Clark, H. O., Jr. 2001. Endangered San Joaquin Kit Fox and Non-native Red Fox: Interspecific Competitive Interactions. MS thesis, California State University, Fresno. 54 pages.

ABSTRACT ENDANGERED SAN JOAQUIN KIT FOX AND NON-NATIVE RED FOX: INTERSPECIFIC COMPETITIVE INTERACTIONS I investigated the interference and exploitation competition between two species of fox: the endangered native San Joaquin kit fox and the non-native red fox. Seven kit foxes and 16 red foxes were radiocollared and tracked via radio telemetry near Lost Hills, California. Home range overlap occurred between the two species; however the activity cores of the individuals did not overlap, indicating that spatial partitioning was occurring. One kit fox was killed, but not eaten, by a red fox, indicating interference competition. Coyotes were the main cause of death for both species of fox. The presence of both the red fox and coyote in kit fox ranging areas may present a negative additive effect on the survivability of the kit fox. The employment of mechanisms by kit foxes, such as year-round den use, may allow coexistence between kit foxes and coyotes, but not necessarily red foxes. Howard Orman Clark, Jr. May 2001 ENDANGERED SAN JOAQUIN KIT FOX AND NON-NATIVE RED FOX: INTERSPECIFIC COMPETITIVE INTERACTIONS by Howard Orman Clark, Jr. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Biology in the College of Science and Mathematics California State University, Fresno May 2001 APPROVED For the Department of Biology David Grubbs (Chair) Biology Brian Tsukimura Biology Patrick Kelly Biology Brian Cypher Biology For the Graduate Committee: Dean, Division of Graduate Studies AUTHORIZATION FOR REPRODUCTION OF MASTER’S THESIS ___________I grant permission for the reproduction of this thesis in part or in its entirety without further authorization from me, on the condition that the person or agency requesting reproduction absorbs the cost and provides proper acknowledgment of authorship. ___________Permission to reproduce this thesis in part or in its entirety must be obtained from me. Signature of thesis writer:_____________________________________________ ACKNOWLEDGMENTS I wish to thank several people and organizations for making my fox research possible. The staff at the Endangered Species Recovery Program were a tremendous help in assisting me with the project; those include Elaine Sheehan, Geoff Gray, Adam Harpster, Steve Clifton, Laurissa Hamilton, Tamra Sandoval, Rachel Zwerdling-Morales, Ron Batie, Mark McFall, Patrick Morrison, Michelle Selmon, Justine Smith, Christine Van Horn Job, Jack McMullin, Cecilia Lopez, and Cristina Lopez. I thank Scott Phillips and Paul Brandy for assisting with telemetry formulae and ArcView® applications. Special thanks to Greg Warrick for constructing the mobile telemetry systems, conducting the necropsies, assisting in the small mammal trapping analyses, data collection and interpretation, and trapping and handling foxes. I thank Daniel F. Williams for securing funds to conduct the research from the following agencies: The Bureau of Reclamation, the US Fish and Wildlife Service, and the California Department of Fish and Game. I thank Katherine Ralls for providing the project with some kit fox radio collars and Linda Spiegel for loaning the project some radio telemetry equipment. I thank California Department of Fish and Game pilots Rich Anthes and Gary Schales and Inland Flight Training Center pilot Monya Constantinescu for performing daytime and v nighttime aerial searches for missing foxes. I thank Robert Warner and Len Marino, from the California Department of Water Resources, for issuing temporary permits to access the California Aqueduct. I thank my Thesis Committee, David Grubbs (Chair), Brian Tsukimura, Patrick Kelly, and Brian Cypher, for their time and patience in assisting me with the thesis manuscript process. I thank California State University, Fresno Dean’s Office for granting a Faculty Sponsored Research Award for the academic year 1999-00 and a Graduate Student Research Award to help fund my research goals. I thank various private land owners for granting permission to survey small mammals on their lands. Lastly, and most importantly, I thank my family for their support. TABLE OF CONTENTS Page LIST OF TABLES . LIST OF FIGURES INTRODUCTION . Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii ix 1 1 6 6 9 15 15 18 18 23 25 27 30 32 METHODS AND MATERIALS . Study Area Methods RESULTS Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . Space Use Overlap (1998) . Space Use Overlap (1999) . Temporal Interactions in 1998 . Temporal Interactions in 1999 . Habitat Selection: Use-Availability Small Mammal Abundance . Diet Analysis . . . . . . . . . vii Page DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 43 45 55 56 CONCLUSIONS REFERENCES CITED . APPENDICES . . . A. CALCULATION OF A FOX LOCATION B. OVERLAPPING MINIMUM CONVEX POLYGON AND 95% ADAPTIVE KERNEL INTERSPECIFIC HOME RANGES FOR ALL FOXES . . . . . . . . . . . . . . . 58 LIST OF TABLES Table 1. Radio-tracking data collection period for all foxes with overlapping home ranges in 1998 . . . . . . . . . . . . 2. Radio-tracking data collection period for all foxes with overlapping home ranges in 1999 . . . . . . . . . . . . 3. Summary of MCP home range data for red foxes and overlap with Kit fox 5575 (1998). . . . . . . . . . . . . . 4. Summary of MCP home range data for kit foxes and overlap with red fox 5761 (1999) . . . . . . . . . . . . . 5. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for kit foxes in 1998 . . . . . . . 6. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for red foxes in 1998 . . . . . . . 7. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for kit foxes in 1999 . . . . . . . 8. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for red foxes in 1999 . . . . . . . 9. Results of single item chi-square independence test . . . . . Page 16 17 20 20 29 29 29 30 34 LIST OF FIGURES Figure 1. 2. 3. 4. 5. 6. 7. 8. 9. Lost Hills Study Area. . . . . . . . . . . . . . . . . . . . . . . . . Page 7 19 21 22 24 26 26 27 28 28 34 Overview of all fox locations gathered in 1998 . Red fox home range vs. overlap with kit fox 5575 in 1998 . Overview of all fox locations gathered in 1999 . . . . Encounter of kit fox 5575 with four juvenile red foxes on 26 August 1998 . . . . . . . . . . . . Temporal interactions between two kit foxes and one red fox on 18 November 1998 . . . . . . . . . . . . . Temporal interactions between a kit fox and red fox on 30 Sept. 1999 . . . . . . . . . . . . Temporal interactions between a kit fox and red fox on 22 Nov. 1999 . . . . . . . . . . . . Red and kit fox use-availability of habitat in 1998 . . . . . . . . . . . . . . . . . . . 10. Red and kit fox use-availability of habitat in 1999 . 11. Food item use for red foxes and kit foxes in 1999 . INTRODUCTION The San Joaquin kit fox (Vulpes macrotis mutica, Mercure and others 1993) is a federally “endangered” and California “threatened” species occurring from the central to southern San Joaquin Valley in California (United States Fish and Wildlife Service 1998). Adult male kit foxes weigh on average 2.2 kg and adult females average 1.9 kg. Kit fox body length averages 788 (range 730 to 840) mm, with an average tail length of 290 mm (McGrew 1979). While habitat loss, pesticide use, and competition with native carnivores are major threats to kit fox survival, competitive interactions with non-native carnivores may exacerbate their ability to survive. It is the aim of this study to examine the ecological relationships between kit foxes and red foxes in order to determine the impact of the latter on the kit fox. Background Historically, the kit fox ranged from Tracy, California, in the north, to Kern County in the south. The east-west range extended from the eastern edge of the San Joaquin Valley to the Salinas Valley (Grinell and others 1937a). The former range of the San Joaquin kit fox has been significantly reduced (Jensen 1972; Swick 1973; Morrell 1975; O’Farrell 1983). Threats to this small fox are loss of habitat due to agriculture, industrialization, and urbanization (Morrell 1972; 2 United States Fish and Wildlife Service 1993). Pesticide and rodenticide use may also constitute a threat to kit fox survival (Schitoskey 1975; Littrell 1990) as well as predator control programs (Bunker 1940; Kitchen and others 1999). Interspecific competition from other predators, such as the coyote (Canis latrans), in the form of interference and exploitation competition (Case and Gilpin 1974), have been identified as a threat to kit foxes (Cypher and Scrivner 1992; Disney and Spiegel 1992; Ralls and White 1995; Cypher and Spencer 1998). Adult male coyotes are usually heavier and larger than adult females; approximately 8 to 20 kg as opposed to 7 to 18 kg, respectively. Coyote body length varies