DESERT KIT ( MACROTIS ARSIPUS) FOOD HABITS AND COMPETITIVE INTERACTIONS WITH

COYOTES ( LATRANS) IN THE

By

Erica Crystal Kelly, B.S.

A Thesis Submitted to the Department of Biology State University, Bakersfield In Partial Fulfillment for the Degree of Masters of Science

Fall 2017

Copyright

By

Erica Crystal Kelly

2017

Desert (Vulpes macrotis arsipus) food habits and competitive interactions with

(Canis latrans) in the Mojave Desert

By Erica Crystal Kelly

This thesis has been accepted on behalf of the Department of Biology by their supervisory committee:

i

Desert Kit Fox (Vulpes macrotis arsipus) food habits and competitive interactions with Coyotes (Canis latrans) in the Mojave Desert

Erica C. Kelly

Department of Biology, California State University, Bakersfield

ABSTRACT

The Kit Fox (Vulpes macrotis) is comprised of two subspecies, the

(V. m. mutica) and the Desert Kit Fox (V. m. arsipus). The San Joaquin Kit Fox is federally listed as endangered and state listed as threatened primarily due to loss while the Desert

Kit Fox is a protected furbearer subspecies in California. Although the kit fox population in the

Mojave Desert has not been heavily impacted by widespread land conversion, this may change as urbanization and the increased support for renewable energy facilities continues.

Anthropogenic changes could result in increased stress, disease, and predation risk for Desert Kit

Foxes due to a rise in competition for food and space. Coyotes (Canis latrans), the main competitor and predator of the kit fox, are habitat generalists and opportunistic predators that typically thrive in human altered environments. Because Desert Kit and Coyotes are sympatric species in the Mojave Desert, increased competitive pressures from the latter due to human encroachment may have negative effects on the former. Although the Desert Kit Fox in

California is considered a common subspecies, interest in the conservation of this is increasing and further knowledge could contribute to future conservation efforts. An increased understanding of this subspecies, including dietary patterns, food preferences, and potential competition with other species, is imperative. Therefore, I investigated Desert Kit Fox dietary habits over time and assessed competition with Coyotes. My objectives were to complete a

ii dietary analysis of Desert Kit Fox scats collected over a five-year period, compare diets between

Desert Kit Foxes and Coyotes to determine exploitative competition, incorporate concurrent prey abundance survey findings to determine the relationship between relative prey abundance and item use by my two study species, and use camera station data to analyze potential competitive interactions. To determine Desert Kit Fox diet, I opportunistically collected scats for three consecutive days per season from fall 2009 to summer 2014 in the western Mojave Desert and subsequently analyzed them. I also incorporated concurrent, annual prey abundance data to assess the effects of relative item availability on Desert Kit Fox diet. Lastly, I used camera station data to assess the relative distribution of Desert Kit Foxes and their potential competitive associations with Coyotes. My results suggest that Desert Kit Foxes in the Mojave Desert in

California prefer to consume and invertebrates, but will prey upon a variety of other items, especially when preferred prey has declined due to drought or seasonal conditions.

Overall, my study supported the scientific consensus that kit foxes are specialists on heteromyid rodents. Invertebrate prey, though, does provide an important supplement when prey is less available. For Coyotes, the most frequently occurring items consumed were rodents and lagomorphs. Thus, Desert Kit Foxes and Coyotes did eat many of the same prey items throughout the study which could promote exploitative competition, but they did so in varying frequencies depending on the year and season. Desert Kit Foxes consistently had lower dietary diversity than Coyotes, indicating that Desert Kit Foxes are more specialist consumers while

Coyotes are more generalist. Dietary specialization by kit foxes may help reduce competitive pressures from Coyotes and lead to sympatric coexistence. Seasonality and changes in yearly precipitation patterns did have an influence on the consumption of prey items by our two study species. diet was always more diverse than that of the Desert Kit Fox, but dietary

iii diversity was lowest for both species in winter. Also, during drought conditions, Desert Kit

Foxes and Coyotes did broaden their diets which is expected based on Optimal Foraging Theory.

The decrease in preferred food items during this time may have resulted in an increase competition for remaining food sources. Also, both study species consumed anthropogenic items at an increased rate when natural foods were less abundant. Coyotes, though, seemed to rely on this food type more than Desert Kit Foxes. Anthropogenic subsidization, especially during times of decreased prey availability, can have an overall detrimental effect on the natural environment by enhancing predation pressures on co-occurring species and remaining food supplies.

Specifically, an unnatural increase in the Coyote population due to anthropogenic subsidization could cause a marked change in den selection, dietary selection, and spatial distribution of sympatric Desert Kit Foxes. In regards to habitat partitioning, I found that Desert Kit Foxes and

Coyotes in my study area did not appear to partition habitat on a landscape scale. Other potential strategies, such as temporal avoidance, resource partitioning, and the daily use of dens by Desert

Kit Foxes, may be sufficient to reduce competition and maintain coexistence. However, this balance may become imperiled as human disturbance continues in the Mojave Desert and potentially affords a competitive advantage to coyotes. Therefore, conservation and management strategies may become necessary to promote a healthy, self-sustaining Desert Kit

Fox population in the Mojave Desert in California. Management strategies should include conserving large tracts of intact, high quality habitat, facilitation of healthy and robust prey populations, and maintaining a variety of native food options in the event of declines in primary prey. Measures to control anthropogenic influences, especially those that promote a subsidized

Coyote population, are also necessary to limit unnecessary competition between Coyotes and

Desert Kit Foxes.

iv

ACKNOWLEDGEMENTS

I would like to thank my advisors Dr. Brian Cypher and Dr. David Germano for all of their support and guidance during this endeavor. I never thought as a first year college student back in 2007 that I would have found such an amazing and fulfilling career with such wonderful people as mentors. I am truly grateful to have worked for and learned from Dr. Cypher since beginning as a volunteer at the Recovery Program over a decade ago. I also thank Dr. Germano for all of his advice and suggestions regarding my thesis. I further thank Dr.

Paul Smith for assisting in identification of arthropod remains and reviewing drafts of my thesis.

I would like to thank the many individuals in my life whom without their encouragement and support this project would not have been possible. I thank my Momma and Grandma for always being there for me and telling me to just keep plugging along, Patrick and Atticus for being so understanding of all of the nights and weekends I spent meticulously analyzing scats and writing, and family and friends for their words of encouragement. I would also like to thank the Endangered Species Recovery Program staff who spent countless hours collecting scats in the blistering Mojave Desert heat as well as providing maps, database outputs, pictures, input on drafts of my chapters, and moral support: Tory Westall, Christine Van Horn Job, and Allie

Madrid. Finally, this study would not have been possible without the generous support provided by the Endangered Species Recovery Program, California State University, Stanislaus, and the

U.S. Fish and Wildlife Service.

v

TABLE OF CONTENTS

CHAPTER 1. INTRODUCTION ...... 1

LITERATURE CITED ...... 11

CHAPTER 2. TEMPORAL VARIATION IN FORAGING PATTERNS OF DESERT KIT FOXES (VULPES MACROTIS ARSIPUS) OVER A FIVE-YEAR PERIOD ...... 17

1. INTRODUCTION ...... 18

2. METHODS ...... 22 2.1 Study area ...... 22 2.2 Study Design ...... 22 2.3 Analytical methods ...... 25

3. RESULTS ...... 27

4. DISCUSSION ...... 30

ACKNOWLEDGEMENTS ...... 36

LITERATURE CITED ...... 36

CHAPTER 3. EXPLOITATIVE COMPETITION BETWEEN DESERT KIT FOXES AND COYOTES IN THE MOJAVE DESERT ...... 55

MATERIALS AND METHODS ...... 60 Study area ...... 60 Study design ...... 61 Analytical methods ...... 63

RESULTS ...... 64

DISCUSSION ...... 68

ACKNOWLEDGMENTS ...... 76

LITERATURE CITED ...... 76

CHAPTER 4. LANDSCAPE PARTITIONING BY DESERT KIT FOXES AND COYOTES IN THE MOJAVE DESERT ...... 92

1. INTRODUCTION ...... 93

2. METHODS ...... 94 2.1 Study area ...... 94 2.2 Study design ...... 94 2.3 Analytical methods ...... 95

vi

3. RESULTS ...... 96

4. DISCUSSION ...... 96

ACKNOWLEDGEMENTS ...... 98

LITERATURE CITED ...... 99

CHAPTER 5. CONCLUSIONS ...... 105 Desert Kit Fox foraging patterns ...... 105 Competitive interactions between Desert Kit Foxes and Coyotes ...... 106 Relative distribution and competitor associations ...... 108 Potential threats to Desert Kit Foxes and recommendations ...... 109

LITERATURE CITED ...... 113

vii

LIST OF TABLES

CHAPTER 2

Table 1. Frequency of occurrence of food items found in Desert Kit Fox (Vulpes macrotis arsipus) scats collected in the Mojave Desert, California, USA, for all years combined (October 2009 to September 2014). Primary food items (> 10% frequency of occurrence) are in bold...... 46 Table 2. Annual frequency of occurrence for item categories in Desert Kit Fox (Vulpes macrotis arsipus) scats collected in the Mojave Desert, California, USA, during October 2009 to September 2014. Years span October-September. P-values in bold are significant...... 49 Table 3. Seasonal frequency of occurrence for item categories in Desert Kit Fox (Vulpes macrotis arsipus) scats collected in the Mojave Desert, California, USA, during October 2009 to September 2014. Seasons were defined as Fall = October-December; Winter = January- March; Spring = April-June; Summer = July-September. P-values in bold are significant. .. 50 Table 4. Spearman-rank correlations analysis on various annual values collected from the Mojave Desert, California, USA from October 2009 to September 2014. Annual frequency of occurrence of prey items, annual precipitation in cm, average annual large and small burrows, and average annual pellet counts were used for the various analysis combinations. P-values in bold are significant...... 51

CHAPTER 3

Table 1. Annual frequency of occurrence (FOO) for item categories in desert kit fox (Vulpes macrotis arsipus) (Kelly 2017) and coyote (Canis latrans) (Cypher et al. in press) scats collected in the Mojave Desert, California, United States, from October 2009 to September 2014. Years span October–September...... 84 Table 2. Seasonal frequency of occurrence (FOO) for item categories in desert kit fox (Vulpes macrotis arsipus) (Kelly 2017) and coyote (Canis latrans) (Cypher et al. in press) scats collected in the Mojave Desert, California, United States, from October 2009 to September 2014. Seasons were defined as Fall = October-December; Winter = January-March; Spring = April-June; Summer = July-September...... 85 Table 3. Annual Shannon diversity indices for coyotes (Canis latrans) and desert kit foxes (Vulpes macrotis arsipus) and Horn’s index of similarity along with prey availability indices and annual precipitation in the Mojave Desert, California, United States, from October 2009 to September 2014. Years span October–September...... 86 Table 4. Shannon diversity indices for coyotes (Canis latrans) and desert kit foxes (Vulpes macrotis arsipus) and Horn’s index of similarity calculated for each season across all years in the Mojave Desert, California, United States, from October 2009 to September 2014. Seasons

viii

were defined as Fall = October-December; Winter = January-March; Spring = April-June; Summer = July-September...... 87 Table 5. Spearman rank correlations analysis to compare rankings of items between coyotes (Canis latrans) and desert kit foxes (Vulpes macrotis arsipus) for each year, season, and total. Scats from both species were collected from the Mojave Desert, California, United States, from October 2009 to September 2014. Years span October–September and seasons were defined as Fall = October-December; Winter = January-March; Spring = April-June; Summer = July-September...... 88

CHAPTER 4

Table 1. The number of camera stations visited by Desert Kit Foxes (Vulpes macrotis arsipus) and Coyotes (Canis latrans) for each survey year and for all years combined in the Mojave Desert, California, USA from 2012 to 2014...... 102

ix

LIST OF FIGURES

CHAPTER 1

Figure 1. A Desert Kit Fox (Vulpes macrotis arsipus; left) and a Coyote (Canis latrans; right) investigating a camera station in the Mojave Desert north of Barstow, California. (Photographs taken by California State University, Stanislaus, Endangered Species Recovery Program)...... 2 Figure 2. Typical Mojave Desert scrub vegetation dominated by Creosote Bush (Larrea tridentate) and perennial plant cover as seen within my study area in the Mojave Desert north of Barstow, California, in April 2014. (Photograph taken by California State University, Stanislaus, Endangered Species Recovery Program)...... 5 Figure 3. My research area in the Mojave Desert in California (inset), which is bound by the Fort Irwin National Training Center to the north and Interstate 15 and State Route 58 to the south. The blue line denotes the boundary of my study area. (Map created by Tory Westall and Erica Kelly)...... 9 Figure 4. A) The first Desert Kit Fox scat I analyzed for this project; B) Comparison of size between Desert Kit Fox scat (top of picture) and a Coyote scat (bottom of picture); C) Camera station in the Mojave Desert, California; D) Conducting spring prey abundance surveys. (Photographs A-C taken by Erica Kelly and photograph D taken by Christine Van Horn Job, California State University, Stanislaus, Endangered Species Recovery Program)...... 10

CHAPTER 2

Figure 1. Study area located in the Mojave Desert in California, USA (inset), which is bound by the Fort Irwin National Training Center to the north and Interstate 15 and State Route 58 to the south. The blue boundary is the specific study area and the black dots denote every location a scat or multiple scats were collected during the five year project. (This is an original map created for this manuscript)...... 52 Figure 2. Yearly proportions of grouped items with the annual precipitation trend from the Mojave Desert, California, USA, during October 2009 to September 2014...... 52

CHAPTER 3

Figure 1. Study area located in the Mojave Desert in California, United States (inset), which is bound by the Fort Irwin National Training Center to the north and Interstate 15 and State Route 58 to the south. The blue boundary is the specific study area and the black dots denote every location a scat or multiple scats were collected during the 5 year project...... 89 Figure 2. Annual Shannon diversity indices for both desert kit foxes (Vulpes macrotis arsipus) and coyotes (Canis latrans) with annual precipitation totals from the Mojave Desert, California, United States, during October 2009 to September 2014...... 89

x

CHAPTER 4

Figure 1. Study area located in the Mojave Desert in California, USA (inset), which is bounded by the Fort Irwin National Training Center to the north and Interstate 15 and State Route 58 to the south. The blue boundary is the specific study area and the black circles denote annual camera station locations...... 103

xi

CHAPTER 1

INTRODUCTION

The family () is comprised of 35 species within 13 genera, although there is disagreement about the number of species in the family (Macdonald and Sillero-Zubiri 2004,

Sillero-Zubiri and Macdonald 2004). Classification of species within the Canidae, conducted using both phenotypic and more recent genotypic analyses, continues to change and disputes within the scientific community regarding classification will probably continue (Xiaoming et al.

2004). Coyotes (Canis latrans), endemic to North America, are one of the eight species within the genus Canis (Macdonald and Sillero-Zubiri 2004, Sillero-Zubiri and Macdonald 2004,

Xiaoming et al. 2004). Foxes, including the -like canids and South American foxes, comprise 23 species within six genera (Bekoff and Gese 2003, Macdonald and Sillero-Zubiri

2004, Xiaoming et al. 2004). The six fox species that reside in North America are in either the genus Vulpes or (Cypher 2003).

Animals within Canidae have extremely diverse ecologies and many are highly adaptable and opportunistic (Kleiman 1967, Bueler 1973, Sheldon 1992). Canids tend to be omnivorous in their diet, ingesting both and plant material such as fruits (Kleiman 1967, Bueler 1973).

Body size is also highly variable between species, ranging from the (Vulpes zerda), which weighs < 1.5 kg, to the Gray Wolf (Canis lupus), which weighs up to 80 kg (Kleiman

1967, Bueler 1973, Mech 1974, Sheldon 1992, Lariviere 2002). Canids are also present on every continent except Antarctica, with many areas having multiple species living in close proximity to one another (Kleiman 1967). Various forms of competition are a common occurrence within this

1 family (Cypher and Spencer 1998, Kamler et al. 2003, Nelson et al. 2007, Thompson and Gese

2007). Throughout their range in southwestern North America, Kit Foxes (V. macrotis) are sympatric with Coyotes resulting in the potential for interference and exploitative competition

(Fig. 1; Ralls and White 1995, Cypher and Spencer 1998, Cypher et al. 2000, Nelson et al. 2007).

The Kit Fox was first described by Merriam in 1888 and currently the general scientific consensus is that the species is comprised of two subspecies, the San Joaquin Kit Fox (V. m. mutica) and the Desert Kit Fox (V. m. arsipus; Waithman and Roest 1977, Mercure et al. 1993).

These two subspecies cannot be distinguished by physical characteristics alone (Waithman and

Roest 1977, Dragoo et al. 1990, Mercure et al. 1993). They can only be differentiated by specific skull measurements and genetic testing (Waithman and Roest 1977, Mercure et al. 1993,

Wayne et al. 2004). While the San Joaquin Kit Fox is federally listed as endangered and state listed as threatened, the Desert Kit Fox populations in California have no formal protections

Figure 1. A Desert Kit Fox (Vulpes macrotis arsipus; left) and a Coyote (Canis latrans; right) investigating a camera station in the Mojave Desert north of Barstow, California. (Photographs taken by California State University, Stanislaus, Endangered Species Recovery Program).

2 aside from being a protected furbearer species (U.S. Fish and Wildlife Service [USFWS] 1998).

The Desert Kit Fox is listed as state endangered in Colorado, state threatened in Oregon, and endangered in Mexico (Moehrenschlager et al. 2004). The San Joaquin Kit Fox only resides in the San Joaquin of and is separated from all other kit foxes by the

Tehachapi Mountains (Mercure et al. 1993, Moehrenschlager et al. 2004). Remaining kit fox populations to the east of the are considered Desert Kit Foxes (Mercure et al.

1993, Moehrenschlager et al. 2004).

Kit foxes in general are characterized by a small and slim appearance, large ears, an overall light yellowish-gray pelt, and a black-tipped tail (Bueler 1973, McGrew 1979, Sheldon

1992). Kit foxes need little water to survive and rely heavily on rodents and for sustenance (Bueler 1973, McGrew 1979, Sheldon 1992, Clark et al. 2005). Females, on average, give birth to four or five pups in February or March and care for their offspring until they disperse in October (Egoscue 1962, McGrew 1979). Kit foxes are primarily socially monogamous animals, a life-history strategy found in many canid species, and both parents care for the young, albeit mothers invest significantly more time and energy than fathers (Egoscue

1962, McGrew 1979, Moehrenschlager et al. 2004, Ralls et al. 2007, Westall 2015). Fox pairs are thought to typically mate for life, but polygyny and extra-pair copulations are a fairly common occurrence with kit foxes (Egoscue 1962, Moehrenschlager et al. 2004, Ralls et al.

2007, Westall 2015).

Scats of this species are easily distinguished from those of other canids that may reside in the area (Egoscue 1962). Kit foxes not only defecate outside dens, but they will also deposit their scat along roads, near or on top of novel objects, and in latrines composed of many feces; sometimes from multiple canid species (Egoscue 1962, Ralls and Smith 2004, Smith et al. 2005).

3

Den use is also an integral part of the daily life of a kit fox (McGrew 1979, O’Farrell and

Gilbertson 1986, Koopman et al. 1998). Kit foxes will construct multiple dens throughout their home range that they use for sleeping, pup rearing, and protection from predators (Morrell 1972,

McGrew 1979, O’Farrell and Gilbertson 1986, Koopman et al. 1998).

Areas with sparse vegetation are preferred to better detect and avoid predators (Zoellick et al. 1989). As a desert-adapted species, Kit foxes do not occur in all habitat types within their range (McGrew 1979, Moehrenschlager et al. 2004). Typical for the kit fox are the dry desert and semi-arid regions of Baja California, northern Mexico, and the bounded by California to the west, Colorado and New Mexico to the east, southern Oregon and

Idaho to the north, and southwestern Texas to the south (Fig. 2; Bueler 1973, McGrew 1979,

Sheldon 1992, Moehrenschlager et al. 2004).

The main competitor and predator of the kit fox is the Coyote (Ralls and White 1995,

Cypher and Spencer 1998, Cypher et al. 2000, Nelson et al. 2007). Coyotes are habitat generalists and opportunistic predators (Bueler 1973, Bekoff 1977, Sheldon 1992). Before

Europeans settled in North America, Coyotes mainly inhabited prairie regions, but their current range has expanded throughout all habitats, including urban areas, in Canada, the entire continental United States, and as far south as Costa Rica (Bueler 1973, Bekoff 1977, Sheldon

1992, Bekoff and Gese 2003). Coyotes are present throughout the range of the kit fox (Ralls and

White 1995, Cypher and Spencer 1998, Nelson et al. 2007). Characteristics of Coyotes such as size, weight, and pelt coloration vary slightly with geographic location and climate (Bekoff 1977,

Sheldon 1992). Overall, Coyotes exhibit a coat comprised of gray, buff, and reddish banded hairs (Bueler 1973, Bekoff 1977, Sheldon 1992). Adult male Coyotes in California, on average, weigh 11.12 kg and females weigh 9.76 kg (Hawthorne 1971, Bekoff 1977). The competitive

4

Figure 2. Typical Mojave Desert scrub vegetation dominated by Creosote Bush (Larrea tridentate) and perennial plant cover as seen within my study area in the Mojave Desert north of Barstow, California, in April 2014. (Photograph taken by California State University, Stanislaus, Endangered Species Recovery Program). nature and larger build of the Coyote makes them formidable predators to smaller canid species, including the kit fox (Ralls and White 1995, Cypher and Spencer 1998, Nelson et al. 2007).

Ecological interactions between Coyotes and Desert Kit Foxes in the Mojave Desert have not been thoroughly investigated (Zoellick et al. 1989). However, it is well documented that interference and exploitative competition occurs between Coyotes and other subspecies of kit fox, often resulting in mortality of the fox (Cypher and Scrivner 1992, Ralls and White 1995,

Cypher and Spencer 1998, Nelson et al. 2007, Kozlowski et al. 2008). Death from predation, primarily by Coyotes, accounts for 57 to 89% of all known kit fox fatalities (Berry et al. 1987,

Ralls and White 1995, White et al. 1995, Cypher et al. 2000, White et al. 2000, USFWS 2010).

It has also been documented that Coyotes will compete with Swift Foxes (V. velox), Gray Foxes

(Urocyon cinereoargenteus), and Red Foxes (V. vulpes), causing habitat avoidance, possible

5 resource competition, and potential death for the fox (Cypher 1993, Kitchen et al. 1999, Kamler et al. 2003). The intensity of competition likely varies depending upon specific habitat attributes such as resource availability, level of habitat fragmentation, and vegetation structure (White et al.

1995, Kamler et al. 2003, Nelson et al. 2007). Desert Kit Foxes may avoid Coyotes, either spatially or temporally, to lessen competitive interactions (Cypher and Spencer 1998, USFWS

2010). Also, den use by foxes is another strategy for avoiding and coexisting sympatrically with

Coyotes (White et al. 1995, Cypher and Spencer 1998, Koopman et al. 1998). Kit foxes have multiple dens within their home range, each having one or more keyhole-shaped entrances too small for a Coyote to enter (Cypher 2003, Egoscue 1962). Den use is also an essential part of everyday life for kit foxes, from predator avoidance to protection from extreme temperatures, while Coyotes may only use an excavated den to and rear pups (Bekoff 1977, Koopman et al. 1998, Cypher 2003).

