The University of San FranciscoUSF Scholarship: a digital repository @ Gleeson Library |Geschke Center

Master's Projects and Capstones Theses, Dissertations, Capstones and Projects

Spring 5-19-2017

Understanding the complex relationships betweenclimate, vegetation, and foraging behavior of aclimate-sensitive alpine mammal in order to explainpatterns of persistenceEvan ColeUniversity of San Francisco, ejcole2@dons.usfca.edu

Follow this and additional works at: https://repository.usfca.edu/capstone

Part of the Other Ecology and Evolutionary Biology Commons

This Project/Capstone is brought to you for free and open access by the Theses, Dissertations, Capstones and Projects at USF Scholarship: a digitalrepository @ Gleeson Library | Geschke Center. It has been accepted for inclusion in Master's Projects and Capstones by an authorized administratorof USF Scholarship: a digital repository @ Gleeson Library | Geschke Center. For more information, please contact repository@usfca.edu.

Recommended CitationCole, Evan, "Understanding the complex relationships between climate, vegetation, and foraging behavior of a climate-sensitive alpinemammal in order to explain patterns of persistence" (2017). Master's Projects and Capstones. 566.https://repository.usfca.edu/capstone/566

This Master's Project

Understanding the complex relationships between climate, vegetation,

and foraging behavior of a climate-sensitive alpine mammal

in order to explain patterns of persistence

by

Evan Cole

is submitted in partial fulfillment of the requirements

for the degree of:

Master of Science

in

Environmental Management

at the

University of San Francisco

Submitted: Received:

5/17/17

...................................……….. ................................………….

Evan Cole Date John Callaway, Ph.D. Date

Table of Contents

Abstract …………………………………………………………………………………………1

Background …………………………………………………………………………………….2

- Mountain Ecosystems - Study Organism: The American Pika - Current Status and Distribution - A Model Organism - Vegetation as a predictive variable: A brief literature review

Research Methods …………………………………………………………………………..10

Results ………………………………………………………………………………………...11

- What plant characteristics are important to pikas, and how do they influence behavior and survival?

- How do pikas alter their local plant community? - How does climate (and climate change) influence pika persistence and alpine

plant communities?

Discussion ……………………………………………………………………………………36

- Climate and Forage Tightly Linked - Conditional Importance of Plant Secondary Chemistry - Interactive Effects of Climate on Forage Selectivity and Availability - Lessons from an Indicator Species

Management Recommendations …………………………………………………………40

- Habitat Restoration and Enhancement - Legal and Regulatory Actions - Adaptive Capacity as a Conservation Metric

Appendix ………………………………………………………………………………………43

Acknowledgements …………………………………………………………………………51

Literature Cited ………………………………………………………………………………51

Page 1 of 62

Abstract

Mountain ecosystems offer substantial ecosystem services but are highly sensitive to

climate change. The American pika (Ochotona princeps) serves as an indicator species

of climate change and a model organism for studying its impacts on mountain

mammals. Certain aspects of plant community composition and structure can function

as predictors of pika distribution, but understanding the links between climate, forage

quality, and foraging behavior is necessary to identify the mediating mechanism. Pika

foraging pressure help shape the local plant community, which can confound modeling

efforts and must be considered when evaluating the influence of vegetation on pika

persistence. Plant Secondary Metabolites (PSMs) appear to be the most important

indicators of forage quality, driving winter diet selection, especially in high elevation

areas with extreme seasonality. Nutritional components, such as protein, nitrogen, and

water content, are more important to summer diet. Foraging selectivity for nutritional

components changes in response to environmental conditions and forage availability,

suggesting increased importance as climate warms and dries. While metrics of heat

stress, cold stress, and water budget appear to be significant divers of pika distribution,

the heterogeneity with which climate impacts pikas across their range calls for a place-

based perspective. The transformation of alpine plant communities in response to

climate change may further stress pika populations as preferred forage items become

scarcer. Pikas highlight the complexity and idiosyncrasy of alpine ecosystems. In order

to address the challenges faced by pikas, managers should identify refugia and

vulnerable populations, install plants high in PSMs and nutritional components, work to

control invasive plants, consider legal and regulatory protections, and expand

monitoring efforts.

Page 2 of 62

Background

Mountain Ecosystems

Covering nearly a quarter of the Earth’s land surface, mountains are quite literally

springs of life and civilization. Functioning as natural water towers, mountains are the

ultimate source of drinking water, hydroelectricity, and irrigation for 40% of the world’s

population (UNESCO 2014). Steep topographic and vegetative gradients promote high

beta and gamma diversity, and most of the protected landscapes across the western

U.S. and the world are in mountains (Beever et al. 2013). Additionally, mountains are

key sources of timber, mineral resources, medicinal plants, and flood control, as well as

significant recreational opportunities and aesthetic value. These vital ecosystem

services, which reach well beyond the isolated geography of mountain ecosystems, are

becoming increasingly jeopardized by global change.

Mountain ecosystems serve as early warning indicators of climate change impacts, but

further attention and research are necessary to adequately interpret the signals. High-

elevation areas are warming at a significantly accelerated rate compared to low-

elevation areas (Figure 1) due to a variety of potential mechanisms and interacting

feedback loops, such as the loss of snow and ice and the associated reduction in

albedo, the increasing release of latent heat in the high atmosphere as water vapor

condenses, and the cooling effect of aerosols at lower elevations (Pepin et al. 2015).

Despite our reliance on mountain ecosystems for ecological, economic, and recreational

benefits, they are poorly monitored due to their isolated nature. Long-term climatic

trends are difficult to characterize for high-elevation areas, as meteorological stations

become increasingly sparse at higher elevations (Pepin 2015). Mountain ecosystems

offer a unique laboratory for studying the ecological implications of climate change, and

understanding how these sentinels of global change respond can inform conservation

strategies and secure essential natural resources for the future.

Page 3 of 62

Figure 1. Rate of warming is greater at high elevations. Note that average temperatures are lower at high

elevations. (Beever 2016)

Study Organism: The American Pika

The American pika (Ochotona princeps) is a small, egg-shaped mammal found in alpine

and subalpine ecosystems of western North America. Though commonly mistaken for a

rodent, the pika is actually in the order Lagomorpha and is related to rabbits and hares.

The pika is a generalist herbivore, feeding upon a variety of available vegetation, and a

habitat specialist, generally found only on rocky outcroppings and talus slopes (Dearing

1996). One defining life history trait of the pika is that it does not hibernate, instead

Page 4 of 62

caching plant material in haypiles, upon which it feeds throughout the winter while most

other vegetation has died back (Dearing 1997a). While the pika is typically considered

diurnal, it can shift its behavior to be crepuscular or even nocturnal based on stressful

environmental conditions, such as high daytime temperatures (Smith 1974). Pikas are

highly territorial, with a single individual occupying its own home range, and maximum

dispersal distances are likely around just 10-20 km (USFWS 2010). Despite its small

size and undeniably adorable disposition, the pika is actually quite hardy and highlights

the rugged adaptability of high elevation species.

Current Status and Distribution

Widely considered an indicator species, or ‘canary in the coal mine,’ foreshadowing

climate change impacts on montane mammals, the American pika has garnered

significant interest from scientists and the environmental community. Due to the isolated

and limited extent of its habitat (talus fields and rocky outcroppings), as well as its small

size, limited mobility, and poor thermoregulation, the pika is extremely sensitive to

climatic changes (Castillo et al. 2014). Recent studies have shown evidence of

widespread distributional loss, including upslope retractions and local extirpations, with

climate change implicated as a likely cause (Beever et al. 2016a). Population shifts

have varied by region, with steeper declines observed in the Great Basin compared to

the Sierra Nevada. In fact, a recent study involving resurveying nearly one thousand

talus patches in three geographic regions – the Sierra Nevada, Great Basin, and

northern Utah – found extirpations of the pika from five of nine sites in the Great Basin

(sites defined as all talus patches within a 3 km radius of a historical occupancy record)

and eleven of twenty-nine sites in northeastern California (sites defined as all talus

patches within a 1 km radius of a historical occupancy record, consistent with earlier

survey methods in that area). Strikingly, even though pikas had been observed in Zion

National Park and Cedar Breaks National Monument in northern Utah as recently as

2011 and 2012, respectively, the study concluded that they are now extirpated from

both locations. Despite growing concern, the American pika was denied listing by the

US Fish and Wildlife Service (USFWS) in 2010, and another petition was rejected in

September 2016; the reasoning in both cases was cited as lack of substantial evidence

Page 5 of 62

demonstrating a need for federal protections based on the range of identified threats.

Further study and monitoring of climate change impacts on pika range, distribution, and

population dynamics can inform management decisions for the benefit of both the

species and the entire ecosystem.

A Model Organism

Due to its distinctive life history, behavioral patterns, and sensitivity to climatic stressors,

the American pika is considered an excellent model organism for studying the impacts

of climate change on montane mammals. Firstly, the pika is uniquely sensitive to rising

temperatures due to its low physiological tolerance for heat – succumbing to

hyperthermia in as little as 6 hours of exposure to 25.5 °C (77.9 °F) (Smith 1974) and 2

hours of exposure to 28 °C (82.4 °F) (MacArthur and Wang 1973). The cage

experiments utilized by Smith (1974) and MacArthur and Wang (1973), though tragic for

the participating pikas, offered a valuable indication of the animal’s upper physiological

tolerance for heat – a metric that would be difficult to obtain by today’s ethics standards.

Second, pikas are locally abundant and easily detectable. While most mammals are

nocturnal (70%; Bennie et al. 2014) and are monitored primarily using camera traps and

tagging, pikas are diurnal or crepuscular, depending upon the locality and season, and

can be easily observed. Their distinct vocalizations, used for conspecific attraction,

announcing territory, and signaling alarm (Ray et al. 1991), as well as their large

haypiles, scat, and urine stains on the talus (personal communication), provide further

evidence of occupation. Additionally, their populations display relatively high interannual

stability, which decreases the likelihood of incorrectly identifying a patch as extirpated

due to a sudden and temporary drop in the local population (Beever et al. 2010). Finally,

the pika is a habitat specialist, relying upon talus patches and rocky outcroppings, which

makes it easier to locate and monitor pika metapopulations. Individual pikas do not

share territories (Peacock 1997), which are typically only about 20 meters in diameter

(Moyer-Horner et al. 2016). Maximum dispersal distances are generally considered to

be just a few hundred meters (Castillo et al. 2014). Due to its isolated nature, talus

habitat has largely escaped direct transformation or degradation by humans and is

relatively stable across biologically relevant timescales, which allows researchers to

Page 6 of 62

evaluate the effects of climate independent of habitat loss (Beever et al. 2010). As an

indicator species within a larger sentinel of global change – mountain ecosystems – the

pika offers a first look at the mechanisms of biological response to climate change.

Understanding the factors that mediate this biological response, either through direct

climatic stress or through indirect impacts like altered food or water availability, may

help managers to preserve the species and get ahead of further ecological

consequences.

Vegetation as a predictive variable: A brief literature review

While climate has been implicated as the primary driver of rapid range shifts and

extirpations of pika populations in the Great Basin (Beever et al. 2010, 2011, 2016) and

in other parts of its range (Erb et al. 2011, Calkins et al. 2012, Jeffress et al. 2013),

vegetation (i.e., plant community composition and structure) has been suggested as a

potential mechanism to either buffer or facilitate climate impacts. Recent attempts to

characterize this relationship between climate, vegetation, and pika persistence have

largely focused upon margins and leading-edges of the pika’s range (Rpdhouse et al.

2010, Varner and Dearing 2014, Varner et al. 2015, 2016). Range margins often feature

atypical habitat or microclimates that allow the species to persist where it otherwise

could not, making them ideal locations for research into the mechanisms and

determinants of range limitations, the effects of extreme environmental conditions and

stress on individuals and populations, and the adaptive capacity of the study organism

(Ray et al. 2016, Varner et al. 2016).

Metrics of forb cover seem to be the most prevalent and significant predictors of pika

occupancy or population density across most studies. Though relatively less abundant

among the talus – more commonly found in nearby alpine meadows or the talus margin

– forbs are selectively collected and cached by pikas and make up a relatively large

portion of the winter haypile (Huntly et al. 1986). The Plant Secondary Metabolites

(PSMs), particularly phenolics, found in many forbs act as preservatives and deter the

pikas from eating the plants until the winter and spring months when the compounds

have degraded (Dearing 1997). At Lava Beds National Monument, a range margin with

Page 7 of 62

atypically warm and arid conditions for the pika, the two best predictors of site-use

intensity have been metrics of graminoid:forb cover and total forb cover (Ray et al.

2016). The ratio of graminoid:forb cover serves as a metric for plant secondary

chemistry, as graminoids do not contain PSMs while forbs are generally high in them,

and indicates the importance of PSMs to pika occupancy dynamics. Forb cover was

also implicated as a key predictor of pika occupancy in the range periphery lava flow

habitat of Craters of the Moon National Monument and Preserve (Rodhouse et al.

2010). In Glacier National Park, a leading-edge of the pika’s range that could represent

an important refugium as the climate warms, average forb height and moss cover – an

evergreen, persistent food source – have been demonstrated as the two strongest

predictors of short-term pika occurrence (Moyer-Horner et al. 2016). Forb height

suggests the importance of plant morphology for facilitating the efficient gathering of

winter haypile components (Millar and Zwickel 1972). Interestingly, relative forb cover

outperformed absolute cover of forbs or graminoids in Glacier National Park, indicating

that there may be an interactive effect between forb and graminoid cover that is more

important than the total cover of either plant type. Studies in the Southern Rocky

Mountains have suggested similar measures of vegetation quality – the diversity and

relative cover of forbs – to be superior predictors of pika population density (Erb et al.