from 1.0 to 1.35 m (Bekoff 1977). The non-native red fox (Vulpes vulpes) was introduced into the Sacramento Valley, California in the 1870s from the Midwest (Grinnell and others 1937b; Lewis and others 1993) and since then has spread as far as San Luis Obispo County (Ralls and White 1995; Cohn 1998), Orange County, and Los Angeles County, California (Jurek 1992). Red foxes weigh 3-7 kg and are about 1 m in length, with a 320-mm tail (Samuel and Nelson 1982). There is not enough information to determine what sort of impact the red fox has on swift fox (Vulpes velox) populations, a closely related species, and this may be the situation for kit foxes as well (Stromberg and Boyce 1986). The rapidly expanding red fox is invading kit fox habitat and appear to be replacing them in some areas (Jurek 1992, Ralls and White 1995). For example, Ralls and White (1995) reported two 3 kit fox mortalities due to the red fox in their study on the Carrizo Plain, California, indicating interference competition between the species. There is little information available concerning interactions between fox species in general (Hersteinsson and Macdonald 1992). However, a number of studies have been conducted that document the competitive interactions between canid species. For example, competitive interactions between coyotes and kit foxes have been studied, concluding that the larger coyote engages in exploitation and interference competition with the kit fox (White and others 1994; Cypher and Spencer 1998). Ecological relationships have been studied between coyotes and swift foxes and interference competition was evident (Kitchen and others 1999). Carbyn (1982) studied coyote population fluctuations in relation to wolf (Canis lupus) territories and found that at high wolf population levels, coyote numbers dropped. Mech (1995) also reported that high densities of wolves reduce coyote populations. In early California, wolves were probably present (Brewer 1974; Schmidt 1991), but drastically declined in numbers with the establishment and expansion of livestock by the Spanish in the Eighteenth century (Paradiso and Nowak 1990). Wolves have since been extirpated from the State (Mech 1995). In this early setting, wolves may have kept coyote numbers relatively low (Parker 1995), or at least stable, because coyotes do not compete well with wolves (Mech 1974; Beckoff 1977), thus reducing the competitive pressures on the kit fox. Once 4 wolves were extirpated, coyotes were the top carnivore in the San Joaquin Valley, increasing in numbers and significantly competing with the kit fox. This is similar to the dynamics described as “mesopredator release:” when large mammalian predators drop in population or go extinct in an area, especially in fragmented landscapes, and smaller carnivores increase in numbers (Soulé and others 1988; Courchamp and others 1999; Crooks and Soulé 1999). Theberge and Wedeles (1989) studied sympatric coyote and red fox populations, reporting that compensatory competitive behaviors, such as habitat partitioning, may allow coexistence. Gese and others (1996) studied interactions between coyotes and red foxes as well and found that at times coyotes tolerated red foxes, but at other times killed them. It has been suggested that coyotes may serve as a buffer between red foxes and kit foxes (Cohn 1998). When coyotes appear in an area, red foxes avoid them. This avoidance behavior may reduce the additive effect of having two larger canids in areas where kit foxes occur. Kit foxes, on the other hand, coevolved with the coyote, and have developed strategies to escape the coyote and reduce interference and exploitation competition. Kit foxes den underground year-round, whereas the coyote (and the red fox) only use dens during natal activities. When the kit fox senses a coyote in the area, it will escape by running to one of the numerous dens (>20) maintained in its home range (Morrell 1972; White and others 1995; Cypher and Spencer 1998). 5 Overlap in food preference may be a key factor in competitive interactions between canid species. Cypher (1993) reported that coyotes, gray foxes (Urocyon cinereoargenteus), and red foxes used similar food items, and the diets of coyotes and red foxes were the most similar. White and others (1995) studied the overlap in food use between coyotes and kit foxes, reporting that high resource overlap between the two canid species led to competition for resources. San Joaquin kit foxes and non-native red foxes also may compete for food resources where they co-occur. Exploring the factors currently impacting San Joaquin kit fox populations is necessary to develop effective management plans and recovery strategies. Resource and spatial overlap between kit foxes and red foxes could indicate that control of red foxes is required for the continued survival of the kit fox. In this study, the roles of exploitation and interference competition by spatial and dietary analyses are examined. The following objectives will be addressed: (1) identify sources of mortality for kit foxes and red foxes, (2) determine space use and habitat use patterns, (3) determine prey availability by habitat type, and (4) determine the diet of both species and dietary overlap. METHODS AND MATERIALS Study Area The study site was located along an approximately 32-km stretch of the California Aqueduct (aqueduct) near the community of Lost Hills, California (Figure 1). This site was selected because a population of kit foxes and red foxes occurred together in one area. Both sides of the aqueduct include a relatively undisturbed strip of land (approximately 60 m wide) typical of the Valley Grassland vegetation type (Heady 1977). Red brome (Bromus madritensis) and filaree (Erodium spp.) dominate the herbaceous vegetation. Common shrubs include desert saltbush (Atriplex polycarpa) and spiny saltbush (Atriplex spinifera). Honey mesquite trees (Prosopis glandulosa) are found within the southern portion of the study area, and occasional almond and pistachio trees are found in areas where the aqueduct borders orchards. Farmland borders both sides of the aqueduct throughout most of the study area. Major crops include cotton, barley, almonds, and pistachios. Less abundant crops include alfalfa, onions, lettuce, watermelon, olives, tomatoes, and vineyards. Annual crops are typically planted in late winter and harvested in the fall. After harvesting, the ground is disked and left bare until the following spring. Pistachio Figure 1. Lost Hills Study Area. 8 and almond groves are drip irrigated and harvested in October of each year. The west side of the aqueduct is bordered with farmland (approx. 1.6 km wide), which is bordered further west with the Lost Hills oil field. The Lost Hills oil field is primarily owned and operated by private oil companies. Although some portions of the Lost Hills oil field are heavily developed, there are significant expanses of natural vegetation typical of the Valley Grassland type throughout much of the field. The study area is predominately flat with elevations ranging from approximately 80 m in the east to approximately 150 m along the Lost Hills anticline. The Lost Hills, forming the western edge of the study area, are gentle, rolling hills that run in a northwest to southeast direction paralleling the aqueduct. Climate for Lost Hills, California is characterized by hot, dry summers, and wet, cool winters, with thick fog during the winter months (National Climatic Data Center 2000). Weather data recorded 40 km east of Lost Hills in Wasco, California indicate that average daily maximum temperatures range from 13.4°C in December to 37.5°C in July and average daily minimum temperatures range from 2.1°C in December to 18.7°C in July. Precipitation averages 13 cm annually. Reported precipitation for 1998 was 16.3 cm and 5.8 cm in 1999. 