Resource partitioning can also lessen the effects of competitive interactions between species (Ralls and White 1995, Cypher and Spencer 1998, Nelson et al. 2007, Kozlowski et al.

2008). Kit foxes may lessen the effects of competition by Coyotes through partitioning of food resources, either by altering the types or proportion of items consumed (White et al. 1995,

Cypher and Spencer 1998). Therefore, as with many species that occur sympatrically, resource partitioning most likely occurs as an adaptive response that leads to coexistence (White et al.

1995, Cypher and Spencer 1998).

Exploitative competition, which occurs when species indirectly compete through use of the same resources, may increase during drought years when prey abundances are lower and prey choices are limited (White et al. 1995, Cypher and Spencer 1998). Dietary overlap of competing species may increase as prey availability decreases during drought conditions (White et al.

6

1995). Also, competition between Coyotes and Desert Kit Foxes may be exacerbated by habitat loss and human encroachment as animals are forced into less space.

Although the kit fox population in the Mojave Desert has not been impacted by such widespread land conversion as the endangered San Joaquin Kit Fox has experienced in the San

Joaquin Valley, this may change in the coming decades (USFWS 1983, Cypher and Spencer

1998, USFWS 1998). Habitat development is occurring at an increasing pace in the Mojave

Desert (Lovich and Ennen 2011). In the United States, the increasing support for renewable energy facilities places large expanses of the Mojave Desert at risk for habitat loss from utility- scale solar plants (Leitner 2009, Lovich and Ennen 2011). Anthropogenic changes could result in increased stress, disease, and predation risk for kit foxes due to a rise in competition for food and space (Cypher and Spencer 1998, Nelson et al. 2007, Kozlowski et al. 2008, Clifford et al.

2013). For some kit foxes, such as the San Joaquin Kit Fox, habitat destruction is a major factor influencing population instability and decline (USFWS 1998). Because the Desert Kit Fox is a widespread subspecies, it is not thought to be threatened in its range, but the continuation of urban sprawl and renewable energy development may jeopardize their ecological standing in

California (McGrew 1979, O’Farrell and Gilbertson 1986). The only regulatory safeguard that

Desert Kit Foxes have was established in 1965 when the state in California listed them as a protected furbearer species (McGrew 1979). In recent years, Desert Kit Foxes in California have been increasingly treated as a Species of Special Concern, and in 2013 a petition was filed by the

Center for Biological Diversity to protect the subspecies under the California Endangered

Species Act (Kadaba et al., unpubl. report). The petition was ultimately rejected by the

California Department of Fish and Wildlife due in large part to a lack of information on Desert

Kit Foxes in California (D. Kadaba, pers. comm.).

7

Interest in the conservation of the Desert Kit Fox is increasing and further knowledge regarding this subspecies, including dietary patterns, food preferences, and potential competition with other species, could contribute to conservation efforts. Therefore, my main project goals were to analyze Desert Kit Fox diet over time and assess competition with Coyotes. The population of Desert Kit Foxes that was part of my study was found within a 1500-km2 study site located in the Mojave Desert north of Barstow, California (Cypher et al., in press). This area is bounded by the Fort Irwin National Training Center and the China Lake Naval Air Weapons

Station to the north and Interstate 15 and State Route 58 to the south (Fig. 3; Cypher et al., in press). From fall 2009 to summer 2014, the Endangered Species Recovery Program conducted a

Coyote food habits study at this location (Cypher et al., in press). During this time, Desert Kit

Fox scats were opportunistically collected for future analysis. The collection of over a thousand of these scats provided the opportunity for me to conduct my research project.

The objectives for my thesis were to: (1) complete a dietary analysis of Desert Kit Fox scats collected from fall 2009 to summer 2014 in natural habitats in the Mojave Desert; (2) compare diets between Desert Kit Fox and Coyotes to determine the degree of exploitative competition; (3) incorporate prey abundance survey findings from concurrent years to determine the relationship between relative prey abundance and item use by kit foxes and Coyotes; and (4) use data from camera stations deployed throughout the study area to determine the spatial overlap of Desert Kit Foxes and Coyotes as a measure of competition (Fig. 4). My goal was to investigate how Desert Kit Fox diet might vary with seasonal and annual fluctuations in the availability of foods, particularly in response to annual rainfall patterns. Also, I wanted to assess the associations of kit foxes with competitors, such as the Coyote, to better understand ecological relationships. Information obtained from this research project may assist in future management

8

Figure 3. My research area in the Mojave Desert in California (inset), which is bound by the Fort Irwin National Training Center to the north and Interstate 15 and State Route 58 to the south. The blue line denotes the boundary of my study area. (Map created by Tory Westall and Erica Kelly).

and conservation of the Desert Kit Fox. Although the Desert Kit Fox has no formal protections at this time, this may change and an increase to the available scientific literature on the subspecies would be beneficial.

My research results are presented in three chapters. I have formatted these chapters (two, three, and four) as manuscripts for the journals to which I will submit each paper. Chapter Two,

“Temporal variation in foraging patterns of Desert Kit Foxes (Vulpes macrotis arsipus) over a five-year period,” provides a strong synopsis of what Desert Kit Foxes consume over time relative to annual precipitation and prey availability. Although there have been various dietary analyses completed on the San Joaquin Kit Fox, a thorough analysis of the Desert Kit Fox diet

9

Figure 4. A) The first Desert Kit Fox scat I analyzed for this project; B) Comparison of size between a Desert Kit Fox scat (top of picture) and a Coyote scat (bottom of picture); C) Camera station in the Mojave Desert, California; D) Conducting spring prey abundance surveys. (Photographs A-C taken by Erica Kelly and photograph D taken by Christine Van Horn Job, California State University, Stanislaus, Endangered Species Recovery Program). had not been conducted until now. My third chapter, “Exploitative competition between Desert

Kit Foxes (Vulpes macrotis arsipus) and Coyotes (Canis latrans) in the Mojave Desert,” is a detailed analysis of foraging overlap between the two species and I also relate it to annual precipitation and prey availability. My study also encompasses a drought cycle, which provides a unique insight into the food habits and competitive interactions between Desert Kit Foxes and

Coyotes in a changing landscape. Chapter Four, “Landscape partitioning by Desert Kit Foxes and Coyotes in the Mojave Desert,” provides presence, relative abundance, and community

10 associations of Desert Kit Foxes based on 30 camera stations deployed in each of three years across my 1500-km2 study site. My final chapter, Chapter Five, provides a summary of the implications my results can have for Desert Kit Fox conservation. Knowing more about this unique species now is necessary to develop effective conservation and/or management strategies to ensure that the Desert Kit Fox does not decline to the point of requiring formal protections.

LITERATURE CITED

Bekoff, M. 1977. Canis latrans. Mammalian Species 79:1-9.

Bekoff, M., and E. M. Gese. 2003. Coyote (Canis latrans). Pages 467–481 in G. A. Feldhamer,

B. C. Thompson, and J. A. Chapman, editors. Wild of North America:

Biology, Management, and Conservation. The Johns Hopkins University Press,

Baltimore, Maryland.

Berry, W. H., J. H. Scrivner, T. P. O’Farrell, C. E. Harris, T. T. Kato, and P. M. McCue. 1987.

Sources and rates of mortality of the San Joaquin Kit Fox, Naval Petroleum Reserve #1,

Kern County, CA, 1980-1986. Rep. No. EGG 10282-2154, EG&G Energy

Measurements, Goleta, California.

Bueler, L. E. 1973. Wild of the World. Stein and Day Publishers, New York, New York.

Clark, Jr., H. O., G. D. Warrick, B. L. Cypher, P. A. Kelly, D. E. Williams, and D. E. Grubbs.

2005. Competitive interactions between endangered Kit Foxes and nonnative Red Foxes.

Western North American Naturalist 65:153-163.

Clifford, D., L. Woods, M. W. Gabriel, J. Rudd, E. J. Dubovi, K. Terio, F. Uzal, A. Nyaoke, A.

De La Mora, S. Diab, et al. 2013. Canine distemper outbreak in free-ranging Desert Kit

11

Foxes inhabiting a solar energy development zone. Abstract, Western Section of the

Wildlife Society 2013 Annual Conference, Reno, Nevada.

Cypher, B. L. 1993. Food item use by three sympatric canids in southern Illinois. Transactions of

the Illinois State Academy of Science 86:139-144.

Cypher, B. L. 2003. Foxes (Vulpes species, Urocyon species, and Alopex lagopus). Pages 511-

546 in G.A. Feldhamer, B. C. Thompson, and J. A. Chapman, editors. Wild Mammals of

North America: Biology, Management, and Conservation. The Johns Hopkins University

Press, Baltimore, Maryland.

Cypher, B. L., E. C. Kelly, T. L. Westall, and C. L. Van Horn Job. In press. Coyote diet patterns

in the Mojave Desert: implications for threatened Desert Tortoises. Pacific Conservation

Biology.

Cypher, B. L., and J. H. Scrivner. 1992. Coyote control to protect endangered San Joaquin Kit

Foxes at the Naval Petroleum Reserves, California. Pages 42–47 in J. E. Borrecco and R.

E. Marsh, editors. Proceedings of the Fifteenth Vertebrate Pest Conference. University of

California-Davis, Davis, California.

Cypher, B. L., and K. A. Spencer. 1998. Competitive interactions between Coyotes and San

Joaquin Kit Foxes. Journal of Mammalogy 79:204-214.

Cypher, B. L., G. D. Warrick, M. R. M. Otten, T. P. O’Farrell, W. H. Berry, C. E. Harris, T. T.

Kato, P. M. McCue, J. H. Scrivner, and B. W. Zoellick. 2000. Population dynamics of

San Joaquin Kit Foxes at the Naval Petroleum Reserves in California. Wildlife

Monographs 145:1-43.

12

Dragoo, J. W., J. R. Choate, T. L. Yates, and T. P. O’Farrell. 1990. Evolutionary and taxonomic

relationships among North American arid-land foxes. Journal of Mammalogy 71:318-

332.

Egoscue, H. J. 1962. Ecology and life history of the Kit Fox in Tooele County, . Ecology

43:481-497.

Hawthorne, V. M. 1971. Coyote movements in Sagehen Creek Basin, northeastern California.

California Fish and Game 57:154-161.

Kamler, J. F., W. B. Ballard, R. L. Gilliland, and K. Mote. 2003. Spatial relationships between

Swift Foxes and Coyotes in northwestern Texas. Canadian Journal of Zoology 81:168-

172.

Kitchen, A. M., E. M Gese., and E.R. Schauster. 1999. Resource partitioning between Coyotes

and Swift Foxes: space, time, and diet. Canadian Journal of Zoology 77:1645-1656.

Kleiman, D. G. 1967. Some aspects of social behavior in the Canidae. American Zoologist

7:365-372.

Koopman, M. E., J. H. Scriver, and T. T. Kato. 1998. Patterns of den use by San Joaquin Kit

Foxes. Journal of Wildlife Management 62:373-379.

Kozlowski, A. J., E. M. Gese, and W. M. Arjo. 2008. Niche overlap and resource partitioning

between sympatric Kit Foxes and Coyotes in the Great Basin Desert of western Utah. The

American Midland Naturalist 160:191-208.

Lariviere, S. 2002. Vulpes zerda. Mammalian Species 714:1-5.

Leitner, P. 2009. The promise and peril of solar power. Wildlife Professional 3:48-53.

Lovich, J. E. and J. R. Ennen. 2011. Wildlife conservation and solar energy development in the

desert southwest, United States. BioScience 61:982-992.

13

Macdonald, D.W., and C. Sillero-Zubiri. 2004. Dramatis personae: Wild Canids-an introduction

and dramatis personae. Pages 3-36 in D. W. Macdonald and C. Sillero-Zubiri, editors.

The Biology and Conservation of Wild Canids. Oxford University Press, Oxford, United

Kingdom.

McGrew, J. C. 1979. Vulpes macrotis. Mammalian Species 123:1-6.

Mech, L. D. 1974. Canis lupus. Mammalian Species 37:1-6.

Mercure, A., K. Ralls, K. P. Koepfli, and R. K. Wayne. 1993. Genetic subdivisions among small

canids: mitochondrial DNA differentiation of Swift, Kit, and Arctic Foxes. Evolution

47:1313-1328.

Moehrenschlager, A., B. L. Cypher, K. Ralls, R. List, and M. A. Sovada. 2004. Swift Foxes and

Kit Foxes: comparative ecology and conservation priorities of Swift and Kit Foxes. Pages

185–198 in D. W. Macdonald and C. Sillero-Zubiri, editors. The Biology and

Conservation of Wild Canids. Oxford University Press, Oxford, United Kingdom.

Morrell, S. 1972. Life history of the San Joaquin Kit Fox. California Fish and Game 58:162-174.

Nelson, J. L., B. L. Cypher, C.D. Bjurlin, and S. Creel. 2007. Effects of habitat on competition

between Kit Foxes and Coyotes. Journal of Wildlife Management 71:1467-1475.

O’Farrell, T., and L. Gilbertson. 1986. Ecology of the Desert Kit Fox, Vulpes macrotis arsipus,

in the Mojave Desert of . Bulletin of the Southern California

Academy of Sciences 85:1-15.

Ralls, K., B. Cypher, and L. K. Spiegel. 2007. Social monogamy in Kit Foxes: formation,

association, duration, and dissolution of mated pairs. American Society of Mammalogy

88:1439-1446.

14

Ralls, K., and D. A. Smith. 2004. Latrine use by San Joaquin Kit Foxes (Vulpes macrotis mutica)

and coyotes (Canis latrans). Western North American Naturalist 64:544-547.

Ralls, K., and P. J. White. 1995. Predation on San Joaquin Kit Foxes by larger canids. Journal of

Mammalogy 76:723-729.

Sheldon, J. W. 1992. Wild Dogs: The Natural History of the Nondomestic Canidae. Academic

Press, San Diego, California.

Sillero-Zubiri, C., and D. W. Macdonald. 2004. Chapter 1: Introduction. Pages 2-7 in C. Sillero-

Zubiri, M. Hoffmann, and D. W. Macdonald, editors. Canids: Foxes, , Jackals,

and Dogs. Status Survey and Conservation Action Plan. IUCN/SSC Canid Specialist

Group, Gland, Switzerland and Cambridge, United Kingdom.

Smith, D. A., K. Ralls, B. L. Cypher, and J. E. Maldonado. 2005. Assessment of scat-detection

dog surveys to determine Kit Fox distribution. Wildlife Society Bulletin 33:897-904.

Thompson, C. M., and E. M. Gese. 2007. Food webs and intraguild predation: community

interactions of a native mesocarnivore. Ecology 88:334-346.

United States Fish and Wildlife Service (USFWS). 1983. San Joaquin Kit Fox Recovery Plan.

United States Fish and Wildlife Service, Portland, Oregon.

United States Fish and Wildlife Service (USFWS). 1998. Recovery Plan for Upland Species of

the San Joaquin Valley, California. United States Fish and Wildlife Service, Portland,

Oregon.

United States Fish and Wildlife Service (USFWS). 2010. San Joaquin Kit Fox (Vulpes macrotis

mutica) 5-year review: summary and evaluation. United States Fish and Wildlife Service,

Region 8, Sacramento, California.

15

Waithman, J., and A. Roest. 1977. A taxonomic study of the Kit Fox, Vulpes macrotis. Journal of

Mammalogy 58:157-164.

Wayne, R. K., E. Geffen, and C. Vila. 2004. Population genetics: population and conservation

genetics of canids. Pages 55–84 in D. W. Macdonald and C. Sillero-Zubiri, editors. The

Biology and Conservation of Wild Canids. Oxford University Press, Oxford, United

Kingdom.

Westall, T. 2015. Parental care in urban San Joaquin Kit Foxes (Vulpes macrotis mutica):

relative roles of parents and helpers. M.S. Thesis. California State University,

Bakersfield, California.

White, P. J., W. H. Berry, J. J. Eliason, and M. T. Hanson. 2000. Catastrophic decrease in an

isolated population of Kit Foxes. Southwestern Naturalist 45:204-211.

White, P. J., K. Ralls, and C. A. Vanderbilt White. 1995. Overlap in habitat and food use

between Coyotes and San Joaquin Kit Foxes. Southwestern Naturalist 40:342-349.

Xiaoming, W., R. H. Tedford, B. Van Valkenburgh, and R. K. Wayne. 2004. Pages 8-20 in C.

Sillero-Zubiri, M. Hoffmann, and D. W. Macdonald, editors. Canids: Foxes, Wolves,

Jackals, and Dogs. Status Survey and Conservation Action Plan. IUCN/SSC Canid

Specialist Group, Gland, Switzerland and Cambridge, United Kingdom.

Zoellick, B. W., N. S. Smith, and R. S. Henry. 1989. Habitat use and movements of Desert Kit

Foxes in western Arizona. The Journal of Wildlife Management 53:955-961.

16

CHAPTER 2

Temporal variation in foraging patterns of Desert Kit Foxes (Vulpes macrotis arsipus) over a

five-year period

Erica C. Kelly1,2,3, Brian L. Cypher2, and David J. Germano1

1Department of Biology, California State University, Bakersfield, CA, 93309, USA

2Endangered Species Recovery Program, California State University, Stanislaus, P.O. Box 9622,

Bakersfield, CA, 93389, USA

3Corresponding author, email: [email protected]

ABSTRACT

Few studies regarding the Desert Kit Fox (Vulpes macrotis arsipus) have been performed and an extensive dietary analysis has never been conducted for the populations in California,

USA. Although the Desert Kit Fox in California does not have any formal protections, aside from being a protected furbearer species, interest in the conservation of this subspecies is increasing. We conducted a five-year multi-season dietary study in the Mojave Desert between the Fort Irwin National Training Center and the city of Barstow, California, USA, in which we collected and analyzed over 1,200 Desert Kit Fox scats. Desert Kit Foxes in this study specialized on heteromyid rodents, even when this preferred prey declined during a drought.

Invertebrates were also regularly consumed and frequency of occurrence of this prey type

17 increased when rodents decreased. Opportunistic items such as Desert Woodrats (Neotoma lepida), native fruits, and Pistachio (Pistacia vera) nuts from small established orchards supplemented the Desert Kit Fox diet. We conclude that Desert Kit Foxes in California are rodent and invertebrate specialists, but they seem to have the ecological plasticity to expand their diet in response to environmental changes, similar to some other canids. Management strategies, if necessary in the future, should thus include steps to maintain healthy prey populations as well as a variety of available food options in the event of declines in primary food items.

Keywords: California, Canidae, dietary analysis, food habits, Mojave Desert, scat

1. Introduction

An important aspect in understanding the biology of a species is to determine the foods that are required for their survival. Biologists have studied the feeding habits of wild animals for more than a century (Korschgen 1980). These investigations provide valuable insight into the life history of wildlife species as well as contribute to management and conservation strategies

(Kalmbach 1934, Korschgen 1980). For species for which there is conservation concern, it may be as important to protect main prey items as it is to protect the species itself. Food-habit analyses are conducted in various ways, including analyzing scat samples or stomach contents

(Bacon et al. 2011, Kalmbach 1934, Korschgen 1980, Litvaitis 2000), using stable isotope or molecular scatology methods (Angerbjörn et al. 1994, Clare et al. 2009, Deagle et al. 2005,

Gannes et al. 1997, Litvaitis 2000, Servello et al. 2005), conducting extensive VHF or GPS telemetry studies to locate predator kill sites (Bacon et al. 2011, Beier et al. 1995, Knopff et al.

18

2009), and observer-based investigations of the target species (Kalmbach 1934, Larson and Craig

2006, Litvaitis 2000).

Dietary studies that pertain to analyzing fecal samples are a common, non-invasive technique for studying wildlife food habits (Klare et al. 2011, Litvaitis 2000). For terrestrial specifically, scat analysis is the most prevalent technique used to determine prey consumption (Klare et al. 2011). By studying food habits, we can understand the effect of geographic location, season, prey availability, and prey preferences on food use (Bakaloudis et al. 2011, Korschgen 1980).

Knowledge of environmental conditions can also be an important factor for interpreting the results of dietary analysis (Anthony 1976, Cypher et al., in press, Kalmbach 1934, Korschgen

1980). For instance, precipitation likely has an impact on the foraging dynamics of animal populations (Anthony 1976, Cypher et al., in press, 1981, Spiegel et al. 1996). The difference in the amount of rainfall from wet to dry years affects vegetation and prey base, which, in turn, may impact what items are consumed and in what frequency (Anthony 1976,

Spiegel et al. 1996). By conducting prey abundance assessments, we can better discern the relationship between prey availability and animal dietary choices (Spiegel et al. 1996). Based on optimal foraging theory, not all prey species available to a predator will be consumed in the same proportions and some prey items may even be preferred over others (Lacher et al. 1982, Perry and Pianka 1997, Pyke et al. 1977, Shoener 1971).

Optimal foraging theory is a concept developed to better understand and explain animal foraging behavior, including dietary preferences (Pyke et al. 1977, Rapport 1971). The theory assumes that there is heritable variation in regards to behavior, that there are multiple foraging behaviors that an animal could exhibit, and that natural selection optimizes specific foraging

19 behaviors in animals based on individuals that contribute the most to the following generations

(Pyke et al. 1977). In general, optimal foraging theory presumes that an animal forages in a way that maximizes food intake or minimizes the time required to obtain nourishment (Perry and

Pianka 1997, Pyke et al. 1977, Schoener 1971). Based on optimal foraging theory, we should also expect that if preferred high energy prey items are more profitable (i.e., benefit of capturing prey item outweighs the cost), then a predator will consume these items over items of lower value even if the lower value prey is more abundant (Lacher et al. 1982, Pyke et al. 1977,

Shoener 1971). It is also predicted that an animal will never specialize on a less preferred food type and as the availability and as consumption of preferred prey increases then the animal will decrease their consumption of less preferred food items (Perry and Pianka 1997, Pyke et al.

1977). Lastly, individuals should either always or never consume a specific food type when they encounter it (i.e., no “partial preferences”) unless there are dietary limitations or random variations in prey availability (Pulliam 1975, Pyke et al. 1977, Rapport 1971, Rapport and Turner

1977). In more recent years, optimal foraging studies have taken into account additional life- history factors, such as predation risk and reproductive state, causing models to become more complex (Perry and Pianka 1997). Although optimal foraging theory is intricate, it is a useful tool in understanding foraging dynamics and dietary preferences of a study species (Perry and

Pianka 1997, Pyke et al. 1977).

Kit Foxes (Vulpes macrotis) are opportunistic foragers that rely heavily on rodents and insects for sustenance (Bueler 1973, Clark et al. 2005, McGrew 1979, Moehrenschlager et al.

2004, Sheldon 1992). Kangaroo rats (Dipodomys spp.), pocket mice (Perognathus and

Chaetodipus spp.), and ground squirrels (Spermophilus spp.) are common rodent prey items and

Coleopterans and Orthopterans are regularly consumed insects (Cypher et al. 2000, Koopman et

20 al. 2001, Morrell 1972, Spiegel et al. 1996). Kit Foxes are also known to consume Leporids

(Lepus californicus and Sylvilagus spp.), birds, reptiles, and a variety of non- invertebrates

(Fisher 1981, Moehrenschlager et al. 2004, Morrell 1972, Spiegel 1996, White et al. 1995).

Multiple dietary studies have been conducted on the Kit Fox, but studies regarding the

Desert Kit Fox (V. m. arsipus) populations in California, USA, are few and a large scale dietary study regarding the Desert Kit Fox in California has never been performed (Cypher and Scrivner

1992, Egoscue 1956, Kozlowski et al. 2008, Morrell 1972, Nelson et al. 2007, Spiegel 1996,

White et al. 1995). Because Desert Kit Foxes in California have not been studied extensively, little information is available on life history, including dietary preferences, of this animal.