2014). This growing evidence of correlations between metrics of forb cover and pika

distribution, considering the established importance of forbs to pika behavior and diet,

indicates a potentially significant link between forage quality and pika persistence and

also raises the question of what other plant qualities and types may be influential.

A few studies in the Great Basin have incorporated broad vegetative components into

larger analyses of climate-mediated extirpations, but these have been limited in spatial

and temporal scope, as well as in the types of vegetative predictors considered. In

agreement with finding from other parts of the pika’s range, relative forb cover has been

demonstrated as a powerful predictor of pika persistence in the Great Basin (Wilkening

2007, Wilkening et al. 2011). Overall vegetation cover, foliar cover, and relative tree

cover negatively correlate with pika persistence, perhaps because a dense canopy

impedes a pika’s ability to surveil its territory and provides perches for predatory birds

Page 8 of 62

(Wilkening 2007). Sites with higher mean summer temperatures and lower relative

cover of forbs show greater rates of extirpation, potentially indicating that pikas in areas

with less forb cover must expose themselves to more heat stress as they spend more

time searching for suitable plant material for their winter haypile (Wilkening 2011).

However, this inverse relationship between mean summer temperature and relative forb

cover may simply reflect the long-term climate’s simultaneous influence on plant

community composition and pika population dynamics, as xeric conditions may limit

both forbs and pikas.

Pikas in atypical habitats spend more time foraging and less time haying, reflecting the

warmer winters and greater winter forage availability found at lower elevations.

Research at the Columbia River Gorge, a range margin characterized by higher

temperatures and lower elevations than those typically occupied by pikas, has

demonstrated a strong relationship between plant community composition and pika

persistence, as well as provided a unique example of climate adaptation through dietary

plasticity – the shifting of diet to accommodate what food items are available. Here,

vegetation species richness and percentage moss cover have been shown to be the

most significant predictors of pika density (Varner et al. 2015). Dietary plasticity through

utilization of moss is a key adaptation that pikas exhibit in order to exist in such atypical,

harsh environments (Varner and Dearing 2014). More than 60% of the pika’s diet

consists of moss in some areas of the Columbia River Gorge, more than has been

observed for any wild mammalian herbivore. In contrast to the seasonal growth and die-

back of grasses and forbs, moss is available year-round, releasing pikas from needing

to collect and store sizeable haypiles for winter during the summer and early fall

months, and thereby minimizing energy expenditure and exposure to extreme heat

(Varner et al. 2016). In fact, the Siberian northern pika (Ochotona hyperborea) actively

avoids evergreen forbs, cushion plants, mosses, and lichens when collecting for its

haypile, as those types of forage are available throughout winter and do not need to be

cached (Gliwicz et al. 2006), further suggesting that the relative value of grazed plants

compared to hayed plants may depend upon local climatic and environmental factors.

Page 9 of 62

The reliance on moss also reduces the need for pikas to venture away from the talus to

forage, mitigating the risk of predation.

As further evidence of adaptive utilization of vegetation, pikas existing in atypical

habitats are more likely to be found in forested habitat away from the talus during

midday compared to pikas existing in the more typical, high-elevation habitat of nearby

Mt. Hood, suggesting the potential for forested stands to act as thermal refuge off the

talus (Varner et al. 2016). Pikas in the atypical habitats of the Columbia River Gorge

exhibit higher population density and smaller home ranges than their high elevation

counterparts, but do not display increased signs of territoriality, suggesting an

abundance of local resources or a decrease in territoriality due to the lack of a need to

defend large winter hay piles. Although it is important to note the potential influence of

genetic differentiation between the two study groups, which would indicate genetic

adaptation rather than behavioral plasticity, evidence from range margins like the

Columbia River Gorge highlight the mediating power of vegetation in pika persistence in

the face of climatic stress.

Though limited, recent research has demonstrated a clear, if difficult to interpret,

relationship between pika occurrence and certain metrics of vegetation quality. In an

attempt to explain the apparent significance of vegetation’s role in determining pika

distribution, three general hypotheses have been proposed: 1) quality of available

forage is important for pika persistence, 2) pikas shape the composition and structure of

the local plant community and so the perceived relationship merely reflects correlation

rather than causation, and 3) microclimate is the true driver affecting pikas and plants

simultaneously (Wilkening et al. 2011, Ray et al. 2016). Though the interactions

between pikas, their respective plant communities, and microclimate are complex,

understanding the mechanisms of persistence and extirpation in the face of a changing

climate is essential for developing useful conservation strategies. Further research with

expanded spatial and temporal scopes, such as in this study, may help rank these three

hypotheses and elucidate the mechanisms of biological response to climate change. In

the following sections, I address these three explanations in order to provide context for

Page 10 of 62

my primary research and to clarify the mechanisms at work that potentially produce the

correlations I have identified.

For the main, secondary research portion of this paper, I seek to identify how climate,

vegetation, and pika foraging behavior interact to produce evident correlations between

pika persistence/extirpation and various plant community features. In order to

characterize this relationship, I pose three sub-questions:

1) What plant characteristics are important to pikas, and how do they influence

behavior and survival?

2) How do pikas alter their local plant community?

3) How does climate (and climate change) influence pika persistence and alpine

plant communities?

Building off these findings, the appendix includes a preliminary investigation into the

vegetative features that potentially predict pika occurrence in the Great Basin. This

primary research complements the main portion of the paper as an example of how

statistical analyses can be utilized to test hypotheses based on pika foraging dynamics.

Research Methods

In order to address my overall research question and subquestions, I employed

secondary research methods through the evaluation of the peer-reviewed literature and

personal communications with researchers in the field. The researchers I conferred with

were able to direct me to pertinent journal articles and studies, as well to advise my

statistical analyses featured in the appendix. I attended a professional conference

where I was able to view presentations on the latest research, participate in working

groups to discuss and address key issues related to pika conservation, and meet with

some of the researchers cited in this paper. In particular, I learned about the research

conducted by Wickhem (2016), which utilized cafeteria-style experiments to assess pika

foraging preferences. This study emphasized previous findings related to the

importance of Plant Secondary Metabolites to pika foraging behavior, and helped guide

Page 11 of 62

my further research. Personal communications with Erik Beever, Ph.D., cited

extensively in this paper, provided valuable perspective that allowed me to recognize

the often variable and sometimes contradictory findings of many studies related to

climatic impacts on the pika across its range. Through synthesis and integration of the

secondary literature, I identify common themes and provide evidence to characterize

the relationship between pika foraging behavior, climate, and plant community

composition and structure.

Results

What plant characteristics are important to pikas, and how do they influence

behavior and survival?

Foraging Behavior

Unlike many montane mammals, the pika does not hibernate, and instead stores plants

in haypiles, upon which it feeds throughout the course of the winter. The American pika

is a generalist herbivore, capable of foraging upon a broad range of local vegetation

depending on relative availability and accessibility, but its complex foraging behavior

and phenology complicates its nutritional requirements. Despite being a dietary

generalist, the pika actively selects plants based on quality rather than prevalence or

abundance (Smith and Erb 2013). The pika largely relies upon metabolic water from

plants for hydration and preferentially collects and stores plants for winter based on

specific nutritional, chemical, and structural characteristics (Huntly et al. 1986, Dearing

1997), prompting some researchers to ironically label it a “selective generalist” (Kristina

Ernest, Ph.D., personal communication, February 7, 2017). Pikas have exceptionally

high metabolic demands, especially at high altitudes, and the availability of adequate

nutrients and water could limit fitness and survival (Sheafor 2003, Erb et al. 2014).

Selectivity is expected to be enhanced in areas of extreme seasonality and relatively

high ambient temperatures (Moyer Horner et al. 2016), such as the isolated mountain

ranges of the Great Basin. The availability of high quality forage, dependent upon local

Page 12 of 62

plant community composition and structure, may very well mediate pika persistence in

these harsh environments.

Grazing and haying behavior serve separate purposes for the pika, and limitations to

either behavior, either via environmental constraints or plant availability, have distinct

consequences on fitness. Grazing sustains continuous metabolic needs and daily

activity, while the sole purpose of haying is for winter survival and to ensure adequate

fitness for spring mating and reproduction (Huntly et al. 1987, Dearing 1997a).

Importantly pikas graze throughout the year but only hay during the summer and early

fall months, coinciding with peak vegetative biomass (Huntly et al. 1987). Therefore,

haying, and thus winter survival and spring fitness, is constrained by the timing and

duration of the growing season, as well as by extreme summer heat, which limits the

amount of time spent foraging each day (Stafl and O’Connor 2015). In addition, pikas

use different parts of plants when grazing versus haying and prefer to cache plants high

in secondary chemical compounds that help preserve them through the winter months

(Dearing 1997), making plant chemical composition especially important to haypile

production.

In agreement with Central Place Foraging Theory, which attempts to explain how an

organism maximizes foraging efficiency by balancing risks and costs with energy won,

pika foraging behavior is heavily influenced by predation risk and environmental

conditions, just as it is also influenced by selectivity for specific plant characteristics

(Huntly et al. 1985, 1986). Pika foraging is concentrated around the talus and decreases

with distance, while forage selectivity becomes more focused further away from the

talus, suggesting the influence of predation risk and energy budgeting (Huntly et al.

1987, Holmes 1991). Females venture further from the talus when lactating compared

to when pregnant, as their priorities shift from predator avoidance when carrying young

to maximization of food when nursing (Holmes et al. 1991). The phenological timing of

plant growth and the availability of high quality forage may be most important during

pregnancy when mothers cannot risk exposing themselves for extended durations.

Pikas are positively correlated with sites of high visibility for predator surveillance (i.e.,

Page 13 of 62

negative correlation between pika persistence and overall vegetation cover, foliar cover,

and relative tree cover), as vegetation structure can inhibit their ability to safely forage

(Wilkening 2007, Moyer-Horner et al. 2016). Human disturbance can also disrupt

foraging behavior, as pikas respond to hikers with antipredator behaviors by reducing

foraging time (Stafl and O’Connor 2015). Trail traffic could be especially deleterious to

pika winter survival and spring fitness during summer and early fall, when pikas are

actively collecting for their haypiles. Similarly, pikas reduce their foraging time at higher

ambient temperatures, which can be particularly limiting to vital summer haying

activities, especially as the climate continues to warm (Stafl and O’Connor 2015). Pika

foraging selectivity can be highly site-specific and variable in response to environmental

heterogeneity (Smith and Erb 2013). Understanding the external factors that influence

foraging behavior is essential to realistically characterize the relationship between plant

community composition and structure and pika persistence.

Considering the complexity of the pika’s foraging behavior and phenological

requirements, combined with the extreme seasonality and unforgiving conditions of the

pika’s high-elevation habitat, metrics of forage quality may be especially important in

buffering or facilitating environmental stressors, and could improve efforts to model pika

distribution and to identify important refugia. As central place foragers, pikas do not

venture far beyond the talus in search of food and are thus limited by the abundance,

variety, and quality of available local forage plants (Huntly et al. 1985, 1986). Forage

quality can be best quantified by considering the relative importance of 1) plant

nutritional components, 2) Plant Secondary Metabolites (PSM), and 3) plant

morphology.

Nutritional Components

Metabolic water content

The metabolic water content of local vegetation may be most important to pika

persistence at sites where climate is drier and warmer, especially at lower elevations

and during periods of drought. Smith and Erb (2013) demonstrated that, while selection

for plants with high water content negatively correlates with elevation (i.e., selectivity is

Page 14 of 62

higher at lower elevations) in the southern Rocky Mountains, the water content of

commonly available vegetation does not correlate across the same elevation gradient,

indicating that pikas do not select for water content based upon differences in available

forage but rather are likely responding to differences in climate and free water

availability at different elevations. One metric of water balance, annual

evapotranspiration, has been shown to be a strong predictor of pika persistence in the

Great Basin; however, it can be mediated by regional and local environmental variables,

such as temperature, water inputs (i.e., precipitation), and soil characteristics, which

makes its influence on vegetation quality and pika persistence difficult to interpret

(Beever et al. 2016a).

Protein and Nitrogen Content

Another important nutritional component of forage plants consumed by pikas can be

defined in terms of digestible protein or digestible nitrogen (Wickhem 2016). High

protein diets can increase reproduction rates and improve overall fitness in mammals

(Smith 1988). Similarly, diets high in digestible nitrogen can improve growth rates in

offspring and increase reproduction rates (DeGabriel et al. 2009). In order to quantify

the protein content of a plant, scientists often measure concentrations of leaf nitrogen,

which makes up amino acids, the building blocks of protein, and thus can serve as a

proxy for protein content (Wickhem 2016). However, tannins in consumed forage can

actually bind to protein and reduce its availability for digestion (Robbins et al. 1987).

This defense mechanism against herbivory by which tannins reduce the digestible

protein content in plants confounds the assessment of what constitutes ‘high quality’

forage for a species like the pika that actively selects for plants containing tannins in

order to help preserve its winter haypile.

Pikas have been shown to preferentially collect and store plant species with higher

crude protein content (Millar and Zwickel 1972). Considering nitrogen a metric of protein

content, and so of forage quality, Smith and Erb (2013) found that selectivity for nitrogen

increases for pika populations with low-quality vegetation and decreases for pika

populations with high quality vegetation, suggesting pikas balance the energetic costs of

Page 15 of 62

haying by increasing selectivity for higher-energy food when it is less available.

Similarly, pika selectivity for nitrogen increases at sites with higher average summer

temperatures, allowing the pika to reduce its exposure to heat stress by spending less

time foraging – focusing on quality over quantity for cached items. However, high-

nitrogen foods may be more important for the pika’s summer diet than for the winter

haypile, as more focus is given to secondary chemistry when selecting for the winter

diet (Dearing 1996).