9 Methods Kit foxes were captured using wire-mesh traps (38 x 38 x 107 cm) baited with canned mackerel, wieners, bacon, or chicken (O’Farrell 1987). Red foxes were captured by plunging them from drainage culverts into a handling bag (plunger consisted of plastic pipe lengths attached together with a foam ball attached to one end). Captured foxes were ear-tagged, measured, weighed, and fitted with radiocollars (Advanced Telemetry Systems, Isanti, MN) containing motion sensors (mortality sensors activated after 8 h of non-movement). Each collar, with battery, weighed approximately 60 g. Radiocollar mass was <5% of the animal body mass (Eberhardt and others 1982). Foxes were released at the capture site. Foxes were radio-tracked for 2 years using two truck-mounted null tracking systems with paired 2-element antennae (White 1985, Kenward 1987). Stations were located along access roads of the aqueduct and separated by approximately 800 m. After setup and calibration of the systems, biologists at two adjacent stations simultaneously took bearings on radiocollared kit foxes and red foxes (Mech 1970, 1983). A telemetry session was initiated approximately 1 h before sunset and continued for approximately 4.5 h. The first 3-5 h after sunset is typically when kit and swift fox activity is highest (Zoellick 1990; Hines and Case 1991; Kitchen and 10 others 1999; Cypher and others 2000). Locations were collected on all collared foxes in the vicinity and successive locations on individual foxes were separated by 10 min. Kit fox and red fox locations were calculated using a model (see Appendix A) described in White and Garrott (1990). Two trucks at known locations collect simultaneous location fixes. Three azimuths (referencing true north) are obtained: (1) the azimuth to the fox from the south truck, (2) the azimuth to the fox from the north truck and (3) the azimuth from the south truck to the north truck. Information gathered from Global Positioning System (GPS) units, United States Geological Survey (USGS) maps, and from ground mapping was entered into a geographic information system (GIS) for analyses of habitat use (ARC/INFO®, Environmental Systems Research Institute, Redlands, CA). A survey grade GPS Pathfinder Pro XR/XRS (Trimble Navigation Limited, Sunnyvale, CA) was used to determine the locations of the telemetry stations, and delineate the boundaries of the aqueduct right-of-way (ROW). Roads, section lines, and other pertinent data were taken from USGS topographic maps. Telemetry locations of foxes were also added as a GIS layer for analyses. “Home range,” as it is used in this work, is defined as “a more or less restricted area within which an animal moves when performing its normal activities” (Harris and others 1990). Home range size was calculated using the 11 minimum convex polygon (MCP) method (Mohr 1947) and the 95% adaptive kernel method (Worton 1989; Kernohan and others 1998). Points collected throughout the year were used in calculating home range size for each fox. The kernel method is used to determine the utilization distribution (UD) of the foxes, providing a probability density estimation (Worton 1989). An ArcView® (Environmental Systems Research Institute, Redlands, CA) program extension was used to calculate the kernel home range (Hooge and Eichenlaub 1997). The smoothing parameter (h) was calculated using least-squares cross-validation (LSCV). The adaptive kernel estimate using LSCV to determine h produced the best results (Worton 1989). A small value of h allows the fine detail of the data to be observed, and a large value of h obscures all but the most prominent features. Four probability contours were assigned, 0.95, 0.75, 0.50, and 0.25 (similar to Seaman and Powell 1996). Home range size was determined only on those foxes with >30 locations (Chamberlain and Leopold 2000). The core area of an animal is defined as “the portion of an animal’s home range that exceeded an equal-use pattern” (Samuel and others 1985; Samuel and Green 1988). Core areas can be used to denote central areas of consistent or intense use (Kaufmann 1962). Using the 95% adaptive kernel method, areas in the home range that fall within the 25% probability contour are considered “core areas.” 12 Accuracy of the telemetry system was determined by having two people gather bearings on radiocollars placed at locations known only by a third person. Locations derived from telemetry were then compared to the actual locations of the radiocollars (recorded using a survey grade GPS unit) to determine the average telemetric error (Springer 1979; Zimmerman and Powell 1995). The average error was 37.9 6.8 m (range = 4-186 m). Eighty percent of the triangulated locations had an error of < 45 m. The telemetry vehicles averaged 552.2 34.7 m (range = 74-1318 m) from the reference transmitters. To determine which habitat types the two fox species have in common, only those kit and red foxes with overlapping home ranges were included in the habitat selection analysis. To ensure data independence, one random location per fox per telemetry session was selected (Swihart and Slade 1985). The fox locations were plotted in ArcView® and each location was assigned a habitat type. Habitat availability was determined as being a 1.6 km buffer east and west of the aqueduct, and 1.6 km from the most southerly and most northerly fox points. The 1.6 km mark is the maximum resolution of the telemetry equipment. Utilizationavailability analysis was conducted using the method described in Neu and others (1974) and Byers and others (1984). Similar land types were grouped together for analysis. Almonds, olives, and pistachios were grouped together as “orchards”; cotton, barley, and tomatoes 13 were grouped together as “row crops,” and aqueduct ROW, vineyard, and grassland (exotic grasses) were allowed to stand alone. The category “other” was used to group small parcels of tilled and miscellaneous land and “residential” referred to any farmhouse, equipment staging area, or farm equipment storage yard. Kit fox and red fox diet in 1999 was determined by analyzing scats (fecal deposits) collected from trapped foxes, and from scats collected at known fox dens. Scats were oven-dried for 24 h at 60oC to facilitate handling and to destroy cysts of zoonotic parasites. Prey remains were identified by macroscopic characteristics of hairs (Mayer 1952, Stains 1958) and through comparison of teeth, bones, scales, skin, exoskeletons, and seeds with reference specimens (Roest 1991). Frequency of occurrence, diversity, and overlap of food diet was determined between species (Horn 1966, Hutcheson 1970, Bower and others 1990, Zar 1999). During necropsies, cause of death was determined based on observed injuries and abnormalities (Roy and Dorrance 1976). If the fox had contusions associated with lethal injuries caused by tooth punctures, predators were considered the cause of death. When possible, distances between marks made by a distinct pair of canines were measured to determine what species caused the death (Disney and Spiegel 1992). If the cause of death could not be determined because 14 the carcass was badly decomposed or scavenged, cause of death was classified as unknown. Rodent abundance was assessed by placing trap lines consisting of 10 traps spaced 10 m apart within the ROW and within some row crops and orchards where permission was granted. A total of 18 trap lines was placed within the ROW, 6 trap lines were placed within almond orchards, and 6 trap lines were placed within cotton fields. Traps were baited with white proso millet seed in the afternoon and checked 2-4 h after sunset. Captured rodents were identified to species, weighed, sexed, and fur-clipped on their rump to differentiate them from newly captured animals on successive nights. Trapping sessions were conducted in December (1998 and 1999) and traps were operated for three consecutive nights. The number of small mammals captured was compared between habitat types and between years using analysis of variance. RESULTS Four adult kit foxes (2 female, 2 male), 3 juvenile male kit foxes, and 16 red fox juveniles (10 females, 6 males) were captured and radiocollared during both years of this study (1998 and 1999). However, only those red foxes and kit foxes having overlapping home ranges were analyzed. The radio-tracking data collection periods for all red foxes and kit foxes with overlapping home ranges during both years are summarized in Tables 1 and 2. Mortality During this 2-year study, 4 radiocollared kit foxes (2 adults, 2 juveniles) were recovered dead. Three kit foxes (1 adult, 2 juveniles) were killed by coyotes, and 1 adult kit fox was killed by a red fox (50% mortality rate due to coyote attacks, 66% due to predator attacks overall). Out of 16 radiocollared red foxes, 11 were found dead. One red fox’s cause of death could not be determined (most likely a predator kill), and 1 red fox may have drowned (collar signal coming from bottom of aqueduct). Nine were verified as coyote kills, a 56% mortality rate due to coyotes. Table 1. Radio-tracking data collection period for all foxes with overlapping home ranges in 1998. ear tag 5575 5561 5563 5564 5565 1 2 species kit fox red fox red fox red fox red fox age/sex adult male juv. male juv. female juv. female juv. male date collared 06/02/98 06/01/98 06/16/98 06/23/98 06/16/98 tot. Jul-98 19 10 39 22 37 127 months monitored (value = locations gathered) Aug-98 Sep-98 Oct-98 Nov-98 18 32 39 6 8 12 6 29 13 10 6 22 5 8 21 18 22 98 80 85 12 tot. 114 36 97 57 98 402 data end date 8/11/19992 10/29/19981 10/11/19991 10/29/19981 11/30/19981 mortality dispersed Table 2. Radio-tracking data collection period for all foxes with overlapping home ranges in 1999. ear tag species 5548 5572 5574 5575 5761 1 2 age/sex date collared kit fox adult female kit fox juv. male kit fox juv. male kit fox adult male red fox juv. female