Although the Desert Kit Fox in California does not have the added protection of being a listed threatened or endangered species, interest in the conservation of this subspecies is increasing (D.

Kadaba, pers. comm., McGrew 1979). Kit Foxes in the Mojave Desert in California have been less impacted by widespread land conversion compared to the endangered San Joaquin Kit Fox

(V. m. mutica), but continuing urban sprawl and renewable energy development may jeopardize them in the future (McGrew 1979, O’Farrell and Gilbertson 1986).

Due to a lack of life-history data on Desert Kit Foxes in California as well as increased interest in conservation of this subspecies, we conducted a multi-year dietary analysis. We opportunistically collected scats of Desert Kit Foxes in the western Mojave Desert from fall 2009 to summer 2014 to investigate temporal patterns of food item use. We also incorporated annual prey abundance data collected concurrently to assess the effects of relative item availability on

Kit Fox diet. Our objectives were to quantify annual and seasonal use of food items by Desert

Kit Foxes in California and to determine the relationship between annual item use and relative

21 prey abundance. This information will help to define relationships and identify resources important for Desert Kit Foxes.

2. Methods

2.1 Study area

We collected our data from a 1500-km2 study site located in the Mojave Desert north of

Barstow, California, USA (Cypher et al., in press). This area is bounded by the Fort Irwin

National Training Center and the China Lake Naval Air Weapons Station to the north and

Interstate 15 and State Route 58 to the south (Fig. 1). The study area was characterized as typical Mojave Desert scrub vegetation dominated by Creosote Bush (Larrea tridentata) and a ground cover ranging from 1 to 29% and consisting of a diversity of forbs and grasses (Esque et al. 2010, Turner 1994, United States Bureau of Land Management [BLM] 1980). Elevation ranged from 500-900 m and the terrain consisted of flat, dry lake beds, alluvial fans, sand dunes, steep, rugged hills, and wide expanses of land dotted with natural soil crusts and sparse vegetation cover (BLM 1980). Consistent with an arid desert environment, the mean annual precipitation for Barstow is only 13.4 cm (U.S. Climate Data 2014). Much of the study area is comprised of public lands managed by the BLM with interspersed private property. Human densities and influences were greatest around Barstow (Population 23,835) followed by the small towns of Hinkley (population 1,915) and Harvard, California (unincorporated community with fewer than 100 people; included in Newberry Springs, an unincorporated community with a population of 2,895), and declined quickly with distance from these towns (Esque et al. 2010).

2.2 Study Design

From fall 2009 to summer 2014, the Endangered Species Recovery Program (ESRP) of

California State University, Stanislaus, USA conducted a food habits study on Coyotes at this

22 location as part of an investigation of predation on Desert Tortoises (Gopherus agassizii;

(Cypher et al., in press). During that study, we opportunistically collected Desert Kit Fox scats for future dietary analysis. In arid environments, food availability and abundance is particularly influenced by precipitation (Noy-Meir 1973). In the Mojave Desert, the wet-season occurs from fall through spring. Thus, to better pair canid foraging patterns with annual prey availability, years were defined as October to September. We determined annual precipitation totals using data from U.S. Climate Data (2014). We defined the seasons as fall (Oct. to Dec.), winter (Jan. to Mar.), spring (Apr. to June), and summer (July to Sept.). To locate scats, we had a crew of two people slowly drive (i.e., < 15 kph) dirt roads within the research site for three consecutive days during each season. Each scat that we located was individually bagged in a small brown paper bag and labeled with the date and corresponding location (Lat/Long WGS 84). We also opportunistically collected scats during prey transect surveys and camera station surveys. Only fresh scats were collected; any scats that were dried out and bleached white were not used in our study.

Back in the laboratory, we heated all scats in a drying oven for 24 h at 60ºC to destroy any eggs and cysts of the zoonotic parasitic Hydatid Tapeworm (Echinococcus multilocularis;

Cypher et al., in press, Spiegel et al. 1996). After drying, we placed each scat inside a nylon pantyhose that was tied with an identification marker. We then put wrapped scats into a mesh

“delicates” laundry bag, washed them in a standard household washing machine, and dried them in a household dryer for 60-120 min. This process removed soluble material leaving undigested food item remains.

To analyze each scat, we spread the remaining undigested material on a paper towel and carefully sorted through to find identifiable food items. If there were mammalian teeth in the

23 scat, we identified the prey item down to species using published guides and reference specimens

(Glass 1981, Roest 1986). We identified other undigested mammalian remains by examining macroscopic (e.g., length, texture, color, banding patterns) hair characteristics, nail characteristics, and bone fragments, and we compared these items to published guides and reference specimens (Adorjan and Kolenosky 1969, Glass 1981, Moore et al. 1974, Roest 1986).

For other items, such as a reptile scales and insect exoskeletons, we identified the remains to the lowest taxonomic level possible. We identified fleshy fruits at least to genus based on seed characteristics (Young and Young 1992). We also classified anthropogenic items based on the presence of domestic animal remains or indigestible items (e.g., plastic, paper, foil).

Personnel of ESRP collected prey availability data during the Coyote study and we also incorporated this data into the analyses for our project (Cypher et al., in press). We used the data to assess the annual abundance of the following prey items: leporids (primarily Black-tailed

Jackrabbits, Lepus californicus, and Desert Cottontails, Sylvilagus audubonii), kangaroo rats

(Dipodomys spp.), pocket mice (Perognathus spp. and Chaetodipus spp.), and squirrels

(Xerospermophilus spp. and the White-tailed Antelope Squirrel, Ammospermophilus leucurus).

To assess annual abundance of prey, ESRP established 60 1-km transects on BLM and California

Department of Fish and Wildlife public lands throughout the study area. Their transects began approximately 25 m from a dirt road and were oriented perpendicular to the road. To increase sampling efficiency, transects were established in pairs with transects oriented parallel to each other and separated by 250 m. Pairs of transects were spaced at least 2 km apart and located in areas with typical habitat conditions for the region. ESRP clearly marked the beginning and end of each transect with a wooden stake and a global positioning system (GPS) point was taken.

24

Personnel of ESRP conducted prey abundance assessments once each spring. They performed the assessments by having two observers slowly walk along each transect. The first observer used a GPS unit to navigate to the end of the transect and also counted all active rodent burrows within 1 m of either side of the transect. ESRP characterized burrows as either large

(burrow opening ≥ 3 cm) or small (burrow opening < 3 cm). Large burrows were typical of those used by kangaroo rats or ground squirrels while small burrows were typical of those used by mice, particularly pocket mice. An inactive burrow, characterized by openings obstructed by vegetation or spider webs, was not counted by the observer. The second observer followed behind the first and counted all fresh lagomorph pellets within 1 m of either side of the transect and recorded all data. Fresh pellets were characterized by ESRP as having a golden to dark brown color and a smooth surface whereas old pellets were characterized by a gray color and a rough, weathered appearance.

2.3 Analytical methods

We determined food items found in Desert Kit Fox scats and the frequency of occurrence of each item (number of scats with a particular item divided by the total number of scats) for all years combined. For statistical analyses, we grouped items into the following seven broad categories: lagomorphs, rodent, bird, reptile, invertebrate, fruit, and anthropogenic items. We also determined the frequency of occurrence of each broad category for all years and seasons.

We used Kendall’s coefficient of concordance (W) to compare rankings of items among seasons and among years. To compare the use of each food category among seasons and among years, we conducted contingency table analyses employing a chi-square test for heterogeneity on the number of scats with and without an item in each season and year (Zar 1984). For instances where the proportions of scats with an item varied among seasons or years, we repeated the chi-

25 square test for each pair of seasons or years. For the 2 × 2 contingency tables, we applied Yates’ correction for continuity and, to adjust for an increased probability of a Type I error, we used

Hochberg’s variation on Holm’s method to correct for P-values (Legendre and Legendre 1998).

We calculated Shannon diversity indices (H’) for seasonal and annual diets by using the equation:

H’ = (N log N - ∑ni log ni)/N where N is the total number of occurrences of all items and ni is the number of occurrences of item i (Brower and Zar 1984).

We conducted Spearman-rank correlation analysis to assess annual item use relative to the availability of that item and also relative to annual precipitation. We examined the relationship between: frequency of occurrence of kangaroo rats relative to average large burrow counts, frequency of occurrence of pocket mice relative to average small burrow counts, and the frequency of occurrence of lagomorphs relative to average pellet counts. To better understand how precipitation affected food habits, we examined the relationship between annual precipitation and all of the broader food categories. We also examined the relationship between annual precipitation and Shannon diversity indices.

We used Minitab statistical software to perform all necessary statistical tests. For all statistical analyses, we considered P-values to be significant at α = 0.1. We chose a more relaxed alpha value to reduce the risk of committing a Type II error, which is considered more detrimental than a Type I error when making wildlife management decisions and within the field of conservation biology (di Stefano 2003, Scherer and Tracey 2011, Taylor and Gerrodette

1993). By relaxing the alpha value we are ensuring that we do not present results that show no difference between the consumption of certain grouped food items over years or seasons when

26 there is a true difference. Because a thorough dietary analysis of Desert Kit Fox scat has never been conducted and we do not have a detailed account of what Desert Kit Foxes in California consume over time, we believed that it is important to compensate for the possibility of failing to reject a false null hypothesis. Although the Desert Kit Fox in not currently a special status species, conservation measures may be necessary in the future and having a thorough understanding of what they consume and when is important.

3. Results

During the five-year study, we collected and analyzed 1,230 Desert Kit Fox scats (range

76-410 per year and 187-636 per season). Overall, we identified 45 different items in the scats, including , birds, various rodent species, and a few anthropogenic items (Table

1). Invertebrates, rodents, reptiles, and birds appeared to be primary food items for Desert Kit

Foxes (frequency of occurrence for all years combined > 10%). Primary invertebrate prey were

Orthopteran and Coleopteran species, solpugids, scorpions, larvae, grasshoppers, and sand treader crickets (Family Stenopelmatidae). Kangaroo rats (Dipodomys spp.) and pocket mice

(Perognathus spp. and Chaetodipus spp.) had the highest frequency of occurrence for rodent consumption. The three kangaroo rat species that potentially occurred in the study area included

Desert Kangaroo Rats (D. deserti), Merriam’s Kangaroo Rats (D. merriami), and Chisel-toothed

Kangaroo Rats (D. microps). Potential pocket mice species included Desert Pocket Mice (C. pencillatus), Long-tailed Pocket Mice (C. formosus), and Little Pocket Mice (P. longimembris).

In general, birds, lizards, snakes, and invertebrate remains were not identifiable to species.

Anthropogenic items, including nuts of Pistachios (Pistacia vera) and man-made materials, had a total frequency of occurrence of 7.97%. A number of items that we found within the scats appeared to be consumed incidentally: twigs, pieces of grass, small amounts of plant material,

27 and dirt. These items were most likely consumed while the Kit Fox was capturing an intended food item.

When we grouped items into broader food categories, rodents were the most frequently occurring items in Years 1-3, while invertebrates were the most frequently occurring items in

Years 4 and 5 (Table 2). The occurrence of birds increased yearly while the frequency of occurrence for reptiles decreased from Years 1-3 and then increased in Years 3-5. We found that fruits were only consumed in Year 4. The occurrence of anthropogenic items increased yearly and by Year 5 we found these items in > 10% of scats. When we grouped scats by season, rodents were the most frequently occurring item in the fall while invertebrates had the highest frequency of occurrence in all other seasons (Table 3). The frequency of occurrence of anthropogenic items, fruit, and lagomorphs were highest in winter, bird was highest in spring, and reptile was highest in summer.

Precipitation may have had an effect on prey abundance (Fig. 2). Annual precipitation fluctuated throughout the study with drought conditions during the last three years of the study.

For Years 1-5, annual precipitation was 16.9 cm, 28.2 cm, 7.3 cm, 7.5 cm, and 8.0 cm, respectively. Annual precipitation for this location on average is 13.2 cm. Precipitation was highest in Year 2 followed by a dramatic decrease in rainfall during the remainder of the study.

As annual precipitation changed, so did dietary patterns of the Desert Kit Foxes. As precipitation decreased in the last three years of the study, the proportional use of rodents decreased while invertebrates increased. Also, when precipitation was low, Desert Kit Foxes consumed a greater variety of items. As precipitation decreased, the proportion of anthropogenic items and birds increased. When precipitation was at its peak, so was the occurrence of rodents while the proportional use of reptiles was at its lowest. As rainfall plummeted from Years 2-5

28 we found that the proportion of invertebrates consumed showed an inverse relationship by steadily increasing through these same years. In all years, regardless of precipitation, rodents and invertebrates comprised over 60% of the items consumed.

Based on the Shannon diversity indices we calculated, Desert Kit Fox diets were less diverse in the first two years of the study and more diverse in the latter two years (Table 2).

Seasonally, Shannon diversity indices fluctuated slightly with fall being less diverse and winter more diverse than the other seasons (Table 3). In all, Desert Kit Fox diets were the most diverse in Year 5 as well as during the winter season.

Among years, the ranks of item categories exhibited significant concordance (W6 = 0.943,

X2 = 28.29, P < 0.001) indicating that the relative use of items was similar across years.

However, the proportional use of each of the item categories varied among years (Table 2). We did not include the fruit category in the proportional use analyses because Desert Kit Foxes did not eat fruit for most of the years and seasons. Rodent use by Desert Kit Foxes fluctuated substantially, from a low frequency of occurrence of 59% to a high of 96%. We found that the proportional use of invertebrates between years was significantly different except between Years

1 and 2 (Table 2).

Among seasons, the ranks of item categories also exhibited significant concordance (W6

= 0.920, X2 = 22.07, P = 0.001) indicating that the relative use of items was similar across seasons. However, only the proportional use of lagomorph and rodent categories lacked variation among seasons (Table 3). The proportional use of reptiles was significantly different for all seasons except between spring and summer (X2 = 1.60, df = 1, P = 0.206). The consumption of anthropogenic items was significantly different between winter and all other seasons (X2 = 14.61-23.11, df = 1, P = <0.001).

29

We conducted 12 combinations of Spearman-rank correlation analyses using various annual values obtained from both our data and the data produced in the previous ESRP study

(Table 4; Cypher et al., in press). From the prey transect data we collected annually over the five year study, the average number of large burrows was 50, 42, 36, 7, and 9; the average number of small burrows was 24, 19, 20, 15, 15; and the average number of lagomorph pellets per transect per year was 77, 130, 1227, 343, and 31, respectively (Cypher et al., in press). There was a strong, positive significant correlation between annual precipitation and frequency of occurrence of pocket mice (r = 1.00, t3 > 38.70, P = < 0.001). All remaining Spearman-rank correlations resulted in varying degrees of correlation, but none were significant (all P-values > 0.100).

4. Discussion

Our results suggest that Desert Kit Foxes in the Mojave Desert in California rely heavily on rodents and invertebrates for food, but will consume a variety of other items especially when the number of preferred prey is low. The primary use of rodents, specifically kangaroo rats and pocket mice, and invertebrates for this Desert Kit Fox population is similar to Desert Kit Foxes in Utah (Kozlowski et al. 2008) and San Joaquin Kit Foxes in California (Clark et al. 2005,

White et al. 1995). Contrary to some studies that have found lagomorphs prevalent in the diet of foxes (Cypher and Spencer 1998, Cypher et al. 2000, Egoscue 1962, Morrell 1972, White and

Garrott 1997), lagomorphs were not a major food item in this kit fox population. This may be due to multiple factors, including a possible low density of lagomorphs, a high number of rodents and invertebrates, or prey avoidance to potentially decrease competition with Coyotes that are also present in the area (Cypher et al., in press, Kozlowski et al. 2008, White et al. 1995).

Relative abundance of kangaroo rats and pocket mice, indicated by our annual prey transect data, decreased from Year 1-5. Rodents remained a primary food item for all years and

30 seasons, despite the apparent decline in this specific prey, suggesting that rodents are a preferred food item for Desert Kit Foxes. Kit foxes are considered specialists on heteromyid rodents

(Cypher et al. 2000, Fisher 1981, Koopman et al. 2001, Laughrin 1970, Morell 1972) and our study supports this observation. However, as the availability of this primary food item declined at our site, kit foxes increased their consumption of other food items, which is an expected response based on optimal foraging theory (Pyke et al. 1977). As the availability of rodents declined from years 3-5, there was a substantial increase in the consumption of invertebrates. An increase in invertebrate predation during drought conditions has been noted in other kit fox dietary studies (Cypher et al., unpubl. report, Spiegel et al. 1996), further indicating that this type of food provides an important supplement when rodent prey is less available. Desert Kit Foxes did increase their consumption of invertebrates, but even before the drought began the frequency of occurrence of invertebrates was that of a primary prey item (> 10% frequency of occurrence).

The use of invertebrates as a primary food source for San Joaquin Kit Foxes (Cypher et al., unpubl. report) and Desert Kit Foxes in Utah has also been reported in several recent studies

(Arjo et al. 2007, Kozlowski et al. 2008). Other arid land fox species also regularly consume invertebrates (Sheldon 1992). In our study, invertebrates were a significant component in the

Desert Kit Fox diet regardless of weather conditions, but reliance on this prey item may increase during instances of lowered rodent abundance.

The frequency of occurrence of birds, reptiles, and anthropogenic material also increased at our site when rodent numbers declined. Based on optimal foraging theory, as the preferred food type decreased (i.e., rodents), it would be expected that Desert Kit Foxes would increase their consumption of a variety of less preferred food items (Perry and Pianka 1997, Pyke et al.

1977). By the final year of our study the Mojave Desert was in the third year of below average

31 rainfall. It was during this year that the frequency of occurrence of rodents was at its lowest and the frequency of occurrence of these other items were at their highest. Also, based on the

Shannon diversity indices, Year 5 was more diverse than any other year. This likely reflected a broadening of the diet as natural food items became less abundant. If drought years continued, we would expect this effect to become more pronounced.

We also found a significant correlation between annual precipitation and frequency of occurrence of pocket mice. As annual precipitation decreased, so did the consumption of pocket mice. We did not find similar significant correlations between precipitation and other items, but this may have been because the effects of rainfall are not necessarily immediate. There can be a lag between the amount of annual precipitation and the response by prey populations to precipitation effects (Brown and Harney 1993, Cypher et al. 2000, Otten and Holmstead 1996).

However, pocket mice may exhibit a more immediate response to changes in annual rainfall patterns than other species (Cypher et al. 2000, Otten and Holmstead 1996).

The use of anthropogenic items was the highest in Year 5 and during winter. The most common anthropogenic item consumed were Pistachio nuts, which had a frequency of occurrence > 10% during both time periods. Small Pistachio orchards were present within portions of our research area. Interestingly, Desert Kit Foxes rarely ate native fruits found within our study location, although these were readily consumed by sympatric Coyotes (Cypher et al., in press). In California, Pistachio nuts tend to be harvested from late August to early September, but this is not when they were most consumed by Desert Kit Foxes (Perry and Sibbett 1998).

Instead, the foxes ate the nuts that remained after harvesting during winter, when preferred prey would be less abundant. During winter, reptiles and many invertebrates are underground in a dormant state, overall prey densities are lower, and preferred prey is harder to obtain. This may

32 cause Kit Foxes to forage more for anthropogenic items. The increased frequency of Pistachio nuts in Year 5 also may be a result of continued drought conditions in the Mojave Desert.

Similar to use of anthropogenic items during winter, Kit Foxes may have foraged more for nuts due to a decrease in available natural prey. This is analogous to San Joaquin Kit Fox diets during a study in western Kern County, California, USA where the use of human-derived foods decreased as rainfall increased (Spiegel et al. 1996).

Our results show that Desert Kit Foxes will consume agricultural and other anthropogenic items at a higher rate when natural prey abundance is altered. Therefore, the kit fox population may be receiving some amount of anthropogenic subsidization during years with decreased prey availability. Although this may help with individual survival, it may contribute to a higher kit fox population than would be expected during lean times. This can result in unnaturally high predation pressure on secondary food items and anthropogenic influences can potentially have a negative effect on the wildlife community as a whole (Cypher et al., in press,

Newsome and van Eeden 2017, Rodewald et al. 2011). Thus, as has been demonstrated between

Common Ravens (Corvus corax) and Desert Tortoises in the Mojave Desert (Boarman 1992),

Western Gullsan (Larus occidentalis) and Steelhead (Oncorhynchus mykiss) near Monterey Bay,

California (Osterback et al. 2015), and Coyotes and Desert Kit Foxes in the Great Basin Desert,

Utah, USA (Arjo et al. 2007), anthropogenic subsidization may have an impact on Desert Kit

Foxes and their predator-prey dynamics (Newsome and van Eeden 2017, Rodewald et al. 2011).

There was significant agreement in the yearly ranks of item categories used by foxes, although the proportional use of items varied among years. This result was somewhat unexpected given the marked fluctuations in precipitation and item availability. However, rainfall affects the abundance of all food categories except for anthropogenic items. Thus, the

33 fluctuations in precipitation and item availability may have been sufficient to alter proportionate use of items between years, but they may not have been sufficient to significantly alter the rankings of items. Even as proportionally the use of items varied among years, Desert Kit Foxes still primarily consumed rodents and insects.

Desert Kit Foxes did not exhibit a strong functional response to varying item availability, as evidenced by the dietary similarity among years and seasons. However, they could have experienced a numerical response, although we did not assess population trends. Many studies show that there is a positive correlation between primary prey densities and relative canid abundance (Cypher and Scrivner 1992, Cypher et al. 1994, Egoscue 1975, White and Ralls 1993,

Spiegel et al. 1996). For example, the population of San Joaquin Kit Foxes residing within the

Carrizo Plain Natural Area of central California had decreased reproductive success and thus lower Kit Fox densities during a time of decreased primary prey availability (Ralls and White

1995, White et al. 1996, White and Ralls 1993). Also, during a time of low primary prey availability for a population of Desert Kit Foxes in Utah, many of the individuals failed to reproduce (Egoscue 1975). In our study, the consumption of invertebrates increased in the dry years when vertebrates presumably were less abundant. Although invertebrates can adequately sustain adult foxes (Cypher et al., unpubl. report), these items may not be deliverable to weaning offspring in a den and might therefore result in poor reproductive success (Geffen and

Macdonald 1992, Poessel and Gese 2013, White and Ralls 1993).

The consumption of some secondary food items exhibited notable seasonal variation.

Reptile use was highest in spring and summer, and this likely was due to when these species are active. Reptiles emerge from their dormant state and are more active during the spring and summer months before reverting back to less above ground activity when daily temperatures fall.

34

Birds were also consumed the most during spring when adults are tending to the nest and young birds are vulnerable, flightless, and easier targets for Desert Kit Foxes. Also, some of the bird species in the Mojave Desert are ground nesting, making them more vulnerable to predation by mammals, such as kit foxes (Degregorio et al. 2016, Söderström et al. 1998).

Coyotes were present at this study site and are a potential competitor for Desert Kit Foxes

(Cypher et al., in press). Coyotes prey upon many of the same food items as kit foxes, increasing the likelihood for resource competition (Arjo et al. 2007, Cypher et al. 1994, Kozlowski et al.

