TABLE OF TABLES
Table 1. Cause of death for all foxes (1985-2004) and foxes with transmitting radio signals at time of death (1997-2004), Bakersfield, CA................................................................................................... .......... 11 Table 2. Frequency of fox-vehicle collisions by road type for all foxes and those with transmitting radio signals at time of death, Bakersfield, CA, 1998-2004......................................................................... 11 Table 3. Fox-vehicle collisions by posted speed limit, Bakersfield, CA, 1998-2004.................................. 12 Table 4. Observed versus random generated fox-vehicle collisions on arterial and collector roads that were within one road width (26m) of an intersection with a linear right-of-way, Bakersfield, CA, 1998- 2004. .......................................................................................................................... ......................... 16
LIST OF FIGURES
Figure 1. Urbanized San Joaquin kit fox study area, Bakersfield, CA. ......................................................... 4 Figure 2. Minimum convex polygon of fox-vehicle collisions for individuals that had transmitting radio collars, Bakersfield, CA, 1998-2004............................................................................................ ......... 6 Figure 3. Nocturnal (a) and Diurnal (b) locations of San Joaquin kit fox in Bakersfield, CA, 1998-2004. 10 Figure 4. Fox-vehicle collisions by traffic volume for all mortalities and mortalities of individuals with transmitting radio collars at time of death, Bakersfield, CA, 1998-2004............................................ 12 Figure 5. Fox-vehicle collisions by gender and month, Bakersfield, CA, 1998-2004................................. 13 Figure 6. Fox-vehicle collisions (1998-2004) versus scent marking rates (Murdoch 2004) for male foxes, Bakersfield, CA................................................................................................................ ................... 14 Figure 7. Fox-vehicle collisions (1998-2004) versus interactions with foxes that were not within the immediate social group (Murdoch), Bakersfield, CA. ........................................................................ 14 Figure 8. Mean (Std. Dev.) traffic volume by month at six control stations, Bakersfield, CA, 2003.......... 15 Figure 9. Fox-vehicle collisions and road segments where mortality occurred, Bakersfield, CA, 1998-2004. ............................................................................................................................... ............................. 16 Figure 10. Number of known dens by distance to nearest centerline of major road, Bakersfield, CA, 2001- 2004. .......................................................................................................................... ......................... 17 Figure 11. Drainage culvert under Highway 99 in Bakersfield, CA that was used as a den by San Joaquin kit fox........................................................................................................................ .......................... 18 Figure 12. Kit fox dens at varying distances from collector and arterial roads in Bakersfield, CA, 2001- 2004. .......................................................................................................................... ......................... 18 Figure 13. Kit fox dens at varying distances from collector and arterial roads, Bakersfield, CA, 2001-2004.
A barrier wall, indicated with dashed line, isolated dens from one roadway with relatively high traffic volume. ........................................................................................................................ ....................... 19 Figure 14. Earthen kit fox dens in urban Bakersfield, CA. a) Single entrance den under building foundation. b) Natal den complex on canal bank. c) Multi-entrance den in drainage basin bordering intersection of two arterial roads. d) Single entrance den in vacant lot............................................... 36 Figure 15. Non-earthen den locations for kit foxes in Bakersfield, CA. a) Fox emerging from stacked pipe. b) Narrow entrances to a kit fox den under a storage container. c) Fox resting beside stacked lumber at a construction site. d) Fox returning to a den within a modular housing unit. ............................... 37 Figure 16. Camouflaged earthen kit fox dens in urban Bakersfield, CA. a) Entrance to den in landscaping at parking lot. b) Thatch covered den entrance in a golf course fairway. c) Dirt mound from den in shrub-island at park........................................................................................................... .................. 38 Figure 17. Artificial subterranean dens for San Joaquin kit fox at Bakersfield, CA. a) Artificial den schematic. b) PVC two-entrance chamber den under construction. c) High-density polyethylene two- entrance den. d) PVC tunnel with floor removed longitudinally. ...................................................... 39 Figure 18. Artificial escape dens for San Joaquin kit fox at Bakersfield, CA. a) Escape den schematic. b) High-density polyethylene escape den under construction. c) Completed den. d) Kit fox entering escape den. .................................................................................................................... ...................... 40 Figure 19. Kit fox pups at artificial subterranean dens at a golf course (a) and drainage basin (b) in Bakersfield, CA................................................................................................................ ................... 41 Figure 20. Railroad underpass on highway 99 in Bakersfield, CA. a) Habitat view. b) Kit fox natal den at the top of the underpass embankment............................................................................................ ..... 41 iv