Plant Secondary Metabolites

Due to the unique foraging behavior and dietary requirements of a non-hibernating

alpine mammal, the driver of the pika’s haypile preferences and selective caching may

be the most critical indicator of forage quality. Plant Secondary Metabolites (PSMs), a

key adaptation of many plants in defense against herbivory, which can produce

negative effects ranging from reduced digestibility of vital nutrients to toxicity, appear to

be a dominant driver of haypile selectivity in both the American pika and the Siberian

pika (Dearing 1996, 1997a, 1997b, Gliwicz et al. 2006). Though classifications vary, the

three major types of PSMs include alkaloids, terpenes, and phenolics. Strikingly, the

concentration of phenolics, the most common class of PSMs, found in all angiosperms

(Wickhem 2016), in cached plants is three to six times higher than that of plants

consumed during summer (Dearing 1996, 1997). Somewhat counterintuitively, pikas

actually prefer plants with high concentrations of PSMs for their haypiles, as they can

provide the dual advantage of preserving plant material through the winter and delaying

consumption until the compounds have degraded sufficiently to make the plant

palatable (Dearing 1997b). Phenolics and terpenes possess antifungal and antimicrobial

properties that resist decomposition (Dearing 1996). The preservative qualities of PSMs

help to maximize the biomass and nutritional content of the haypile during winter and

early spring when forage availability is highly limited, allowing for approximately 70 more

days of food than if it were comprised of plants typical of the summer diet (Dearing

1997b). By early spring, the concentration of secondary compounds in the haypile

becomes as low as in summer forage, just in time for the crucial period of mating and

reproduction when proper diet and fitness are essential. The unpalatability of tannins (a

Page 16 of 62

common form of phenolic) and their associated decreased digestibility of proteins and

carbohydrates is reduced by the drying of the plant material, and terpenes lose their

harmful volatiles following harvest, allowing pikas to utilize plant resources that were

previously unavailable for consumption (Gliwicz et al. 2006, Wickhem 2016). In fact,

pikas are able to monitor the concentration of secondary compounds in the vegetation,

testing the palatability of the plants until concentrations are acceptable for consumption

(Dearing 1997b). Rather than developing chemical-specific, physiological adaptations,

the generalist pika is able to overcome a variety of plant chemical defenses through

simple behavioral modifications (Dearing 1996). This preference for PSMs allows pikas

to circumvent extreme seasonal changes in forage availability and climate in their harsh,

high-elevation habitat, potentially indicating a significant link between metrics of forage

quality and persistence.

In a recent study, in which ten plant species commonly present in pika habitats were

offered to pikas through a series of cafeteria-style experiments, Wickhem (2016) found

alkaloid and tannin content to be the top drivers of pika foraging preference. Of the

plants tested, the top species included Douglas fir (Pseudotsuga menziesii), Sitka alder

(Alnus viridis), willow (Salix spp.), and black cottonwood (Populus balsamifera

trichocarpa). Interestingly, pikas preferred plants with either high concentrations of

tannins or alkaloids, but not both, potentially in response to an interaction between the

two compounds (Figure 3). Protein and water content did not appear to be strong

drivers of pika foraging preference, further supporting the hypothesis that PSMs dictate

summer caching behavior and winter diet (Dearing 1997b). Incorporating a variety of

PSMs from different plant species may help prevent a single detoxification pathway

from becoming overloaded (Freeland and Janzen 1974). Understanding the relative

levels and combinations of PSMs present in a given plant community may help

characterize forage quality and inform restoration approaches.

Page 17 of 62

Figure 3. Pika foraging preference for tannins is negatively associated with alkaloid concentration,

suggesting pikas prefer one or the other secondary compound. (Wickhem 2016).

Plant Morphology and Structure

Plant morphology does correspond with pika foraging preferences, but morphological

bias is not consistent across the pika’s range and may not be as influential or useful of a

predictor as nutritional or secondary chemical content. According to Central Place

Foraging Theory, which highlights how organisms maximize foraging rates and

efficiency when venturing from a home base (e.g., talus patch), pikas should

preferentially collect larger plant species for their winter haypiles in order to reduce the

time and energy spent foraging. In Alberta, Millar and Zwickel (1972) concluded that

pikas preferentially harvest shrubs and clumped herbs in order to maximize the amount

of cached plant material for minimal effort. Similarly, Huntly et al. (1986) found that

pikas preferentially cache tall bunch grasses over shorter graminoids, forb flowering

stalks over forb leaves, and more woody plants compared to its summer grazing diet

(Huntly et al. 1986). However, both of these studies occurred before Dearing (1996) first

provided evidence of pika selectivity based on Plant Secondary Metabolites (PSMs),

and neither study considered the role of secondary chemistry in defining the observed

Page 18 of 62

foraging preferences. The generally higher concentrations of secondary compounds in

forbs and woody plants compared to graminoids could partly account for these

observed relationships. While mean forb height positively correlates with short-term pika

abundance (defined as the number of occupants per pika home range) in Glacier

National Park, this could just as well reflect local microclimate (i.e., mesic habitat) or

foraging patterns (i.e., late harvest of forbs in response to the extreme seasonality and

relatively high summer temperatures of the park) and may not represent convincing

evidence of forage quality (in this case, defined by plant morphology) driving pika

abundance (Moyer-Horner et al. 2016). Looking to other species of Ochotona fails to

offer any further clarity on the matter; collared pikas (Ochotona collaris) in the Yukon

Territory actively select plant species with larger leaf sizes for their haypiles (Hudson et

al. 2008), while the Siberian northern pika (Ochotona hyperborea) does not appear to

have a clear preference for certain plant morphologies (Gliwicz et al. 2006). Although

the size of plants and plant structures may contribute to pika forage selection, the

tendency of plant morphology to reflect other factors, from nutritional and chemical

components to environmental conditions and herbivory pressure, makes it unreliable at

best as a predictor of foraging preferences.

In addition to the forage implications of plant morphology, vegetation structure is an

important habitat characteristic that may influence foraging behavior and survival.

Dense canopy cover may obstruct visibility and impair the pika’s ability to surveil the

talus, a vital behavior by which pikas maintain their territories and avoid predators

(Tyser 1980, Wilkening 2007). Trees can serve as perches for avian predators, which

would enhance the threat of predation. The persistence of pygmy rabbit (Brachylagus

idahoensis) populations in the Great Basin has been negatively associated with pinyon

and juniper encroachment, as these trees gradually reduce the understory and provide

perches for predators (Larrucea 2007). The height of forbs, which has been positively

associated with pika abundance, can indicate shade for foraging pikas, reducing heat

stress and allowing for prolonged activity during the crucial haying period of summer

and early fall (Moyer-Horner et al. 2016). Further research on the influence of trees and

plant community structure on pika persistence via alteration of microclimate and

Page 19 of 62

predator dynamics is needed to clarify the importance of these predictors in the context

of other, potentially collinear, variables.

Place-based Foraging Behavior

Despite general patterns in dietary preferences, pika forage selection appears to be

highly site-dependent (Hudson et al. 2008). Microclimate and local plant phenology play

important roles in determining pika foraging behavior and preference. While selectivity

for high-nitrogen plants relates to both the availability of nitrogen within the local plant

community and abiotic environmental factors, the selectivity for water content appears

to be solely based upon abiotic environmental factors (Smith and Erb 2013). Pika

preference for specific plant nutritional qualities may be location-specific and

heterogeneous across its range, depending upon local plant community composition

and environmental conditions. Further, summer and winter dietary preferences can vary

between locations. While pikas avoid caching lichens and evergreen cushion plants,

such as mosses, they still frequently graze them during spring and presumably during

winter as a haypile-supplement (Millar and Zwickel 1972, Conner 1983, Gliwicz et al.

2006). In warmer, drier areas where haypiles are less advantageous, pikas take this a

step further and shift their diet to become much more reliant on cushion plants (Varner

and Dearing 2014). Even when experimentally presented with harvested cushion plants,

pikas still avoid incorporating them into haypiles, suggesting they do not avoid the

cushion plant because of the difficulty and high-energy cost of harvesting a matted plant

(Dearing 1996). Instead, pikas may instinctually avoid caching plants that are available

throughout winter and spring (Hudson et al. 2008). This careful budgeting of resources

based upon seasonal availability may be most beneficial in high-elevation, montane

habitats forage resources can be restricted by extreme seasonality (Gliwicz et al. 2006).

Local plant phenology in relation to seasonal microclimate may outweigh the importance

of plant secondary chemistry to pika persistence in some areas, and could be especially

important in mediating the effects of a changing climate.

Page 20 of 62

How do pikas alter their local plant community?

The interpretation of the seeming significance of plant community composition and

structure to pika persistence is complicated by the effects that the pika has on its local

plant community. In other words, does vegetation actually mediate pika persistence, or

does the pika mediate the plant community, giving rise to spurious correlations?

Pika foraging pressure can confound modeling efforts but does not contradict

importance of forage quality. Well understood mechanisms of foraging pressure,

including lessons offered by Central Place Foraging Theory, can be compared against

established vegetative predictors of pika occupancy and persistence in order to assess

the relative role that foraging behavior plays in creating these linkages. Central Place

Foraging Theory correctly predicts that an increasing gradient of plant abundance and

species richness follows a declining gradient of grazing pressure away from the home

base – in this case, the talus (Huntly et al. 1986) (Figure 4). Concentration of foraging

near the talus results in a local decline in biomass and plant species richness (Huntly

1985), which disagrees with the positive relationship between pika occupancy and

vegetation species richness (Varner et al. 2015), suggesting forage quality is a more

consequential predictor than foraging pressure. Similarly, while pika foraging selectivity

tends to cause an increase in plants of lower calorie and protein content (Huntly 1985),

pika occupation and persistence are positively associated with relative forb cover

compared to graminoids (Wilkening 2007, Erb et al. 2014, Ray et al. 2016), despite

forbs generally having higher energy and protein value than graminoids (Holechek

1984). Heightened selectivity in areas of extreme seasonality may cause pikas to graze

more graminoids and delay forb collection until late in the season, resulting in a greater

forb-to-graminoid ratio and taller forbs, both of which have been shown to be strong

predictors of pika occupancy and abundance (Moyer-Horner et al. 2016). A comparable

argument could be made regarding the negative relationship between pika persistence

and total vegetation cover, foliar cover, and relative tree cover (Wilkening 2007). Pika

foraging pressure can suppress vegetation and biomass (Huntly 1986), which could

explain the negative correlation. However, the pika’s tendency to remove, but not

Page 21 of 62

consume, obstructive vegetation in areas of dense cover (Markham and Whicker 1973)

– a behavior that has been shown to improve predator avoidance in other small

mammals (Hansen and Gold 1977) – suggests that the negative relationship between

canopy cover and pika presence may actually reflect a plant-to-pika effect rather than

the other way around. Gradated plant community transformations caused by intense

foraging pressure, as well as the pika’s ability to actively engineer its habitat to promote

its own survival, complicates efforts to model pika-vegetation relationships and must be

considered when developing hypotheses and evaluating results.

Figure 4. Change in vegetation cover and vegetation grazed by pikas in relation to distance from talus

patch. (Huntly et al. 1986)

How does climate (and climate change) influence pika persistence and alpine

plant communities?

What has (pre)history shown us?

The paleontological record indicates that the American pika has experienced significant

range expansions and contractions over previous millennia, largely in response to

Page 22 of 62

glacial and interglacial cycles. Near the end of the Pleistocene epoch (c. 12,000

radiocarbon years ago), pika populations existed at much lower elevations – about 1750

m on average – than at present and did not exclusively utilize talus habitat (Grayson

2006). The cooler conditions of the Pleistocene epoch facilitated gene flow between

different high elevation pika populations, promoting a regional genetic cohesion that still

characterizes pika lineages today (Galbreath et al. 2009). Since the last glacial

maximum of the late Wisconsinan (c. 40,000-10,000 radiocarbon years ago), the

average elevation of pika populations in the Great Basin has risen by 783 m (Figure 5),

and the distance between now-extirpated populations and currently extant populations

declined dramatically as the range contracted around the high elevation mountain

ranges that transect the xeric valleys (Grayson 2005) (Figure 6). Low-elevation

populations were extirpated from most of the Great Basin by or during the mid-

Holocene, a period marked a warm period known as the climatic optimum (Grayson

2000). Losses were most severe in the southern Great Basin and along the Utah-

Nevada border, where conditions became too hot and dry to support viable pika

populations (Grayson 2006). These rapid distributional shifts and extirpations in

response to Holocene warming have produced considerable demographic decline

across all five genetic lineages of pika (Figure 7) since the Last Glacial Maximum; the

potential for further demographic retraction, population fragmentation, and even lineage

extinction may threaten the pika’s ability to cope with continued warming (Galbreath et

al. 2009). Comparable demographic impacts caused by postglacial warming can

provide valuable insight into the trajectory of pika populations in the future, though the

unprecedented rate of contemporary climate change and associated range contraction

elicit considerable uncertainty and concern.

Page 23 of 62

Figure 5. Rise in elevation of pikas in the Great Basin since the late Wisconsinan. (Grayson 2005)

Figure 6. Extirpations and range retraction across the Great Basin since the late Winsconsinan. Closed

circles indicate prehistorically occupied (and now extirpated) sites. Open circles with crosses indicate

historically occupied (and now extirpated) sites. Open circles indicate currently extant sites. The gray

dotted line indicates the boundary of the hydrographic Great Basin. (Grayson 2005)

Page 24 of 62

Figure 7. Major mitochondrial DNA lineages of the American pika. The broken line represents the

approximate southern extent of the continental ice sheet during the LGM. The numbers indicate sampling

locations. (Galbreath et al. 2009)

Just as past climate change has shaped the pika’s current distribution, climate has also

shaped the evolution of plant communities and the pika’s dietary niche. The

evolutionary history of the pika family, Ochotonidae, and its current geographic range

has been closely tied to climate and associated shifts in dominant vegetation.