2008, White et al. 1995). Lagomorphs tend to be the primary prey item used by Coyotes, but they also frequently consume rodents and insects (Cypher and Spencer 1998, Cypher et al. 1994,

Kozlowski et al. 2008). Desert Kit Foxes did not tend to consume lagomorphs as a primary food item at our study site even though other studies have shown them being consumed more often

(Cypher and Spencer 1998, Cypher et al. 2000, Egoscue 1962, Morrell 1972, White and Garrott

1997). This may be due to exploitative competition between Coyotes and Desert Kit Foxes within the Mojave Desert (Kelly 2017).

A greater understanding regarding the diet of Desert Kit Foxes in California is pertinent additional life-history knowledge that may contribute to future management and conservation needs of this subspecies. As human development increases in the Mojave Desert and more of the

Desert Kit Fox habitat is converted and fragmented, we may find the need for management strategies to protect these animals. Although they are not currently considered threatened or endangered, this may change in the future. Desert Kit Foxes rely primarily on rodents and insects for sustenance, but they have the ability to expand their diet and will consume a variety of items when primary items are less abundant, such as during drought conditions. Therefore, effective management strategies should include steps to maintain healthy prey populations and a

35 variety of available food options in the event of declines in primary food items. This may best be achieved by conserving large tracts of intact, high quality habitat. Management strategies should also include minimizing the availability of anthropogenic food items to discourage unnatural changes in predator-prey relationships between Desert Kit Foxes and their food items.

Acknowledgements

We would like to thank the Endangered Species Recovery Program of California State

University, Stanislaus, for providing work space and the tools necessary to prepare and analyze scats. We would like to specifically thank Alexandria Y. Madrid, Erin N. Tennant, Christine L.

Van Horn Job, and Tory L. Westall for their assistance in collecting Desert Kit Fox scats. We also thank Tory L. Westall for her assistance with graphics. We would like to further thank Paul

T. Smith for reading an earlier draft of the manuscript and providing comments and editing suggestions. Support for this project was generously provided by the Endangered Species

Recovery Program, California State University, Bakersfield, and the U.S. Fish and Wildlife

Service.

Literature Cited

Adorjan, A. S. and G. B. Kolenosky. 1969. A manual for the identification of hairs of selected

Ontario Mammals. Ontario Department of Lands and Forests, Ontario, Canada.

Angerbjörn, A., P. Hersteinsson, K. Lidén, and E. Nelson. 1994. Dietary variation in Arctic

Foxes (Alopex lagopus): an analysis of stable carbon isotopes. Oecologia 99:226-232.

Anthony, R. G. 1976. Influence of drought on diets and numbers of desert deer. Journal of

Wildlife Management 40:140-144.

36

Arjo, W. M., E. M. Gese, T. J. Bennett, and A. J. Kozlowski. 2007. Changes in Kit Fox-Coyote-

prey relationships in the Great Basin Desert, Utah. Western North American Naturalist

67:389-401.

Bakaloudis, D. E., C. G. Vlachos, M. A. Papakosta, V. A. Bontzorlos, and E. N. Chatzinikos.

2011. Diet composition and feeding strategies of the Stone (Martes foina) in a

typical Mediterranean ecosystem. Scientific World Journal 2012:1-11.

Bacon, M. M., G. M. Becic, M. T. Epp, and M. S. Boyce. 2011. Do GPS clusters really work?

Carnivore diet from scat analysis and GPS telemetry methods. Wildlife Society Bulletin

35:409-415.

Beier, P., D. Choate, and R. H. Barrett. 1995. Movement patterns of Mountain Lions during

different behaviors. Journal of Mammalogy 76:1056-1070.

Boarman, W. I. 1992. Problems with management of a native predator on a threatened species:

raven predation on Desert Tortoises. Proceedings of the Vertebrate Pest Conference

15:48-52.

Brower, J. E., and J. H. Zar. 1984. Field and Laboratory Methods for General Ecology. Wm. C.

Brown Publishers, Dubuque, Indiana.

Brown, J. H., and B. A. Harney. 1993. Populations and community ecology of heteromyid

rodents in temperate habitats. Pages 618-651 in H. H. Genoways and J. H. Brown,

editors. Biology of the Heteromyidae. American Society of Mammalogists, Special

Publication No. 10.

Bueler, L. E. 1973. Wild Dogs of the World. Stein and Day Publishers, New York, New York.

37

Clare, E. L., E. E. Fraser, H. E. Braid, M. Brock Fenton, and P. D. N. Hebert. 2009. Species on

the menu of a generalist predator, the Eastern Red Bat (Lasiurus borealis): using a

molecular approach to detect arthropod prey. Molecular Ecology 18:2532-2542.

Clark, Jr., H. O., G. D. Warrick, B. L. Cypher, P. A. Kelly, D. E. Williams, and D. E. Grubbs.

2005. Competitive interactions between endangered Kit Foxes and nonnative Red Foxes.

Western North American Naturalist 65:153-163.

Cypher, B. L., E. C. Kelly, T. L. Westall, and C. L. Van Horn Job. In press. Coyote diet patterns

in the Mojave Desert: implications for threatened Desert Tortoises. Pacific Conservation

Biology.

Cypher, B. L., and J. H. Scrivner. 1992. Coyote control to protect endangered San Joaquin Kit

Foxes at the Naval Petroleum Reserves, California. Proceedings of the Vertebrate Pest

Conference 15:42-47.

Cypher, B. L., and K. A. Spencer. 1998. Competitive interactions between Coyotes and San

Joaquin Kit Foxes. Journal of Mammalogy 79:204-214.

Cypher, B. L., K. A. Spencer, and J.H. Scrivner. 1994. Food-item use by Coyotes at the Naval

Petroleum Reserves in California. Southwestern Naturalist 39:91-95.

Cypher, B. L., G. D. Warrick, M. R. M. Otten, T. P. O’Farrell, W. H. Berry, C. E. Harris, T. T.

Kato, P. M. McCue, J. H. Scrivner, and B. W. Zoellick. 2000. Population dynamics of

San Joaquin Kit Foxes at the Naval Petroleum Reserves in California. Wildlife

Monographs 145:1-43.

Deagle, B. E., D. J. Tollit, S. N. Jarman, M. A. Hindell, A. W. Trites, and N. J. Gales. 2005.

Molecular scatology as a tool to study diet: analysis of prey DNA in scats from captive

Steller Sea Lions. Molecular Ecology 14:1831-1842.

38

Degregorio, B. A., S. J. Chiavacci, T. J. Benson, J. H. Sperry, and P. J. Weatherhead. 2016. Nest

predators of North American birds: continental patterns and implications. Bioscience

20:1-11.

Di Stefano, J. 2003. How much power is enough? Against the development of an arbitrary

convention for statistical power calculations. Functional Ecology 17:707-709.

Egoscue, H. J. 1956. Preliminary studies of the Kit Fox in Utah. Journal of Mammalogy 37:351-

357.

Egoscue, H. J. 1962. Ecology and life history of the Kit Fox in Toole County, Utah. Ecology

43:481-497.

Egoscue, H. J. 1975. Population dynamics of the Kit Fox in western Utah. Bulletin of the

Southern California Academy of Sciences 74:122-127.

Esque, T. C., K. E. Nussear, K. K. Drake, A. D. Walde, K. H. Berry, R. C. Averill-Murray, A. P.

Woodman, W. I. Boarman, P. A. Medica, J. Mack, et al. 2010. Effects of subsidized

predators, resource variability, and human population density on Desert Tortoise

populations in the Mojave Desert, USA. Endangered Species Research 12:167-177.

Fisher, J. L. 1981. Kit Fox diet in south-central Arizona. M.S. thesis, The University of Arizona,

Tucson, Arizona.

Gannes, L. Z., D. M. O’Brien, and C. Martinez Del Rio. 1997. Stable isotopes in animal ecology:

assumptions, caveats, and a call for more laboratory experiments. Ecology 78:1271-1276.

Geffen, E., and D. W. Macdonald. 1992. Small size and monogamy: spatial organization of

Blanford’s Foxes, Vulpes cana. Animal Behaviour 44:1123-1130.

Glass, B. P. 1981. Key to the skulls of North American mammals. Oklahoma State University,

Stillwater, Oklahoma.

39

Kalmbach, E. R. 1934. Field observations in economic ornithology. The Wilson Bulletin 46:73-

90.

Kelly, E. C. 2017. Desert Kit Fox (Vulpes macrotis arsipus) food habits and competitive

interactions with Coyotes (Canis latrans) in the Mojave Desert. M.S. Thesis. California

State University, Bakersfield, California.

Klare, U., J. F. Kamler, and D. W. Macdonald. 2011. A comparison and critique of different

scat-analysis methods for determining diet. Review 41:294-312.

Knopff, K. H., A. A. Knopff, M. B. Warren, and M. S. Boyce. 2009. Evaluating global

positioning system telemetry techniques for estimating predation parameters.

Journal of Wildlife Management 73:586-597.

Koopman, M. E., B. L. Cypher, and D. R. McCullough. 2001. Factors influencing space and prey

use by San Joaquin Kit Foxes. Transactions of the Western Section of The Wildlife

Society 37:77-83.

Korschgen, L .J. 1980. Procedures for food-habit analyses. Pages 113-127 in S. D. Schemnitz,

editor. Wildlife Management Techniques Manual. The Wildlife Society, Washington,

D.C..

Kozlowski, A. J., E. M. Gese, and W. M. Arjo. 2008. Niche overlap and resource partitioning

between sympatric Kit Foxes and Coyotes in the Great Basin Desert of western Utah. The

American Midland Naturalist 160:191-208.

Lacher, T. E., Jr., M. R. Willig, and M. A. Mares. 1982. Food preference as a function of

resource abundance with multiple prey types: an experimental analysis of optimal

foraging theory. The American Naturalist 120:297-316.

40

Larson, K., and D. Craig. 2006. Digiscoping vouchers for diet studies in bill-loaded holding

birds. Waterbirds: The International Journal of Waterbird Biology 29:198-202.

Laughrin, L. 1970. San Joaquin Kit Fox: its distribution and abundance. California Department

of Fish and Game, Wildlife Management Branch, Administrative Report 70-2,

Sacramento, California.

Legendre, P., and L. Legendre. 1998. Numerical Ecology, second edition. Elsevier Science,

Amsterdam, The Netherlands.

Litvaitis, J. A. 2000. Investigating food habits of terrestrial vertebrates. Pages 165-190 in L.

Boitani and T. K. Fuller, editors. Research techniques in animal ecology: controversies

and consequences. Columbia University Press, New York, New York.

McGrew, J. C. 1979. Vulpes macrotis. Mammalian Species 123:1-6.

Moehrenschlager, A., B. L. Cypher, K. Ralls, R. List, and M. A. Sovada. 2004. Swift Foxes and

Kit Foxes: comparative ecology and conservation priorities of Swift and Kit Foxes. Pages

185–198 in D. W. Macdonald and C. Sillero-Zubiri, editors. The Biology and

Conservation of Wild Canids. Oxford University Press, Oxford, United Kingdom.

Moore, T. D., L. E. Spence, C. E. Dugnolle. 1974. Identification of the dorsal guard hairs of

some mammals of Wyoming. W.G. Hepworth, editor. Wyoming Game and Fish

Department, Bulletin No. 14, Cheyenne, Wyoming.

Morrell, S. 1972. Life history of the San Joaquin Kit Fox. California Fish and Game 58:162-174.

Nelson, J. L., B. L. Cypher, C. D. Bjurlin, and S. Creel. 2007. Effects of habitat on competition

between Kit Foxes and Coyotes. Journal of Wildlife Management 71:1467-1475.

Newsome, T. M., and L. M. van Eeden. 2017. The effects of food waste on wildlife and humans.

Sustainability 9:1-9.

41

Noy-Meir, I. 1973. Desert ecosystems: environments and producers. Annual Review of Ecology

and Systematics 4:25-51.

O’Farrell, T., and L. Gilbertson. 1986. Ecology of the Desert Kit Fox, Vulpes macrotis arsipus,

in the Mojave Desert of southern California. Bulletin of the Southern California

Academy of Sciences 85:1-15.

Osterback, A. K., D. M. Frechette, S. A. Hayes, S. A. Shaffer, and J. W. Moore. 2015. Long-

term shifts in anthropogenic subsidies to gulls and implications for imperiled fish.

Biological Conservation 191:606-613.

Otten, M. R. M., and G. L. Holmstead. 1996. Effect of seeding burned lands on the abundance of

rodents and leporids on Naval Petroleum Reserve No. 1, Kern County, California.

Southwestern Naturalist 41:129-135.

Perry, G., and E. R. Pianka. 1997. Animal foraging: past, present, and future. Trends in Ecology

and Evolution 12:360-364.

Perry, E., and G. S. Sibbett. 1998. Harvesting and storing your home orchard’s nut crop:

almonds, walnuts, pecans, pistachios, and chestnuts. University of California, Division of

Agriculture and Natural Resources, Publication 8005, Oakland, California, USA.

Poessel, S. A., and E. M. Gese. 2013. Den attendance patterns in Swift Foxes during pup rearing:

varying degrees of parental investment within the breeding pair. Journal of Ethology

31:193-201.

Pulliam, H. R. 1975. Diet optimization with nutrient constraints. American Naturalist 109:765-

768.

Pyke, G. H., H. R. Pulliam, and E.L. Charnov. 1977. Optimal foraging: a selective review of

theory and tests. Quarterly Review of Biology 52:137-154.

42

Ralls, K., and P. J. White. 1995. Predation on San Joaquin Kit Foxes by larger canids. Journal of

Mammalogy 76:723-729.

Rapport, D. J. 1971. An optimization model of food selection. American Naturalist 105:575-587.

Rapport, D. J., and J. E. Turner. 1977. Economic models in Ecology. Science 195:367-373.

Rodewald, A. D., L. J. Kearns, and D. P. Shustack. 2011. Anthropogenic resource subsidies

decouple predator-prey relationships. Ecological Applications 21:936-943.

Roest, A. I. 1986. A key-guide to mammal skulls and lower jaws. Mad River Press, Eureka,

California.

Scherer, R. D., and J. A. Tracey. 2011. A power analysis for the use of counts of egg masses to

monitor Wood Frog (Lithobates sylvaticus) populations. Herpetological Conservation and

Biology 6:81–90.

Schoener, T. W. 1971. Theory of feeding strategies. Annual Review of Ecology and Systematics

11:369-404.

Sheldon, J. W. 1992. Wild Dogs: The Natural History of the Nondomestic Canidae. Academic

Press, San Diego, California.

Servello, F. A., E. C. Hellgren, and S. R. McWilliams. 2005. Techniques for wildlife nutritional

ecology, sixth edition. Pages 554-590 in C. E. Braun, editor. Techniques for Wildlife

Investigations and Management. The Wildlife Society, Bethesda, Maryland.

Söderström, B., T. Pärt, and J. Rydén. 1998. Different nest predator faunas and nest predation

risk on ground and shrub nests at forest ecotones: an experiment and a review. Oecologia

117:108-118.

Spiegel, L., B. L. Cypher, and T. C. Dao. 1996. Diet of the San Joaquin Kit Fox (Vulpes macrotis

mutica) at three sites in western Kern County, California. Pages 39-52 in L. Spiegel,

43

editor. Studies of San Joaquin Kit Fox in undeveloped and oil developed areas. California

Energy Commission, Sacramento, California.

Taylor, B. L., and T. Gerrodette. 1993. The uses of statistical power in conservation biology: the

vaquita and northern spotted owl. Conservation Biology 7:489-500.

Turner, R. M. 1994. Mojave Desert scrub. Pages 157-168 in D. E. Brown, editor. Biotic

communities, southwestern United States and northwestern Mexico. University of Utah

Press, Salt Lake City, Utah.

United States Bureau of Land Management (BLM). 1980. The California Desert conservation

area plan, 1980. U.S. Bureau of Land Management, California Desert District, Riverside,

California.

United States Climate Data. 2014. Precipitation data for Barstow, California.

http://www.usclimatedata.com/climate/barstow/california/united-states/usca0069/2014/8

Accessed September 2014.

White, P. J., and R. A. Garrott. 1997. Factors regulating Kit Fox populations. Canadian Journal

of Zoology 75:1982-1988.

White, P. J., and K. Ralls. 1993. Reproduction and spacing patterns of Kit Foxes relative to

changing prey availability. Journal of Wildlife Management 57:861-867.

White, P. J., K. Ralls, and C. A. Vanderbilt White. 1995. Overlap in habitat and food use

between Coyotes and San Joaquin Kit Foxes. Southwestern Naturalist 40:342-349.

White, P. J., C. A. Vanderbilt White, and K. Ralls. 1996. Functional and numerical responses of

Kit Foxes to a short-term decline in mammalian prey. Journal of Mammalogy 77:370-

376.

44

Young, J. A., and C. G. Young. 1992. Seeds of woody plants in North America. Dioscorides

Press, Portland, Oregon.

Zar, J. H. 1984. Biostatistical analysis. Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

45

Table 1. Frequency of occurrence of food items found in Desert Kit Fox (Vulpes macrotis arsipus) scats collected in the Mojave Desert, California, USA, for all years combined (October

2009 to September 2014). Primary food items (> 10% frequency of occurrence) are in bold.

Frequency of occurrence (%)

All Years

Food item (n = 1230)

Kangaroo rats (Dipodomys spp.) 30.16

Pocket mice (Perognathus spp. and Chaetodipus spp.) 20.65

Deer Mouse (Peromyscus maniculatus) 0.57

House Mouse (Mus musculus) 0.24

Western Harvest Mouse (Reithrodontomys megalotis) 0.08

Squirrels (Xerospermophilus spp. and Ammospermophilus leucurus) 3.25

Desert Woodrat (Neotoma lepida) 0.49

Valley Pocket Gopher (Thomomys bottae) 0.08

Unknown rodent 22.36

Lagomorphs (Lepus californicus and Sylvilagus audubonii) 9.02

Unknown small mammal 2.52

46

Unknown mammal 2.52

Birds (Class Aves) 14.47

Snakes (Order Squamata) 8.86

Common Chukwalla (Sauromalus ater) 0.08

Desert Iguana (Dipsosaurus dorsalis) 0.08

Desert Horned Lizard (Phrynosoma platyrhinos) 0.41

Other lizards (Order Squamata) 10.65

Desert Tortoise (Gopherus agassizii) 0.57

Unknown reptile 1.46

Eggshells 0.41

Unknown vertebrate 1.95

Field crickets (Family Gryllidae) 0.16

Jerusalem crickets (Family Stenopelmatidae) 6.34

Sand treader crickets (Family Stenopelmatidae) 10.24

Grasshoppers (Order Orthoptera) 11.63

Unknown Orthopterans 12.36

Scarab beetles (Family Scarabaeoidea) 9.35

47

Tenebrionid beetles (Family Tenebrionidae) 3.74

Weevils (Family Curculionidae) 0.08

Other beetles (Order Coleoptera) 27.56

Earwigs (Forficula auricularia) 0.41

True bugs (Order Hemiptera) 0.16

Unknown insect 26.91

Bot fly larvae (Family Oestridae) 0.08

Larvae 12.85

Solpugids (Order Solifugae) 20.08

Scorpions (Order Scorpiones) 11.38

Invertebrate 0.24

Boxthorn (plant) (Lycium spp.) 0.08

Plant material 0.65

Pistachio nuts (plant) (Pistacia vera) 6.10

Domestic animal (pets, farm animals) 0.16

Domestic animal gut contents/feces 0.41

Man-made material (cloth, paper) 2.11

48

Table 2. Annual frequency of occurrence for item categories in Desert Kit Fox (Vulpes macrotis arsipus) scats collected in the Mojave Desert, California, USA, during October 2009 to

September 2014. Years span October-September. P-values in bold are significant.

Frequency of occurrence (%)

Year 1 Year 2 Year 3 Year 4 Year 5

2 Item category (n = 127) (n = 76) (n = 229) (n = 388) (n = 410) Χ 6 P

Lagomorphs 1.57B 1 3.95 AB 6.99 AB 13.40 A 9.27 A 21.22 <0.001

Rodent 92.91 A 96.05 A 89.96 A 62.37 B 59.27 B 135.88 <0.001

Bird 7.09 B 9.21 B 10.92 B 12.11 B 21.95 A 29.91 <0.001

Reptile 23.62 A 7.89 B 12.23 B 20.62 A 25.85 A 25.14 <0.001

Invertebrate 53.54 D 46.05 D 69.00 C 77.58 B 83.66 A 82.93 <0.001

Fruit 0.00 0.00 0.00 0.26 0.00 n/a n/a

Anthropogenic 0.79 B 1.32 B 2.18 B 9.79 A 12.93 A 39.48 <0.001

Diversity2 0.50 0.48 0.54 0.64 0.66

1 Means with similar letters were not significantly different. 2 Shannon diversity index

49

Table 3. Seasonal frequency of occurrence for item categories in Desert Kit Fox (Vulpes macrotis arsipus) scats collected in the Mojave Desert, California, USA, during October 2009 to

September 2014. Seasons were defined as Fall = October-December; Winter = January-March;

Spring = April-June; Summer = July-September. P-values in bold are significant.

Frequency of occurrence (%)

Fall Winter Spring Summer

2 Item category (n = 209) (n = 636) (n = 198) (n = 187) Χ 6 P

Lagomorphs 7.18 A1 10.69 A 5.56 A 9.09 A 5.93 0.115

Rodent 74.64 A 71.38 A 71.72 A 69.52 A 1.36 0.715

Bird 9.57 B 14.47 AB 19.19 A 14.97 AB 7.66 0.054

Reptile 18.18 B 12.58 C 31.31 A 37.43 A 72.72 <0.001

Invertebrate 65.55 C 72.17 BC 76.26 AB 84.49 A 19.77 <0.001

Fruit 0.00 0.16 0.00 0.00 n/a n/a

Anthropogenic 3.83 B 13.36 A 1.52 B 1.07 B 53.53 <0.001

Diversity2 0.58 0.63 0.60 0.60

1 Means with similar letters were not significantly different. 2 Shannon diversity index

50

Table 4. Spearman-rank correlations analysis on various annual values collected from the

Mojave Desert, California, USA from October 2009 to September 2014. Annual frequency of occurrence of prey items, annual precipitation in cm, average annual large and small burrows, and average annual pellet counts were used for the various analysis combinations. P-values in bold are significant.

Annual values tested rs t P

Kangaroo rat vs. large burrows 0.60 1.30 0.285

Pocket mice vs. small burrows 0.20 0.35 0.747

Lagomorphs vs. pellets 0.20 0.35 0.747

Precipitation vs. kangaroo rat -0.20 -0.35 0.873

Precipitation vs. pocket mice 1.00 > 38.70 < 0.001

Precipitation vs. lagomorphs -0.60 -1.30 0.873

Precipitation vs. bird -0.50 -1.00 0.873

Precipitation vs. reptile -0.10 -0.17 0.873

Precipitation vs. invertebrate -0.60 -1.30 0.873

Precipitation vs. fruit -0.35 -0.65 0.873

Precipitation vs. anthropogenic -0.50 -1.00 0.873

Precipitation vs. H’ -0.60 -1.30 0.873

51

Figure 1. Study area located in the Mojave Desert in California, USA (inset), which is bound by the Fort Irwin National Training Center to the north and Interstate 15 and State Route 58 to the south. The blue boundary is the specific study area and the black dots denote every location a scat or multiple scats were collected during the five year project. (This is an original map created for this manuscript).

Figure 2. Yearly proportions of grouped items with the annual precipitation trend from the

Mojave Desert, California, USA, during October 2009 to September 2014.