Figure 21. Barrier wall to encourage use of road crossing structures for the San Joaquin kit fox in urban environments. The roadside embankment allows foxes that are trapped on the roadway to clear the barrier wall. a) Cross-section view. B. Three-dimensional rendering............................................... 42 Figure 22. Buena Vista Road golf cart crossing structure, Bakersfield, CA. a) Aerial view of roadkill and radio-tagged kit fox locations at crossing structure. b) Habitat view from east entrance of crossing structure. c) Approach view to east entrance. d) Passage interior...................................................... 43 Figure 23. Truxtun Extension horse culvert, Bakersfield, CA. a) Aerial view of radio-tagged kit fox locations at horse culvert. b) View from pedestrian trail at south culvert entrance (canal in background). c) North culvert entrance. Yellow line depicts extent of fencing on north side of road. d) Culvert interior............................................................................................................ .................... 44 Figure 24. Coffee Road overpass at Kern River bicycle trail, Bakersfield, CA. a) Aerial view of kit fox roadkill and radio-tagged night locations at crossing structure. b) Habitat view from west entrance. c) East entrance to bicycle path on southern embankment of Kern River. d) View of roadway from East with location of kit fox roadkill in red. No fences prevent fox access to roadway on eastern side. ............................................................................................................................... ............................. 45 Figure 25. Highway 58/Fruitvale Avenue canal bridge, Bakersfield, CA. a) Aerial view of radio-tagged kit fox night locations at crossing structure. b) East entrance to Fruitvale Bridge obstructed by water. c) South entrance to highway 58 bridge. d) Dirt embankment under Highway 58 Bridge. .................... 46 v
Urban Roads and the Endangered San Joaquin Kit Fox
DISCLOSURE
This research was done under contract 65A0136. The contract total was $170,700.
DISCLAIMER
The contents of this report reflect the views of the authors who are responsible for the facts and accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of the State of California or the Federal Highway Administration. This report does not constitute a standard, specification, or regulation.
vi
Urban Roads and the Endangered San Joaquin Kit Fox
EXECUTIVE SUMMARY
The San Joaquin kit fox (Vulpes macrotis mutica) is at risk of extinction primarily due to profound habitat degradation, fragmentation, and loss. Industrial and urban developments are contributing factors to habitat destruction. Based on recent research, however, kit foxes appear to have persistent populations in some urban and industrial lands. While many factors regulate urban fox populations, roads are a dominant feature of these areas, and a potential threat to kit fox populations. Roads may affect fox survival, den use, movements, and other essential functions. Furthermore, the potential for negative impact is proportional to road width, traffic volume, and speed limit.
Projections for the San Joaquin Valley indicate that the human population and associated urbanization and transportation networks will increase for the foreseeable future. Roads will grow wider, carry higher traffic volumes, and become more inhospitable to wildlife.
As urban lands spread and natural lands diminish, urbanized fox populations also may become more valuable to species recovery. Consequently, the influence of transportation networks on the persistence of urban fox populations will grow simultaneously to the importance of those populations. Identifying ways to reduce the impact of urban roads on the kit fox is therefore important to species recovery.
The goal of this project was to examine effects of transportation networks (particularly roads) on urbanized San Joaquin kit foxes. Fox-vehicle collision was the primary cause of mortality for urban kit foxes and most strikes occurred on arterial roads that had relatively higher traffic volumes and posted speed limits. Fox-vehicle collision also showed temporal and spatial variability. Foxes more frequently were struck near intersections between major roads and other linear rights-of-way, including railroads, golf courses, canals, rivers, and other roads of all sizes. Furthermore, male kit foxes died disproportionately on roads during winter when mating, territory defense, and exploratory movements were common.
