P, Lamb CT, Robson KM, Russello MA (2013) Novel genomic resources for a climate change sensitive mammal: characterization of the American pika transcriptome. BMC Genomics, 14, 311.  Lenoir J, Gegout JC, Marquet PA, de Ruffray P, Brisse H (2008) A significant upward shift in plant species optimum elevation during the 20th century. Science, 320, 1768-1771.   81Lutton LM (1975) Notes on territorial behavior and response to predators of the pika, Ochotona princeps. Journal of Mammalogy, 56.  Macarthur RA, Wang LCH (1974) Behavioral thermoregulation in pika Ochotona princeps- field study using radiotelemetry. Canadian Journal of Zoology, 52, 353-358.  Manning T, Hagar JC (2011) Use of nonalpine anthropogenic habitats by American pikas (Ochotona princeps) in Western Oregon. Western North American Naturalist, 71, 106-112.  McCarty JP (2002) Ecological consequences of recent climate change. Conservation Biology, 15, 320-331.  McKee JK, Sciulli PW, Fooce CD, Waite TA (2003) Forecasting global biodiversity threats associated with human population growth. Biological Conservation, 115, 161-164.  McKinney ML (2001) Role of human population size in raising bird and mammal threats among nations. Animal Conservation, 4, 45-57.  Merideth SJ (2002) The impact of habitat spatial structure on pika (Ochotona princeps) dispersal dynamics. Unpublished Master's thesis, The University of Nevada Reno.  Millar CI, Westfall RD (2010) Distribution and climatic relationships of the American pika (Ochotona princeps) in the Sierra Nevada and western Great Basin, U.S.A.; periglacial landforms as refugia in warming climates. Arctic, Antarctic, and Alpine Research, 42, 76-88.  Moilanen A, Smith AT (1998) Long-term dynamics in a metapopulation of the American pika. American Naturalist, 152, 530-542.  Nielsen EE, Hemmer-Hansen J, Larsen PF, Bekkevold D (2009) Population genomics of marine fishes: identifying adaptive variation in space and time. Molecular Ecology, 18, 3128-3150.  Niu YF, Wei M, Li X, Feng Z (2004) Phylogeny of pikas (Lagomorpha, Ochotona) inferred from mitochondrial cytochrome b sequences. Folia Zoologica, 53, 141-155.  Pamilo P (1985) Effect of inbreeding on genetic relatedness. Hereditas, 103, 195-200.  Parmesan C (1996) Climate and species range. Nature, 382, 765-766.  82 Parmesan C (2005) Biotic response: range and abundance changes. Yale University Press, New Haven.  Parmesan C, Yohe G (2003) A globally coherent fingerprint of climate chage impacts across natural systems. Nature, 421, 37-42.  Peacock MM, Kirchoff VS (2009) Identification and characterization of 19 polymorphic microsatellite loci in the American pika, Ochotona princeps. Unpublished report, 5.  Peacock MM, Smith AT (1997a) The effect of habitat fragmentation on dispersal patterns, mating behavior, and genetic variation in a pika (Ochotona princeps) metapopulation. Oecologia, 112, 524-533.  Peacock MM, Smith AT (1997b) Nonrandom mating in pikas Ochotona princeps: evidence for inbreeding between individuals of intermediate relatedness. Molecular Ecology, 6, 801-811.  Peacock MM, Kirchoff VS, Merideth SJ (2002) Identification and characterization of nine polymorphic microsatellite loci in the North American pika, Ochotona princeps. Molecular Ecology Notes, 2, 360-362.  Peakall R, Smouse PE (2006) GenAlEx 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes, 6, 288-295.  Peck L, S. (2011) Organisms and responses to environmental change. Marine Genomics, 4, 237-243.  Piggott MP, Bellemain E, Taberlet P, Taylor AC (2004) A multiplex pre-amplification method that significantly improves microsatellite amplification and error rates for faecal DNA in limiting conditions. Conservation Genetics, 5, 417-420.  Primmer CR (2009) From conservation genetics to conservation genomics. Annals of the New York Academy of Sciences, 1162, 357-368.  Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics, 155.  Queller DC, Goodnight KF (1989) Estimating relatedness using genetic markers. Evolution, 43, 258-275.  83 Rambaut A, Drummond AJ (2007) Tracer v1.4.  Raymond M, Rousset F (1995) GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity, 86, 248-249.  Rice WR (1989) Analyzing tables of statistical tests. Evolution, 43, 223-225.  Rodhouse TJ, Beever EA, Garrett LK, Irvine KM, Jeffress MR, Munts M, Ray C (2010) Distribution of American pikas in a low-elevation lava landscape: conservation implications from the range periphery. Journal of Mammalogy, 91, 1287-1299.  