52

Figure 1

53

Figure 2

1000/o 30 ,-....._ ~ 90% 25 s s 80% (.) <1) "-" ~ 70% . 20 ~ ~ ·- 60% (.) 0 ·-<1) ~ 50% 15 ~ 0 40% ~ ·-t 10 ca 0 30% -;::::$ ~ 20% ~ 0 5 ~ 10% ~ ~ 00/o 0 1 2 3 4 5 Year - Rodent - Invertebrate - Reptile - Bird

- Rabbit - Fruit - Anthropogenic - Annual precip.

54

1 CHAPTER 3 2 Erica Kelly 3 Endangered Species Recovery Program 4 P.O. Box 9622 5 Bakersfield, California 93389 6 [email protected]

7

8 Running Header: Competition between sympatric canids

9

10 Exploitative competition between desert kit foxes and coyotes in the Mojave Desert 11

12 Erica C. Kelly,* Brian L. Cypher, and David J. Germano

13

14 Department of Biology, California State University, Bakersfield, California 93309 (ECK, DJG)

15 Endangered Species Recovery Program, California State University, Stanislaus, P.O. Box 9622,

16 Bakersfield, California 93389 (ECK, BLC)

17

18 Exploitative competition is one of the potential interactions between species that may

19 have an influence on the complex dynamics of an ecological community. We assessed potential

20 exploitative competition between desert kit foxes (Vulpes macrotis arsipus) and coyotes (Canis

21 latrans) by comparing food habits of the 2 species from 2009–2014 on a study site in the Mojave

22 Desert in California, United States. Desert kit foxes in this study specialized on heteromyid

23 rodents and insects while the most frequently occurring items consumed by coyotes were rodents

24 and lagomorphs. Both species did consume a variety of prey species throughout the study, but

25 they did so in varying frequencies depending on the year and season. Also, precipitation most

26 likely had an effect on prey abundance and thus prey consumption by the 2 species. Our results

55

27 showed that overall, the diets of desert kit foxes and coyotes were very similar, which indicates

28 the possibility of exploitative competition. Desert kit foxes consistently had lower dietary

29 diversity than coyotes, indicating that desert kit foxes are more specialist consumers while

30 coyotes are more generalists when it comes to food item selection. Dietary specialization by kit

31 foxes, especially on smaller available items, may help reduce competitive pressure from coyotes

32 and lead to sympatric coexistence. Both of our study species also consumed anthropogenic

33 material, but coyotes did so at much higher frequencies than desert kit foxes. This may cause

34 increased competitive pressures on desert kit foxes especially as habitat conversion and human

35 influences increase in the Mojave Desert. Desert kit foxes in California are not a listed protected

36 species, but management strategies may be necessary in the future as human encroachment into

37 the desert increases. Management strategies should not only include steps to maintain healthy

38 prey populations and a variety of available food options, but also should include measures to

39 limit increased competition from coyotes resulting from anthropogenic subsidization.

40

41 Key words: Canis latrans, food habits, resource partitioning, sympatric, Vulpes macrotis arsipus

42

43 *Correspondent: [email protected]

44

45 Interactions between species, including competition, are one of the many factors that influence

46 the complex dynamics of an ecological community. Not all species within a community will

47 have an effect on each other through competition, but if they do the outcome can be detrimental

48 to 1 or both of the species involved (Pianka 1978). Competition is not an ecologically

49 advantageous relationship for animals, and thus it is an important driver in the evolution of

56

50 strategies that result in niche separation through specialization on different resources or

51 diversification in resource use (Pianka 1978). Exploitative competition, which occurs when 2

52 species are using the same resources, is 1 of the 2 primary forms of competition that can occur

53 between sympatric species, with the other being interference competition (Pianka 1978; Maurer

54 1984; White et al. 1995).

55 Coyotes (Canis latrans) are the main predator and competitor of the kit fox (Vulpes

56 macrotis) and occur sympatrically across much of the range of kit foxes, including the Mojave

57 Desert in California, United States (Cypher and Spencer 1998; Cypher 2003; Arjo et al. 2007).

58 Both canids consume many of the same foods, resulting in the potential for exploitative

59 competition. The competitive exclusion principal emphasizes that no 2 competing species can

60 reside in the same community if they both rely on the same resources (Lotka 1925; Volterra

61 1926; Gause 1934; Hardin 1960). If this event occurs, then the weaker competitor will become

62 extinct, at least within the territory of the more dominant species (Hardin 1960). For coyotes and

63 desert kit foxes (V. m. arsipus) to occur together there must be sufficient differentiation in their

64 ecological niches.

65 Resource partitioning is one strategy that can decrease the effects of competitive

66 interactions between species (Cypher and Spencer 1998; Nelson et al. 2007; Kozlowski et al.

67 2008). Kit foxes may lessen the effects of competition with coyotes through partitioning of

68 available prey by either altering the types or proportions of items consumed (White et al. 1995;

69 Cypher and Spencer 1998). For example, a study conducted at the Naval Petroleum Reserves in

70 California found that San Joaquin kit foxes (V. m. mutica) in this population consumed many of

71 the same prey items as coyotes, but did so in different proportions, reflecting differences in

72 preferred prey (Cypher and Spencer 1998). Conversely, in the Great Basin Desert of Western

57

73 Utah, United States, resource partitioning between desert kit foxes and coyotes was not very

74 pronounced (Kozlowski et al. 2008). Competitive pressures on the kit fox population in this area

75 were very high, although this may have been mitigated somewhat by spatial partitioning

76 (Kozlowski et al. 2008).

77 The intensity of exploitative competition through use of common food items is likely to

78 increase during times of decreased resource availability (White et al. 1995; Cypher and Spencer

79 1998). For instance, exploitative competition may become exacerbated during drought years

80 when prey abundances are lower and prey choices are limited (White et al. 1995; Cypher and

81 Spencer 1998). Results from a multi-year study of coyotes and San Joaquin kit foxes on the

82 National Monument, California, indicated that dietary overlap may increase as prey

83 availability decreases during drought conditions (White et al. 1995). Competition between

84 coyotes and desert kit foxes may also be heightened by habitat loss and human encroachment as

85 animals are forced into smaller habitat patches.

86 For the San Joaquin kit fox, a federally listed endangered and state listed threatened

87 subspecies, habitat destruction is a major factor influencing population instability and decline

88 (United States Fish and Wildlife Service 1998). While the Mojave Desert kit fox population has

89 not yet been impacted by such widespread land conversion as the San Joaquin kit fox, this may

90 change in the coming decades (United States Fish and Wildlife Service 1983; Cypher and

91 Spencer 1998; United States Fish and Wildlife Service 1998). In the United States, the

92 increasing support for renewable energy facilities places large expanses of the Mojave Desert at

93 risk for habitat loss from utility-scale solar plants (Leitner 2009; Lovich and Ennen 2011).

94 Anthropogenic changes could result in increased stress, disease, and predation risk for kit foxes

58

95 due to a rise in competition for food and space (Nelson et al. 2007; Kozlowski et al. 2008;

96 Clifford et al. 2013).

97 Currently, desert kit foxes in California are considered a common subspecies and

98 therefore lack the protections granted to the San Joaquin kit fox (United States Fish and Wildlife

99 Service 1998). The desert kit fox is state listed as endangered in Colorado, state listed as

100 threatened in Oregon, and listed as endangered in Mexico, whereas populations in California

101 have no formal protections aside from being a protected furbearer species (United States Fish

102 and Wildlife Service 1998; Moehrenschlager et al. 2004). However, in recent years desert kit

103 foxes in California have been increasingly treated as a Species of Special Concern, and in 2013 a

104 petition was filed by the Center for Biological Diversity to protect the subspecies under the

105 California Endangered Species Act (Kadaba et al., unpubl. report). The petition was ultimately

106 rejected by the California Department of Fish and Wildlife due to lack of literature regarding

107 desert kit foxes in California (D. Kadaba, pers. comm.).

108 Multiple studies on competitive interactions between coyotes and kit foxes have been

109 conducted, but none of these have addressed interactions between coyotes and the desert kit fox

110 populations in California. Interest in the conservation of the desert kit fox is increasing and

111 additional information on factors affecting this subspecies, including potential exploitative

112 competition with sympatric coyotes, could contribute to future conservation efforts. Our

113 extensive collection of coyote and desert kit fox scats from the Mojave Desert in California

114 allowed us to study competitive interactions through prey consumption. Our objectives were to

115 compare diets between desert kit foxes and coyotes to determine the degree of dietary overlap

116 and to assess prey use relative to prey abundance to determine the effects of food availability on

117 competition intensity between desert kit foxes and coyotes.

59

118 MATERIALS AND METHODS

119 Study area.—We collected our data from a 1500-km2 study site located in the Mojave

120 Desert north of Barstow, California (Cypher et al., in press). This area is bounded by the Fort

121 Irwin National Training Center and the China Lake Naval Air Weapons Station to the north and

122 Interstate 15 and State Route 58 to the south (Fig. 1). The study area was characterized as

123 typical Mojave Desert scrub vegetation dominated by Creosote Bush (Larrea tridentata) and a

124 ground cover of subshrubs, forbs, and grasses ranging from 1–29% (United States Bureau of

125 Land Management 1980; Turner 1994; Esque et al. 2010). Elevation ranged from 500–900 m

126 and the terrain consisted of flat, dry lake beds, alluvial fans, sand dunes, steep, rugged hills, and

127 wide expanses of land dotted with natural soil crusts and sparse vegetation cover (United States

128 Bureau of Land Management 1980). Consistent with an arid desert environment, the mean

129 annual precipitation for Barstow is only 13.4 cm (United States Climate Data 2014). Much of

130 the study area is comprised of public lands managed by the United States Bureau of Land

131 Management with interspersed private property. Human densities and influences were greatest

132 around Barstow (Population 23,835) followed by the small towns of Hinkley, California

133 (Population 1,915), and Harvard, California (unincorporated community with population < 100;

134 included in Newberry Springs, California (unincorporated community with a population of

135 2,895), and declined quickly with distance from these towns (Esque et al. 2010).

60

136 Study design.—From Fall 2009 to Summer 2014, we collected coyote and desert kit fox

137 scats as part of an investigation of coyote predation on desert tortoises (Gopherus agassizii;

138 Cypher et al., in press). In arid environments, food availability and abundance is particularly

139 influenced by precipitation (Noy-Meir 1973). In the Mojave Desert, the wet-season occurs from

140 fall through spring. Thus, to better pair canid foraging patterns with annual prey availability,

141 years were defined as October to September. We determined annual precipitation totals using

142 data from United States Climate Data (2014). We defined the seasons as fall (October to

143 December), winter (January to March), spring (April to June), and summer (July to September).

144 To locate scats, a crew of 2 people slowly drove (i.e., < 15 kph) along dirt roads within the

145 research site for 3 consecutive days during each season. Each scat that we located was

146 individually bagged in a small brown paper bag and labeled with the date and corresponding

147 location (Lat/Long WGS 84). We also opportunistically collected scats during prey transect

148 surveys and camera station surveys. Only fresh scats were collected; any scats that were dried

149 out and bleached white were not used in our study.

150 Back in the laboratory, we heated all scats in a drying oven for 24 hours at 60ºC to

151 destroy any eggs and cysts of the zoonotic parasitic hydatid tapeworm (Echinococcus

152 multilocularis) (Spiegel et al. 1996; Cypher et al., in press). After drying, we placed each scat

153 inside a nylon pantyhose that was tied with an identification marker. We then put wrapped scats

154 into a mesh laundry bag, washed them in a standard household washing machine, and dried them

155 in a household dryer for 60–120 minutes. This process removed soluble material leaving

156 undigested food item remains.

157 To analyze each scat, we spread the remaining undigested material from each scat on a

158 paper towel and carefully sorted through to find different food items. If there were mammalian

61

159 teeth in the scat, we identified the prey item down to species using published guides and

160 reference specimens (Glass 1981; Roest 1986). We identified other undigested mammalian

161 remains by examining macroscopic (e.g., length, texture, color, banding patterns) hair

162 characteristics, nail characteristics, and bone fragments, and we compared these items to

163 published guides and reference specimens (Adorjan and Kolenosky 1969; Moore et al. 1974;

164 Glass 1981; Roest 1986). For other items, such as a reptile scales and insect exoskeletons, we

165 identified the remains to the lowest taxonomic level possible. We identified fleshy fruits at least

166 to genus based on seed characteristics (Young and Young 1992). We also classified

167 anthropogenic items based on the presence of domestic animal remains or indigestible items

168 (e.g., plastic, paper, foil).

169 As part of the investigation of predation on desert tortoises, we collected prey availability

170 data (Cypher et al., in press). We used these data to assess the annual abundance of the

171 following prey items: leporids (primarily black-tailed jackrabbits, Lepus californicus, and desert

172 cottontails, Sylvilagus audubonii), kangaroo rats (Dipodomys spp.), pocket mice (Perognathus

173 spp. and Chaetodipus spp.), and squirrels (Xerospermophilus spp. and the white-tailed antelope

174 squirrel, Ammospermophilus leucurus). To assess prey abundance, we conducted annual surveys

175 along 60 1-km transects on United States Bureau of Land Management and California

176 Department of Fish and Wildlife public lands throughout the study area. The transects began

177 approximately 25 m from a dirt road and were oriented perpendicular to the road. To increase

178 sampling efficiency, transects were established in pairs with transects oriented parallel to each

179 other and separated by 250 m. Pairs of transects were spaced at least 2 km apart and located in

180 areas with typical habitat conditions for the region. We clearly marked the beginning and end of

181 each transect with a wooden stake and a global positioning system (GPS) point was taken.

62

182 We conducted prey abundance assessments once each spring. We performed the

183 assessments by having 2 observers slowly walk along each transect. The first observer used a

184 GPS unit to navigate to the end of the transect and also counted all active rodent burrows within

185 1 m of either side of the transect. We characterized burrows as either large (burrow opening ≥ 3

186 cm) or small (burrow opening < 3 cm). Large burrows were typical of those used by kangaroo

187 rats or ground squirrels while small burrows were typical of those used by mice, particularly

188 pocket mice. Inactive burrows, characterized by openings obstructed by vegetation or spider

189 webs, were not counted. The second observer followed behind the first and counted all fresh

190 lagomorph pellets within 1 m of either side of the transect and recorded all data. Fresh pellets

191 were characterized as having a golden to dark brown color and a smooth surface whereas old

192 pellets were characterized by a gray color and a rough, weathered appearance.

193 Analytical methods.—Kelly (2017) determined the food items found in desert kit fox

194 scats and the frequency of occurrence (FOO) of each item (number of scats with a particular item

195 divided by the total number of scats) for all years combined while Cypher et al. (in press) did the

196 same with coyote scats. For statistical analyses, we grouped items into the following 7 broad

197 categories: lagomorphs, rodent, bird, reptile, invertebrate, fruit, and anthropogenic items. We

198 calculated Shannon diversity indices (H’) for both desert kit foxes and coyotes for each year,

199 each season by year, and the total project length by using the equation:

200 H’ = (N log N - ∑ni log ni)/N

201 where N is the total number of occurrences of all items and ni is the number of occurrences of

202 item i (Brower and Zar 1984). We used a two-sample t-test, after determining parametric

203 assumptions were met, to compare both mean annual and seasonal H’ between species. We also

204 determined Horn’s index of similarity between both species for each year, for each season, and

63

205 for all years combined (Horn 1966). To compare mean Horn’s index of similarity values

206 between seasons, we used a 1-way analysis of variance (ANOVA) with multiple range tests if a

207 significant difference was found.

208 We performed Spearman-rank correlation analysis to compare rankings of items between

209 species for each season, year, and the total project length. To examine the effect of annual

210 precipitation, we conducted Spearman-rank correlation analysis on precipitation vs. annual

211 Spearman-rank correlation coefficients and precipitation vs. annual Horn index of similarity

212 values. We also compared annual Horn indices to annual counts of rabbit pellets, large holes,

213 and small holes using Spearman-rank correlation analyses.

214 To adjust for an increased probability of a Type I error, we used Hochberg’s variation on

215 Holm’s method to correct for P-values when necessary (Legendre and Legendre 1998). Also, for

216 all statistical analyses, we considered P-values to be significant at α = 0.1. We chose a more

217 relaxed alpha value to reduce the risk of committing a Type II error, which is considered more

218 detrimental than a Type I error when making wildlife management decisions and within the field

219 of conservation biology (Taylor and Gerrodette 1993; di Stefano 2003; Scherer and Tracey

220 2011). By relaxing the alpha value, we aimed to reveal potential ecological relationships that

221 can be more fully explored through further investigation. We used Minitab statistical software to

222 perform all necessary statistical tests.

223 RESULTS

224 During the 5 year study, we collected and analyzed 1,230 desert kit fox scats (range 76–410 per

225 year and 187–636 per season). Also, 3,246 coyote scats (range 474–801 per year and 738–845

226 per season) were collected concurrently and analyzed (Cypher et al., in press). Overall, 45

227 different items were identified in the desert kit fox scats and 50 distinct items were identified in

64

228 the coyote scats (Kelly 2017; Cypher et al., in press). We found many of the same items in the

229 scats of both species.

230 Desert kit foxes and coyotes both regularly preyed upon rodents. The 3 kangaroo rat

231 species that potentially occurred in the study area included desert kangaroo rats (D. deserti),

232 Merriam’s kangaroo rats (D. merriami), and chisel-toothed kangaroo rats (D. microps). Potential

233 pocket mice species included desert pocket mice (C. pencillatus), long-tailed pocket mice (C.

234 formosus), and little pocket mice (P. longimembris). Desert kit foxes and coyotes both

235 consumed lagomorphs as well, of which 2 species occurred in the Mojave Desert: black-tailed

236 jackrabbit and desert cottontail. The squirrel species that were potentially preyed upon included

237 round-tailed ground squirrel (X. tereticaudus), Mohave ground squirrel (X. mohavensis), and

238 white-tailed antelope squirrel.

239 Invertebrate prey mostly consisted of Orthopteran and Coleopteran species, solpugids,

240 scorpions, larvae, and sand treader crickets (Family Stenopelmatidae). In general, bird, lizard,

241 and snake remains within the scats were not identifiable to species. Anthropogenic items

242 consumed varied by species, including nuts of pistachios (Pistacia vera) and other agricultural

243 crops, man-made materials (e.g., cloth, plastic, food wrappers), domestic pets (i.e., cats and

244 dogs), and livestock and their waste. We suspect that a number of items that we found within the

245 scats were ingested incidentally: twigs, pieces of grass, small amounts of plant material, and dirt.

246 These items were most likely ingested while a kit fox or coyote was capturing or consuming an

247 intended food item.

248 When grouped into broader categories, differences occurred between the annual FOO of

249 items consumed by desert kit foxes and coyotes (Table 1). For desert kit foxes, rodents were the

250 most frequently occurring items in Years 1–3, while invertebrates were the most frequently

65

251 occurring items in Years 4 and 5. For coyotes, rodents were the most frequently occurring item

252 for Years 1 and 2. In all other years, lagomorphs occurred most frequently. Invertebrates were a

253 primary food item for coyotes as well (FOO > 10%), but FOO never exceeded 26% while FOO

254 of this same item category in desert kit foxes never fell below 46%. When all years were

255 combined, rodents and invertebrates had the highest FOO in desert kit fox diets while rodents

256 and lagomorphs had the highest FOO in coyote diets. Also, desert kit foxes only consumed fruits

257 in Year 4 while coyotes consumed fruits every year. Overall, coyotes ate more anthropogenic

258 material than desert kit foxes; frequently they had a FOO of anthropogenic material that was 2

259 times higher than that of desert kit foxes. Interestingly, both species exhibited the highest FOO

260 of anthropogenic material in Year 5.

261 In desert kit fox diets, the occurrence of birds increased yearly while the FOO of birds in

262 coyote diets fluctuated from year to year with Year 2 having the lowest FOO and Year 5 having

263 the highest. The FOO of reptiles in desert kit fox scats was lowest in Years 2 and 3 and highest

264 in Year 5. The occurrence of reptiles in coyote scats did not have an apparent pattern, but they

265 consumed the least amount of reptiles in Year 4 and had the highest consumption in Year 2.

266 Seasonal FOO of grouped food items for desert kit foxes and coyotes also exhibited

267 similarities and differences (Table 2). For both desert kit foxes and coyotes, rodents were the

268 most frequently occurring of all the items consumed in the Fall. In the remaining seasons,

269 invertebrates had the highest FOO in desert kit fox scats while lagomorphs were the most

270 frequently occurring item in coyote scats. When we considered when a particular item had the

271 highest FOO relative to itself, the FOO of anthropogenic items, fruit, and lagomorphs in desert

272 kit fox scats was highest in Winter, FOO of birds was highest in Spring, and FOO of reptiles was

273 highest in Summer. The FOO of anthropogenic items in coyote scats was also the highest in

66

274 Winter, but the FOO of birds was highest in Spring, and the FOO of fruits, invertebrates, and

275 reptiles was highest in Summer. In desert kit fox scats, the FOO of birds and invertebrates were

276 lowest in Fall, the FOO of reptiles was lowest in Winter, the FOO of lagomorphs was lowest in

277 Spring, and the FOO of rodents and anthropogenic material were lowest in Summer. For

278 coyotes, the FOO of lagomorphs and anthropogenic items in their scats were lowest in Fall while

279 the FOO of all other grouped items were lowest in Winter.

280 The annual H’ of both species varied slightly between years (Table 3). The H’ for desert

281 kit foxes and coyotes were both lowest in Year 2 (H’ = 0.48 and 0.65, respectively) and highest

282 in Year 5 (H’ = 0.66 and 0.80, respectively). Overall, coyotes exhibited a higher H’ than desert

283 kit foxes (H’ = 0.75 and 0.62, respectively). The mean annual H’ for desert kit foxes was

284 significantly lower than that for coyotes (t8 = -3.50, P = 0.008). The overall trend in the annual

285 H’ for our 2 study species was that annual H’ increased as precipitation decreased (Fig. 2).

286 Mean H’ for desert kit foxes was highest in Spring and lowest in Fall (Table 4). Mean H’

287 for coyotes was highest in Summer and lowest in Winter. Mean H’ for desert kit foxes was

288 significantly lower than that for coyotes in Fall, Spring, and Summer, but similar between

289 species in Winter (Table 4).

290 Horn’s similarity indices were high in all years and seasons (Tables 3 and 4), indicating

291 high overlap between desert kit fox and coyote diets. The diets were most similar in Years 2 and

292 5 (Table 3). Seasonally, mean Horn’s similarity indices were not significantly different (F3,16 =

293 0.32, P = 0.808; Table 4).

294 Rankings of food categories in desert kit fox and coyote diets were not correlated in any

295 year, season, or in the total diet (Table 5). Precipitation was not related to either annual Horn’s

296 similarity indices (r = 0.90, t3 = 3.58, P = 0.187) or annual correlation coefficients between

67

297 desert kit fox and coyote diets (r = 0.30, t3 = 0.54, P = 0.873). Furthermore, annual Horn’s

298 similarity indices were not correlated with annual abundance of rabbit pellets (r = -0.70, t3 = -

299 1.70, P = 0.752), large rodent burrows (r = 0.30, t3 = 0.54, P = 0.873), or small rodent burrows (r

300 = -0.10, t3 = -0.17, P = 0.873).