Kit foxes did not show preference for or avoidance of dens with respect to proximity to major roads (arterial, collector, highway, and highway ramp). A 100ft buffer (Fish and Wildlife Service exclusion zone) from the centerline of major roads included 9.5% of known kit fox dens at the study site. Proximity of dens (especially natal sites) to major roads, however, may increase risk of vehicle strike. Furthermore, earth moving associated with road construction and improvement may endanger kit foxes that inhabit roadside dens. Finaly, land that is converted to roadway is permanently lost as kit fox habitat. Therefore, while the selection of den sites by kit fox did not appear influenced by road proximity, there were several impacts of roads on the availability and suitability of habitat for dens.
We report on 21 sessions of vehicle-based intensive monitoring with 15 different focal kit foxes. In all, we recorded 4,633 minutes of observation over 44,121m of minimum straight line distance (MSLD) between consecutive points. Using MSLD and observation time, we calculated that foxes traveled at an average rate of 0.5958 km/hr during intensive monitoring. We directly observed foxes crossing local roads on 10 occasions and arterial and collector roads on one occasion, each. MSLD predicted that foxes made vii
Urban Roads and the Endangered San Joaquin Kit Fox
108 road crossings during monitoring. The majority (79 crossings) occurred on local roads, while 12 and 17 crossings occurred on arterial and collector roads, respectively.
Rate of road crossing per road type was in direct contrast to frequency of fox-vehicle collision per road type. Foxes most frequently crossed local roads, but most frequently died on arterial roads.
In addition to analyzing the effects of roads on kit foxes, we provide detailed recommendations on protection and compensation measures for urban road projects.
Specifically, we discuss the criteria for and construction of artificial dens and road crossing structures. These measures, when judiciously applied, may increase fox survival, habitat connectivity, and likelihood of persistence of urban fox populations.
ACKNOWLEDGMENTS
In addition to California Department of Transportation funding, the U.S. Bureau of Reclamation, California Department of Fish and Game, Great Valley Center, ARCO Western Energy, and Chevron USA provided valuable financial support for the urbanized kit fox ecology research project. Additional support was provided by the Smithsonian Institute, U.S. Department of Energy, U.S. Fish and Wildlife Service, and City of Bakersfield. J. Storlie was integral to the early development of this project. We thank S.
Bremner-Harrison, A. Brown, J. Murdoch, and E. Tennant for ongoing field assistance.
All schematics in this report came from the pen of S. Harrison. S. Phillips’ expertise with GIS and report preparation were invaluable. Finaly, we thank H. Hunt (Caltrans), S.
Jones (USFWS), and T. Marshall (Caltrans) for reviewing draft versions of this report.
viii
Urban Roads and the Endangered San Joaquin Kit Fox
INTRODUCTION
INFORMATION NEED
The San Joaquin kit fox historically occupied arid upland habitats throughout the San Joaquin Valley. Former and current conversion of these lands to agriculture, urban, and industrial uses has resulted in profound habitat degradation, fragmentation, and loss. As a result, the San Joaquin kit fox was listed as Federally Endangered in 1967 and California Threatened in 1973 (U.S. Fish and Wildlife Service 1998).
Kit foxes currently persist in a metapopulation consisting of 3 large “core” populations (Carrizo Plain National Monument, LoKern Natural Area, Panoche Hills region) and a number of smaller “satellite” populations. Movement of foxes between these populations is important for maintaining gene flow and avoiding inbreeding effects. Furthermore, kit fox populations exhibit marked fluctuations in the number of individuals as a result of natural and anthropogenic processes (e.g., Cypher et al. 2000). Risk of extinction due to catastrophic or random demographic events is increased for small populations in general and kit foxes specifically (White et al 2000). Thus, dispersal between populations may be necessary to prevent local extinctions or to recolonize lands where foxes are extirpated. The probability of long-term persistence of this species will increase with the number and stability of populations and the ability of animals to move between these areas.