Ochotonids strongly prefer C3 plants, while many leporids (rabbits and hares) are able

to incorporate significant amounts of C4 plants into their diet (Ge et al. 2013). Plants

using the C4 metabolic pathway for carbon fixation attained their current dominant

status in the late Miocene to early Pliocene (4-8 million years ago), coinciding with a

decline in global ambient temperatures. C4 plants, most of which are monocots in the

Poaceae family, are largely fire adapted and characterized by a high carbon-to-nitrogen

ratio and low protein (Ge et al. 2012). This poorer nutritional quality, as well as the fact

that not all herbivores are equally capable to digest C4 plant material, resulted in large-

scale faunal shifts after the C4 expansion of the late Miocene (Osborne 2008). In

Page 25 of 62

parallel with the shift from C3 dominated plant communities to mixed C3 and C4, and

C4 dominated, drove large scale extinctions and range contractions of Ochotonids,

resulting in today’s single extant genus (Ochotona), which includes 28 species of pika

found in talus habitats of North America and Asia (Ge et al. 2013) (Figure 8). While the

current increase in atmospheric carbon dioxide is expected to promote primary

production and plant biomass, it may offer a competitive advantage to C3 plants over

C4 plants, with potentially positive implications for Ochotona spp.; however, the

relationship between carbon dioxide levels and C3 and C4 plant community dynamics is

exceptionally complex, and the associated consequences of future climate change are

still poorly understood (Wand et al. 1999). The pika’s current dietary niche does not

appear to have been altered by the last century of climate change, with the greatest

pattern of dietary variation occurring in individuals between seasons rather than in

populations across decades or latitudes (Westover 2017). The question that remains is

how plant communities will shift in response to contemporary climate change, and

whether pikas will have the capacity to adapt and evolve along with these vegetative

transformations.

Figure 8. Changes in climate and vegetation in relation to the number of ochotonid and leporid genera

from the Paleocene to the early Pleistocene. (Ge et al. 2013)

Impacts of Climate Change on Alpine Plant Communities

Alpine plant communities may be some of the most impacted by climate change, as

rising temperatures and increasing aridity across steep elevation gradients force

Page 26 of 62

species to either tolerate or avoid environmental stressors. The elevational distributions

of many plants species are rising dramatically in response to recent climate change

(Kelly and Goulden 2008). Alpine species face a greater risk of localized extinction, as

they have a narrower elevation tolerance than subalpine species and species with

ranges spanning lower elevations (Guisan and Theurillat 2000). Due to the conic shape

of mountains, habitat area becomes reduced as species migrate higher to cope with

shifting environmental conditions. Some high elevation plant communities in the western

U.S. have experienced significant compositional transformations, including more bare

ground and graminoids and shrubs, and fewer forbs (Zorio et al. 2016).

Forbs appear to be particularly vulnerable to climate change, and alpine plant

communities may see a trend towards more xeric shrubs, with unknown consequences

for the pika’s winter dietary requirements. Concerningly, climate change impacts on

water balance may mediate changes in vegetation more significantly than the direct

consequences of rising temperatures (Franklin et al. 2016). Based upon the severe,

long-term drought that occurred across the western United States in conjunction with

the Medieval Warm Period (900-1300 AD), the continued rise in temperatures can be

expected to trigger a shift towards extreme, long-term aridity in the region (Cook et al.

2004). Increased warming promotes aboveground biomass (AGB) production of shrubs

while reducing AGB of forbs in montane meadows, indicating a potential shift in

dominant vegetation of montane plant communities (Harte and Shaw 1995). Warming

may favor evergreen shrubs compared to deciduous forbs via changes in soil moisture

and inhibitions to the light-capturing protein complex, photosystem II (PSII) (Loik et al.

2000).

A reduction in forb biomass may negatively impact forage quality and availability for pika

haypiles. In addition, reduction in annual snowpack caused by global warming may

expose overwintering perennial forbs in montane meadows to more extreme cold,

potentially delaying bloom and decreasing seed production (Inouye and McGuire 1991).

Low population density and altered phenology of forbs might limit haypile production

and influence foraging behavior, forcing pikas to expose themselves to further heat

Page 27 of 62

stress and predation risk in order to cache sufficient plant material for winter. Though

forbs are frequently cited as significant predictors of pika persistence (Rodhouse et al.

2010, Moyer-Horner et al. 2016, Ray et al. 2016), some shrub species have also proven

beneficial (Ray et al. 2016). Improved understanding of the dietary preferences of pikas

at the plant species level, as well as how alpine plant communities transform in

response to climate change, will be necessary to inform long-term conservation

strategies.

Climatic Predictors of Extirpation Explored

The direct effects of specific climatic variables on pika distribution must be considered in

conjunction with the indirect effects mediated through vegetation so as to distinguish the

mechanisms of change and account for synergistic effects. In order to assess the full

impact of climate change on species distributions, scientists have largely focused on

investigating the realized niche of a species through the development of species

distribution models (SDMs) (Schwalm et al. 2016). SDMs seek to estimate potential

range shifts and the vulnerability of populations by assessing the relationship between

occupancy and a range of biotic and abiotic variables. When a model has a good fit

(i.e., the chosen predictor variables adequately explain the variation in the response

variable), it can then be extrapolated to evaluate future distributions and identify the key

limiting factors. The pika’s unique position as a model organism threatened by climate

has spurred research focusing on developing realistic SDMs based on climatic variables

in order to better understand how montane mammals react to environmental change.

Temperature

Given the pika’s physiological sensitivity to extreme heat, rising temperatures have

unsurprisingly been widely implicated in the observed collapse of pika populations

throughout much of the species’ range. Recent studies have focused on resurveying

historically occupied sites and comparing shifts in abundance, elevation, or evidence of

extirpation to climatic data, such as ambient vs. sub-talus temperatures (Beever et al.

2010, 2013, 2016a). Researchers can place thermal sensors below the talus and

compare temperature trends to ambient temperature derived from weather stations in

Page 28 of 62

order to hindcast temperatures and search for correlations between temperature and

patterns of persistence. Testing alternative models of climate-mediated extirpations in

the Great Basin, researchers found chronic heat stress and acute cold stress to be the

strongest predictors of pika persistence (Beever et al. 2010). Importantly, chronic heat

stress often appears to be more important than acute heat stress (Beever et al. 2013),

though this is not consistent across the pika’s range. While pikas may largely be able to

avoid acute heat stress through behavioral adaptations, such as seeking refuge under

the talus during the hottest part of the day, the sub-lethal impacts of chronic heat stress

on foraging behavior and fecundity may actually be the more important driver of

extirpation (Beever et al. 2010). Chronic heat stress may be more influenced by climate

change, given that minimum temperatures increased several times more than maximum

temperatures during the 20th century (Beever et al. 2011), a trend that may become

more critical as temperatures continue to rise.

Projections of further warming suggest a dramatic decline in habitat for the pika across

its range. While average global temperatures are predicted to increase by 2.4-6.4°C by

the year 2100, a 4°C increase is expected to reduce suitable habitat for the 31

traditional American pika subspecies (which, through genetic analyses, have now been

synonymized into five distinct lineages) by, on average, 73% across their range (Calkins

et al. 2012) (Figure 2). Further, a 7°C increase could cause 19 of the 31 pika

subspecies to lose more than 98% of their habitat. The consequences of high elevation

warming are not restricted to the pika, as a scenario of warming by 3°C could cause

more than 20% of high-elevation small mammals in the Great Basin to go extinct

(McDonald and Brown 1992).

Page 29 of 62

Figure 2. Reduction in pika habitat under different scenarios of warming. (Calkins et al. 2012)

Importantly, the impacts of temperature are not consistent across the pika’s range.

While average temperature has been strongly correlated with pika persistence in the

Great Basin and northern Utah, it was outperformed by habitat area as a predictor

variable in northeastern California, reflecting the heterogeneous and place-based

relative importance of different climatic impacts (Beever et al. 2016a). The presence of

pikas in atypically warm habitat, such as at Craters of the Moon National Monument

where maximum July temperatures regularly exceed the pika’s upper physiological

temperature threshold, further suggests that temperature is not consistently or definitely

the most important limiting factor at work across their range (Rodhouse et al. 2010). The

interactive effects of temperature with other environmental factors, as well as the pika’s

significant adaptive capacity, require a place-based perspective, despite the temptation

to charge temperature as the “smoking gun” driving extirpations.

Page 30 of 62

Snowpack

One of the most surprising relationships linking climate and pika persistence is that of

acute and chronic cold stress and snowpack. While most research has focused upon

the effects of rising temperatures in relation to the pika’s upper physiological tolerance

for heat, cold stress also appears to be a growing concern in relation to climate change.

Since pikas are so small (i.e., high surface-to-volume ratio leads to rapid loss of heat)

and do not hibernate (i.e., faster metabolism enhances heat loss), they rely on the

snowpack for insulation from the coldest winter temperatures (Beever et al. 2010). The

mountain snowpack has displayed a declining trend since the mid-20th century,

exposing pikas to increasingly extreme cold. Consistent exposure of pikas to lower

temperature thresholds (-5°C and -10°C), as well as to their upper physiological

thresholds, can strongly predict pika extirpation risk (Beever et al. 2011)

Possible ways in which snowpack, commonly measured as the snow-water equivalent

(SWE), or the theoretical depth of water if the snowpack was melted, may influence pika

persistence include: 1) mediating the number of days in which pikas are exposed to c

vhnb of insulation, and 2) controlling how far into the year that freshwater would be

available (Beever et al. 2013). Maximum SWE has been demonstrated as an excellent

indicator of pika density in contemporary surveys, with a strong link between number of

acutely cold days and pika extirpations. The synergistic relationship between SWE,

growing-season precipitation, and average temperature suggests potential indirect

mechanisms of climatic influence, reinforced by the strong correlation between relative

pika density and total aboveground gross annual productivity (Beever et al. 2013).

Hydrology

In combination with temperature, hydrologic variables – including metrics of

precipitation, evapotranspiration, humidity, and water balance – have been implicated

as determinants of pika persistence and extirpation across much of its range. In the

Southern Rocky Mountains, annual precipitation is an important limiting factor of pika

persistence, which contributed to the extirpation of several populations within the region

during the last few decades (Erb et al. 2011). Extirpations in this region, though

Page 31 of 62

relatively few compared to other parts of the pika’s range (e.g., the Great Basin), have

only been documented in areas that were consistently dry throughout the 20th century.

It can be difficult to confidently identify evidence of climate change impacts, particularly

for precipitation, which is far more variable and difficult to predict than temperature, but

the relationship between low precipitation and pika extirpations in this region helps to

inform future management through prioritization of conservation efforts in areas of

extended drought.

To evaluate the impacts of hydrology on pika persistence, it is important to consider

biologically relevant ‘ecohydrological variables,’ such as actual evapotranspiration,

SWE, growing-season precipitation, and climatic water-deficit. Together, these variables

reflect the climatic water balance in an ecologically relevant context. Drought results

from an imbalance between water inputs (measured as precipitation) and water demand

(measured as evapotranspiration, or the amount of water transpired through vegetation

and evaporated from the soil), and can stress pikas directly (i.e., via the lack of available

drinking water) or indirectly (i.e., via reduced primary productivity and available forage).

In fact, ecosystem productivity and plant population dynamics have been shown to be

highly sensitive to changes in water balance, and the impact of climate change on water

balance could be even more consequential for vegetation than the direct effects of rising

temperatures (Franklin et al. 2016). When assessing the impacts of water imbalance on

a system, it is important to distinguish between potential evavotranspiration (PE) and

actual evapotranspiration (AE); whereas PE represents the maximum amount of water

that could be evapotranspired given adequate water inputs and is used to represent the

water demand of a system, AE represents the actual amount of water transpired given

the limited inputs (NOAA National Centers for Environmental Information). Therefore,

AE is always equal to or less than PE, and can reflect the degree of water stress in a

system. In the Great Basin, annual evapotranspiration, calculated as the sum of the

monthly actual evapotranspiration, was found to be the strongest predictor of pika

occupancy in an assessment of multiple competing models based on a range of climatic

variables (Beever et al. 2016a). Annual actual evapotranspiration is regularly mediated

by precipitation, temperature, and soil moisture storage, and can offer a more realistic,

Page 32 of 62

ecologically meaningful measure of water balance than annual precipitation. Similarly,

maximum snow-water equivalent and average growing-season precipitation more

strongly predict pika population size than average annual precipitation, as they more

precisely influence water and forage availability in relation to the phenological patterns

of the pika (Beever et al. 2013). The relationship between water balance and pika

persistence and extirpation may be especially pronounced in the driest, hottest reaches

of the pika’s range and where severe droughts may occur.