301 DISCUSSION

302 Our results suggest that desert kit foxes and coyotes in the Mojave Desert in California consume

303 many of the same food items and exhibit high dietary overlap, which indicates the potential for

304 exploitative competition. Because our study was not manipulative (i.e., removing coyotes,

305 changing available foods), we can only infer that competition was occurring in regard to prey

306 consumption (Schoener 1974). According to Lotka (1925), Volterra (1926), and Gause (1934),

307 species cannot coexist if they rely on the same resources for survival (Shoener 1974). Therefore,

308 competition may be mitigated between desert kit foxes and coyotes by a disproportionate use of

309 prey items, as has been demonstrated in previous studies (White et al. 1995; Cypher and Spencer

310 1998).

311 Although desert kit foxes and coyotes in the Mojave Desert rely heavily on rodents for

312 food, desert kit foxes also extensively consume invertebrates while coyotes preferentially

313 consume lagomorphs. Coyotes are markedly larger than kit foxes, and thus it is expected that the

314 former would prefer larger prey items while the latter would prefer relatively smaller prey

315 (Rosenzweig 1966; Shoener 1971). Past dietary studies between San Joaquin kit foxes and

316 coyotes at the Naval Petroleum Reserves and Carrizo Plain National Monument in California

317 also found that coyotes consume a higher proportion of larger prey items, such as lagomorphs,

318 while kit foxes rely more heavily on smaller prey items like rodents and invertebrates (White et

319 al. 1995; Cypher and Spencer 1998). The consumption of invertebrates as a primary food source

68

320 has also been documented for San Joaquin kit foxes (Cypher et al., unpubl. report), desert kit

321 foxes in Utah (Arjo et al. 2007; Koslowski et al. 2008), and many other arid land fox species

322 (Sheldon 1992).

323 Shannon diversity indices indicated that coyotes within our study exhibited an overall

324 more generalized and diverse diet than desert kit foxes for all years and seasons. In coyote scats,

325 rodents and lagomorphs generally occurred the most often, but birds, reptiles, invertebrates, and

326 anthropogenic material were frequently found at a FOO of > 10% for any given year or season.

327 Conversely, the desert kit foxes in our study relied heavily on rodents and invertebrates and

328 consumed other items (i.e., birds, reptiles) more opportunistically. Our results are consistent

329 with past studies that indicated that coyotes are opportunistic predators, but with some preference

330 for lagomorphs (Bueler 1973; Bekoff 1977; Sheldon 1992) while kit foxes are more rodent and

331 insect specialists that opportunistically feed upon other prey (White et al. 1995; Clark et al. 2005;

332 Kozlowski et al. 2008). When rodent consumption was analyzed in more detail, desert kit foxes

333 in the Mojave Desert specialized on heteromyid rodents (Kelly 2017) which is a typical food

334 group (Morrell 1972; Fisher 1981; Koopman et al. 2001). Desert kit foxes did consume birds,

335 reptiles, lagomorphs, fruit, and anthropogenic material, but typically at a lower frequency than

336 coyotes.

337 Our overall findings are also consistent with the concept of optimal foraging theory in

338 that both of our species are most likely consuming items that maximize energy input while

339 minimizing energy output (Pyke et al. 1977; Pianka 1978). Item selection by our 2 study species

340 is not only based on the size of the prey item, but also on its abundance and ease of capture

341 (Pyke et al. 1977; Pianka 1978). For example, a desert kit fox can catch and eat a black-tailed

342 jackrabbit, thereby maximizing their foraging gain with this one item, but these large rabbits are

69

343 difficult for a kit fox to dispatch and can potentially injure them (Egoscue 1962). Likewise, a

344 coyote can probably easily capture invertebrates and rodents, but they would have to eat a higher

345 proportion of these items to equate to predating upon a larger, more nutrient rich jackrabbit.

346 Coyotes are also known to prey upon much larger items, such as ungulates (Brown and

347 Conover 2011) and livestock (Mitchell et al. 2004), when the opportunity is available. Cypher et

348 al. (in press) found that coyotes in the Mojave Desert did consume domestic livestock, albeit

349 infrequently, while Kelly (2017) found that desert kit foxes rarely (< 0.5%) ate these prey items.

350 We do not know if livestock were consumed as a result of predation or scavenging (i.e., the

351 result of human dumping), but coyotes are widely known to kill domestic livestock (Mitchell et

352 al. 2004) while kit foxes would only be able to consume animals of this size as carrion. Coyotes

353 in our study also opportunistically consumed native fruits, a common behavior of this species

354 (Cypher 1993; Cypher et al. 1994; White et al. 1995), while desert kit foxes rarely ate fruits

355 despite their apparent availability.

356 The FOO of lagomorphs in coyote scats was always > 40%, even when our prey transect

357 surveys indicated that lagomorph numbers were low. Conversely, lagomorph consumption by

358 desert kit foxes only reached that of a primary prey item during Year 4 as well as during the

359 winter season. Hence, the FOO of lagomorphs in coyote scats were, on average, more than 5

360 times higher than what was found in desert kit fox scats. This implies that even when lagomorph

361 densities declined, coyotes still preferentially consumed this prey type (Cypher et al., in press).

362 A marked preference for lagomorphs by coyotes is similar to results from other coyote dietary

363 studies performed in Utah (Kozlowski et al. 2008), southeastern Idaho (MacCracken and Hansen

364 1987), southern Illinois (Cypher 1993), northeastern North Carolina (McVey et al. 2013), and

365 other locales in California (Ferrel et al. 1953; Cypher et al. 1994; Cypher and Spencer 1998).

70

366 Also, some past studies on kit foxes found that lagomorphs were prevalent in the diet (Egoscue

367 1962; Morrell 1972; Cypher et al. 2000), but that was not the case in our study. Therefore, a

368 difference in the use of lagomorphs by desert kit foxes and coyotes may constitute resource

369 partitioning. The specialization by desert kit foxes on smaller items and the infrequent predation

370 on larger prey, such as lagomorphs, is a strategy that may decrease competition with coyotes and

371 lead to coexistence.

372 By not foraging for lagomorphs, desert kit foxes may also be less likely to come into

373 direct contact with coyotes, thereby reducing the potential for interference competition. It is well

374 documented that interference competition between coyotes and kit foxes often results in the

375 mortality of the fox (Ralls and White 1995; Cypher and Spencer 1998; Nelson et al. 2007).

376 Death from predation, primarily by coyotes, accounts for 57–89% of all known kit fox fatalities

377 (Berry et al. 1987; Ralls and White 1995; White et al. 1995; Cypher et al. 2000; White et al.

378 2000; United States Fish and Wildlife Service 2010). Thus, desert kit foxes may not only prey

379 less upon lagomorphs, but they also may avoid areas where lagomorphs are more prevalent to

380 reduce encounters with coyotes. Desert kit foxes in the Great Basin Desert of Western Utah

381 were found to infrequently visit or avoid areas preferred by coyotes (Kozlowski 2008). Nelson

382 et al. (2007) found that San Joaquin kit foxes selectively used habitats with few shrubs where

383 lagomorphs were less abundant, and only infrequently used habitats with shrubs where both

384 lagomorphs and coyotes were more abundant.

385 The daily use of dens by kit foxes is another strategy that can ease interference

386 competition with coyotes (White et al. 1995; Cypher and Spencer 1998). Den entrances of kit

387 foxes are usually too small for a coyote to enter (Egoscue 1962; Cypher 2003) and thus allows an

388 opportunity for kit foxes to escape interactions with their main predator. Behaviors such as den

71

389 use and resource and habitat partitioning may be especially important during times of decreased

390 prey abundance due to seasonality and drought.

391 Seasonality had an influence on the consumption of prey items by our 2 study species,

392 similar to what has been reported in other canid dietary studies (White et al. 1995; Kozlowski et

393 al. 2008; Schrecengost et al. 2008; McVey et al. 2013). Desert kit foxes and coyotes experienced

394 high dietary overlap across all seasons, but proportional use of items was different. Coyote diet

395 was always more diverse than that of the desert kit fox in all seasons although dietary diversity

396 was lowest for both species in winter. During the winter months, reptiles and some invertebrates

397 are underground in a dormant state, many bird species migrate to warmer climates, most species

398 are not reproducing, and overall prey densities are lower. At this time of the year, prey in

399 general and preferred items in particular may be more difficult to obtain and exploitative

400 competition for remaining resources may increase. Desert kit foxes also had the highest FOO of

401 lagomorphs during the winter season, which may have caused increased competition with

402 coyotes due to their preference for this prey item. Additionally, it is possible that as both species

403 spend more time and effort foraging during these periods of decreased prey availability the

404 potential for interference competition was increased.

405 Precipitation had an effect on prey abundance on our study site. Drought conditions

406 occurred during the last 3 years of the study. As annual precipitation decreased, Shannon

407 diversity indices increased for both desert kit foxes and coyotes. As the availability of primary

408 food items declined at our site, both species increased their consumption of other food items,

409 which is an expected response based on optimal foraging theory (Pyke et al. 1977). Although

410 our study species broadened their diets, the overall decrease in more preferred food items may

411 have resulted in an increase in competition for remaining food resources. Also, the FOO of

72

412 lagomorphs in desert kit fox scats increased during the drought, which may have enhanced

413 exploitative competition with coyotes during this time.

414 In all years, regardless of precipitation, rodents and invertebrates comprised over 70% of

415 the items consumed by desert kit foxes, but in different proportions (Kelly 2017). However, as

416 the availability of rodents declined during the latter years of our study, the consumption of

417 invertebrates increased substantially (Kelly 2017). An increase in use of invertebrate during

418 drought conditions has been noted in other kit fox dietary studies (Spiegel et al. 1996, Cypher et

419 al., unpubl. report), further indicating that this type of food provides an important supplement

420 when rodent prey is less available. For coyotes, rodents and lagomorphs comprised over 60% of

421 the items consumed for Years 1–3, but use of these items declined and use of other items

422 increased during the remaining 2 years when drought conditions prevailed (Cypher et al., in

423 press). Our findings are consistent with previous studies in that coyotes appear better able to

424 expand their diet in response to changing conditions while kit foxes continue to preferentially

425 consume their preferred prey items with some subsidization with other foods (White et al. 1995;

426 Cypher and Spencer 1998).

427 Both desert kit foxes and coyotes consumed man-made items that would not have

428 naturally occurred in the Mojave Desert and thus would not normally be available for them to

429 eat. The FOO of anthropogenic material by coyotes in the Mojave Desert was frequently 2 times

430 higher than that of desert kit foxes. Among anthropogenic items, desert kit foxes mainly

431 consumed nuts of pistachios (Kelly 2017) while coyotes ate a plethora of items including nuts of

432 Pistachios, domestic farm animals, dogs, cats, and man-made refuse (cloth, dog leashes, paper,

433 etc.; Cypher et al., in press). By consuming items that are not normally found in a natural

73

434 environment both of our study species may be receiving some amount of anthropogenic

435 subsidization, especially during times of drought when natural prey may be in decline.

436 Both species consumed anthropogenic items at an increased rate when natural foods were

437 less abundant. However, coyotes seemed to rely on anthropogenic subsidization more than

438 desert kit foxes. For example, in Year 5 when the drought was in its third consecutive year, the

439 FOO of anthropogenic items in desert kit fox scats was 13% while coyotes exhibited a FOO of

440 29%. Also, both species had the highest FOO of anthropogenic items during the winter season

441 when many prey items are less available or harder to find due to life history traits (i.e.,

442 hibernation, torpor, seasonal migration). Anthropogenic subsidization during times of decreased

443 prey availability may increase individual survival resulting in a higher predator population than

444 would be expected during lean times. This can enhance predation pressure on co-occurring

445 species and remaining food supplies (Rodewald et al. 2011; Newsome and van Eeden 2017;

446 Cypher et al., in press). Arjo et al. (2007) documented an increase in available surface water at

447 their study location due to human activities and this most likely contributed to an increase in

448 coyote abundance. Due in part to the increase in coyotes, desert kit foxes showed a marked

449 change in den selection, spatial distribution, and dietary selection (Arjo et al. 2007). The

450 increase in the coyote population could also increase direct interactions with desert kit foxes and

451 increase direct mortality of the latter species. In the Mojave Desert, anthropogenic subsidization

452 of coyotes could increase predation pressure on federally listed threatened desert tortoises (Esque

453 et al. 2010; Cypher et al., in press). An unnatural increase in coyote abundance could ultimately

454 increase exploitative competition with desert kit foxes. Because this is a relatively recent and

455 rapidly emerging phenomenon, desert kit foxes may not be able to adapt to such abrupt changes

74

456 in their environment and their population in the Mojave Desert could become at risk, at least in

457 areas where subsidization of coyotes is occurring.

458 A greater understanding regarding the diets and potential competitive interactions

459 between desert kit foxes and coyotes in the Mojave Desert is pertinent additional life-history

460 knowledge that may contribute to future management and conservation needs of the desert kit

461 fox. As human development increases in the Mojave Desert and more habitat is converted and

462 fragmented, management strategies may become necessary to protect these animals. Although

463 the desert kit fox in California is not currently considered threatened or endangered, this may

464 change in the future and a management plan may be necessary to help ensure that they persist in

465 this area. Desert kit foxes exhibit many adaptive responses, including resource and habitat

466 partitioning, that lead to coexistence with sympatric coyotes (White et al. 1995; Cypher and

467 Spencer 1998), but changes to the landscape as a result of human influences may threaten this

468 balance. Coyotes, due to their generalistic tendencies, are able to rapidly adapt and thrive in

469 human altered environments (Bekoff and Gese 2003; Arjo et al. 2007; Gehrt 2007). Also, the

470 introduction of or increase in anthropogenic material in the environment can unnaturally increase

471 the coyote population and thus place higher competitive pressures on the desert kit fox.

472 Management strategies should also include facilitation of a healthy, robust prey

473 population to ensure that desert kit foxes continue to reside and reproduce in the Mojave Desert.

474 Desert kit foxes rely primarily on rodents and insects for sustenance, but they have the ability to

475 expand their diet and will consume a variety of items when primary items are less abundant, such

476 as during drought conditions. Therefore, effective management strategies should not only

477 promote an overall robust prey population, but also include steps to maintain a variety of

478 available food options in the event of declines in preferred food items. Maintaining a healthy

75

479 prey population and mitigating the effects of exploitative and interference competition between

480 sympatric desert kit foxes and coyotes may best be achieved by conserving large tracts of intact,

481 high quality habitat. Management strategies should also include minimizing the availability of

482 anthropogenic food items to discourage unnatural changes in predator-prey relationships between

483 desert kit foxes, coyotes, and their prey.

484 ACKNOWLEDGMENTS

485 We would like to thank the Endangered Species Recovery Program of California State

486 University, Stanislaus, for providing work space and the tools necessary to prepare and analyze

487 scats. We would like to specifically thank Alexandria Y. Madrid, Erin N. Tennant, Christine L.

488 Van Horn Job, and Tory L. Westall for their assistance in collecting and analyzing desert kit fox

489 and coyote scats. We also thank Tory L. Westall for her assistance with graphics and data entry.

490 We would like to further thank Paul T. Smith for reading an earlier draft of the manuscript and

491 providing comments and editing suggestions. Support for this project was generously provided

492 by the Endangered Species Recovery Program, California State University, Stanislaus, and the

493 United States Fish and Wildlife Service.

494 LITERATURE CITED

495 ADORJAN, A. S. AND G. B. KOLENOSKY. 1969. A manual for the identification of hairs of selected

496 Ontario Mammals. Ontario Department of Lands and Forests, Ontario, Canada.

497 ARJO, W. M., E. M. GESE, T. J. BENNETT, A. J. KOZLOWSKI. 2007. Changes in kit fox-coyote-

498 prey relationships in the Great Basin Desert, Utah. Western North American Naturalist

499 67:389–401.

500 BEKOFF, M. 1977. Canis latrans. Mammalian Species 79:1–9.

76

501 BEKOFF, M., AND E. M. GESE. 2003. Coyote (Canis latrans). Pp. 467–481 in Wild Mammals of

502 North America: Biology, Management, and Conservation (G. A. Feldhamer, B. C.

503 Thompson, and J. A. Chapman, eds.). The Johns Hopkins University Press, Baltimore,

504 Maryland.

505 BERRY, W. H., J. H. SCRIVNER, T. P. O’FARRELL, C. E. HARRIS, T. T. KATO, AND P. M. MCCUE.

506 1987. Sources and rates of mortality of the San Joaquin kit fox, Naval Petroleum Reserve

507 #1, Kern County, CA, 1980–1986. Rep. No. EGG 10282-2154, EG&G Energy

508 Measurements, Goleta, California.

509 BROWER, J. E., AND J. H. ZAR. 1984. Field and Laboratory Methods for General Ecology. Wm. C.

510 Brown Publishers, Dubuque, Indiana.

511 BROWN, D. E., AND M. R. CONOVER. 2011. Effects of large-scale removal of coyotes on

512 pronghorn and productivity and abundance. Journal of Wildlife Management

513 75:876–882.

514 BUELER, L. E. 1973. Wild Dogs of the World. Stein and Day Publishers, New York, New York.

515 CLARK, JR., H. O., G. D. WARRICK, B. L. CYPHER, P. A. KELLY, D. E. WILLIAMS, AND D. E.

516 GRUBBS. 2005. Competitive interactions between endangered kit foxes and nonnative red

517 foxes. Western North American Naturalist 65:153–163.

518 CLIFFORD, D., L. WOODS, M. W. GABRIEL, J. RUDD, E. J. DUBOVI, K. TERIO, F. UZAL, ET AL.

519 2013. Canine distemper outbreak in free-ranging desert kit foxes inhabiting a solar

520 energy development zone. Abstract, Western Section of the Wildlife Society 2013

521 Annual Conference, Reno, Nevada.

522 CYPHER, B. L. 1993. Food item use by three sympatric canids in Southern Illinois. Transactions

523 of the Illinois State Academy of Science 86:139–144.

77

524 CYPHER, B. L. 2003. Foxes (Vulpes species, Urocyon species, and Alopex lagopus). Pp. 511–546

525 in Wild Mammals of North America: Biology, Management, and Conservation (G.A.

526 Feldhamer, B. C. Thompson, and J. A. Chapman, eds.). Johns Hopkins University Press.

527 Baltimore, Maryland.

528 CYPHER, B. L., E. C. KELLY, T. L. WESTALL, AND C. L. VAN HORN JOB. In press. Coyote diet

529 patterns in the Mojave Desert: implications for threatened Desert Tortoises. Pacific

530 Conservation Biology.

531 CYPHER, B. L., AND K. A. SPENCER. 1998. Competitive interactions between coyotes and San

532 Joaquin kit foxes. Journal of Mammalogy 79:204–214.

533 CYPHER, B. L., K. A. SPENCER, AND J. H. SCRIVNER. 1994. Food-item use by coyotes at the Naval

534 Petroleum Reserves in California. Southwestern Naturalist 39:91–95.

535 CYPHER, B. L., G. D. WARRICK, M. R. M. OTTEN, T. P. O’FARRELL, W. H. BERRY, C. E. HARRIS,

536 T. T. KATO, ET AL. 2000. Population dynamics of San Joaquin kit foxes at the Naval

537 Petroleum Reserves in California. Wildlife Monographs 145:1–43.

538 DI STEFANO, J. 2003. How much power is enough? Against the development of an arbitrary

539 convention for statistical power calculations. Functional Ecology 17:707–709.

540 EGOSCUE, H. J. 1962. Ecology and life history of the kit fox in Toole County, Utah. Ecology

541 43:481–497.

542 ESQUE, T. C., K. E. NUSSEAR, K. K. DRAKE, A. D. WALDE, K. H. BERRY, R. C. AVERILL-

543 MURRAY, A. P. WOODMAN, ET AL. 2010. Effects of subsidized predators, resource

544 variability, and human population density on desert tortoise populations in the Mojave

545 Desert, United States. Endangered Species Research 12:167–177.

78

546 FERREL, C. M., H. R. LEACH, AND D. R. TILLOTSON. 1953. Food habits of the coyote in

547 California. California Fish and Game 39:301–341.

548 FISHER, J. L. 1981. Kit fox diet in south-central Arizona. M.S. thesis, The University of Arizona.

549 Tucson, Arizona.

550 GAUSE, G. F. 1934. The struggle for existence. Williams and Wilkins, Baltimore, Maryland.

551 GEHRT, S. D. 2007. Ecology of coyotes in urban landscapes. Proceedings of the Wildlife Damage

552 Management Conference 63:303–311.

553 GLASS, B. P. 1981. Key to the skulls of North American mammals. Oklahoma State University,

554 Stillwater, Oklahoma.

555 HARDIN, G. 1960. The competitive exclusion principle. Science 131:1292–1297.

556 HORN, H. S. 1966. Measurement of “overlap” in comparative ecological studies. American

557 Naturalist 100: 419–424.

558 KOOPMAN, M. E., B. L. CYPHER, AND D. R. MCCULLOUGH. 2001. Factors influencing space and

559 prey use by San Joaquin kit foxes. Transactions of the Western Section of the Wildlife

560 Society 37:77–83.

561 KELLY, E. C. 2017. Desert kit fox (Vulpes macrotis arsipus) food habits and competitive

562 interactions with coyotes (Canis latrans) in the Mojave Desert. M.S. Thesis, California

563 State University, Bakersfield. Bakersfield, California.

564 KOZLOWSKI, A. J., E. M. GESE, AND W. M. ARJO. 2008. Niche overlap and resource partitioning

565 between sympatric kit foxes and coyotes in the Great Basin Desert of western Utah. The

566 American Midland Naturalist 160:191–208.

567 LEGENDRE, P., AND L. LEGENDRE. 1998. Numerical Ecology, second edition. Elsevier Science,

568 Amsterdam, The Netherlands.

79

569 LEITNER, P. 2009. The promise and peril of solar power. Wildlife Professional 3:48–53.

570 LOTKA, A. J. 1925. Elements of physical biology. Williams and Wilkins, Baltimore, Maryland.

571 LOVICH, J. E. AND J. R. ENNEN. 2011. Wildlife conservation and solar energy development in the

572 desert southwest, United States. BioScience 61:982–992.

573 MACCRACKEN, J. G., AND R. M. HANSEN. 1987. Coyote feeding strategies in southeastern Idaho:

574 optimal foraging by an opportunistic predator? Journal of Wildlife Management 51:278–

575 285.

576 MAURER, B. A. 1984. Interference and Exploitation in bird communities. Wilson Bulletin

577 96:380–395.

578 MCVEY, J. M., D. T. COBB, R. A. POWELL, M. K. STOSKOPF, J. H. BOHLING, L. P. WAITS, AND C.

579 E. MOORMAN. 2013. Diets of sympatric red wolves and coyotes in northeastern North

580 Carolina. Journal of Mammalogy 94:1141–1148.

581 MITCHELL, B. R., M. M. JAEGER, AND R. H. BARRETT. 2004. Coyote depredation management:

582 current methods and research needs. Wildlife Society Bulletin 32:1209–1218.

583 MOEHRENSCHLAGER, A., B. L. CYPHER, K. RALLS, R. LIST, AND M. A. SOVADA. 2004. Swift

584 foxes and kit foxes: comparative ecology and conservation priorities of swift and kit

585 foxes. Pp. 185–198 in The Biology and Conservation of Wild Canids (D. W. Macdonald

586 and C. Sillero-Zubiri, eds.). Oxford University Press. Oxford, United Kingdom.

587 MOORE, T. D., L. E. SPENCE, C. E. DUGNOLLE. 1974. Identification of the dorsal guard hairs of

588 some mammals of Wyoming. W. G. Hepworth, editor. Wyoming Game and Fish

589 Department, Bulletin No. 14, Cheyenne, Wyoming.

590 MORRELL, S. 1972. Life history of the San Joaquin kit fox. California Fish and Game 58:162–

591 174.