Bjurlin and Cypher (2003) identified numerous potential impacts of roads on kit fox populations. While vehicle strike is the most obvious and documented, roads also may degrade or fragment habitat, increase chemical or noise pollution, alter prey or predator abundance, introduce invasive species, cause changes to fire regime, and increase human presence and development. The aforementioned have a variety of potential impacts on foxes, including mortality or morbidity, disrupted social ecology, reduced productivity, displacement, altered space use, inhibited dispersal, reduced genetic exchange, and decreased carrying capacity.
One study (Cypher et al., in prep) found few impacts of two-lane highways on kit fox ecology in natural lands. Probability of fox-vehicle collision was low, perhaps as a result of low traffic volumes (approximately 8,000 vehicles per day) and few vehicles during periods of peak fox activity. Similarly, there was no evidence that foxes avoided territories that bordered or crossed the roads, had decreased reproductive output along roadsides, encountered changes in availability or utilization of prey, or were more at risk from predation. Road presence, and associated human activity did significantly alter fire regime, increasing the frequency of wildfire, and changing native Atriplex spp. shrubland to non-native grassland. This change appeared to alter relative abundance of several Dipodomys spp., the primary prey of resident foxes. Furthermore, distribution (and perhaps abundance) of the coyote (Canis latrans) also was altered, with coyotes preferentially utilizing shrub remnants (Nelson et al., in prep). Clearly, potential impacts by roads are diverse, and perhaps mediated by other ecosystem processes (i.e., fire).
Furthermore, as road width, traffic volume, and speeds increase, the potential for direct impacts, such as vehicle strike will increase. Roads throughout the range of the San 1 Urban Roads and the Endangered San Joaquin Kit Fox
Joaquin kit fox are increasing in traffic volume. Therefore, lack of direct impact on San Joaquin kit fox as reported above is not necessarily indicative of future relationships between the kit fox and human transportation networks. For a synthesis of the literature on road impacts on San Joaquin kit fox refer to Cypher (2000) and Bjurlin and Cypher (2003).
The highest density of roads, peak traffic volume, and greatest potential for impact to foxes occurs in urban areas. Urban kit fox populations in California minimally occur in Bakersfield, Taft, Santa Nella, and Coalinga. Reports from the public and carcasses collected opportunistically indicate that kit foxes are dying by vehicle strike in the human developments of the Southern San Joaquin Valley. Urbanized foxes may encounter roads while foraging, looking for mates, defending territories, and dispersing. Little is known, however, about the factors that influence where, when, and how vehicle strikes occur.
Furthermore, it is unknown whether there are changes to kit fox use or placement of dens, patterns of movement, susceptibility to predation, prey selection, or other elements of fox ecology as a consequence of proximity to roads in urban areas.
Why is the impact of urban roads on kit fox populations important? First, urban areas have rapidly grown throughout the range of the kit fox. In some cases, development has occurred in agricultural lands that long have been devoid of kit foxes. But in many other cases, urbanization has displaced or altered fox populations occurring in natural lands.
Conversion of natural lands permanently decreases availability of habitat, which is a major obstacle to species recovery. Second, the network of three core- and several satellite-populations identified in the recovery plan for this species is vulnerable to fragmentation associated with transportation networks. For this metapopulation to remain functional, individuals must be able to move between sub-populations. The potential for negative impact of roads on the ability of foxes to utilize corridors may be proportional to road width, traffic volume, and speed limit. Projections for the San Joaquin Valley indicate that the human population and associated urbanization and transportation networks will increase for the foreseeable future. The highest potential for fragmentation will occur in urbanized areas where roads are most prevalent. Third, as urban lands spread and natural lands diminish, urbanized fox populations may become more valuable to species recovery. Urban fox populations are relatively less susceptible to meteorological stochasticity, and may prove useful as a repository for genetic diversity and/or individuals for reintroduction efforts. Consequently, the influence of transportation networks on San Joaquin kit fox recovery is likely to grow simultaneously to the importance of urbanized fox populations where transportation networks are most prevalent. Identifying ways to reduce the impact of urban roads on the kit fox is therefore a conservation priority.