months monitored (value = locations gathered) Jan-99 Feb-99 Mar-99 Apr-99 May-99 Jun-99 Jul-99 Aug-99 Sep-99 Oct-99 Nov-99 Dec-99 tot. 07/01/97 10 5 16 40 4 8 7 8 11 18 5 132 05/05/99 4 12 7 18 13 1 55 05/03/99 13 14 12 39 06/02/98 35 40 23 42 3 9 12 164 05/27/99 20 11 10 20 10 22 93 tot. 45 45 39 82 11 49 26 42 45 44 28 27 483 data end date n/a 1 10/4/1999 11/03/19992 8/11/19992 n/a mortality dispersed 18 Space Use Overlap (1998) In 1998, there were 3 radiocollared kit foxes (2 female, 1 male) north of State Route 46, however no radiocollared red foxes were present. South of State Route 46, there was 1 red fox family group and a single male kit fox. Four of the red fox juveniles (2 female, 2 male) from the family group were radiocollared and tracked (Figure 2). No adult red foxes from this family group were captured, although they were spotted on occasion in the study area. All home range figures for red foxes and kit foxes for both years are in Appendix B (both MCP and 95% adaptive kernel). No fox pair had overlap in the 25% kernel contours, however there was overlap in the other contours (50% 95%). Overlap areas in the MCP home range for both years are summarized in Tables 3 and 4. The home range area (km2) in 1998 for each of the four red foxes was plotted against the home range overlap area with kit fox 5575 (Figure 3). Space Use Overlap (1999) In 1999, two red fox family groups were in the study area, one in the vicinity of Twisselman Road, and one in the vicinity of GP Road. Nine red foxes (4 female, 5 male) from the GP family group were radiocollared, however their home ranges did not overlap with the northern kit foxes (Figure 4). Three of the Figure 2. Overview of all fox locations gathered in 1998. 20 Table 3. Summary of MCP home range for red foxes and overlap with kit fox 55751 (1998) ear tag 5561 5563 5564 5565 1 home range (sq km) 2.5 7.0 3.4 3.1 overlap (sq km) 1.0 3.3 1.7 2.5 % overlap area KF RF 14 40 48 47 25 50 36 81 % pnts in overlap KF RF 11 8 17 78 10 67 15 82 5575 home range = 6.9 km2 Table 4. Summary of MCP home range for kit foxes and overlap with red fox 57612 (1999) ear tag 5548 5572 5574 5575 2 home range (sq km) 3.8 5.1 0.9 2.7 overlap (sq km) 0.8 0.7 0.3 0.6 % overlap area KF RF 21 14 14 12 36 5 24 11 % pnts in overlap KF RF 10 56 14 12 7 49 32 9 5761 home range = 5.7 km2 21 3.5 Overlap with kit fox (sq km) 3 2.5 2 1.5 1 0.5 2 3 4 5 6 7 Red fox home range (sq km) overlap (sq km) Linear (overlap (sq km)) y = 0.4x + 0.5 R2 = 0.71 Figure 3. Red fox home range vs. overlap with kit fox 5575 in 1998. This correlation shows that the larger the red fox home range, the more it overlaps with the kit fox home range. Figure 4. Overview of all fox locations gathered in 1999. 23 red foxes (all female) from the Twisselman family group were radiocollared. One of the radiocollared red foxes was killed 6 days after the radiocollar was fitted and before telemetry data were collected. Only 22 locations were gathered on juvenile female red fox 5571, too few to conduct analysis (Chamberlain and Leopold 2000). Only 1 red fox, juvenile female 5761, had enough telemetry locations to warrant analysis. In the center of these two red fox family groups was a kit fox family group. During the winter of 1998, adult female kit fox 5548 and adult male kit fox 5575 pair bonded. In February of 1999, three male kit fox pups were born. In May, they were live-trapped in box traps and radiocollared. One of the juvenile kits was killed by a coyote 37 days after fitting the radiocollar. Data were collected on the remaining two kit fox juveniles, along with the two adult kit foxes. Temporal Interactions in 1998 There were instances where temporal data on red foxes and kit foxes were gathered in real time. On 26 August 1998, all five foxes were located near the aqueduct ROW (Figure 5). Twenty minutes later, the kit fox (5575) maneuvered his way through four red foxes, and continued to head south. One red fox (5561) headed south as well, but not to the extent that the kit fox traveled. Twenty minutes after the encounter with the red foxes, the kit fox was nearly 1.6 km south. The kit fox probably traveled south to avoid the encroaching red foxes. Although 24 Figure 5. Encounter of kit fox 5575 with four juvenile red foxes on 26 August 1998. The kit fox tends to move away from the group of red foxes. 25 the red foxes were juveniles, they outnumbered the kit fox, perhaps constituting a threat. On 18 November 1998, an encounter between two kit foxes and one red fox was recorded via radio telemetry. Adult male kit fox 5538 and adult female kit fox 5548 were within 250 m of red fox 5563 in a twenty minute window. Kit fox 5548 approached the red fox, while the red fox and the male kit fox are moved away from each other (Figure 6). In this instance, kit foxes out-numbered the red fox, perhaps invoking an avoidance measure on the part of the juvenile red fox. Temporal Interactions in 1999 There were examples of temporal interactions between red foxes and kit foxes in 1999. Juvenile red fox 5761 apparently took an avoidance measure from an approaching adult kit fox on the 30th of September, 1999 (Figure 7). The shortest distance between the two foxes was approximately 300 m within a 1minute window. An encounter between the same foxes occurred on the 22nd of November, 1999 (Figure 8). Both foxes moved away from each other between 18:51 and 18:54. The shortest distance between the two foxes was approximately 100 m. 26 Figure 6. Temporal interactions between two kit foxes and one red fox on 18 Nov. 1998. Figure 7. Temporal interactions between a kit fox and a red fox on 30 Sept. 1999. 27 Figure 8. Temporal interactions between a kit fox and red fox on 22 Nov.1999 Habitat Selection: Use-Availability The habitats used by each fox species and amount of habitat available to them is shown in Figures 9 and 10. To test if foxes utilized each habitat category in proportion to its occurrence within the available area, the chi-square method described in Neu et al. (1974) was used (Tables 5-8). Some of the habitat types are different for each year, i.e. vineyards in 1999, and grasslands in 1998, because the fox locations analyzed occurred in different areas of the study site. 28 Red and kit fox use-availability of habitat in 1998 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% ROW orchard residential row crop grassland vineyard other available red fox actual kit fox actual Figure 9. Red and kit fox use-availability of habitat in 1998. Note that orchards are used significantly by both species, whereas row crops, although comprise most of the study area, are not preferred by either species. Red and kit fox use-availability of habitat in 1999 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% ROW orchard residential row crop grassland vineyard other available red fox actual kit fox actual Figure 10. Red and kit fox use-availability of habitat in 1999. Note that row crops comprise most of the study area, but are not preferred by either species. Orchards and aqueduct right-of-way tend to be selected for by both species. 29 Table 5. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for kit foxes in 1998 (DF = 3, χ2 = 20.01, P < 0.01). location expected actual (pi) ROW 0.043 0.053 orchard 0.361 0.842 row crop 0.442 0.105 other 0.154 0.000 *Indicates a difference at the 0.05 level of significance. confidence interval -0.079≤ p1 ≤0.185 0.627≤ p2 ≤1.058* -0.076≤ p3 ≤0.287* 0.000≤ p4 ≤0.000* Table 6. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for red foxes in 1998 (DF = 5, χ 2 = 86.4, P = < 0.01). location expected actual (pi) ROW 0.043 0.233 orchard 0.361 0.617 residential 0.016 0.017 row crop 0.442 0.017 grassland 0.131 0.117 other 0.006 0.000 *Indicates a difference at the 0.05 level of significance. confidence interval 0.144≤ p1 ≤0.323* 0.513≤ p2 ≤0.720* -0.011≤ p3 ≤0.044 -0.011≤ p4 ≤0.044* 0.048≤ p5 ≤0.185 0.000≤ p6 ≤0.000* Table 7. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for kit foxes in 1999 (DF = 3, χ 2 = 240.6, P = < 0.01). location expected actual (pi) ROW 0.050 0.362 orchard 0.130 0.319 other 0.180 0.000 row crop 0.640 0.319 *Indicates a difference at the 0.05 level of significance. confidence intervals 0.234≤ p1 ≤0.490* 0.195≤ p2 ≤0.443* 0.000≤ p3 ≤0.000* 0.195≤ p4 ≤0.443* 30 Table 8. Simultaneous confidence intervals using Bonferroni inequality for land type utilization for red foxes in 1999 (DF = 4, χ 2 = 88.4, P = < 0.01). location expected actual (pi) ROW 0.058 0.393 orchard 0.148 0.214 residential 0.008 0.071 vineyard 0.040 0.143 row crop 0.746 0.179 *Indicates a difference at the 0.05 level of significance. Small Mammal Abundance During 900 trap nights in December 1998, 183 individuals of six species of small mammals were captured 219 times, for an overall trapping success of 24.3%. Deer mice (Peromyscus maniculatus) were captured most frequently (63.4% of individuals), followed by house mice (Mus musculus; 31.1%), San Joaquin pocket mice (Perognathus inornatus; 1.6%), Heermann’s kangaroo rats (Dipodomys heermanni; 1.6%), short-nosed kangaroo rats (Dipodomys nitratoides brevinasus; 1.6%), and harvest mice (Reithrodontomys megalotis; 0.6%). The average number of rodents captured per trap line was significantly different among habitats (F[2] = 6.52, P < 0.01) and varied from 9.3 (SE = 1.77, range = 0-21) in the ROW to 2.2 (SE = 0.48, range = 0-3) in cotton fields and 0.5 (SE = 0.34, range = 0-2) in almond orchards. The average number of rodents captured per trap line was significantly higher (P = 0.04) within the ROW than within the almond orchards. Average number of rodents captured per line did not confidence interval 0.241≤ p1 ≤0.545* 0.087≤ p2 ≤0.342 -0.009≤ p3 ≤0.151 0.034≤ p4 ≤0.252 0.060≤ p5 ≤0.298* 31 differ between the ROW and cotton fields (P = 0.12) or between cotton fields and almond orchards (P = 0.88). All six species of rodents were captured along the ROW, whereas only deer mice and house mice were captured within cotton fields and only house mice were captured within almond orchards. The average number of species captured per line was significantly different among habitats (F[2] = 7.57, P < 0.01) and varied from 1.8 (SE = 0.21, range = 0-3) within the ROW to 1.2 (SE = 0.31, range = 0-2) in cotton fields and 0.3 (SE = 0.21, range = 0-1) in almond orchards. The average number of species captured per trap line was significantly higher within the ROW than within almond orchards (P = 0.01). Average number of species captured per trap line did not differ between the ROW and cotton fields (P = 0.40) or between cotton fields and almond orchards (P = 0.19). In December 1999, all trap lines within the ROW and 3 trap lines within almond orchards were set in the same locations as 1998. However, because of access problems, 3 trap lines in almond orchards were moved to new locations. Trap lines were not set within the cotton fields, because the fields had been plowed under. During 840 trap nights in December 1999, 16 individuals (13 deer mice and 3 Heermann’s kangaroo rats) were captured 17 times for an overall trapping success of 2.0 %. Because all captures were within the ROW habitat, results were not compared statistically. 32 Diet Analysis Kit Fox Diet In 1999, 207 kit fox scats were gathered from active dens, captured animals and during necropsies. Most (204) scats were collected from pupping dens in April (32.4%), June (64.3%), and July (1.9%). The remaining 3 scats (1.4%) were collected from known individuals in July, August, and October. Rodents were the most frequently occurring item in kit fox scats (88.4%), followed by insects (18.4%), other arthropods (11.6%), leporids (8.7%), human derived items (6.3%), and birds (1.9%). Species of rodents that occurred in kit fox scats included house mice (34.3%), deer mice (17.9 %), pocket gophers (Thomomys bottae; 9.7%), California voles (Microtus californicus; 3.9%), harvest mice (3.4%), and San Joaquin pocket mice (1.5%). In addition, 27.0 % of the scats contained murid rodents that could not be identified to species and 4.8% of the scats contained rodents that could not be identified to species. Insect species included field crickets (9.7%), grasshoppers (4.4%), ants (4.4%) and beetles (2.9%). Other arthropod remains were not identifiable. Bird remains in scats typically consisted of a few feathers and were not identified to species. Humanderived items included plastic (1.9%), string (1.9%), paper (1.5%), and rubber (1.0%). 33 Red Fox Diet In 1999, 140 red fox scats were gathered from known red fox dens in February (10%), June (67%), and September (23%). Murids were the most frequently occurring item in red fox scats (91.4%), followed by insects (16.4%) leporids (11.4%), birds (7.1%), and human-derived items (4.9%). Species of rodents that occurred in red fox scats included California voles (31.4%), house mice (28.6%), deer mice (4.3%), pocket gophers (2.9%), and harvest mice (0.7%). In addition, 27.1% of the scats contained murid rodents that could not be identified to species and 6.4% of the scats contained rodents that could not be identified to species. Insect species included ants (7.9%), field crickets (7.1%), and beetles (1.4%). Bird remains in scats typically consisted of a few feathers and were not identified to species. Human-derived items included paper (2.8%), plastic (0.7%), string (0.7%), and rubber (0.7%). Most of the scats contained some sort of vegetation, grass and seeds from brome. Four scats (2.8%) contained almonds, and one scat contained a barley seed head. Diet Overlap Analysis Items were grouped to simplify analyses (Table 9 lists the analyzed groups, Figure 11 shows all items prior to grouping). Horn’s index, R0, was calculated to determine the amount of overlap between diets. A value of 1.0 means the diets are identical, and a value of zero (0) means the diets have nothing in common. The 34 Table 9. Results of single item chi-square independence test. Food item voles house mice murids deer mice pocket gophers leporids orthoptera birds arthropods other 1 not significant kit fox 8 71 56 37 20 18 29 4 39 45 red fox 44 40 38 6 4 16 10 10 16 18 χ2 74.4 5.41 8.5 14.9 11.7 12.6 9.5 19.0 8.2 8.0 Food item use in 1999 40.0% 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% unk mammal unk murid human-derived pocket gopher California vole harvest mice other arthropods grasshopper pocket mice house mice deer mice 0.0% leporid unk rodent field cricket beetles birds ants kit fox red fox Figure 11. Food item use for red foxes and kit foxes in 1999. Note that red foxes prefer California voles over other rodent species, but do prefer murids in general. Vole use by red foxes may indicate diurnal foraging activities. 35 calculated R0 value was 0.87, indicating the diet overlap between the fox species was high. A Shannon index of diversity, H’, was calculated for each species. The Shannon index for kit foxes was 0.91 and 0.90 for red foxes. There is no significant difference between these two indices. A 2 x 10 contingency table chi-square test was conducted on the dietary data. The test result was significant (χ2 = 78, DF = 9, P < 0.01). The frequencies of occurrence of items for one species was significantly different than the frequencies of occurrence of items for the other species. A 2 x 2 contingency table chi-square test was conducted on each item to determine if specific use of an item by the two species was significant (DF = 1 for all tests, P = 0.01; Table 9). All but house mice were significant. DISCUSSION In this study, I set out to do the following: identify sources of mortality for kit foxes and red foxes, determine space use and habitat use patterns, determine prey availability by habitat type, and determine dietary overlap and the diet of both species. A red fox killed a kit fox during the study, indicating that kit foxes do experience interference competition from red foxes. However, both fox species experience interference competition from a larger canid, the coyote. Coyotes were seen throughout the study area during the 2-year study period. Red foxes were more abundant than kit foxes on the study site, and most likely encountered coyotes more often. Simultaneous and sequential radio telemetry demonstrated spatial interaction between red foxes and kit foxes. Kit fox avoidance of red foxes was documented in this study. Core activity areas did not overlap, which suggests that some spatial partitioning does exist between the species. However, portions of the home ranges that are infrequently used do overlap, indicating a degree of spatial competition. Some habitat types seemed to be selected over other habitat types. Kit foxes used habitat in orchards and the aqueduct ROW more than expected and 37 used row crops less than expected, based on availability. Likewise, red foxes used orchards and aqueduct ROW more than expected, based on availability. In addition, red foxes also used grasslands, vineyards, and residential areas, but not significantly. This may indicate that red foxes are more adaptable and use other habitat types to their advantage. Kit foxes and red foxes frequently used the aqueduct ROW and the orchards, indicating that exploitation competition for these limited habitat types appears likely. Orchards and the aqueduct ROW may be selected by foxes for a couple reasons. First, the ROW has a higher small mammal density due to the availability of recovered native vegetation for leporids and rodents. This provided a prey base for the foxes. Second, orchards are generally open areas where foxes can easily see an approaching predator, like coyotes or other foxes, and can quickly move out of the area, avoiding an ambush. Orchards do harbor rodents, like house mice, gophers, and possibly voles (Clark 