80

592 NELSON, J. L., B. L. CYPHER, C. D. BJURLIN, AND S. CREEL. 2007. Effects of habitat on

593 competition between kit foxes and coyotes. Journal of Wildlife Management 71:1467–

594 1475.

595 NEWSOME, T. M., AND L. M. VAN EEDEN. 2017. The effects of food waste on wildlife and

596 humans. Sustainability 9:1–9.

597 NOY-MEIR, I. 1973. Desert ecosystems: environments and producers. Annual Review of Ecology

598 and Systematics 4:25–51.

599 PIANKA, E. R. 1978. Evolutionary Ecology, 2nd ed. Harper and Row Publishers. New York, New

600 York.

601 PYKE, G. H., H. R. PULLIAM, AND E. L. CHARNOV. 1977. Optimal foraging: a selective review of

602 theory and tests. Quarterly Review of Biology 52:137–154.

603 RALLS, K., AND P. J. WHITE. 1995. Predation on San Joaquin kit foxes by larger canids. Journal

604 of Mammalogy 76:723–729.

605 RODEWALD, A. D., L. J. KEARNS, AND D. P. SHUSTACK. 2011. Anthropogenic resource subsidies

606 decouple predator-prey relationships. Ecological Applications 21:936–943.

607 ROEST, A. I. 1986. A key-guide to mammal skulls and lower jaws. Mad River Press. Eureka,

608 California.

609 ROSENZWEIG, M. L. 1966. Community structure in sympatric carnivore. Journal of Mammalogy

610 47:602–621.

611 SCHERER, R. D., AND J. A. TRACEY. 2011. A power analysis for the use of counts of egg masses

612 to monitor wood frog (Lithobates sylvaticus) populations. Herpetological Conservation

613 and Biology 6:81–90.

614 SCHOENER, T. W. 1974. Resource partitioning in ecological communities. Science 185:27–39.

81

615 SHELDON, J. W. 1992. Wild Dogs: The Natural History of the Nondomestic Canidae. Academic

616 Press. San Diego, California.

617 SCHRECENGOST, J. D., J. C. KILGO, D. MALLARD, H. S. RAY, AND K. V. MILLER. 2008. Seasonal

618 food habits of the coyote in South Carolina coastal plain. Southeastern Naturalist 7:135–

619 144.

620 SPIEGEL, L., B. L. CYPHER, AND T. C. DAO. 1996. Diet of the San Joaquin kit fox (Vulpes

621 macrotis mutica) at three sites in western Kern County, California. Pp. 39–52 in Studies

622 of San Joaquin kit fox in undeveloped and oil developed areas (L. Spiegel, ed.).

623 California Energy Commission, Sacramento, California.

624 TAYLOR, B. L., AND T. GERRODETTE. 1993. The uses of statistical power in conservation biology:

625 the vaquita and northern spotted owl. Conservation Biology 7:489–500.

626 TURNER, R. M. 1994. Mojave Desert scrub. Pages 157–168 in Biotic communities, southwestern

627 United States and northwestern Mexico (D.E. Brown, ed.). University of Utah Press. Salt

628 Lake City, Utah.

629 UNITED STATES BUREAU OF LAND MANAGEMENT. 1980. The California Desert conservation area

630 plan, 1980. United States Bureau of Land Management, California Desert District,

631 Riverside, California.

632 UNITED STATES CLIMATE DATA. 2014. Precipitation data for Barstow, California.

633 http://www.usclimatedata.com/climate/barstow/california/united-states/usca0069/2014/8

634 Accessed September 2014.

635 UNITED STATES FISH AND WILDLIFE SERVICE. 1983. San Joaquin kit fox recovery plan. United

636 States Fish and Wildlife Service, Portland, Oregon.

82

637 UNITED STATES FISH AND WILDLIFE SERVICE. 1998. Recovery plan for upland species of the San

638 Joaquin Valley, California. United States Fish and Wildlife Service, Portland, Oregon.

639 UNITED STATES FISH AND WILDLIFE SERVICE. 2010. San Joaquin kit fox (Vulpes macrotis

640 mutica) 5–year review: summary and evaluation. United States Fish and Wildlife Service,

641 Region 8, Sacramento, California.

642 VOLTERRA, V. 1926. Fluctuations in the abundance of a species considered mathematically.

643 Nature 188:558–560.

644 WHITE, P. J., W. H. BERRY, J. J. ELIASON, AND M. T. HANSON. 2000. Catastrophic decrease in an

645 isolated population of kit foxes. Southwestern Naturalist 45:204–211.

646 WHITE, P. J., K. RALLS, AND C. A. VANDERBILT WHITE. 1995. Overlap in habitat and food use

647 between coyotes and San Joaquin kit foxes. Southwestern Naturalist 40:342–349.

648 YOUNG, J. A., AND C. G. YOUNG. 1992. Seeds of woody plants in North America. Dioscorides

649 Press, Portland, Oregon.

83

650 TABLES

651 Table 1.—Annual frequency of occurrence (FOO) for item categories in desert kit fox (Vulpes macrotis arsipus; Kelly 2017) and

652 coyote (Canis latrans; Cypher et al., in press) scats collected in the Mojave Desert, California, United States, from October 2009 to

653 September 2014. Years span October–September.

Desert kit fox FOO (%) Coyote FOO (%)

Item category Year 1 Year 2 Year 3 Year 4 Year 5 Total Year 1 Year 2 Year 3 Year 4 Year 5 Total

Scat total (n) 127 76 229 388 410 1230 625 474 631 801 715 3246

Lagomorphs 1.6 4.0 7.0 13.4 9.3 9.0 48.3 42.0 58.2 67.5 41.4 52.5

Rodent 92.9 96.1 90.0 62.4 59.3 71.7 53.1 65.4 46.3 24.3 26.4 40.6

Bird 7.1 9.2 10.9 12.1 22.0 14.5 10.4 6.5 12.8 10.9 19.0 12.3

Reptile 23.6 7.9 12.2 20.6 25.9 20.3 20.3 26.2 19.3 17.9 21.8 20.7

Invertebrate 53.5 46.1 69.0 77.6 83.7 73.6 14.4 15.6 16.6 20.4 25.7 19.0

Fruit 0.0 0.0 0.0 0.3 0.0 0.1 3.5 0.4 2.9 4.5 4.5 3.4

Anthropogenic 0.8 1.3 2.2 9.8 12.9 8.0 7.2 4.9 14.9 14.7 28.7 14.9

84

654 Table 2.—Seasonal frequency of occurrence (FOO) for item categories in desert kit fox (Vulpes macrotis arsipus; Kelly 2017) and

655 coyote (Canis latrans; Cypher et al., in press) scats collected in the Mojave Desert, California, United States, from October 2009 to

656 September 2014. Seasons were defined as Fall = October-December; Winter = January-March; Spring = April-June; Summer = July-

657 September.

Desert kit fox FOO (%) Coyote FOO (%)

Item category Fall Winter Spring Summer Fall Winter Spring Summer

Scat total (n) 209 636 198 187 845 834 738 829

Lagomorphs 7.2 10.7 5.6 9.1 46.8 54.2 52.2 57.1

Rodent 74.6 71.4 71.7 69.5 51.1 34.4 38.8 37.8

Bird 9.6 14.5 19.2 15.0 11.7 10.1 14.5 13.3

Reptile 18.2 12.6 31.3 37.4 19.5 15.0 24.1 24.6

Invertebrate 65.6 72.2 76.3 84.5 16.6 15.6 20.3 23.6

Fruit 0.0 0.2 0.0 0.0 3.8 2.8 3.4 3.6

Anthropogenic 3.8 13.4 1.5 1.1 12.2 17.0 14.2 16.3

85

658 Table 3.—Annual Shannon diversity indices for coyotes (Canis latrans) and desert kit foxes

659 (Vulpes macrotis arsipus) and Horn’s index of similarity along with prey availability indices and

660 annual precipitation in the Mojave Desert, California, United States, from October 2009 to

661 September 2014. Years span October–September.

Year 1 Year 2 Year 3 Year 4 Year 5 Total

Diversity

kit foxes 0.50 0.48 0.54 0.64 0.66 0.62

coyotes 0.70 0.65 0.72 0.72 0.80 0.75

Similarity 0.83 0.88 0.80 0.82 0.86 0.85

Annual Precipitation (cm) 16.9 28.2 7.3 7.5 8.0

Average large burrows 50 42 36 7 9

Average small burrows 24 19 20 15 15

Average lagomorph pellets 77 130 1227 343 31

86

662 Table 4.—Shannon diversity indices for coyotes (Canis latrans) and desert kit foxes (Vulpes

663 macrotis arsipus) and Horn’s index of similarity calculated for each season across all years in the

664 Mojave Desert, California, United States, from October 2009 to September 2014. Seasons were

665 defined as Fall = October-December; Winter = January-March; Spring = April-June; Summer =

666 July-September.

Fall Winter Spring Summer

Diversity

kit fox 0.49 ± 0.07 0.56 ± 0.04 0.57 ± 0.02 0.53 ± 0.03

Coyote 0.70 ± 0.02 0.65 ± 0.04 0.72 ± 0.03 0.73 ± 0.02

T -2.58 -1.77 -3.50 -4.92

P -0.033 0.115 0.008 0.001

Similarity 0.81 ± 0.02A1 0.80 ± 0.03A 0.82 ± 0.02A 0.83 ± 0.02A

667 1 Means with similar letters were not significantly different.

87

668 Table 5.—Spearman rank correlations analysis to compare rankings of items between coyotes

669 (Canis latrans) and desert kit foxes (Vulpes macrotis arsipus) for each year, season, and total.

670 Scats from both species were collected from the Mojave Desert, California, United States, from

671 October 2009 to September 2014. Years span October–September and seasons were defined as

672 Fall = October-December; Winter = January-March; Spring = April-June; Summer = July-

673 September.

Time Frame rs t P

Year 1 0.75 2.54 0.261

Year 2 0.68 2.07 0.281

Year 3 0.54 1.43 0.430

Year 4 0.71 2.25 0.281

Year 5 0.07 0.16 0.879

Fall 0.68 2.07 0.375

Winter 0.21 0.48 0.645

Spring 0.54 1.43 0.588

Summer 0.46 1.16 0.588

All Years/Seasons 0.46 1.16 0.294

88

674 FIGURE LEGENDS

675 Fig. 1.—Study area located in the Mojave Desert in California, United States (inset), which is

676 bound by the Fort Irwin National Training Center to the north and Interstate 15 and State Route

677 58 to the south. The blue boundary is the specific study area and the black dots denote every

678 location a scat or multiple scats were collected during the 5 year project.

679 Fig. 2.—Annual Shannon diversity indices for both desert kit foxes (Vulpes macrotis arsipus)

680 and coyotes (Canis latrans) with annual precipitation totals from the Mojave Desert, California,

681 United States, during October 2009 to September 2014.

89

682 FIGURES

683 Fig. 1

Fort Irwin

D

N 0 5 10 30 40 Km 684 A

90

685 Fig. 2

0.9 30 00 8 0.8 r-- · ~ 25 ~ r-- ~ "'0 0. 7 r-- .- s (.) .s ~ r-- ~ 0.6 20 '-:" .~ r-- .----- v .~ 00 0.5 (.) $..-4 .- 15 (!) ...... -~ 0.4 $..-4 ·~ A "'0 ~ 0.3 10 -ro;::::$ 0 ~ ~ ~ 0.2 5 ~ .,.q 0.1 r:fJ. 0 0 Year1 Year2 Year 3 Year4 Year5 Year

c::::J Desert Kit fo x c:::JCoyote -Annual Precip. 686

91

CHAPTER 4

Landscape partitioning by Desert Kit Foxes and Coyotes in the Mojave Desert

Erica C. Kelly1,2,3, Brian L. Cypher2, and Tory L. Westall2

1Department of Biology, California State University, Bakersfield, CA, 93309, USA

2Endangered Species Recovery Program, California State University, Stanislaus, P.O. Box 9622,

Bakersfield, CA, 93389, USA

3Corresponding author, email: [email protected]

ABSTRACT

We assessed the relative distribution of Desert Kit Foxes (Vulpes macrotis arsipus) and potential competitive associations with Coyotes (Canis latrans) within a study site in the Mojave

Desert in California, USA. With the use of camera stations, we found that Desert Kit Foxes were present in the majority of areas surveyed. We also found that Desert Kit Foxes and Coyotes at our study location do not appear to partition habitat on a landscape scale. Temporal avoidance, resource partitioning, and daily den use by Desert Kit Foxes may be sufficient to reduce competition and facilitate coexistence. This balance, though, may become imperiled as human disturbance continues in the Mojave Desert and potentially affords a competitive advantage to

Coyotes.

Keywords: Camera trapping, Canis latrans, landscape partitioning, resource partitioning, Vulpes macrotis arsipus

92

1. Introduction

Desert Kit Foxes (Vulpes macrotis arsipus) and Coyotes (Canis latrans) are sympatric in the Mojave Desert in California, USA (Arjo et al. 2007, Cypher 2003). Coyotes are the main competitor and predator of kit foxes, and both interference and exploitative competition potentially occur between the two species (Cypher and Spencer 1998, Cypher et al. 2000, Ralls and White 1995). Kit foxes may be able to mitigate competition with Coyotes and coexist by the daily use of dens as well as resource and habitat partitioning (Kelly 2017, Nelson et al. 2007,

White et al. 1995). Nelson et al. (2007) found that San Joaquin Kit Foxes (V. m. mutica) in the

Lokern Natural Area of Kern County, California, USA, partitioned habitat because they inhabited areas that were used less frequently by Coyotes. Conversely, White et al. (1994) found that all home ranges of San Joaquin Kit Foxes at their study site had > 30% overlap by Coyotes.

Thus, spatial partitioning between the two species appears to vary with location.

Unlike the San Joaquin Kit Fox, which is federally listed as endangered and state listed as threatened, Desert Kit Foxes in California are considered a common subspecies (United States

Fish and Wildlife Service 1998). Within the last few years, Desert Kit Foxes in California have increasingly been treated as a Species of Special Concern, and in 2013, a petition was submitted to the California Department of Fish and Wildlife (CDFW) to afford them increased protections

(Kadaba et al., unpubl. report). The petition, though, was denied by CDFW due in large part to a lack of information regarding Desert Kit Foxes in California (D. Kadaba, pers. comm.).

Due to an increased interest in the conservation of the Desert Kit Fox and the potential for further human encroachment into their habitat (Leitner 2009), additional information regarding the relative distribution of this subspecies could contribute to future conservation measures.

Also, by determining habitat use by Desert Kit Foxes in the Mojave Desert in relation to Coyote

93

presence, we can gain a better understanding of the competitive interactions between these two sympatric species. Therefore, we conducted an investigation of landscape-level spatial partitioning between Desert Kit Foxes and Coyotes over a 3-year period in the Mojave Desert.

2. Methods

2.1 Study area

Our 1500-km2 study site was located in the Mojave Desert north of Barstow, California,

USA (Cypher et al., in press). This area is bounded by Fort Irwin National Training Center and the China Lake Naval Air Weapons Station to the north and Interstate 15 and State Route 58 to the south (Fig. 1). The study area has Mojave Desert scrub vegetation dominated by Creosote

Bush (Larrea tridentata) and a sparse ground cover consisting of a diversity of forbs and grasses

(Esque et al. 2010, Turner 1994). Elevation ranged from 500-900 m and the terrain consisted of flat, dry lake beds, alluvial fans, sand dunes, steep, rugged hills, and wide expanses of land dotted with natural soil crusts and a sparse cover of vegetation (Esque et al. 2010, Turner 1994).

Consistent with arid desert environments, the mean annual precipitation for Barstow is only 13.4 cm (U.S. Climate Data 2014). Much of the study area is comprised of public lands managed by the U.S. Bureau of Land Management with interspersed private property. Populated areas included Barstow (population 23,835), Hinkley (population 1,915), and Harvard (unincorporated community with fewer than 100 people; included in Newberry Springs, an unincorporated community with a population of 2,895).

2.2 Study design

We conducted camera surveys in winter during 2012, 2013, and 2014 to document presence and relative abundance of Coyotes and Desert Kit Foxes. At each camera station, we secured an automated digital field camera (Cuddeback Digital Attack IR, Model 1156, Non

94

Typical Inc. Green Bay, Wisconsin, USA; Stealth Cam 3.0 MP Digital Scouting Cameras, Model

STC-AD2/AD2RT, Stealth Cam LLC, Bedford, Texas, USA) to a 1.2-m U-post with zip ties and duct tape. We used 12-in nails to stake down a 155.92 g can of cat food paté approximately 2 m in front of each camera. Lastly, we dripped scent lure (Carman’s Canine Call Lure, Russ

Carman, New Milford, Pennsylvania, USA) onto the can, the surrounding ground, and on the tallest immediate shrub as an additional attractant for carnivores. We deployed camera stations near prey abundance transects established as part of companion studies (Cypher et al., in press,

Kelly 2017). We deployed 30 cameras each session at stations spaced at least 2 km apart (Fig.

1). We deployed the cameras in early December of each year and collected the cameras 8-10 weeks later.

2.3 Analytical methods

For each year, we determined which camera stations were visited by each species. We conducted chi-square analyses by year to test against equal visitation by Desert Kit Foxes only,

Coyotes only, or by both species to determine whether Desert Kit Foxes were avoiding areas used by Coyotes. For instances where the results showed a significant difference, we repeated the chi-square analysis to test visitation by each species individually versus both species, and applied Yates’ correction for continuity. To adjust for an increased probability of a Type I error, we used Hochberg’s variation on Holm’s method to correct for P-values (Legendre and

Legendre 1998). Also, for all statistical analyses, we considered P-values to be significant at α =

0.1 to reduce the risk of committing a Type II error, which is considered more detrimental than a

Type I error when making wildlife management decisions and within the field of conservation biology (di Stefano 2003, Scherer and Tracey 2011, Taylor and Gerrodette 1993). By relaxing the alpha value, we aimed to reveal potential ecological relationships that can be more fully

95

explored through further investigation. We also calculated the proportion of cameras visited by

Coyotes that were also visited by Desert Kit Foxes.

3. Results

During the three-year study, we obtained hundreds of images from the 87 operable camera stations (Table 1). Totaled across all years, 52 (60%) stations had pictures of both study species,

26 (30%) captured only Desert Kit Foxes, and 9 (10%) had only Coyotes. Among years, 2014 had the most stations visited by both species, 2012 had the most stations visited by only Coyotes, and 2013 had the most stations visited by only Desert Kit Foxes. The proportion of cameras visited by Coyotes that also were visited by Desert Kit Foxes in each year was 77.3%, 89.5%, and 90.0%, respectively.

We found that there was a significant difference among species visiting camera stations for each year (2012: X2 = 7.80, df = 2, P = 0.020; 2013: X2 = 11.66, df = 2, P = 0.003; 2014: X2 =

14.00, df = 2, P = 0.001). For each year, there was no significant difference between stations visited by Desert Kit Foxes only and stations visited by both Desert Kit Foxes and Coyotes

(2012: X2 = 2.56, df = 1, P = 0.220; 2013: X2 = 1.33, df = 1, P = 0.248; 2014: X2 = 3.12, df = 1, P

= 0.220). Also, for each year, the number of stations visited by Coyotes only was significantly fewer than the number of stations visited by both species (2012: X2 = 5.50, df = 1, P = 0.076;

2013: X2 = 10.32, df = 1, P = 0.005; 2014: X2 = 11.25, df = 1, P = 0.005).

4. Discussion

Our results suggest that over a landscape scale, Desert Kit Foxes on our study site did not appear to partition habitat as a strategy to avoid competition with Coyotes. The proportion of camera stations visited by either species individually was similar or less than the proportion of camera stations visited by both species. Specifically, there were more camera stations visited by

96

both species than would be expected if Desert Kit Foxes were avoiding areas with Coyotes.

Interestingly, about a third of the total operable cameras were only visited by Desert Kit Foxes.

In all, Desert Kit Foxes used the majority of our study site. Although we did not quantify habitat attributes sufficiently to determine specific habitat use, it would be expected, based on competition theory, that the Desert Kit Fox would have a wider niche breadth than Coyotes because they are the lesser competitor (Nelson et al. 2007).

The proportion of stations visited by both species was highest in 2014. This was the third consecutive drought year and abundance of prey on our study sites had declined (Kelly

2017). Also, the proportion of cameras that were visited by Coyotes that were also visited by

Desert Kit Foxes was > 89% when the drought was in its second and third consecutive years.

The two species may have expanded foraging ranges to find adequate food, and therefore may have encountered more of our stations. The potential for interference competition may have also increased during this time.

Our results did not suggest that habitat partitioning was occurring over a landscape scale, but competitive interactions between Desert Kit Foxes and Coyotes in the Mojave Desert may be mitigated in other ways. Although both species are using the same habitat, Desert Kit Foxes may exhibit temporal partitioning in response to interference competition by Coyotes (List and

Macdonald 2003). Kit foxes also use dens to avoid direct interactions with Coyotes (White et al.

1995). When encounters do occur, Coyotes may kill Desert Kit Foxes as a means to decrease competition by reducing the number of kit foxes (White et al. 1995). Resource partitioning by

Desert Kit Foxes is another way that may reduce competition with Coyotes, which would promote coexistence (Cypher and Spencer 1998, White et al. 1995). For example, Kelly (2017)

97

found that although both species rely heavily on rodents for sustenance, Desert Kit Foxes also extensively consume invertebrates while Coyotes preferentially prey upon larger lagomorphs.

Evolved strategies by Desert Kit Foxes to mitigate competition with Coyotes in the

Mojave Desert currently allow the two species to coexist sympatrically. However, this balance may become threatened if human development continues to encroach on remaining habitat. For example, anthropogenic subsidization of Coyotes or destruction of kit fox dens could provide a competitive advantage to Coyotes and necessitate the implementation of management strategies to conserve Desert Kit Foxes. Therefore, a greater understanding of Desert Kit Foxes in

California now may be necessary to develop effective conservation and management strategies to ensure that this subspecies in California is not affected to the point of requiring formal protections as has occurred with the San Joaquin Kit Fox.

Acknowledgements

We would like to thank the Endangered Species Recovery Program of California State

University, Stanislaus, for providing a work space to analyze the camera data. We would like to specifically thank Alexandria Y. Madrid and Christine L. Van Horn Job for their assistance with camera surveys and data analysis. We would like to further thank David J. Germano for statistical assistance and editing guidance. Additionally, we would like to thank Paul T. Smith for reading an earlier draft of the manuscript and providing comments and editing suggestions.

Support for this project was provided by the Endangered Species Recovery Program, California

State University, Stanislaus, and the U.S. Fish and Wildlife Service.

98

Literature Cited

Arjo, W. M., E. M. Gese, T. J. Bennett, and A. J. Kozlowski. 2007. Changes in Kit Fox-Coyote

relationships in the Great Basin Desert, Utah. Western North American Naturalist

67:389–401.

Cypher, B. L. 2003. Foxes (Vulpes species, Urocyon species, and Alopex lagopus). Pages 511-

546 in G. A. Feldhamer, B. C. Thompson, and J. A. Chapman, editors. Wild Mammals of

North America: Biology, Management, and Conservation. The Johns Hopkins University

Press, Baltimore, Maryland.

Cypher, B. L., E. C. Kelly, T. L. Westall, and C. L. Van Horn Job. In press. Coyote diet patterns

in the Mojave Desert: implications for threatened Desert Tortoises. Pacific Conservation

Biology.