Fire Regime

Given the evident relationship between pika occurrence and environmental variables

like heat, dryness, and plant community composition and structure, one might expect

fire disturbance to have huge consequences for pika survival and fecundity. While

limited research has been conducted to evaluate the potential impacts of increased fire

severity and frequency on montane mammals, Varner et al. (2015) had the unique

opportunity to study these impacts after their pika monitoring project at Mt. Hood was

interrupted by wildfire at the study location. Thermal sensors had been placed at the

talus surface and interstices in order to compare microclimate and thermoregulatory

behavior with other, already surveyed sites at nearby low elevations. One week after

placing the sensors, a nearby lightning strike ignited a wildfire that would burn 2430

hectares over the course of twenty-two days, impacting four of the nine previously

surveyed sites and two of the three new sites with thermal sensors. The researchers

seized this opportunity to analyze the data from the thermal sensors and conduct

subsequent survey monitoring of the locations in order to look for potential relationships

between pika occupancy and abundance and different metrics of fire severity. Resulting

statistical analyses demonstrated a surprisingly muted importance of the burn severity

indices in predicting pika density in the burned areas (Varner et al. 2015). On the

contrary, indices of vegetation and habitat structure strongly explained patterns of

density. The apparent quick recovery of vegetation at burned sites facilitated the

persistence and widespread distribution of pikas in the area. Moss, a component of the

highest-weighted model, can serve as a persistent and abundant food source for the

pika when other plants are limited, and can also allow pikas to avoid venturing out from

Page 33 of 62

the stable microclimate of the talus to forage. Thermal sensors placed within the talus

showed that the temperature in talus interstices never exceeded 19°C, demonstrating

the importance of talus as a thermal refuge for small animals during and after fire

disturbance (Varner et al. 2015). The resilience and behavioral adaptability displayed by

the pika, as well as the buffering capacity of an atypical forage source, in the face of fire

disturbance emphasizes the need for further, species-specific study of biotic response

to altered fire regimes.

Heterogeneity and Nuance of Climate Impacts

As evidence continues to accumulate associating rapid distributional losses of the pika,

including upslope retractions and local extirpations, with climate change, a unifying

theme has emerged in the scientific literature – that of the heterogeneity and nuance

with which climate acts upon a species’ range (Jeffress et al. 2013, Beever et al. 2016a,

Schwalm et al. 2016). The importance of different climatic variables in determining pika

persistence and extirpation is not consistent across the pika’s range. In some areas,

pikas appear to have a wider tolerance for climatic stress, and factors of habitat quality

seem to be the most important determinants of pika distribution. Rock-ice feature till

functions as an important landform refuge in warmer conditions of the southwestern

U.S., allowing pikas to tolerate a wider range of elevation, temperature, and average

precipitation conditions than expected (Millar and Westfall 2010). In fact, habitat

composition and connectivity can be even more important than climatic variables in

modeling pika occurrence (Schwalm et al. 2016). However, sites where pikas have

been extirpated do display significantly higher average maximum temperatures,

suggesting a potential limit to either the adaptive capacity of the species or to the

buffering capacity of certain habitat features.

The variability in importance of climate and other predictors of pika distribution can be

readily distinguished when comparing models across different regions of the pika’s

range. Under Island Biogeography Theory, the Sierra-Cascade extent of the pika’s

range in northeastern California could be considered a ‘mainland’ habitat, while the

isolated mountains of the Great Basin and nearby Zion National Park and Cedar Breaks

Page 34 of 62

National Monument in northern Utah represent ‘island’ habitat. Species richness in

‘island’ habitats, a subset of the species richness found in ‘mainland’ habitats, reflects

the balance between extinction and colonization rates, which in turn are mediated by

habitat area and magnitude of isolation (Beever et al. 2016a). Localized extinctions

should theoretically be more extensive in montane ‘island’ habitats due to the smaller

habitat area and population size (i.e., genetic pool), as well as the narrower elevation

gradients compared to the ‘mainland.’ As predicted by Island Biogeography Theory,

localized pika extirpations are far more severe in the Great Basin and northern Utah

compared to northeastern California (Beever et al. 2016a). Interestingly, metrics of

water balance and mean temperature strongly predict pika occupancy in the two ‘island’

regions, but habitat area outperforms these variables in the ‘mainland’ region, further

suggesting the reduced limitations imposed by climatic stressors in areas of substantial,

ecologically-connected habitat.

The heterogeneity and nuance of climate impacts can be distinguished by comparing

models at different scales, as well. At macroecological scales, impacts of climate on

pika distribution appear unequivocal (Beever et al. 2010, 2013, 2016a). At the local

level, however, pika occurrence varies widely across bioclimatic gradients (Jeffress et

al. 2013). Local-scale differences in the realized niche of the pika makes the relative

importance of various predictor variables highly heterogeneous across the species'

range; therefore, a place-based approach is important in modeling pika distribution and

density. In addition, the interactions between climatic variables can often mediate their

relative influence on pika populations. In a study of the impacts of a suite of climatic

factors on local patterns of pika distribution across bioclimatic gradients in eight U.S.

National Parks, precipitation alone was not found to be a strong predictor, but heat

stress performed better in the driest parks, suggesting an interaction between the

magnitude of temperature effects and position along precipitation gradients (Jeffress et

al 2013).

In addition to the spatial heterogeneity of climate’s importance, the primary drivers of

extirpations appear to be changing over time, with elevation best explaining historical

Page 35 of 62

extirpations and metrics of temperature best explaining recent extirpations in the Great

Basin (Beever et al. 2011). However, the magnitude and rate of contemporary climatic

changes do not appear to be significant drivers of extirpation compared to the long-

term, microclimatic conditions at extirpated sites (Erb et al. 2011). Though a clear signal

implicating rapid change as the cause of extirpation has not yet been identified, the

clear importance of microclimate (particularly temperature and water balance) suggests

limiting effects posed by future climate change.

Mechanisms of Climate-driven Extirpation

Despite mounting support for climate as a primary driver of recent distributional shifts of

the American pika, the precise mechanisms of fatality and extirpations are still poorly

understood, as well as how these mechanisms will respond to a changing climate.

Climate may impact pika persistence in a number of ways, including through

transformations of available forage, shifting timing of forage and water availability,

reduced ability to forage and reproduce due to chronic stress, and acute stress like

hyperthermia. However, food resource decline has been cited as the most common

mechanism through which contemporary climate change has caused extirpations (Cahill

et al. 2013). Adult survival rate of the collared pika (Ochotona collaris) in the Ruby

Range of western Canada has been positively linked to effects of the Pacific Decadal

Oscillation (PDO), which tends to result in below-average snowpack and earlier timing

of spring snowmelt in that region, suggesting the importance of plant phenology and

forage availability (Morrison and Hik 2007). An earlier and longer growing season would

allow more time to collect vegetation for haypiles and could lead to improved fitness

during the spring breeding season and the climatically stressful summer. Conversely, a

declining snowpack could result in advanced exposure to acute and chronic cold stress

during the winter, reducing survival rate (Beever et al. 2010, 2011). Climatic conditions

have also been linked to reproduction rates of O. princeps, as reproduction may be

significantly reduced during periods of extreme weather (Millar 1974). Pika foraging

behavior is closely mediated by temperature, as time spent foraging by O. princeps

decreases by 3% for ever 1°C increase in average daily temperature (Stafl and

O’Connor 2015). Potentially, rising temperatures could produce extirpations by limiting

Page 36 of 62

foraging and haying time, causing a reduction in fitness that might be exacerbated by

alternative mechanisms like cold stress. One might expect the benefits of an extended

growing season to be most offset by chronic heat and cold stress in trailing edges of the

pika’s range where conditions are most hot and arid, such as in the Great Basin.

Discussion

The relationship between climate, local plant community composition and structure, and

pika distribution is exceptionally complex and place-based, highlighting the

idiosyncrasies of alpine ecosystems. Climatic influence on pika persistence is

heterogenous across its range, though metrics of heat stress, cold stress, and water

budget appear to explain recent shifts in distribution (Jeffress et al. 2013, Beever et al.

2016a). Pika foraging behavior, as well as forage availability, are, in turn, influenced by

microclimate and cyclic weather patterns (Smith and Erb 2013, Gliwicz et al. 2006). The

accelerated warming and drying of the climate in western North America that is

expected to occur as a result of climate change may further stress pikas and limit the

quality of their diet (Beever et al. 2013, Stafl and O’Connor 2015), putting their

demonstrated adaptive capacity to the test. Further research into the direct and indirect

mechanisms by which climate induces fatality and extirpations is necessary to elucidate

the relative role of vegetation.

Climate and Forage Tightly Linked

In order to characterize the relationship between climate, vegetation, and pika

occurence, and to evaluate the significance of forage quality in predicting pika

persistence, I considered the three general explanations for apparent correlations

between various vegetative metrics and pika occurrence: 1) quality of available forage is

important for pika persistence, 2) pikas shape the composition and structure of the local

plant community and so the perceived relationship merely reflects correlation rather

than causation, and 3) microclimate is the true driver affecting pikas and plants

simultaneously (Wilkening et al. 2011, Ray et al. 2016). In regards to the second

Page 37 of 62

explanation, although pika foraging pressure can confound modeling efforts and should

be considered when interpreting results, it is likely not as significant as the effects of

climate or forage quality in determining persistence patterns. Positive associations with

forb cover and plant species richness (Wilkening 2007, Erb et al. 2014, Varner et al.

2015) contradict expectations that pika selective foraging would promote a less diverse

local plant community dominated by graminoids and plants of poor nutritional quality

(Huntly 1985). However, the pika’s dynamic dietary niche across seasons (Westover

2017) suggest that the complex relationship between microclimate, foraging behavior,

and forage availability – a combination of the first and third explanation – may, in fact,

have significant implications for pika persistence in the face of environmental change.

Conditional Importance of Plant Secondary Chemistry

Despite being generalist herbivores, pikas commonly exhibit selective foraging behavior

in order to capitalize on available plant resources and ensure fitness and survival. Pikas

select for Plant Secondary Metabolites (PSMs), nutritional composition (e.g., nitrogen

and digestible protein), water content, and structural characteristics that facilitate

efficient collection (Dearing 1997a, 1997b, Gliwicz et al. 2006). PSMs appear to be the

most important driver of winter diet composition, while nutritional composition is more

predictive of summer diet (Dearing o91996). Winter diet quality is highly important to

winter survival and to fitness for the spring breeding season, especially in high elevation

areas where seasonality is extreme (i.e., hot, dry summers and freezing winters), as

summer heat stress can limit vital caching time and harsh winters can limit forage

availability and cause cold stress (Wilkening 2011, Beever et al. 2016a). PSMs ensure

adequate food storage and allow pikas to utilize plant resources that would be otherwise

unavailable (Gliwicz et al. 2006, Wickhem 2016). However, this selectivity is largely

influenced by other environmental factors, including microclimate and local seasonality,

and can vary dramatically between regions and populations. Pikas display impressive

adaptive capacity, able to shift their diets to evergreen, persistent resources like moss in

order to tolerate atypical, low elevation habitats where haypiles are less essential and

summer heat is more limiting (Varner et al. 2016). Understanding the limits of pika

adaptive capacity may inform efforts to gauge pika response to future climate change.

Page 38 of 62

Interactive Effects of Climate on Forage Selectivity and Availability

Pikas become more selective and reliant on metrics of high quality forage (i.e., PSMs,

metabolic water content, and digestible protein) as an adaptive response to dynamic

climatic stressors; however, these same stressors threaten to transform alpine plant

communities, making them less appealing and beneficial to pikas. The importance of

forage quality is spatially variable relative to local forage availability and environmental

constraints on foraging behavior and fitness, adding a layer of complexity to the

heterogeneity of climatic limitations on pika distribution (Smith and Erb 2013). Pikas

prefer plants with high water content at low elevations where heat and aridity are more

limiting, just as temperature is a stronger predictor of pika occurrence when coupled

with aridity (Smith and Erb 2013, Jeffress et al. 2013). Concerningly, climate change is

expected to transform montane plant communities, resulting in drier, warmer, habitats

favoring xeric shrubs and grasses and discouraging PSM-rich forbs (Inouye and

McGuire 1991, Franklin et al. 2016). The interactive effects of climatic variables with

foraging preferences and dietary requirements constitute indirect mechanisms by which

climate may mediate pika persistence through a simultaneous increased demand and

reduced availability of high quality forage. Understanding the specific plant community

composition and structure that benefit pikas at the local level can help prioritize areas

for conservation and inform restoration efforts.

Lessons from an Indicator Species

As an indicator species, the pika offers valuable insight into the mechanisms of

biological response to climate change and reflects pervasive challenges faced by a

range of montane mammals. Marmots, large squirrels common at high elevations, face

similar threats from climate change, as water availability impacts survival and fecundity

and reduced snowpack induces further cold stress in the winter months (Armitage

2013). Just as tree cover negatively correlates with pika persistence, as it may reflect a

reduction in available understory forage and can limit their ability to surveil their territory

and avoid avian predators (Wilkening 2007), tree encroachment into subalpine meadow

habitat has been linked to population fragmentation in the Olympic marmot (Marmota

Page 39 of 62

Olympus) in Washington (Griffin et al. 2009). The upslope migration of trees and shrubs

associated with climate change is also reducing habitat for the Alaska marmot (Marmota

broweri) in Alaska’s Brooks Range (Gunderson et al. 2012). Similarly, juniper-pinyon

encroachment has contributed to extirpations of pygmy rabbits (Brachylagus

idahoensis) in Nevada and California (Larrucea 2007), just as juniper-pinyon percent

cover appears to be a significant predictor of pika absence in the Great Basin (see

Appendix). Mirroring the hardiness and complexity of pika response to atypical climatic

conditions, some populations of marmots appear to be better adapted to warmer

climates than others, though the rate of warming will likely outpace that adaptive

capacity; widespread extinction seems unlikely, but localized extirpations focused at low

elevation, marginal habitat are expected (Armitage 2013). Pikas exemplify the fragility

and limitations of high elevation ecosystems.