1984), which have proved to be a food source for foxes. In comparison, row crops are often densely packed with a monoculture crop, limiting visibility and maneuverability, leaving a fox vulnerable to coyote attack. In addition, row crops appear low in prey base numbers, but may contain voles (Clark 1984). Another reason row crops lack small mammal diversity, and are not readily used by foxes, is the lack of stability; row crops are often rotated (i.e., cotton 1 year, barley the next). This rotation results in frequent soil 38 disturbance due to plowing. In comparison, the aqueduct ROW and orchards are more stable, with little soil disturbance, and have a relatively intact ecosystem throughout the year. Both fox species had diverse diets. There was no significant difference between the two Shannon indices of diet diversity. Horn's similarity index indicated high overlap between red foxes and kit foxes, potentially enhancing resource competition. However, frequencies of occurrence of food items between species was significantly different (except for house mice), indicating that both species prey on similar items, but did not consume them in the same proportions. The difference in proportions implies some dietary partitioning; the two species are competing for similar prey items, but shift their preferences in a certain direction, thereby reducing competition. This may be an adaptive mechanism to offset competition. Red foxes favored California voles, which is consistent with their food habits throughout their range (Samuel and Nelson 1990). Voles were not captured during small mammal surveys, probably because the bait type was not effective in capturing voles. Voles prefer a wetter environment, which may be provided in crop fields and orchards. Red foxes displayed a more diverse use of habitat, and therefore can take advantage of vole populations throughout the study area. 39 Kit foxes seemed to prefer insects, but this may be an artifact of gathering scat samples at pupping dens, where most of the scat could possibly belong to pups. Pups are not very experienced at capturing prey, and seem to practice on crickets and grasshoppers to possibly improve their hunting skills. Red foxes are very effective at competing with smaller canids. Red foxes were used as biological control agents for arctic foxes on Alaskan islands (Bailey 1992). Red foxes possibly excluded arctic foxes from the best feeding areas, rather than outright killing them (Rudzinski and others 1982). When environmental conditions limit resources, red foxes dominated the smaller arctic foxes (Hersteinsson and Macdonald 1992). During times of low rainfall and mammalian prey decline, kit foxes experience a decrease of reproductive success and a decrease in home range overlap, leading to spatial partitioning (White and Ralls 1993). In addition, White and others (1996) reported that kit foxes did not switch to another prey type during a short-term decline of their preferred prey type. Red foxes may be more adaptable and readily use other prey types during times of prey decline. However, there is evidence that kit foxes are capable of switching prey preferences during natural predator-prey cycles (Cypher and others 2000). Red foxes do not use dens year-round, and have no immediate way to escape an approaching coyote. Coyotes are believed to influence red fox 40 abundance and distribution (Dekker 1989, Sargeant and Allen 1989). Sargeant and others (1987) suggested that red foxes express an avoidance mechanism in which red foxes will move out of areas where coyotes are moving in, and thus interactions may not result in the actually killing of the red fox. However, coyotes may not be able to totally displace red foxes over large areas (Voigt and Earle 1983). Gese and others (1996) reported that although coyotes were tolerant of red foxes, levels of prey resources may influence the tolerance level. During years of low prey abundance, interspecific aggression by coyotes towards red foxes would be higher, resulting in avoidance of coyotes by red foxes. The key to the red fox’s success is staying out of the coyote’s way. In contrast, kit foxes use dens year-round as an escape mechanism from approaching coyotes. Kit foxes have coevolved with coyotes and developed mechanisms to enhance their survival, such as resource partitioning and greater dietary breadth, facilitating a level of coexistence (Cypher and Spencer 1998). Hence, kit foxes and coyotes can use the same area (White and others 1994). However, these mechanisms may not be as effective against the non-native red fox. Although red foxes are larger than kit foxes, there are kit fox dens that red foxes can fit into. White and others (2000) reported that red foxes have usurped several dens that were used by kit foxes during previous years at one study site. 41 Therefore, the larger red fox may pose a greater threat to the kit fox in some areas (Ralls and White 1995). Red foxes are able to survive in a multitude of environments. They range in more habitats that any other canid (Samuel and Nelson 1990). Red foxes are able to survive in intermixed cropland, farmland, shrubland, mixed hardwood stands, and edges of open areas. They take particular advantage of edge habitats (Henry 1996a, 1996b), like agricultural fields abutting against canals and aqueducts. Interespecific interactions with coyotes can drive red foxes to live closer to human habitation, another frontier that red foxes have adapted very well to (Dekker 1983, Baker and Harris 2000). Many human activities create suitable edges for red foxes to exploit (Henry 1996a, 1996b). Due to their ability to readily adapt to altered habitats, they have an advantage over the native kit fox, which may be more selective in the habitat types it lives in. Red foxes require free water, unlike the kit fox (Golightly and Ohmart 1983). Human habitation of the semi-arid and arid regions of California, where kit foxes are normally found, result in the existence of anthropogenic water sources, like canals, stock ponds, aqueducts, and irrigated agricultural crops (Reisner 1993). These readily available water sources facilitate colonization by red foxes. In addition, red foxes commonly engage in surplus killing and egg caching, giving 42 them an advantage over other canids during periods of low food resource availability (Schmidt 1985). Red foxes have been known to harass swift foxes, by urinating outside swift fox burrows (Stewart 1999). These sorts of behaviors may be prevalent between kit foxes and red foxes as well. If red foxes exclude kit foxes from their established home ranges and known escape dens, the kit fox is more exposed to interference competition with the coyote. Ironically, the coyote may be an effective management tool. Coyotes have been suggested as a biological control strategy for red foxes in coastal areas of California where the foxes are preying on endangered California least terns (Sterna antillarum browni) and California light-footed clapper rails (Rallus longirostris levipes, Jurek 1992). Coyotes have also been recommended in controlling red fox presence in the Prairie Pothole Region of North America to reduce red fox predation on duck nests (Sargeant and Arnold 1984). White and others (2000) cautioned against the removal of coyotes in kit fox habitat where red foxes are also present. Coyote removal via management practices may increase the abundance and distribution of red foxes, increasing the potential for competitive interactions with kit foxes. CONCLUSIONS Coyotes were the major source of mortality for both red foxes and kit foxes. However, red foxes have been identified in this study as a source of mortality for the kit fox, indicating interference competition. Both species of foxes prefer using the aqueduct ROW and orchards, and do not seem to prefer using row crops like cotton and barley. The aqueduct ROW has a high prey base, indicating that competition for this limited habitat type may exist between the species. Both species of fox preferred murid rodents, especially house mice, and the red fox preferred California voles specifically. Both species have diverse diets, which is expected for two opportunistic, generalist foragers. Dietary overlap was high, implying that both species prey on similar items, but did not consume them in the same proportions, indicating the presence of dietary resource partitioning. Exploitation competition between the two fox species is evident from these dietary and habitat use overlap analyses. More studies are required to fully understand the dynamics between native and non-native canids in the Central Valley of California. The conservation of large blocks of habitat is a paramount goal for the recovery of kit foxes and a number of other rare species that occur sympatrically with kit foxes (United States Fish and Wildlife Service 1998). Management of