A less tangible, but still important, reason to reduce the impact of roads on urban fox populations is the interaction between urbanized kit foxes and the public. Standardized surveying shows that the citizens of Bakersfield (where the largest population of urbanized foxes occurs) generally have seen a kit fox on one or more occasions and support protection of the urban population (Bjurlin and Cypher, in press). Furthermore, respondents expressed widespread appreciation of urbanized foxes (in some cases becoming strongly protective). The greatest concern identified by respondents was that kit foxes would be killed by vehicle strike. Furthermore, fox-vehicle collision, or the 2 Urban Roads and the Endangered San Joaquin Kit Fox
attempt to avoid it, may in some cases contribute to traffic accidents (unconfirmed citizen report) or unsafe road conditions. Measures designed to minimize kit fox crossing of busy roads at grade are likely to receive strong public support. To the extent that pedestrian and fox crossing structures can be combined for multiple use, the perceived benefits of this investment in infrastructure are likely to increase.
GOAL
Our goal was to determine whether roads affect the ecology of the urbanized San Joaquin kit fox. From 1997-2004, we outfitted kit foxes with radio transmitters and monitored survival, den use patterns, and movements. We examined these data for statistical and spatial associations to urban road networks. Based upon these results we provide detailed recommendations for the protection of San Joaquin kit foxes in urban environments during and after road projects. If implemented, these recommendations are likely to contribute to the long-term viability of urban fox populations and species recovery, receive strong public support, and may even increase public safety.
STUDY AREA
The study site is located in the southwestern quarter of the Bakersfield metropolitan area (Figure 1). Bakersfield lies at the southern end of the San Joaquin Valley in central California and is the Kern County seat of government. The metropolitan area as of January 2003 was approximately 580 km2 (224 mi2) with 394,234 residents (http://www.bakersfieldchamber.org/pdf/community-profile.pdf). The urbanized area of Bakersfield grew by 23.6% from 1990-2000 and continued growth is expected (http://factfinder.census.gov/home/saff).
Hot, dry summers, and cool, wet winters characterize the climate of the southern San Joaquin Valley. Average daily maximum temperatures range from 14°C in December to 37°C in July and average daily minimum temperatures range from 4°C in December to 21°C in July (Meadows Field Airport weather station). Annual precipitation averages
14.3 cm, but varies greatly between years.
3 Urban Roads and the Endangered San Joaquin Kit Fox
METHODS
KIT FOX CAPTURE AND MONITORING
From May 1997 to July 2004 we captured kit foxes in live traps, outfitted them with radio transmitters, and monitored them for sources of mortality, nightly movements, daily resting places, and reproductive output. We captured foxes with wire-mesh box traps (38 x 38 x 107 cm) baited with assorted meats. Traps were covered with cloth tarps to protect animals from inclement weather and direct sun. Captured foxes were examined for injury and parasites, ear-tagged, sexed, aged, and fitted with a radio collar (40g, Advanced Telemetry Systems, Isanti, Minnesota) containing a mortality sensor that altered signal pulse rate if the animal was motionless for four hours. We released foxes at site of capture after 0.5-1 hour of handling. Capture and handling protocols were in accordance with permits TE023496-1 and TE825573-2 from the U.S. Fish and Wildlife Service and a Memorandum of Understanding from the California Department of Fish and Game.
Figure 1. Urbanized San Joaquin kit fox study area, Bakersfield, CA. 4 Urban Roads and the Endangered San Joaquin Kit Fox
At a minimum, we attempted to record one day-resting location and two night-activity locations for each animal with a functioning radio collar during each week. In some cases fewer locations were recorded because of collar failure, fox dispersal from typical home ranges, use of resting sites that attenuated radio signal, or other factors. Radio collar transition to mortality signal allowed us to promptly retrieve deceased foxes and increased the likelihood of confirming cause of mortality. Fox carcasses also were collected opportunistically during the study when spotted on roads, reported by citizenry, or collected by other organizations. Each spring, we conducted natal den observations to estimate litter size for all radio-tagged females. Finally, we collected kit fox feces for dietary analysis throughout the study duration. Locations of activity, mortality, resting, and pup whelping and rearing were recorded with GPS and entered in a GIS for spatial analysis.