Roest AJ (1953) Notes on pikas from the Oregon Cascades. Journal of Mammalogy, 34, 132-133.  Root TL, Price JT, Hall KR, Schneider SH, Rosenzweig C, Pounds JA (2003) Fingerprints of glabal warming on wild animals and plants. Nature, 421, 57-60.  Rousset F (2008) Genepop '007: a complete reimplementation of the Genepop software for Windows and Linux. Molecular Ecology Resources, 8, 103-106.  Rozen S, Skaletsky HJ (2000) Primer3 on the WWW for general users and for biologist programmers, 365-386.  Russello MA, Amato G (2004) Ex situ population management in the absence of pedigree information. Molecular Ecology, 13, 2829-2840.  Safriel UN, Volis S, Kark S (1994) Core and peripheral populations and global climate change. Israel Journal of Plant Sciences, 42, 331-345.  Schipper J, Chanson JS, Chiozza F, Cox NA, Hoffmann M, Katariya V, Lamoreux J, Rodrigues AS, Stuart SN, Temple HJ, Baillie J, Boitani L, Lacher TEJ, Mittermeier RA, et al. (2008) The status of the world's land and marine mammals: diversity, threat, and knowledge. Science, 322, 225-230.  Schuelke M (2000) An economic method for the fluorescent labeling of PCR fragments. Nature Biotechnology, 18, 233-234.  Sgro CM, Lowe AJ, Hoffmann AA (2011) Building evolutionary resilience for conserving biodiversity under climate change. Evolutionary Applications, 4, 326-337.  84 Simpson WG (2009) American pikas inhabit low-elevation sites outside the species' previously described bioclimatic envelope. Western North American Naturalist, 69, 243-250.  Smith AT (1974a) The distribution and dispersal of pikas: consequences of insular population structure Ecology, 55, 1112-1119.  Smith AT (1974b) The distribution and dispersal of pikas: influences of behavior and climate. Ecology, 55, 1368-1376.  Smith AT (1978) Comparative demography of pikas (Ochotona): effect of spatial and temporal age-specific mortality. Ecology, 59, 133-139.  Smith AT, Ivins BL (1983) Reproductive tactics of pikas: why have two litters? Canadian Journal of Zoology, 61, 1551-1559.  Smith AT, Weston ML (1990) Ochotona princeps. Mammalian Species, 352, 1-8.  Soul? ME (1991) Conservation: tactics for a constant crisis. Science, 253, 744-750.  Steiner CC, Weber JN, Hoekstra HE (2007) Adaptive variation in beach mice produced by two interacting pigmentation genes. PLoS Biology, 5, 219.  Thomas CD, Lennon JJ (1999) Birds extend their ranges northwards. Nature, 399, 213.  Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Colingham YC, Erasmus BFN, de Siquiera MF, Grainger A, Hannah L, Hughes L, Huntley B, van Jaarsveld AS, Midgley GF, Miles L, Ortega-Huerta MA, Townsend Peterson A, Phillips OL, Williams SE (2004) Extinction risk from climate change. Nature, 427, 145-148.  United Nations Department of Economic and Social Affairs (2011) World populaiton prospects: the 2010 revision. United Nations Press Release, New York.  USFWS (2010) 12-month finding on a petition to list the American pika as threatened or endangered.  (ed. Federal register), pp. 6438-6471.  Van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P (2004) MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology, 4, 535-538.  85 Waits LP, Paetkau D (2005) Noninvasive genetic sampling tools for wildlife biologists: a review of applications and recommendations for accurate data collection. Journal of Wildlife Management, 69, 1419-1433.  Walther GR, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, Fromentin JM, Hoegh-Guldberg O, Bairlein F (2002) Ecological responses to recent climate change. Nature, 416, 389-395.  Weir BS, Cockerham CC (1984) Estimating f-statistics for the analysis of population-structure. Evolution, 38, 1358-1370.  Wilson GA, Rannala B (2003) Bayesian inference of recent migration rates using multilocus genotypes. Genetics, 163, 1177-1191.  Wuethrich B (2000) How climate change alters rhythms of the wild. Science, 287, 793-795.  Yingzhong Y, Yue C, Guoen J, Zhenzhong B, Lan M, Haixia Y, Rili G (2007) Molecular cloning and characterization of hemoglobin ? and ? chains from plateau pika (Ochotona curzoniae) living at high altitude. Gene, 403, 118-124.  Zgurski JM, Hik DS (2012) Polygynandry and even-sexed dispersal in a population of collared pikas, Ochotona collaris. Animal Behaviour, 83, 1075-1082.   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0074070/manifest

Comment

Related Items