the non-native red fox in these 44 blocks of habitat may be necessary to assist in kit fox survival. The presence of coyotes in these conserved habitat blocks may be a significant factor in controlling red fox numbers. Red foxes may experience higher rates of interference competition from coyotes because they are formidable competitors and occur in multiple habitats. They also may have a higher dietary overlap, increasing the potential for exploitation competition. The more effort the coyote expends on red foxes, the less it spends competing with the kit fox. Once the threat of the red fox is lowered, the kit fox can coexist with the coyote by continuing to use the various mechanisms evolved to enhance their survival. In addition, providing artificial dens capable of excluding both red foxes and coyotes may be key for the continued survival of the endangered San Joaquin kit fox. REFERENCES CITED REFERENCES CITED Bailey EP. 1992. Red foxes, Vulpes vulpes, as biological control agents for introduced arctic foxes, Alopex lagopus, on Alaskan islands. The Canadian Field Naturalist 106:200-205. Baker PJ, Harris S. 2000. Interaction rates between members of a group of red foxes (Vulpes vulpes). Mammal rev. 30:239-242. Bekoff M. 1977. Canis latrans. Mammalian Species 79:1-9. Bower JE, Zar JH, von Ende CN. 1990. Field and laboratory methods for general ecology. Wm. C. Brown Publishers, Dubuque, IA. 237 pp. Brewer WH. 1974. Up and Down California in 1860-1864: The Journal of William H. Brewer. (Ed. F. P. Farquhar.) University of California Press, Berkeley, California. 583 pp. Bunker CD. 1940. The kit fox. Science 92:35-36. Byers CR, Steinhorst RK, Krausman PR. 1984. Clarification of a technique for analysis of utilization-availability data. J. Wildl. Manage. 48:1050-1053. Carbyn LN. 1982. Coyote population fluctuations and spatial distribution in relation to wolf territories in Riding Mountain National Park, Manitoba. Canadian Field-Naturalist. 96:176-183. Case TJ, Gilpin ME. 1974. Interference competition and niche theory. Proc. Nat. Acad. Sci. 71:3073-3077. Chamberlain MJ, Leopold BD. 2000. Spatial use patterns, seasonal habitat selection, and interactions among adult gray foxes in Mississippi. J. Wildl. Manage. 64:742-751. Clark JP. 1984. Vole control in field crops. Pages 5-6 in: Proceedings of the 11th vertebrate pest control conference (DO Clark, Ed.). University of California, Davis. 47 Cohn J. 1998. A dog-eat-dog world? BioScience 48:430-434. Courchamp F, Langlais M, Sugihara G. 1999. Cats protecting birds: modeling the mesopredator release effect. Journal of Animal Ecology 68:282-292. Crooks KR, Soulé ME. 1999. Mesopredator release and avifaunal extinctions in a fragmented system. Nature 400:563-566. Cypher BL. 1993. Food item use by three sympatric canids in southern Illinois. Transactions of the Illinois State Academy of Science 83:139-144. Cypher BL, Scrivner JH. 1992. Coyote control to protect endangered San Joaquin kit foxes at the Naval Petroleum Reserves, California. Pages 42-47 in: Proceedings of the 15th vertebrate pest conference (J. E. Borrecco and R. E. Marsh, Eds.). University of California, Davis. Cypher BL, Spencer KA. 1998. Competitive interactions between coyotes and San Joaquin kit foxes. Journal of Mammalogy 79:204-214. Cypher BL, Warrick GD, Otten MR, O'Farrell TP, Berry WH, Harris CE, Kato TT, McCue PM, Scrivner JH, Zoellick BW. 2000. Population dynamics of San Joaquin kit foxes at the Naval Petroleum Reserves in California. Wildlife Monographs 145. 43pp. Dekker D. 1983. Denning and foraging habits of red foxes, Vulpes vulpes, and their interaction with coyotes, Canis latrans, in Central Alberta, 1972-1981. Canadian Field-Naturalist 97:303-306l. Dekker D. 1989. Population fluctuations and spatial relationships among wolves, Canis lupus, coyotes, Canis latrans, and red foxes, Vulpes vulpes, in Jasper National Park, Alberta. Canadian Field-Naturalist 103:261-164. Disney M, Spiegel LK. 1992. Sources and rates of San Joaquin kit fox mortality in western Kern County, California. Transactions of the western section of the wildlife society 28:73-82. Eberhardt LE, Hanson WC, Bengtson JL, Garrot RA, Hanson EE. 1982. Arctic fox home-range characteristics in an oil-development area. J. Wildl. Manage. 46:183-190. 48 Gese EM., Stotts TE, Grothe S. 1996. Interactions between coyotes and red foxes in Yellowstone National Park, Wyoming. Journal of Mammalogy 77:377-382. Golightly RT, Ohmart RD. 1983. Metabolism and body temperature of two desert canids: coyotes and kit foxes. Journal of Mammalogy 64:624-635. Grinnell J, Dixon J, Linsdale J. 1937a. Kit foxes. Pages 399-420 in: Fur bearing animals of California. Vol. 2. (Eds: J. Grinnell, J. Dixon, and J. Linsdale) Univ. Calif. Press, Berkeley, CA. Grinnell J, Dixon J, Linsdale J. 1937b. Red foxes. Pages 377-398 in: Fur bearing animals of California. Vol. 2. (Eds: J. Grinnell, J. Dixon, and J. Linsdale) Univ. Calif. Press, Berkeley, CA. Harris S, Cresswell WJ, Forde PG, Trewhella WJ, Woollard T, Wray S. 1990. Home-range analysis using radio-tracking data - a review of problems and techniques particularly as applied to the study of mammals. Mammal Rev. 20:97123. Heady HF. 1977. Valley Grassland. Pages 491-514 in: M. C. Barbour and J. Major, eds. Terrestrial vegetation of California. John Wiley and Sons, New York, N. Y. Henry JD. 1996a. Living on the Edge: Foxes. NorthWord Press, Minocqua, Wisconsin. 143 pp. Henry JD. 1996b. Red Fox: The Catlike Canine. Smithsonian Institution Press, Washington, DC. 174 pp. Hersteinsson P, Macdonald DW. 1992. Interspecific competition and the geographical distribution of red and arctic foxes Vulpes vulpes and Alopex lagopus. Oikos 64:505-515. Hines TD, Case RM. 1991. Diet, home range, movements, and activity periods of swift fox in Nebraska. Prairie Naturalist 23:131-138. Hooge PN, Eichenlaub B. 1997. Animal movement extension to ArcView. ver. 1.1. Alaska Biological Science Center, U.S. Geological Survey, Anchorage, AK, USA. 49 Horn HS. 1966. Measurement of “overlap” in comparative ecological studies. The American Naturalist 100:419-424. Hutcheson K. 1970. A test for comparing diversities based on the Shannon formula. Journal of Theoretical Biology 29:151-154. Jensen CC. 1972. San Joaquin kit fox distribution. U. S. Fish and Wildlife Services, Bureau of Sport Fisheries and Wildlife, Division of Wildlife Services, Sacramento, California. 22 pp. Jurek RM. 1992. Nonnative red foxes in California. California Department of Fish and Game, Nongame Bird and Mammal Section Report 92-04. 15 pp. Kaufmann JH. 1962. Ecology and social behavior of the coati, Nasua narica on Borrow Colorado Island, Panama. University of California, Publications in Zoology. 60:95-222. Kenward R. 1987. Wildlife Radio Tagging. Academic Press, San Diego, California. 222 pp. Kernohan BJ, Millspaugh JJ, Jenks JA, Naugle DE. 1998. Use of an adaptive kernel home-range estimator in a GIS environment to calculate habitat use. Journal of Environmental Management 53:83-89. Kitchen, AM, Gese EM, Schauster ER. 1999. Resource partitioning between coyotes and swift foxes: spaces, time, and diet. Canadian Journal of Zoology 77:1645-1656. Lewis JC, Sallee KL, Golightly LT Jr. 1993. Introduced red fox in California. California Department of Fish and Game, Nongame Bird and Mammal Section Report 93-10. 70 pp. Littrell EE. 1990. Effects of field vertebrate pest control on nontarget wildlife (with emphasis on bird and rodent control). Pages 59-61 in: Proceedings of the 14th vertebrate pest conference (L. R. Davis and R. E. Marsh, Eds.). University of California, Davis. McGrew JC. 1979. Vulpes macrotis. Mammalian Species 123:1-6. 50 Mayer WV. 1952. The hair of California mammals with keys to the dorsal guard hairs of California mammals. The American Midland Naturalist 48:480-512 Mech LD. 1970. Current techniques in the study of elusive wilderness carnivores. Pages 315-322 in: XIth International Congress of Game Biologists. Stockholm, September 3-7. Mech LD. 1974. Canis lupus. Mammalian Species 37:1-6. Mech LD. 1983. Handbook of animal radio-tracking. The University of Minnesota Press. 107 pp. Mech LD. 1995. The wolf: The ecology and behavior of an endangered species. University of Minnesota press, Minneapolis, Minnesota. 384 pp. Mercure A, Ralls K, Koepfli KP, Wayne RK. 1993. Genetic subdivisions among small canids: mitochondrial DNA differentiation of swift, kit, and arctic foxes. Evolution 47:1313-1328. Mohr CO. 1947. Table of equivalent populations of North American small mammals. American Midland Naturalist. 37:223-249. Morrell SH. 1972. Life History of the San Joaquin Kit Fox. California Fish and Game 58:162-174. Morrell SH. 1975. San Joaquin Kit Fox distribution and abundance in 1975. Administrative Report, 75-3, California Department of Fish and Game, Wildlife Management Branch, Sacramento, CA. 27 pp. National Climatic Data Center. 