MORTALITY
We conducted necropsy for all recovered fox carcasses. We limited our analysis of cause of death to animals that were actively being monitored with radio telemetry in order to decrease bias caused by differences in visibility of carcasses. We also report, however, on all carcasses retrieved during the study. Cause of death was determined using the pattern of injury, including external abrasion, broken bones, internal hemorrhaging, and bite wounds. The location that the carcass was retrieved from also was used in determining cause of death (e.g., within road right-of-way, buried at construction site, in carnivore scrape, etc.).
For all animals (transmitting or not) that died by vehicle strike, we assigned road type (local, collector, arterial, highway), traffic volume, and posted speed limit. Local roads had the primary purpose of providing access to abutting residential property, did not exceed 12.2m (40ft) in width, had no more than one traffic lane in each direction, and were not longer than 0.8km (0.5mile) without interruption. Speed limit was 25mph on all local roads, but was not always posted. Collector roads generally conducted local road traffic to the arterial road network, but sometimes were residential in nature. Speed limits (generally 30-45 mph) for collector roads were based upon speed surveys and were posted. Arterial roads had 1-3 lanes of traffic in each direction (typically more than one), carried the majority of city traffic, and connected the local and collector road networks to the State highway system. Arterial road speed limits (35-55mph) were always based on speed surveys and posted. Lastly, highways were State numbered roads with the primary purpose of conducting traffic into and away from the urban area. They included highways 58, 99, 178, and 119 and in some cases passed through residential, commercial, and industrial areas. Vehicle access to highway 99 and portions of 178 was by ramp only.
We compared road type where foxes died to an expected distribution based on road availability using contingency table analysis. Road availability was calculated by summing the length of three road types (local, collector, arterial) within the minimum convex polygon of the locations of vehicle strike (Figure 2). Vehicle strikes also were examined for temporal patterns. For this analysis we limited the time period to 6 complete years, 07/01/1998-06/30/2004 so that months wouldn’t be disproportionately represented. Strikes were separated by fox age (adult, juvenile) and gender (male, 5 Urban Roads and the Endangered San Joaquin Kit Fox
female) and compared with other sources of morality by month. Temporal distribution in strikes was compared with frequency of fox scent marking, interaction with foxes outside of social group (Murdoch 2004), and traffic volume by month.
Finally, we examined spatial patterns of kit fox vehicle strikes that occurred on arterial or collector roads. We had insufficient sample size to conduct similar analyses for local and highway strikes. We limited the extent of this analysis to southwest Bakersfield where we had the greatest knowledge of roads and fox habits. Strikes were examined for association with intersections of the road where the mortality occurred and other linear rights-of-way. Linear rights-of-way for this analysis were defined as other roads (local, collector, arterial, highway), canals, railroads, rivers, or golf course crossings.
Figure 2. Minimum convex polygon of fox-vehicle collisions for individuals that had transmitting radio collars, Bakersfield, CA, 1998-2004. 6 Urban Roads and the Endangered San Joaquin Kit Fox
In order to conduct this spatial analysis, we defined vehicle strike road segments. These segments were the shortest sections of road that were bounded by intersection with other arterial or collector roads and contained the vehicle strike location. This rule was modified when the mortality occurred at an intersection with arterial or collector road. In these cases, the segment was extended in both directions to the next, nearest intersection with a major road. We then assigned 100 randomly distributed points to the road segment for each vehicle strike. For example, if three strikes occurred on the segment, 300 randomly distributed points were assigned. We calculated the distance to nearest intersection with linear right-of-way for all points (confirmed strike and random).
Distance was coded into three categories, near (x <= 26m), intermediate (26m < x <= 52m), and far (52m < x) based on a mean road width of 26m. Therefore, strikes assigned as “near” occurred within one road width of the intersection, as “intermediate” within two road widths, and as “far” greater than two road widths.
Observed strikes versus random locations were examined with this coding scheme in a contingency table analysis for a departure from what would be expected by chance.
Lastly, we evaluated the type of intersections where vehicle strikes occurred for strikes coded as “near”. We then compared the frequency of observed mortalities and random points among intersection types.