Cypher, B. L., and K. A. Spencer. 1998. Competitive interactions between Coyotes and San

Joaquin Kit Foxes. Journal of Mammalogy 79:204-214.

Cypher, B. L., G. D. Warrick, M. R. M. Otten, T. P. O’Farrell, W. H. Berry, C. E. Harris, T. T.

Kato, P. M. McCue, J. H. Scrivner, and B. W. Zoellick. 2000. Population dynamics of

San Joaquin Kit Foxes at the Naval Petroleum Reserves in California. Wildlife

Monographs 145:1-43. di Stefano, J. 2003. How much power is enough? Against the development of an arbitrary

convention for statistical power calculations. Functional Ecology 17:707-709.

Esque, T. C., K. E. Nussear, K. K. Drake, A. D. Walde, K. H. Berry, R. C. Averill-Murray, A. P.

Woodman, W. I. Boarman, P. A. Medica, J. Mack, et al. 2010. Effects of subsidized

predators, resource variability, and human population density on Desert Tortoise

populations in the Mojave Desert, USA. Endangered Species Research 12:167-177.

99

Kelly, E. C. 2017. Desert Kit Fox (Vulpes macrotis arsipus) food habits and competitive

interactions with Coyotes (Canis latrans) in the Mojave Desert. M.S. Thesis. California

State University, Bakersfield, California.

Legendre, P., and L. Legendre. 1998. Numerical Ecology, second edition. Elsevier Science,

Amsterdam, The Netherlands.

Leitner, P. 2009. The promise and peril of solar power. Wildlife Professional 3:48-53

List, R. and D. Macdonald. 2003. Home range and habitat use of the Kit Fox (Vulpes macrotis)

in a Prairie Dog (Cynomys ludovicianus) complex. Journal of Zoology 259:1-5.

Nelson, J. L., B. L. Cypher, C. D. Bjurlin, and S. Creel. 2007. Effects of habitat on competition

between Kit Foxes and Coyotes. Journal of Wildlife Management 71:1467-1475.

Ralls, K., and P. J. White. 1995. Predation on San Joaquin Kit Foxes by larger canids. Journal of

Mammalogy 76:723-729.

Scherer, R. D., and J. A. Tracey. 2011. A power analysis for the use of counts of egg masses to

monitor Wood Frog (Lithobates sylvaticus) populations. Herpetological Conservation and

Biology 6:81–90.

Taylor, B. L., and T. Gerrodette. 1993. The uses of statistical power in conservation biology: the

Vaquita and Northern Spotted Owl. Conservation Biology 7:489-500.

Turner, R. M. 1994. Mojave Desert scrub. Pages 157-168 in D. E. Brown, editor. Biotic

communities, southwestern United States and northwestern Mexico. University of Utah

Press, Salt Lake City, Utah.

United States Climate Data. 2014. Precipitation data for Barstow, California.

http://www.usclimatedata.com/climate/barstow/california/united-states/usca0069/2014/8

Accessed September 2014.

100

United States Fish and Wildlife Service. 1998. Recovery Plan for Upland Species of the San

Joaquin Valley, California. United States Fish and Wildlife Service, Portland, Oregon.

White, P. J., K. Ralls, and R. A. Garrott. 1994. Coyote-Kit Fox interactions as revealed by

telemetry. Canadian Journal of Zoology 72:1831-1836.

White, P. J., K. Ralls, and C. A. Vanderbilt White. 1995. Overlap in habitat and food use

between Coyotes and San Joaquin Kit Foxes. Southwestern Naturalist 40:342-349.

101

Tables

Table 1. The number of camera stations visited by Desert Kit Foxes (Vulpes macrotis arsipus) and Coyotes (Canis latrans) for each survey year and for all years combined in the Mojave

Desert, California, USA, from 2012 to 2014.

Stations visited by

Year Kit foxes only Coyotes only Both species Total operable cameras

2012 8 5 17 30

2013 10 2 17 29

2014 8 2 18 28

All years 26 9 52 87

102

Figure 1. Study area located in the Mojave Desert in California, USA (inset), which is bounded by the Fort Irwin National Training Center to the north and Interstate 15 and State Route 58 to the south. The blue boundary is the specific study area and the black circles denote annual camera station locations.

103

Figure 1

Fort Irwin

0

'· ~ l

Hinkley

N 0 5 10 30 40 Km A

104

CHAPTER 5

CONCLUSIONS

Desert Kit Fox foraging patterns

I found that the foraging patterns of Desert Kit Foxes (Vulpes macrotis arsipus) at my study site were similar to those of Desert Kit Foxes in Utah (Kozlowski et al. 2008) and to San

Joaquin Kit Foxes (Vulpes m. mutica) in California (White et al. 1995, Clark et al. 2005). In particular, rodents and invertebrates were the primary items consumed. I also found that despite a drought-related decrease in rodent abundance in the last three years of the study, Desert Kit

Foxes continued to rely heavily on rodents, primarily kangaroo rats (Dipodomys spp.) and pocket mice (Perognathus and Chaetodipus spp.). This further supports past research indicating that kit foxes are specialists on heteromyid rodents (Morell 1972, Fisher 1981, Cypher et al. 2000). As drought conditions progressed in the Mojave Desert, Desert Kit Foxes increased their consumption of other food items, including reptiles and birds. This broadening of the diet was an expected response based on optimal foraging theory (Pyke et al. 1977). Although invertebrates were always regularly consumed, a notable increase in the consumption of these prey during the drought indicated that invertebrates also provide an important supplement during periods of decreased rodent abundance.

I also found that the consumption of some food items by Desert Kit Foxes exhibited notable seasonal variation and this was most likely influenced by life-history traits of the prey.

Reptiles are most active during spring and summer and therefore use by kit foxes were higher during these times. Similarly, birds were consumed most frequently during spring when adults are tending to the nest and young birds are vulnerable to kit fox predation. 105

I periodically found anthropogenic material in the Desert Kit Fox scats that most likely came from areas near the few small towns and interspersed private inholdings within the study area. I found that the use of anthropogenic items was highest in the last year (third year of drought) and during winter, most likely due to a decline in overall prey abundance during these times. The most common anthropogenic item that the Desert Kit Foxes consumed was Pistachio

(Pistacia vera) nuts from the small Pistachio orchards present within portions of my research area. Interestingly, although the foxes consumed Pistachio nuts, they rarely ate native fruits that were present within my study area.

In summary, my results indicated that Desert Kit Foxes are rodent and insect specialists, but become more generalists when these items are less abundant. Thus, Desert Kit Foxes can exhibit considerable ecological plasticity, which may help them respond and adjust to seasonality and drought. This plasticity may also allow Desert Kit Foxes to better adapt to anthropogenic changes in their environment that might affect prey populations.

Competitive interactions between Desert Kit Foxes and Coyotes

My research also suggests that the potential for exploitative competition does exist between Desert Kit Foxes and Coyotes (Canis latrans) in the Mojave Desert in California. Both species consumed many of the same food items and exhibited high dietary overlap throughout my five-year study. Exploitative competition, though, may have been mitigated to some extent by a disproportionate use of prey items. Desert Kit Foxes and Coyotes in my study relied heavily on rodents for food, but Desert Kit Foxes also extensively consumed invertebrates, while

Coyotes preferentially consumed lagomorphs. The Coyotes within my study had an overall more generalized and diverse diet than Desert Kit Foxes regardless of the year or season. Besides

106

lagomorphs, Coyotes also commonly consumed birds, reptiles, fruit, and anthropogenic material, but Desert Kit Foxes consistently relied mostly on rodents and invertebrates and consumed other items less frequently than did Coyotes.

My results were consistent with optimal foraging theory in that both of the study species are most likely consuming items that maximize energy input while minimizing energy output

(Pyke et al. 1977, Pianka 1978). Item selection by predators is influenced by the size of the predator as well as the size, abundance, and ease of capture of the prey item (Pyke et al. 1977,

Pianka 1978). Thus, Coyotes regularly preferred larger prey items while Desert Kit Foxes focused on relatively smaller prey. For instance, the occurrence of lagomorphs in Coyote scats was typically more than five times higher than that in Desert Kit Fox scats, including during the drought when relative abundance of lagomorphs had dramatically declined.

Differential use of food items by Coyotes and Desert Kit Foxes also contributed to resource partitioning. The specialization by Desert Kit Foxes on heteromyid rodents and invertebrates coupled with the more infrequent predation on lagomorphs may be a strategy that decreases competition with Coyotes and lead to coexistence. Additional behaviors, such as habitat partitioning and the daily use of dens by kit foxes may further promote coexistence. Both species, though, broadened their diets as the availability of primary food items declined and this may have resulted in an increase in competition for remaining food resources. Furthermore, it is possible that as both species spend more time and effort foraging during these periods, the potential for interference competition was heightened. However, I did not measure levels of competition in my study.

I also found that when natural foods were less abundant, both Desert Kit Foxes and

Coyotes increased their consumption of anthropogenic material. However, Coyotes seemed to

107

rely on anthropogenic subsidization more than Desert Kit Foxes. In Year 5, when the drought was in its third consecutive year, the occurrence of anthropogenic items in Coyote scats was more than double that of Desert Kit Foxes. Also, the use of anthropogenic foods by both species was highest during winter when prey items declined in density or became more difficult to obtain because many prey hibernate, go into torpor, or migrate during the winter. Unfortunately, the use of unnatural food sources has the potential to cause adverse interactions between co- occurring species and their prey.

In summary, exploitative competition potentially occurs between Desert Kit Foxes and

Coyotes and the intensity of this competition may vary annually and seasonally with fluctuations in the abundance of food items. Competition may be decreased through resource partitioning, habitat avoidance, and den use by Desert Kit Foxes. The availability of anthropogenic food items can pose a threat to the Mojave Desert ecosystem. Anthropogenic subsidization may promote larger predator populations than normal, thus enhancing competition between co- occurring species and predation pressure on remaining food supplies.

Relative distribution and competitor associations

I found that the proportion of camera stations visited by both species was similar or more than the proportion of camera stations visited by either species individually. Thus, Desert Kit

Foxes did not appear to partition habitat on a landscape scale to avoid competition with Coyotes.

I also found that Desert Kit Foxes inhabited a majority of my study site.

The proportion of camera stations visited by both species was highest in 2014, which coincided with the third consecutive drought year in the Mojave Desert and a decrease in abundance of both rodents and lagomorphs. Therefore, due to drought conditions, both Desert

108

Kit Foxes and Coyotes may have broadened their foraging area and encountered more of my camera stations. As a result, the potential for interference competition may have also increased during this time.

Competition between Desert Kit Foxes and Coyotes may have been mitigated in other ways including partitioning habitat on a temporal scale (List and Macdonald 2003), daily den use

(White et al. 1995), and partitioning resources by Desert Kit Foxes (Cypher and Spencer 1998,

White et al. 1995, Kelly 2017). These evolved strategies potentially used by kit foxes to mitigate competition with Coyotes would permit the two species to coexist sympatrically in the Mojave

Desert in California without the need for habitat partitioning on a landscape level. This possible balance, though, may become threatened if habitat conversion continues and Coyotes gain more of a competitive advantage due to anthropogenic subsidization. If so, conservation and management strategies may be necessary to protect the Desert Kit Fox in the future.

Potential threats to Desert Kit Foxes and recommendations

Desert Kit Foxes in California are not currently protected under the Federal or State

Endangered Species Acts, but this could change as research reveals emerging threats and as human encroachment increases. Desert Kit Foxes in California have been increasingly treated as a Species of Special Concern, but a petition to protect the subspecies under the California

Endangered Species Act was rejected by the California Department of Fish and Wildlife predominately due to insufficient data (Kadaba et al., unpubl. report, D. Kadaba, pers. comm.).

My study provided data on dietary patterns and food preferences of Desert Kit Foxes in

California spanning multiple years, seasons, and drought. My research also examined potential competitive interactions with Coyotes and how drought conditions and anthropogenic

109

subsidization may enhance these interactions. I found that there is a high potential for Desert Kit

Foxes to experience exploitative competition from Coyotes, although this may be mitigated by partitioning of available food items, habitat avoidance, and den use by Desert Kit Foxes. By conserving large areas of healthy, intact habitat, Desert Kit Foxes can mitigate the effects of competition with Coyotes and thereby persist in the Mojave Desert.

Coyotes are a highly adaptable species that are not only expanding their range, but regularly thrive within human-altered landscapes (Bekoff 1977, Sheldon 1992). Thus, human presence and disturbance may threaten kit fox populations because of resource subsidization of

Coyotes. Arjo et al. (2007) found that man-made water sources increased Coyote abundance and therefore placed higher competitive pressures on the local Desert Kit Fox population. During a multi-year drought, Cypher and Spencer (1998) similarly noted that Coyotes increased their consumption of anthropogenic food items when native prey declined. Anthropogenic subsidization during times of decreased prey availability can not only cause negative consequences for sympatric predators by potentially intensifying exploitative and interference competition, but also increase predation pressure on the remaining prey populations resulting in even further declines in food resources (Rodewald et al. 2011, Newsome and van Eeden 2017).

Therefore, if habitat conversion continues in the Mojave Desert, a proactive approach should be taken to limit the availability of human food and water resources that could artificially enhance

Coyote abundance.

An increase in Coyote abundance resulting from an increase in human resources may also pose a disease threat to kit foxes. Coyotes are capable of carrying multiple diseases that can infect and potentially kill kit foxes. In Bakersfield, California, Coyotes are suspected of infecting the San Joaquin Kit Fox population with Sarcoptic Mange (Sarcoptes scabei; Cypher et

110

al. 2017) and as a result, dozens of kit foxes have died (B. L. Cypher, pers. comm.). Coyotes with mange have also been observed on camera stations in the Mojave Desert and the Carrizo

Plains National Monument (B. L. Cypher, pers. comm.). Lastly, a mange outbreak of epidemic proportions is occurring among a Coyote population that congregates at the Fort Irwin National

Training Center Sanitary Landfill (J. L. Rudd, pers. comm.) and transmission of mange to the local Desert Kit Fox population is possible.

Rabies is also a potential threat to kit fox populations. During a study at Camp Roberts,

San Miguel, California, 6.3% of the research kit foxes died from rabies and was a possible factor that contributed to their local extirpation (Standley et al. 1992, White et al. 2000). Canine distemper and canine parvovirus are additional potential threats to kit fox populations. In 2011, a canine distemper outbreak of unknown origin occurred in a Desert Kit Fox population causing multiple deaths and a localized population reduction (Clifford et al. 2013). Although disease has not been found to be a primary factor in the decline of most kit fox populations, stress due to habitat modifications and possible overcrowding can increase their susceptibility to diseases

(Acevedo-Whitehouse and Duffus 2009, Clifford et al. 2013).

In the United States, the increasing support for renewable energy facilities places large expanses of the Mojave Desert at risk for habitat loss from utility-scale solar plants (Leitner

2009, Lovich and Ennen 2011). Also, as the human population continues to grow, cities and towns further expand into remaining intact ecosystems (Lovich and Bainbridge 1999). In addition to the possible artificial increase in the Coyote population, the effects of livestock, grazing, roadways, off-road vehicles, pipelines, landfills, and other human activities have had and continue to have detrimental effects on this delicate ecosystem (Lovich and Bainbridge

1999). Lastly, with human development comes the potential use of toxic chemicals such as

111

rodenticides (United States Fish and Wildlife Service [USFWS] 1998, McMillin et al. 2008).

Most likely due to their dietary specialization on rodents, kit foxes are especially susceptible to secondary exposure and lethal poisoning from rodenticides (McMillin et al. 2008).

Currently, the Desert Kit Fox population in the Mojave Desert has not been as severely impacted by human encroachment and land conversion as the endangered San Joaquin Kit Fox

(USFWS 1983, USFWS 1998) and the Desert Kit Fox in Colorado (Meaney et al. 2006).

Therefore, we have the opportunity to employ strategies to protect the Desert Kit Fox populations in California and mitigate human impacts to the extent possible. The U.S. Fish and

Wildlife Service recovery plan for upland species of the San Joaquin Valley of California provides information and recovery strategies for multiple species, including the listed San

Joaquin Kit Fox (USFWS 1998). Conservation strategies in this plan could potentially be adapted for Desert Kit Foxes and thus help keep them from declining to a threatened or endangered state. For example, maintaining large areas of intact habitat may allow for the conservation of multiple robust Desert Kit Fox populations and enhance the probability of maintaining healthy populations in the Mojave Desert. This would also promote and possibly maintain a diverse and healthy population of prey species, which is critical in the event of declines in primary food items. Desert Kit Foxes are rodent and invertebrate specialists, but they do have the ability to expand their diet when primary items are less abundant, such as during drought conditions (White et al. 1995, Cypher and Spencer 1998, Kelly 2017). In addition, the recovery plan suggests that adequate land corridors should be established to maintain gene flow between populations and potentially lessen the detrimental effects of habitat loss and fragmentation (USFWS 1998). It is important to have a robust gene pool to maintain a healthy

112

population and if habitat conversion continues in the Mojave Desert then habitat corridors between populations is crucial.

With the possibility of further human encroachment and modification of the Mojave

Desert ecosystem, the need for increased conservation and management strategies for the Desert

Kit Fox may be necessary in the coming decades. To assist in the development of these strategies, further research on this relatively understudied subspecies is necessary to provide a robust and comprehensive understanding of Desert Kit Foxes in California. In particular, information on demographic (e.g., survival, causes of mortality, reproductive rates) and ecological (e.g., space use, movements, habitat selection) patterns as well as the effects of various anthropogenic activities (e.g., solar farms, roads, military exercises, urbanization) would be invaluable in the development of such strategies.

LITERATURE CITED

Acevedo-Whitehouse, K., and A. L. J. Duffus. 2009. Effects of environmental change on wildlife

health. Philosophical Transactions of the Royal Society B 364:3429-3438.

Arjo, W. M., E. M. Gese, T. J. Bennett, and A. J. Kozlowski. 2007. Changes in Kit Fox-Coyote-

prey relationships in the Great Basin Desert, Utah. Western North American Naturalist

67:389-401.

Bekoff, M. 1977. Canis latrans. Mammalian Species 79:1-9.

Clark, Jr., H. O., G. D. Warrick, B. L. Cypher, P. A. Kelly, D. E. Williams, and D. E. Grubbs.

2005. Competitive interactions between endangered Kit Foxes and nonnative Red Foxes.

Western North American Naturalist 65:153-163.

113

Clifford, D., L. Woods, M. W. Gabriel, J. Rudd, E. J. Dubovi, K. Terio, F. Uzal, A. Nyaoke, A.

De La Mora, S. Diab, et al. 2013. Canine distemper outbreak in free-ranging Desert Kit

Foxes inhabiting a solar energy development zone. Abstract, Western Section of the

Wildlife Society 2013 Annual Conference, Reno, Nevada.

Cypher, B. L., J. L. Rudd, T. L. Westall, L. W. Woods, N. Stephenson, J. E. Foley, D.

Richardson, et al. 2017. Sarcoptic mange in endangered Kit Foxes (Vulpes macrotis

mutica): case histories, diagnoses, and implications for conservation. Journal of Wildlife

Diseases 53:46-53.

Cypher, B. L., and K. A. Spencer. 1998. Competitive interactions between Coyotes and San

Joaquin Kit Foxes. Journal of Mammalogy 79:204-214.

Cypher, B. L., G. D. Warrick, M. R. M. Otten, T.P. O’Farrell, W. H. Berry, C. E. Harris, T. T.

Kato, P. M. McCue, J. H. Scrivner, and B. W. Zoellick. 2000. Population dynamics of

San Joaquin Kit Foxes at the Naval Petroleum Reserves in California. Wildlife

Monographs145:1-43.

Fisher, J. L. 1981. Kit Fox diet in south-central Arizona. M.S. thesis, University of Arizona,

Tucson, Arizona.

Kelly, E. C. 2017. Desert Kit Fox (Vulpes macrotis arsipus) food habits and competitive

interactions with Coyotes (Canis latrans) in the Mojave Desert. M.S. Thesis. California

State University, Bakersfield, California.

Kozlowski, A. J., E. M. Gese, and W. M. Arjo. 2008. Niche overlap and resource partitioning

between sympatric Kit Foxes and Coyotes in the Great Basin Desert of western Utah.

American Midland Naturalist 160:191-208.

Leitner, P. 2009. The promise and peril of solar power. Wildlife Professional 3:48-53.

114

List, R., and D. Macdonald. 2003. Home range and habitat use of the Kit Fox (Vulpes macrotis)

in a Prairie Dog (Cynomys ludovicianus) complex. Journal of Zoology 259:1-5.

Lovich, J. E., and D. Bainbridge. 1999. Anthropogenic degredation of the southern California

desert ecosystem and prospects for natural recovery and restoration. Environmental

Management 24:309-326.

Lovich, J. E., and J. R. Ennen. 2011. Wildlife conservation and solar energy development in the

desert southwest, United States. BioScience 61:982-992.

McMillin, S. C., R. C. Hosea, B. F. Finlayson, B. L. Cypher, and A. Mekebri. 2008.

Anticoagulant rodenticide exposure in an urban population of San Joaquin Kit Fox. Pp.

163-165 in Proceedings of the 23rd Vertebrate Pest Conference. University of California,

Davis.

Meaney, C. A., M. Reed-Eckert, and G. P. Beauvais. 2006. Kit Fox (Vulpes macrotis): A

Technical Conservation Assessment. USDA Forest Service, Rocky Mountain Region,

Golden, Colorado.

Morrell, S. 1972. Life history of the San Joaquin Kit Fox. California Fish and Game 58:162-174.

Newsome, T. M., and L. M. van Eeden. 2017. The effects of food waste on wildlife and humans.

Sustainability 9:1–9.

Pianka, E. R. 1978. Evolutionary Ecology, 2nd ed. Harper and Row Publishers, New York, New

York.

Pyke, G. H., H. R. Pulliam, and E. L. Charnov. 1977. Optimal foraging: a selective review of

theory and tests. Quarterly Review of Biology 52:137-154.

Rodewald, A. D., L. J. Kearns, and D. P. Shustack. 2011. Anthropogenic resource subsidies

decouple predator-prey relationships. Ecological Applications 21:936–943.

115

Sheldon, J. W. 1992. Wild Dogs: The Natural History of the Nondomestic Canidae. Academic

Press, San Diego, California.

Standley, W. G., W. H. Berry, T. P. O’Farrell, and T. T. Kato. 1992. Mortality of San Joaquin

Kit Fox (Vulpes velox macrotis) at Camp Roberts Army National Guard Training Site,

California. United States Department of Energy Topical Report Number EGG 10617-

2157, EG&G/EM Santa Barbara Operations, National Technical Information Service,

Springfield, Virginia.

United States Fish and Wildlife Service (USFWS). 1983. San Joaquin Kit Fox Recovery Plan.

United States Fish and Wildlife Service, Portland, Oregon.

United States Fish and Wildlife Service (USFWS). 1998. Recovery Plan for Upland Species of

the San Joaquin Valley, California. United States Fish and Wildlife Service, Portland,

Oregon.

White, P. J., W. H. Berry, J. J. Eliason, and M. T. Hanson. 2000. Catastrophic decrease in an

isolated population of Kit Foxes. Southwestern Naturalist 45:204-211.

White, P. J., K. Ralls, and C. A. Vanderbilt White. 1995. Overlap in habitat and food use

between Coyotes and San Joaquin Kit Foxes. Southwestern Naturalist 40:342-349.

116