2000. Local climatological data, Wasco, California, USA. National Climatological Data Center, Asheville, North Carolina. Neu CW, Byers CR, Peek JM. 1974. A technique for analysis of utilizationavailability data. J. Wildl. Manage. 38:541-545. O'Farrell TP. 1983, San Joaquin Kit Fox Recovery Plan, U.S. Fish and Wildlife Service, Sacramento, CA. 51 O'Farrell TP. 1987. Kit fox. Pages 422-431 in: Wild Furbearer Management and Conservation in North America. (Eds: M. Novak, J. A. Baker, M. E. Obbard, and B. Malloch) Ministry of Natural Resources, Ontario, Canada. Paradiso JL, Nowak RM. 1990. Wolves. Pages 460-474 in: Wild mammals of North America: Biology, Management, Economics (Eds: J. A. Chapman and G. A. Feldhamer). The John Hopkins University Press, Baltimore, Maryland. Parker G. 1995. The Eastern coyote: The story of its success. Nimbus Publishing, Halifax, Nova Scotia. 254pp. Ralls K, White PJ. 1995. Predation on San Joaquin kit foxes by larger canids. Journal of Mammalogy 76:723-729. Reisner M. 1993. Cadillac Desert: The American west and its disspearing water. Penguin Books USA Inc, New York, New York. 582 pp. Roest AI. 1991. A key-guide to mammal skulls and lower jaws. Mad River Press, Inc. Eureka, California. 39 pp. Roy LD, Dorrance MJ. 1976. Methods of investigating predation of domestic livestock. Alberta Agriculture, Edmonton. 53 pp. Rudzinski DR, Graves HB, Sargeant AB, Storm GL. 1982. Behavioral interactions of penned red and arctic foxes. J.Wildl. Manage. 46:877-884. Samuel DE, Nelson BB. 1990. Foxes. Pages 475-490 in: Wild mammals of North America: Biology, Management, Economics (Eds: J. A. Chapman and G. A. Feldhamer). The John Hopkins University Press, Baltimore, Maryland. Samuel MD, Green RE. 1988. A revised test procedure for identifying core areas within the home range. Journal of Animal Ecology 57:1067-1068. Samuel MD, Pierce DJ, Garton EO. 1985. Identifying areas of concentrated use within the home range. Journal of Animal Ecology 54:711-719. Sargeant AB, Allen SH. 1989. Observed interactions between coyotes and red foxes. Journal of Mammalogy 70:631-633. 52 Sargeant AB, Allen SH, Hastings JO. 1987. Spatial relations between sympatric coyotes and red foxes in North Dakota. J. Wildl. Manage. 51:285-293. Sargeant AB, Arnold PM. 1984. Predator management for ducks on waterfowl production areas in the northern plains. Pages 161-167 in: Proceedings of the 11th vertebrate pest control conference (D. O. Clark, Ed.). University of California, Davis. Schitoskey F Jr. 1975. Primary and secondary hazards of three rodenticides to kit fox. J. Wildl. Manage. 39:416-418. Schmidt RH. 1985. Controlling arctic fox populations with introduced red foxes. Wildl. Soc. Bull. 13:592-594. Schmidt RH. 1991. Gray wolves in California: their presence and absence. California Fish and Game 77:79-85. Seaman DE, Powell RA. 1996. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology 77:2075-2085. Soulé ME, Bolger DT, Alberts AC, Wright J, Sorice M, Hill S. 1988. Reconstructed dynamics of rapid extinctions of chaparral-requiring birds in urban habitat islands. Conservation Biology 2:75-92. Springer JT. 1979. Some sources of bias and sampling error in radio triangulation. J. Wildl. Manage. 43:926-935. Stains HJ. 1958. Field key to guard hair of middle western furbearers. J. Wildl. Manage. 22:95-97. Stewart D. 1999. Caught in a dog fight. Competing with coyotes and other larger canines, the swift fox struggles to survive in its grassland home. National Wildlife June/July. Stromberg MR, Boyce MS. 1986. Systematics and conservation of the swift fox, Vulpes velox, in North America. Biological Conservation 35:97-110. Swick CD. 1973, Determination of San Joaquin Kit Fox Range in Contra Costa, Alameda, San Joaquin, and Tulare Counties, Special Wildlife Investigation 53 Progress Report No. w-54-R4, California Department of Fish and Game, Sacramento, CA. 15 pp. Swihart RK, Slade NA. 1985. Testing for independence of observations in animal movements. Ecology 66:1176-1184. Theberge JB, Wedeles CHR. 1989. Prey selection and habitat partitioning in sympatric coyote and red fox populations, southwestern Yukon. Can. J. Zool. 67:1285-1290. United States Fish and Wildlife Service. 1993. San Joaquin kit fox recovery plan. United States Fish and Wildlife Service, Portland, Oregon. 84 pp. United States Fish and Wildlife Service. 1998. Recovery plan for upland species of the San Joaquin Valley, California. Region 1, Portland, Oregon. 319 pp. Voigt DR, Earle BD. 1983. Avoidance of coyotes by red fox families. J. Wildl. Manage. 47:852-857. White GC. 1985. Optimal locations of towers for triangulation studies using biotelemetry. J. Wildl. Manage. 49:190-196. White GC, Garrott RA. 1990. Analysis of Wildlife Radio-Tracking Data. Academic Press, San Diego, California. 383 pp. White PJ, Ralls K. 1993. Reproduction and spacing patterns of kit foxes relative to changing prey availability. J. Wildl. Manage. 57:861-867. White PJ, Ralls K, Garrott RA. 1994. Coyote - kit fox interactions as revealed by telemetry. Canadian Journal of Zoology 72:1831-1836. White PJ, Berry WH, Eliason JJ, Hanson MT. 2000. Catastrophic decrease in an isolated population of kit foxes. Southwestern Naturalist 45:204-211. White PJ., Ralls K, Vanderbilt White CA. 1995. Overlap in habitat and food use between coyotes and San Joaquin kit foxes. The Southwestern Naturalist 40:342349. 54 White PJ, Vanderbilt White CA, Ralls K. 1996. Functional and numerical responses of kit foxes to a short-term decline in mammalian prey. Journal of Mammalogy 77:370-376. Worton BJ. 1989. Kernel methods for estimating the utilization distribution in home range studies. Ecology 70:164-168. Zar JH. 1999. Biostatistical Analysis. Prentice-Hall, Englewood Cliffs, N.J., 4th edition. 663 pp. + Appendices Zimmerman JW, Powell RA. 1995. Radiotelemetry error: location error method compared with error polygons and confidence ellipses. Can. J. Zool. 73:11231133. Zoellick BW. 1990. Activity of kit foxes in western Arizona and sampling design of kit fox resource use. Pages 151-155 in: Managing Wildlife in the Southwest. (Eds: P. R. Krausman and N. S. Smith), Ariz. Chapter of the Wildl. Soc., Phoenix, AZ. APPENDICES APPENDIX A CALCULATION OF A FOX LOCATION 57 (x1,y1) α1 ∆X α2 d1 α If you plug: d 2 = ∆Y − (∆X )(tan(α 1 − π 2 ) d5 d3 2 ∆Y D d4 (x3,y3) tan α 2 = d 5d 3 tan α 3 = d 5 d 4 Into: d3 + d4 = d2 Then: d5 = d2 1   α3 (x2,y2) tan α 3 + 1 tan α 2    So, Definitions: (x,y) UTM locations: 1 - north truck, 2 - south truck, 3 - Fox α - measured angle in radians D, d, ∆, h- calculated distance x3 = x2 + d 5 y3 = y2 + d 4 Figure A. calculation of a fox location. Assumptions: (1)Two trucks at known locations collect simultaneous location fixes. (2)Three azimuths (referencing true north) obtained: azimuth to the fox from the south truck, azimuth to the fox from the north truck and the azimuth from the south truck to the north truck. APPENDIX B OVERLAPPING MINIMUM CONVEX POLYGON AND 95% ADAPTIVE KERNEL INTERSPECIFIC HOME RANGES FOR ALL FOXES Figure A. MCP home range and overlap for kit fox 5575 and red fox 5561 in 1998. Figure B. MCP home range and overlap for kit fox 5575 and red fox 5563 in 1998. Figure C. MCP home range and overlap for kit fox 5575 and red fox 5564 in 1998. Figure D. MCP home range and overlap for kit fox 5575 and red fox 5565 in 1998. Figure E. Kernel home range for kit fox 5575 in 1998. Figure F. Kernel home range for red fox 5561 in 1998. Figure G. Kernel home range for red fox 5563 in 1998. Figure H. Kernel home range for red fox 5564 in 1998. Figure I. Kernel home range for red fox 5565 in 1998. Figure J. Kernel home range overlap for kit fox 5575 and red fox 5561 in 1998. Figure K. Kernel home range overlap for kit fox 5575 and red fox 5563 in 1998. Figure L. Kernel home range overlap for kit fox 5575 and red fox 5564 in 1998. Figure M. Kernel home range overlap for kit fox 5575 and red fox 5565 in 1998. Figure N. MCP home range and overlap for kit fox 5548 and red fox 5761 in 1999. Figure O. MCP home range and overlap for kit fox 5572 and red fox 5761 in 1999 (inset showing detail). Figure P. MCP home range and overlap for kit fox 5574 and red fox 5761 in 1999. Figure Q. MCP home range and overlap for kit fox 5575 and red fox 5761 in 1999. Figure R. Kernel home range for red fox 5761 in 1999. Figure S. Kernel home range for kit fox 5548 in 1999. Figure T. Kernel home range for kit fox 5572 in 1999. Figure U. Kernel home range for kit fox 5574 in 1999 Figure V. Kernel home range for kit fox 5575 in 1999. Figure W. Kernel home range overlap for kit fox 5548 and red fox 5761 in 1999. Figure X. Kernel home range overlap for kit fox 5572 and red fox 5761 in 1999. Figure Y. Kernel home range overlap for kit fox 5574 and red fox 5761 in 1999. Figure Z. Kernel home range overlap for kit fox 5575 and red fox 5761 in 1999.