DENS
Association between confirmed kit fox dens and major roads (highway, highway ramp, arterial, and collector) was examined for dens used by radio-collared foxes from 1 July 2001 to 30 June 2004. Distance to nearest road centerline was calculated for each known den and compared across distance categories (34.5m, 69m, etc.). Distance categories corresponded with Fish and Wildlife exclusion zone distances for known kit fox dens (34.5m = 100ft).
MOVEMENTS
Kit fox space use was studied using vehicle-based monitoring of focal animals during 2003. Focal sessions typically began 1 or more hours after nightfall (full dark) and lasted for approximately four hours. Every 15 minutes we attempted to collect a new location for the focal animal. We described fox locations as accurately as possible using a combination of triangulation and spotlighting. With these data we calculated minimum straight-line distance (MSLD) between consecutive locations. We also calculated rate of travel (MSLD km/hour). Finally, we estimated number of road crossings (arterial, collector, local) using MSLD and visual sightings. We then adjusted these data for crossing rate per hour of observation and MSLD of movement. We used shapefile coverages of city roads and geo-referenced aerial photography to assign road crossings for this analysis.
7 Urban Roads and the Endangered San Joaquin Kit Fox
SPATIAL AND STATISTICAL ANALYSIS
Spatial analysis was conducted using ArcView version 3.2 (ESRI 1996). Data for roads and traffic volumes were generated with shapefile coverages from City of Bakersfield GIS and Traffic Engineering Departments, respectively (City of Bakersfield 2000-2004).
All spatial data were projected in Lambert Conformal Conic, California State Plane, NAD27. In some cases data on road location, size, or posted speed limit was unavailable for the year in which the kit fox feature was described (i.e., vehicle strike, den, movements). In these instances data from the closest available time period were used (usually less than one year distant). Geo-referenced aerial photography (AirPhoto USA) was used to assign the location of road intersections with other linear rights-of-way. All statistical analyses were performed using SPSS version 11.5 (SPSS Inc. 2002) and results were considered significant at P <= 0.05.
San Joaquin Kit Fox
8 Urban Roads and the Endangered San Joaquin Kit Fox
RESULTS
URBAN KIT FOX ECOLOGY
From May 1997 to July 2004 we captured a total of 358 individual foxes (109 adults, 206 juveniles, and 43 juveniles that were captured again as adults). Of these, we outfitted 229 individuals with radio collars (84 adults, 79 juveniles, and 66 juveniles that continued to be monitored as adults). We monitored radio-tagged animals until they died, transmitters failed, or they were lost. We collected 8,115 night activity locations and 6,647 daytime resting locations during the study (Figure 3a-b). Kit foxes primarily rested diurnally within dens, of which we described 471 over the study period. We located kit fox litters on 54 separate occasions. With these data we will conduct additional analyses on survival rates, reproductive rates, den site selection, space requirements, preferred habitats, and foraging patterns. We will present these data in a comprehensive monograph of kit fox demographics and ecology in urban environments. California Department of Transportation (Caltrans) will be supplied with a copy of the monograph and associated recommendations for the conservation of urban fox populations.
MORTALITY
We collected 156 kit fox carcasses from September 1985 to August 2004. Carcasses collected prior to 1997 were stored by California State University, Bakersfield. A near majority of mortalities were attributed to vehicle strike (48.1%, Table 1). These data, however, include animals that were received from a variety of opportunistic sources. By limiting analysis to animals that were actively being monitored with radio telemetry at time of death (N = 78) we reduce bias (e.g., differential reporting and accessibility to carcasses). Of this subset, confirmed vehicle strike accounted for 26.9% of mortality.
An additional 7.7% of foxes may have died from vehicle strike, but evidence, while suggestive, was not conclusive. Predation (16.7%) was the second most common cause of death for transmitting animals. Other sources of mortality included entombment, poison, drowning, gunshot, complications during parturition, and entanglement.