Despite the similar challenges, researchers must use caution when comparing the

pika’s response to climate with other montane species, as its unique life history and

behavioral plasticity may limit its applicability. Montane mammals employ different

strategies to respond to the challenges posed by the dynamic seasonality and harsh

winters of mountain ecosystems, each with their own unique obstacles posed by climate

change. While pikas cache haypiles for the winter, a behavior that is limited by high

summer temperatures and shifting plant communities (Smith and Erb 2013, Stafl and

O’Connor 2015), marmots and other mountain squirrels hibernate to endure the low

temperatures and limited plant availability of winter. Marmots are emerging from

hibernation earlier in response to warmer spring temperatures, out of sync with the

timing of snowmelt and the growing season, with potential impacts on their fitness and

survival (Inouye et al. 2000). In this case, the timing of warming may be more important

than the warming itself, as hibernating animals depend upon environmental cues to

maximize the efficiency of their adaptive response. When considering mutual climatic

stressors facing different montane species, it is important to understand the unique

limiting factors associated with their life histories and evolutionary strategies.

Page 40 of 62

Management Recommendations

Developing a conservation strategy for an indicator species at the frontier of climate

change impacts offers both a novel challenge and a significant learning opportunity. The

main question, though, is how we protect a species whose primary threat is climate

change – a global phenomenon that cannot be directly or significantly mediated by local

or regional management actions. Most likely, the answer is an integrated approach

informed by a place-based, comprehensive understanding of the pika’s life history and

foraging behavior as it relates to changes in local climate. First, we must prioritize

conservation efforts by identifying genetically significant populations, high-quality sites

that could function as habitat refugia, and areas of vital habitat connectivity to allow for

adequate gene flow and population stability. Species Distribution Models (SDMs) can

tell us where pikas are the most vulnerable to shifts in temperature or water availability.

Field surveys and remote sensing can be used to characterize habitat quality and

assess plant community composition and structure. Though little can be done to combat

the source of the issue (i.e., temperature stress, altered hydrologic regimes, etc.), there

are some opportunities to address the mechanisms and mitigate the ecological impacts.

Habitat Restoration and Enhancement

Understanding the vegetation types and characteristics that have either promoted or

limited pika persistence can inform restoration decisions and buffer future climate

stressors. Land managers could potentially install plants high in PSMs, particularly in

areas where seasonality is extreme. Plants with high metabolic water content might be

selected for low-elevation sites where moisture is limiting, and plants with significant

nutritional components (e.g., digestible protein, nitrogen) may be added at locations

where the nutritional quality of available forage is poor. Habitat restoration, including the

installation of preferred plant species and beneficial habitat features (e.g., fallen logs),

can be used to support vulnerable populations and establish habitat connectivity and

wildlife corridors. Habitat features relevant to pika dispersal and recolonization are

currently being tested along Interstate-90 in the Cascade Range of Washington

(Personal communication). Areas seeing rapid successional changes in plant

Page 41 of 62

community, such as through juniper-pinyon encroachment or the replacement of alpine

meadows or aspen woodlands with conifer forests, should be prioritized for monitoring

and potential restoration, as these changes may reflect a reduction in forage diversity

and quality or a shift in microclimatic conditions. Restoration of alpine meadows and

other high elevation plant communities can benefit a range of species beyond the pika,

including birds and other montane mammals.

Researchers are now considering novel experiments to test restoration methods and to

evaluate limiting factors. Hamster bottles are being places among the talus in order to

bolster stressed populations, which will be followed by careful monitoring to assess the

importance of water availability as a limiting factor (Personal communication).

Additionally, targeted plantings of shrubs like fernbush that have been shown to

positively correlate with pika persistence (Ray et al. 2016) are being pursued. Field trials

like these are necessary to provide ground-truthed data that validates models and

informs conservation strategies.

Legal and Regulatory Actions

Conservation strategies can also involve legal action and regulation. Simple

management decisions like placing trail restrictions in areas where vulnerable pika

populations are known to exist, as heavy trail traffic may disrupt foraging behavior (Stafl

and O’Connor 2015). On the larger scale, the Endangered Species Act (ESA) may be

utilized as a tool to protect the pika. This could prevent further habitat degradation or

fragmentation, and allocate resources and attention to restoring populations. However,

federal and even state listing is a highly contentious decision, as protections to such a

widely distributed species could create significant restrictions on activities like mining,

timber production, and watershed management. Due to the isolation and inaccessibility

of mountain areas, the listing of the pika does not pose many of the same problems

presented by other species at lower elevation where urban development and agriculture

can be interrupted; however, mountain ecosystems are hotspots for recreation, and

activities like skiing, snowboarding, camping, hiking, and tourism could be inhibited.

One important question is whether the natural resource agencies (e.g., U.S. Forest

Page 42 of 62

Service, Bureau of Land Management) will take the lead in pursuing further, objective

research to determine whether listing is necessary, despite these potential challenges

and disputes. Legal protections for the pika may very well begin at the local level with

Ecologically Significant Unit designations and state agency regulations. Although efforts

to list the pika under the ESA have been deterred twice in the last decade (2010 and

2016 due to lack of evidence, the rapidly growing body of research may soon facilitate

legal action. Whether or not the species and habitat protections afforded by the ESA

can adequately balance climate impacts in the face of the infamous bureaucracy and

limitations associated with the contentious rule remains unknown.

Adaptive Capacity as a Conservation Metric

Natural resource managers can enhance their ability to address the issues of

biodiversity loss and reduced ecosystem services related to the ecological impacts of

climate change by considering species’ adaptive capacity (AC) when developing

conservation strategies. Reflecting the ability of a species to tolerate stress (such as

climate change), AC corresponds with phenotypic plasticity, genetic diversity, and

dispersal ability, and is not to be confused with a species’ sensitivity to stress. However,

a firm distinction must be made between fundamental AC and realized AC (Beever et al.

2016b). Realized AC is the actual expression of AC after accounting for the limitations

of extrinsic factors, such as land use and land cover, pollution, interspecific competition,

and habitat loss or fragmentation. For example, the fundamental ability of pikas to shift

their diet to a more readily available and persistent food source like moss could be

hampered by anthropogenic disturbance that alters moss cover in the landscape

(Beever et al. 2016b). Furthermore, dietary plasticity can only go as far as the

availability of adequate vegetation. Conservation strategies can work to reduce the

vulnerability of the target species, such as the pika, by taking measures that either

improve the expression of the fundamental AC or reduce extrinsic constraints limiting

that expression.

Page 43 of 62

Appendix:

Preliminary Vegetation Data Analyses

for the Great Basin

Page 44 of 62

Background

Objective

The results of my secondary research presented in the main portion of this paper

suggest that the quality and availability of forage, including its interactions with climate,

may influence the persistence or extirpation of pika populations. A diverse understory of

forbs and other plants high in secondary metabolites and nutritional components may

promote pika persistence, while extensive tree cover and graminoid-dominated plant

communities may drive extirpation, especially in areas of extreme seasonality where

warm, dry summers may limit foraging time and where winter haypiles are essential to

survival and fitness. However, these relationships are highly location-specific and vary

across the pika’s range.

In order to test the conclusions generated by my secondary research and to shed light

on the role of vegetation in the Great Basin – the part of the pika’s range most heavily

affected by recent extirpations (Beever et al. 2016a) – I utilized statistical methods to

test various plant species and types as predictors of pika presence. Specifically, I am

interested in finding if any of the vegetative variables in question can serve as

significant predictors of pika occurrence in the Great Basin, which may help further

characterize the ultimate drivers of pika persistence and extirpation. An information-

theoretic framework was employed to assess multiple competing hypotheses based on

a model set constructed a priori. These preliminary analyses help demonstrate what can

be learned from statistical exploration and modeling of data collected in the field, and

also lay the foundation for a forthcoming, more robust investigation.

Hypotheses Development

For my preliminary occupancy analyses, I chose to investigate the following predictors

(independently and in combination), measured as continuous variables (percent cover):

- Juniper (Juniperus spp.)

- Single-leaf pinyon (Pinus monophyla)

Page 45 of 62

- Cheatgrass (Bromus tectorum)

- Aspen (Populus tremuloides)

- Fernbush (Chamaebatiaria millefolium)

- Oceanspray (Holodiscus discolor)

- Forb total (sum of percent cover for all forb species)

Juniper and pinyon encroachment has been identified as one of the most significant

determinants of pygmy rabbit (another lagomorph found at mid-to-high elevations)

extirpation in the Great Basin (Larrucea 2007). Fire suppression and other land-use

changes have allowed juniper-pinyon woodland communities to expand into sagebrush

and grass-dominated plant communities, increasing their distribution approximately 2.5

times over the last 150 years, and resulting in a rapid, drastic reduction in understory

vegetation and habitat. Pollen records suggest that a dramatic decline in pygmy rabbits

during the mid-Holocene correspond with an expansion of juniper-pinyon woodlands

(Grayson 2006). Similarly, cheatgrass, an invasive grass that has ravaged native

grasslands across the Great Basin, has been negatively associated with pygmy rabbit

presence (Larrucea 2007), and may result in a reduction in forage quality and diversity

for pikas. In addition to altering fire regimes and disturbing plant community

successional dynamics, cheatgrass, in its later phenological stages, is difficult to digest

and offers little in nutritional value to foraging pikas (Beever et al. 2008). Representing a

sort of inverse of juniper-pinyon woodlands, aspen woodlands, which are negatively

associated with juniper-pinyon woodlands and have seen dramatic reductions across

the western U.S. over the past century, provide a complex understory and vital habitat

for an array of birds and mammals (Wall et al. 2006, Stam et al. 2008), potentially

offering high quality forage and habitat for pikas, as well. Fernbush is being considered

as a potential predictor of pika occupancy because it was recently implicated as a

significant predictor of distribution at an interior range margin, Lava Beds National

Monument (Ray et al. 2016). Anecdotal evidence has suggested oceanspray as another

shrub that potentially serves as an important source of high-quality forage for pikas

(Personal communication). Finally, forb total – the summed percent cover of all forb

species present – is included as a potential predictor, since forb cover and species

Page 46 of 62

richness has been demonstrated as a significant predictor of pika occupancy and

density in a variety of recent studies across the pika’s range (Rodhouse et al. 2010, Erb

et al. 2014, Ray et al. 2016).

I hypothesize that juniper, pinyon, and cheatgrass will negatively predict pika occupancy

(or positively predict pika absence), and that aspen, fernbush, oceanspray, and forb

total will positively predict pika occupancy (or negatively predict pika absence).

Following my preliminary findings, I plan to incorporate more community-level

hypotheses (e.g., relative tree/shrub and forb/graminoid cover) in later analyses.

Methods

Dataset

The dataset, covering transect data from hundreds of locations in the Great Basin

across twelve years, was imported into R as a CSV file. Occupancy, a binary variable

(0=absent, 1=present), was defined as any talus patch where unequivocal evidence of

pika presence was found, including actually seeing pika individuals, hearing pika calls,

and locating fresh haypiles. Old haypiles (i.e., desiccated or degraded haypile that has

clearly been abandoned) were recorded in the dataset but insufficient for classification

of a patch as occupied. Percent cover was recorded for the five plant species with the

greatest cover value along the transects, assessed using ocular cover estimations. As a

potential source of error, this methodology may underestimate the importance of some

plant species that are not as abundant but function as preferred, high quality sources of

nutrition, metabolic water, or PSMs. Additionally, percent cover was recorded as zero if

the plant was not included in the five species, even though it may still have been

present at low density.

Statistical Analyses

The R Console was used to conduct a series of logistic regressions, with pika

occupancy as the dependent (response) variable and percent cover as the independent

(predictor) variables. In order to test the normality of the distribution of the predictor

Page 47 of 62

variables, I constructed histograms of each. Due to the size of the dataset and the

number of zero values present for percent cover, each of the predictor variables were

left skewed and needed to be log-transformed to correct this. Next, I ran correlation

functions between each of the predictor variables to determine if any were covariates of

each other, in which case they should not be included in the same model. Since none of

the predictors were significantly correlated with each other, I was able to proceed with

running my model sets. I ran generalized linear model (glm) functions on my predictors

and called summaries of the resulting models, which gave me the Akaike’s Information

Criterion (AIC) value for each. Akaike’s Information Criterion (AIC) is a metric for

comparing model fit between a set of models constructed a priori. Lower AIC values

indicate better fit, and models with differences in AIC values (ΔAIC) great than two are

considered significantly different in fit (Burnham and Anderson 2002).

Results

The most predictive model of the set includes the additive variables of juniper, pinyon,

and cheatgrass (Table 1). Juniper and/or pinyon show up in each of the best fitting

models. The null model features the highest AIC score and so the poorest fit, followed

by forb total). Juniper, pinyon, cheatgrass, and forb total each display negative

relationships with pika occupancy in their respective models, while aspen, fernbush, and

oceanspray each display positive relationships. The strongest models are composed of

negative predictors, while the weakest models feature positive predictors with the

exception of the weakest model (exluding the null model), forb total. AIC values range

from 1624.4 (strongest model) to 1727.6 (null model), and ΔAIC values range from 0 to

103.2. The two strongest models are the combinations with the most predictors (3

independent variables and 1 response), and thus the least parsimony.

Table 1. Relative support for selected models of pika occupancy in the hydrogeographic Great Basin. K

indicated the number of fitted parameters. Models are ranked in order of increasing AIC value (lower

value indicated better fit). ΔAIC indicated the difference between the AIC value of the indicated model

(AICi) and the model with lowest AIC value (AIC0).

Page 48 of 62

Model: Predictor (effect sign) K AIC ΔAIC

Juniper (-), Pinyon (-), Cheatgrass (-) 4 1624.4 0

Juniper (-), Pinyon (-), Forb Total (-) 4 1636 11.6

Juniper (-), Pinyon (-) 3 1638 13.6

Pinyon (-) 2 1655 30.6

Juniper (-) 2 1695.8 71.4

Cheatgrass (-), Forb Total (-) 3 1707.1 82.7

Cheatgrass (-) 2 1709.3 84.9

Fernbush (+) 2 1711.9 87.5

Aspen (+), Forb Total (+) 3 1717.2 92.8

Aspen (+) 2 1718.4 94

Oceanspray (+) 2 1722.1 97.7

Forb Total (-) 2 1726.2 101.8

Null (Intercept-only) 1 1727.6 103.2

Discussion

Juniper and pinyon appear to be significant predictors of pika absence in the

hyrdrogeographic Great Basin (Figure1). Although the precise mechanisms are unclear,

the supplemental effect of cheatgrass in the strongest model may indicate that

successional shifts in plant community composition and structure, and resulting decline

in forage quality, may be a driver. These negative variables exhibit the strongest

predictive power of the model set, potentially suggesting that loss of adequate forage or

habitat is a more important driver of extirpation than high quality forage is of

persistence. However, since juniper, pinyon, and cheatgrass are each associated with

xeric conditions, the negative relationship between these plant species and pika

occupancy may simply reflect the negative effects of warmer, drier microclimates.

Further analyses of how juniper-pinyon and cheatgrass have changed at each site over

time in relation to changes in pika density could offer insight into how these plant

community shifts affect pika populations.

Page 49 of 62

Figure 1. Plots of pika occupancy versus total percent cover of juniper and pinyon. On the y-axis, 1

indicates occupied talus patches (presence) and 0 indicates unoccupied patches (absence).

Surprisingly, total forb cover did not offer a significantly better fit compared to the null

model and actually exhibited a weakly negative relationship with pika occupancy.

Metrics of forb cover have received strong support as determinants of pika distribution

in the Great Basin and elsewhere, potentially due to their relatively high nutritional

content and prevalence of PSMs, as well as a reflection of a more favorable, mesic

microclimate (Wilkening 2007, Rodhouse et al. 2010, Erb et al. 2014, Ray et al. 2016).

Part of my hypothesis for why juniper-pinyon encroachment might negatively impact

pikas – the reduction in understory and forage quality – is at odds with the lack of

support for forbs as a predictor of pika occupancy. Inserting total forb cover as an

additional predictor to models featuring juniper and pinyon did not contribute

significantly to model fit, nor did it strengthen models of cheatgrass or aspen. Perhaps

total forb cover is not as essential as the cover of particular, preferred forb species.

Further analysis at a finer taxonomic resolution, exploring particular forb species based

on common haypile components at the sites, might better represent the influence of forb

cover as a predictor. Forb diversity would be another interesting metric to explore, but

the transect methodology, which only recorded the five most pervasive species, does

Page 50 of 62

not lend itself to species richness estimates, since species found at lower densities

would not be represented.

Puzzlingly, parsimony does not seem to matter much for my model set. While AIC is

meant to favor models balancing the number of variables with goodness of fit (i.e., more

variables only strengthen the explanatory power of the model up to a certain point

before it becomes profligate), higher K values consistently equate to greater fit in each

case (Table 1), as well as in other tested combinations not shown. Further examination

of the plots of occupancy against each variable and combination may help explain

patterns of model ranking. Additionally, calculations of the AICc values (AIC corrected

for small sample size) and Akaike weights for the different variables and models may

provide more accurate estimates of model fit and prediction strength.

These results represent preliminary findings that will precede a more robust exploration

and analysis of this dataset, with the intention of eventual publication in the peer-

reviewed literature. Subsequent analyses will cover a range of levels, comparing

averages across different sites (defined as all talus patches within a 3 km radius of a

historical pika record) in the Great Basin, as well as comparing individual occupied and

unoccupied patches within sites. These analyses will consider issues of persistence

rather than just occupancy in order to potentially identify relationships between changes

in occupancy or population density and shifts in plant community composition at the

locations across the twelve years of monitoring. Population density will be approximated

using estimated number of individual pikas present and corresponding number of

transects conducted, which will offer a higher-resolution perspective on population

dynamics and mechanisms of extirpation. The preliminary analyses have facilitated my

exploration of the dataset and provided a baseline upon which to base further

investigation; however, the results of the final analyses should be more credible and

interpretable in order to pass peer review and to justify publication.

Page 51 of 62

Acknowledgements

John Callaway, Ph.D., Professor and Graduate Program Director for M.S. in

Environmental Management at the University of San Francisco, served as my advisor

for this project and offered countless hours of assistance and invaluable feedback on

my paper as it was being developed. I received guidance from Erik Beever, Ph.D.,

Research Ecologist with the U.S. Geological Survey (USGS), and Jennifer Wilkening,

Ph.D., Wildlife Biologist and Habitat Conservation Plan Coordinator with the U.S. Fish

and Wildlife Service (USFWS), both of whom are experts in the field of pika ecology and

have published relevant articles in peer-reviewed journals, many of which are cited in

this paper, and were able to direct me to many other pertinent studies. Dr. Beever

provided the data and R script used in my data analyses in the appendix, and Dr.

Wilkening advised me on the statistical implementation procedures. Finally, Amalia

Kokkinaki, Ph.D., Assistant Professor at the University of San Francisco, worked with

me to prepare my data and to develop my approach for the statistical analyses.

Page 52 of 62

Literature Cited

Beever, E.A., Wilkening, J.L., McIvor, D.E., Weber, S.S., Brussard, P.F., 2008.

American pikas (Ochotona princeps) in northwestern Nevada: A newly discovered

population at a low-elevation site. West. North Am. Nat. 68, 8–14.

doi:10.3398/1527-0904(2008)68[8:APOPIN]2.0.CO;2

Beever, E. A, Ray, C., Mote, P.W., Wilkening, J.L., 2010. Testing alternative models of

climate-mediated extirpation. Ecol. Appl. 20, 164–178. doi:10.1890/08-1011.1

Beever, E.A., Ray, C., Wilkening, J.L., Brussard, P.F., Mote, P.W., 2011. Contemporary

climate change alters the pace and drivers of extinction. Glob. Chang. Biol. 17,

2054–2070. doi:10.1111/j.1365-2486.2010.02389.x

Beever, E.A., Dobrowski, S.Z., Long, J., Mynsberge, A.R., Piekielek, N.B., 2013.

Understanding relationships among abundance, extirpation, and climate at

ecoregional scales. Ecology 94, 1563–1571. doi:10.1890/12-2174.1

Beever, E.A., 2016. Understanding the heterogeneity and nuance with which climate

change is acting on mountain ecosystems: insights from 1994-2016 research.

USGS-Northern Rocky Mountain Science Center. Dept. of Ecology, Montana State

Univ.

Beever, E.A., Perrine, J.D., Rickman, T., Flores, M., Clark, J.P., Waters, C., Weber,

S.S., Yardley, B., Thoma, D., Chesley-Preston, T., Goehring, K.E., Magnuson, M.,

Nordensten, N., Nelson, M., Collins, G.H., 2016a. Pika (Ochotona princeps) losses

from two isolated regions reflect temperature and water balance, but reflect habitat

area in a mainland region. J. Mammal. 97, 1495-1511.

doi:10.1093/jmammal/gyw128

Page 53 of 62

Beever, E.A., O’Leary, J., Mengelt, C., West, J.M., Julius, S., Green, N., Magness, D.,

Petes, L., Stein, B., Nicotra, A.B., Hellmann, J.J., Robertson, A.L., Staudinger,

M.D., Rosenberg, A.A., Babij, E., Brennan, J., Schuurman, G.W., Hofmann, G.E.,

2016b. Improving conservation outcomes with a new paradigm for understanding

species’ fundamental and realized adaptive capacity. Conserv. Lett. 9, 131–137.

doi:10.1111/conl.12190

Bennie, J.J., Duffy, J.P., Inger, R., Gaston, K.J., 2014. Biogeography of time partitioning

in mammals. PNAS 111, 13727–32. doi:10.1073/pnas.1216063110

Burnham, K. P. and Anderson, D.R.., 2002. Model selection and multimodel inference: a

practical information-theoretic approach, 2nd edition. Springer-Verlag, New York.

Burnham, K.P., Anderson, D.R., Huyvaert, K.P., 2011. AIC model selection and

multimodel inference in behavioral ecology: Some background, observations, and

comparisons. Behav. Ecol. Sociobiol. 65, 23–35. doi:10.1007/s00265-010-1029-6

Cahill, A.E., Aiello-lammens, M.E., Fisher-Reid, M.C., Hua, X., Karanewsky, C.J., Ryu,

H.Y., Sbeglia, G.C., Spagnolo, F., Waldron, J.B., Warsi, O., John, J., Yeong Ryu,

H., Sbeglia, G.C., Spagnolo, F., Waldron, J.B., Warsi, O., Wiens, J.J., 2012. How

does climate change cause extinction? Proc. R. Soc. B Biol. Sci. 280, 1890–1890.

doi:10.1098/rspb.2012.1890

Calkins, M.T., Beever, E.A., Boykin, K.G., Frey, J.K., Andersen, M.C., 2012. Not-so-

splendid isolation: Modeling climate-mediated range collapse of a montane

mammal Ochotona princeps across numerous ecoregions. Ecography (Cop.). 35,

780–791. doi:10.1111/j.1600-0587.2011.07227.x

Castillo, J.A., Epps, C.W., Davis, A.R., Cushman, S.A., 2014. Landscape effects on

gene flow for a climate-sensitive montane species, the American pika. Mol. Ecol.

23, 843–856. doi:10.1111/mec.12650

Page 54 of 62

Chamberlin, T.C., 1965. The method of multiple working hypotheses. Science 15, 92–

96. doi:10.1086/629752

Conner, D.A., 1983. Seasonal changes in activity patterns and the adaptive value of

haying in pikas (Ochotona princeps). Canadian Journal of Zoology 61, 411–416.

doi:10.1139/z83-054

Cook, E.R., 2004. Long-term aridity changes in the Western United States. Science

306, 1015–1018. doi:10.1126/science.1102586

Crimmins, S.M., Dobrowski, S.Z., Greenberg, J.A., Abatzoglou, J.T., Mynsberge, A.R.,

2011. Changes in climatic water balance drive downhill shifts in plant species’

optimum elevations. Science 331, 324–327. doi:10.1126/science.1199040

Dearing, M.D., 1996. Disparate determinants of summer and winter diet selection of a

generalist herbivore, Ochotona princeps. Oecologia 108, 467–478.

doi:10.1007/bf00333723

Dearing, M.D., 1997a. The function of haypiles of pikas (Ochotona princeps). J.

Mammal. 78, 1156-1163. doi:10.2307/1383058

Dearing, M.D., 1997b. The manipulation of plant toxins by a food-hoarding herbivore,

Ochotona princeps. Ecol. 78, 774-781. doi:10.2307/2266057

DeGabriel, J.L., Moore, B.D., Foley, W.J., Johnson, C.N., 2009. The effects of plant

defensive chemistry on nutrient availability predict reproductive success in a

mammal. Ecol. 90, 711-719.

Department of the Interior, Fish and Wildlife Service (US FWS), 2010. Endangered and

Threatened Wildlife and Plants, 12-month Finding on a Petition to List the American

Page 55 of 62

Pika as Threatened or Endangered, Proposed Rule. Federal Register 75, 26. 50

CFR Part 17

Erb, L.P., 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.

doi:10.1890/11-0175.1

Ernest, K.A., 2016. Increasing pika connectivity across an interstate highway in the

Washington Cascades: update of early post-construction monitoring. North

American Pika Consortium.

Franklin, J., Serra-Diaz, J.M., Syphard, A.D., Regan, H.M., Asner, G.P., Turner, M.G.,

Vitousek, P.M., 2016. Global change and terrestrial plant community dynamics.

PNAS 113, 3725-3734. doi:10.1073/pnas.1519911113

Freeland, W.J., Janzen, D.H., 1974. Strategies in herbivory by mammals: the role of

plant secondary compounds. The American Naturalist 108, 269–289.

doi:10.1086/282907

Galbreath, K.E., Hafner, D.J., Zamudio, K.R., 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.

doi:10.1111/j.1558-5646.2009.00803.x

Ge, D., Zhang, Z., Xia, L., Zhang, Q., Ma, Y., Yang, Q., 2012. Did the expansion of C4

plants drive extinction and massive range contraction of micromammals?

Inferences from food preference and historical biogeography of pikas.

Palaeogeography, Palaeoclimatology, Palaeoecology 326-328, 160–171.

doi:10.1016/j.palaeo.2012.02.016

Page 56 of 62

Ge, D., Wen, Z., Xia, L., Zhang, Z., Erbajeva, M., Huang, C., Yang, Q., 2013.

Evolutionary history of lagomorphs in response to global environmental change.

PLoS One 8. doi:10.1371/journal.pone.0059668

Grayson, D.K., 2000. Mammalian responses to Middle Holocene climatic change in the

Great Basin of the western United States. Journal of Biogeography 27, 181–192.

doi:10.1046/j.1365-2699.2000.00383.x

Grayson, D.K., 2005. A brief history of Great Basin pikas. Journal of Biogeography 32,

2103–2111. doi:10.1111/j.1365-2699.2005.01341.x

Grayson, D.K., 2006. The Late Quaternary biogeographic histories of some Great Basin

mammals (western USA). Quaternary Science Reviews 25, 2964–2991.

doi:10.1016/j.quascirev.2006.03.004

Guisan, A, and Theurillat, J., 2000. Assessing alpine plant vulnerability to climate

change: a modeling perspective. Integrated Assessment 1, 207-320.

Hansen, R.M., Gold, I.K., 1977. Blacktail prairie dogs, desert cottontails and cattle

trophic relations on shortgrass range. Journal of Range Management 30, 210.

doi:10.2307/3897472

Harte, J., Shaw, R., 1995. Shifting dominance within a montane vegetation community:

results of a climate-warming experiment. Science 267, 876–880.

doi:10.1126/science.267.5199.876

Holechek, J.L., 1984. Comparative contribution of grasses, forbs, and shrubs to the

nutrition of range ungulates. Rangelands 6, 261–263.

Page 57 of 62

Holmes, W.G., 1991. Predator risk affects foraging behaviour of pikas: observational

and experimental evidence. Anim. Behav. 42, 111–119. doi:10.1016/S0003-

3472(05)80611-6

Hudson, J.M.G., Morrison, S.F., Hik, D.S., 2008. Effects of leaf size on forage selection

by collared pikas, Ochotona collaris. Arctic, Antarctic, and Alpine Research 40,

481–486. doi:10.1657/1523-0430(07-069)[hudson]2.0.co;2

Huntly, N.J., Smith, A.T., Ivins, B.L., 1986. Foraging behavior of the pika (Ochotona

princeps), with comparisons of grazing versus haying. J. Mammal. 67, 139–148.

doi:10.2307/1381010

Inouye, D.W., Mcguire, A.D., 1991. Effects of snowpack on timing and abundance of

flowering in Delphinium nelsonii (Ranunculaceae): Implications for climate change.

American Journal of Botany 78, 997. doi:10.2307/2445179

Jeffress, M.R., Rodhouse, T.J., Ray, C., Wolff, S., Epps, C.W., 2013. The idiosyncrasies

of place: Geographic variation in the climate-distribution relationships of the

American pika. Ecol. Appl. 23, 864–878. doi:10.1890/12-0979.1

Kelly, A.E., Goulden, M.L., 2008. Rapid shifts in plant distribution with recent climate

change. Proceedings of the National Academy of Sciences 105, 11823–11826.

doi:10.1073/pnas.0802891105

Loik, M.E., Redar, S.P., Harte, J., 2000. Photosynthetic responses to a climate-warming

manipulation for contrasting meadow species in the Rocky Mountains, Colorado,

USA. Functional Ecology 14, 166–175. doi:10.1046/j.1365-2435.2000.00411.x

MacArthur, R.A., Wang, L.C.H., 1973. Physiology of thermoregulation in the pika. Can.

J. Zool. 51, 11–16.

Page 58 of 62

Markham, O.D., Whicker, F.W., 1973. Notes on the behavior of the pika (Ochotona

princeps) in captivity. American Midland Naturalist 89, 192. doi:10.2307/2424146

Mccain, C.M., King, S.R.B., 2014. Body size and activity times mediate mammalian

responses to climate change. Glob. Chang. Biol. 20, 1760–1769.

doi:10.1111/gcb.12499

Mcdonald, K.A., Brown, J.H., 1992. Using montane mammals to model extinctions due

to global change. Conservation Biology 6, 409–415. doi:10.1046/j.1523-

1739.1992.06030409.x

Millar, J.S., Zwickel, F.C., 1972. Characteristics and ecological significance of hay piles

of pikas. Mammalia 36. doi:10.1515/mamm.1972.36.4.657

Millar, J.S., 1974. Success of reproduction in pikas, Ochotona princeps (Richardson). J.

Mammal. 55, 527–542. doi:10.2307/1379544

Millar, C.I., Westfall, R.D., 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. Reply. Arctic, Antarct. Alp.

Res. 42, 493–496. doi:10.1657/1938-4246-42.4.493

Millar, C.I., Westfall, R.D., Delany, D.L., 2013. New records of marginal locations for

American pika (Ochotona princeps) in the Western Great Basin. Source West.

North Am. Nat. Brigham Young Univ. 73, 457–476. doi:10.3398/064.073.0412

Monson, R.K., Rosenstiel, T.N., Forbis, T.A., Lipson, D.A., Jaeger, C.H., 2006. Nitrogen

and carbon storage in alpine plants. Integr Comp Biol 46, 35-48.

doi:10.1093/icb/icj006

Page 59 of 62

Morrison, S.F., Hik, D.S., 2007. Demographic analysis of a declining pika Ochotona

collaris population: Linking survival to broad-scale climate patterns via spring

snowmelt patterns. J. Anim. Ecol. 76, 899–907. doi:10.1111/j.1365-

2656.2007.01276.x

Moyer-Horner, L., Beever, E.A., Johnson, D.H., Biel, M., Belt, J., 2016. Predictors of

current and longer-term patterns of abundance of American pikas (Ochotona

princeps) across a leading-edge protected area. PLoS ONE 11, 1-25.

doi:10.1371/journal.pone.0167051

Nichols, L.B., Klingler, K.B., Peacock, M.M., 2016. American pikas (Ochotona Princeps)

extirpated from the historic masonic mining district of Eastern California. West.

North Am. Nat. 76, 163–171.

NOAA National Centers for Environmental Information, 2016. Potential

Evapotranspiration. Accessed 15 December 2016. [Available online at

https://www.ncdc.noaa.gov/monitoring-references/dyk/potential-evapotranspiration]

Osborne, C.P., 2007. Atmosphere, ecology and evolution: what drove the Miocene

expansion of C4 grasslands? Journal of Ecology. doi:10.1111/j.1365-

2745.2007.01323.x

Peacock, M.M., 1997. Determining natal dispersal patterns in a population of North

American pikas (Ochotona princeps) using direct mark-resight and indirect genetic

methods. Behav. Ecol. 8, 340–550.

Pepin, N., Bradley, R.S., Diaz, H.F., Baraer, M., Caceres, E.B., Forsythe, N., Fowler, H.,

Greenwood, G., Hashmi, M.Z., Liu, X.D., Miller, J.R., Ning, L., Ohmura, A., Palazzi,

E., Rangwala, I., Schöner, W., Severskiy, I., Shahgedanova, M., Wang, M.B.,

Williamson, S.N., Yang, D.Q., 2015. Elevation-dependent warming in mountain

Page 60 of 62

regions of the world. Nature Climate Change 5, 424–430.

doi:10.1038/nclimate2563

Promislow, D.E.L., Harvey, P.H., 1990. Living fast and dying young: A comparative

analysis of life-history variation among mammals. J. Zool., Lond. 220 417–437.

Ray, C., Beever, E.A., Rodhouse, T.J., 2016. Distribution of a climate-sensitive species

at an interior range margin. Ecosphere 7. doi:10.1002/ecs2.1379

Ray, C., Gilpin, M., Smith, A.T., 1991. The effect of conspecific attraction on

metapopulation dynamics. Biol. J. Linnean. Soc. 42, 123–134.

Robbins, C.T., Hanley, T.A., Hagerman, A.E., Hjeljord, O., Baker, D.L., Schwartz, C.C.,

Mautz, W.W., 1987. Role of tannins in defending plants against ruminants:

Reduction in protein availability. Ecology 68, 98–107. doi:10.2307/1938809

Rodhouse, T.J., Beever, E. A., Garrett, L.K., Irvine, K.M., Jeffress, M.R., Munts, M.,

Ray, C., 2010. Distribution of American pikas in a low-elevation lava landscape:

Conservation implications from the range periphery. J. Mammal. 91, 1287–1299.

doi:10.1644/09-MAMM-A-334.1

Sarukhán, J., 2006. Conservation biology: Views from the ecological sciences. Conserv.

Biol. 20, 674–676. doi:10.1111/j.1523-1739.2006.00461.x

Schwalm, D., Epps, C.W., Rodhouse, T.J., Monahan, W.B., Castillo, J.A., Ray, C.,

Jeffress, M.R., 2016. Habitat availability and gene flow influence diverging local

population trajectories under scenarios of climate change: A place-based approach.

Glob. Chang. Biol. 22, 1572–1584. doi:10.1111/gcb.13189

Page 61 of 62

Scott, R.L., Huxman, T.E., Barron-Gafford, G.A., Darrel Jenerette, G., Young, J.M.,

Hamerlynck, E.P., 2014. When vegetation change alters ecosystem water

availability. Glob. Chang. Biol. 20, 2198–2210. doi:10.1111/gcb.12511

Sheafor, B. A., 2003. Metabolic enzyme activities across an altitudinal gradient: an

examination of pikas (genus Ochotona). Journal of Experimental Biology 206,

1241–1249.

Smith, J.F., 1988. Influence of nutrition on ovulation rate in the ewe. Aust. J. Bioi. Sci.

41, 27–36.

Smith, A.T., 1974. The distribution and dispersal of pikas: Influences of behavior and

climate. Ecology 55, 1368–1376. doi:10.2307/1935464

Smith, J.A., Erb, L.P., 2013. Patterns of selective caching behavior of a generalist

herbivore , the American Pika (Ochotona princeps). Arctic, Antarct. Alp. Res. 45,

396–403. doi:10.1657/1938-4246-45.3.396

Stafl, N., O'connor, M.I., 2015. American pikas' (Ochotona princeps) foraging response

to hikers and sensitivity to heat in an alpine environment. Arctic, Antarctic, and

Alpine Research 47, 519–527. doi:10.1657/aaar0014-057

Stam, B., Malechek, J., Bartos, D., Bowns, J., and Godfrey, E., 2008. Effect of conifer

encroachment into aspen stands on understory biomass. Rangeland Ecology &

Management. 61, 93–97

Stewart, J.A.E., Perrine, J.D., Nichols, L.B., Thorne, J.H., Millar, C.I., Goehring, K.E.,

Massing, C.P., Wright, D.H., 2015. Revisiting the past to foretell the future: Summer

temperature and habitat area predict pika extirpations in California. J. Biogeogr. 42,

880–890. doi:10.1111/jbi.12466

Page 62 of 62

Torregrossa, A.-M., Dearing, M.D., 2009. Caching as a behavioral mechanism to reduce

toxin intake. J. mammal. 90, 803–810. doi:10.1644/08-MAMM-A-261.1

Tyser, R.W., 1980. Use of substrate for surveillance behaviors in a community of talus

slope mammals. American Midland Naturalist 104, 32. doi:10.2307/2424955

UNESCO, 2014. Mountains provide early warning of climate change. United Nations

Educational, Scientific, and Cultural Organization, Natural Sciences Sector.

http://www.unesco.org/new/en/media-services/single-

view/news/mountains_provide_early_warning_of_climate_change/

Varner, J., Dearing, M.D., 2014. Dietary plasticity in pikas as a strategy for atypical

resource landscapes. J. Mammal. 95, 72–81. doi:10.1644/13-MAMM-A-099.1

Varner, J., Lambert, M.S., Horns, J.J., Laverty, S., Dizney, L., Beever, E.A., Dearing,

M.D., 2015. Too hot to trot? Evaluating the effects of wildfire on patterns of

occupancy and abundance for a climate-sensitive habitat specialist. Int. J. Wildl.

Fire 24, 921–932. doi:10.1071/WF15038

Varner, J., Horns, J.J., Lambert, M.S., Westberg, E., Ruff, J.S., Wolfenberger, K.,

Beever, E.A., Dearing, M.D., 2016. Plastic pikas: Behavioural flexibility in low-

elevation pikas (Ochotona princeps). Behav. Processes 125, 63–71.

doi:10.1016/j.beproc.2016.01.009

Wall, T., Miller, R., Svejcar, T., 2006. Juniper encroachment in aspen in the Northwest

Great Basin. Journal of Range Management 54. doi:10.2458/azu_jrm_v54i6_wall

Wand, S.J.E., Midgley, G.F., Jones, M.H., Curtis, P.S., 1999. Responses of wild C4 and

C3 grass (Poaceae) species to elevated atmospheric CO2 concentration: a meta-

analytic test of current theories and perceptions. Global Change Biology 5, 723–

741. doi:10.1046/j.1365-2486.1999.00265.x

Page 63 of 62

Wendler, G. Fahl, C., Corbin, S., 1972. Mass balance studies on McCall Glacier Brooks

Range, Alasaka. Arct. Alp. Res. 32, 211–222. doi:10.1657/1523-0430(07-069)

Westover, M.L., Smith, F.A., 2017. Dietary variation of the American pika in the Rrocky

Mountains across a century of climate change. Fourth Annual North American Pika

Consortium. Western Section of The Wildlife Society.

Wickhem, C., 2016. Testing the forage preference of the American pika (Ochotona

princeps) for use in connectivity corridors in the Washington Cascades. M.S.

Thesis. Central Washington University. Paper 452.

Wilkening, J.L., 2007. Effects of vegetation and temperature on pika (Ochotona

princeps) extirpations in the Great Basin. M.S. Thesis. University of Nevada, Reno.

Wilkening, J.L., Ray, C., Beever, E.A., Brussard, P.F., 2011. Modeling contemporary

range retraction in Great Basin pikas (Ochotona princeps) using data on

microclimate and microhabitat. Quat. Int. 235, 77–88.

doi:10.1016/j.quaint.2010.05.004

Wilson, R.J., Gutiérrez, D., Gutiérrez, J., Martínez, D., Agudo, R., Monserrat, V.J., 2005.

Changes to the elevational limits and extent of species ranges associated with

climate change. Ecol. Lett. 8, 1138–1146. doi:10.1111/j.1461-0248.2005.00824.x

Zorio, S.D., Williams, C.F., Aho, K.A., 2016. Sixty-five years of change in montane plant

communities in Western Colorado, U.S.A. Arctic, Antarctic, and Alpine Research

48, 703–722. doi:10.1657/aaar0016-011