Due to advanced state of decay or lack of gross injury, we were unable to determine cause of death for 28.2% of carcasses. However, 73% of the 22 individuals with unknown cause of death were retrieved from fox dens. Pertinent to this report, animals retrieved from dens were unlikely to have died of vehicle strike (they lacked broken bones or contusions), and therefore the percentages calculated for vehicle strike are considered accurate when unknown cause of death is included. Overall, 30.8% of the 78 mortalities of foxes with operational radio collars were retrieved from dens. In natural lands where most research has been conducted and mortality generally is attributed to predation, carcasses are rarely retrieved from dens. There may be sources of mortality in urban environments (e.g., disease or poisoning) that have yet to be identified, but that do not leave clearly diagnostic evidence with our necropsy techniques. 9 Urban Roads and the Endangered San Joaquin Kit Fox
Figure 3. Nocturnal (a) and Diurnal (b) locations of San Joaquin kit fox in Bakersfield, CA, 1998-2004.
10 Urban Roads and the Endangered San Joaquin Kit Fox
Table 1. Cause of death for all foxes (1985-2004) and foxes with transmitting radio signals at time of death (1997-2004), Bakersfield, CA.
All Foxes Transmitting Foxes Cause of Death
N % N %
Entombed - construction 5 3.2 3 3.8
Entombed - natural 1 0.6 1 1.3
Poison 3 1.9 0 0.0
Possible poison 5 3.2 4 5.1
Predation 16 10.3 13 16.7
Possible predation 5 3.2 4 5.1
Vehicle 75 48.1 21 26.9
Possible vehicle 8 5.1 6 7.7
Other 6 3.8 4 5.1
Unknown 32 20.5 22 28.2
Total 156 100.0 78 100.0
Table 2. Frequency of fox-vehicle collisions by road type for all foxes and those with transmitting radio signals at time of death, Bakersfield, CA, 1998-2004.
All Strikes Transmitting Strikes Road Type
N % N %
2-lane local 2 4.2 1 5.6
2-lane collector 8 16.7 1 5.6
2-lane arterial 2 4.2 1 5.6
4-lane arterial 17 35.4 7 38.9
6-lane arterial 16 33.3 7 38.9
6-lane highway 1 2.1 0 0.0
Not on road 2 4.2 1 5.6
Total 48 100.0 18 100.0
From January 1998 to August 2004, we had the necessary information on road status to describe association between road types and vehicle strikes. We retrieved 48 vehicle strike kit foxes in Bakersfield during this period. Of these, 18 were actively being monitored with radio telemetry at the time of death. Arterial roads accounted for 72.9% and 83.3% of strikes for all animals and actively monitored animals, respectively (Table 2). Collector roads, local roads, and highways were less significant sources of mortality. Within the 100% MCP of actively monitored animals (Figure 2), we identified 327.4 km of local roads, 49.9 km of collector roads, and 50.9 km of arterial roads. There was a significant difference between the observed frequency of roadkill by road type, and that which would be expected based on availability (χ2
2 = 82.127, P < 0.001). Arterial
roads were far more often a source of vehicle strike than can be explained by chance occurrence.
11 Urban Roads and the Endangered San Joaquin Kit Fox
Roads with four or more lanes accounted for 70.8% and 77.8% of roadkill for all animals and actively monitored animals, respectively (Table 2). Approximately 90% of roadkill were retrieved from roads with posted speed limits of greater than 45 mph, irrespective of whether the animal was transmitting at time or death or was collected opportunistically (Table 3). Roads with a 55 mph posted speed limit accounted for over half of all vehicle strikes. Mean traffic volume at vehicle strike was 21,861 and 23,426 vehicles per day for all strikes and actively monitored strikes, respectively (Figure 4).
Table 3. Fox-vehicle collisions by posted speed limit, Bakersfield, CA, 1998-2004.
All Strikes Transmitting Strikes Speed Limit
N % N %
25 mph 1 2.1 0 0
35 mph 1 2.1 0 0
40 mph 1 2.1 1 5.6
45 mph 9 18.8 2 11.1
50 mph 6 12.5 5 27.8
55 mph 27 56.3 9 50.0
65 mph 1 2.1 0 0.0
Not on road 2 4.2 1 5.6
Total 48 100.0 18 100.0
Figure 4. Fox-vehicle collisions by traffic volume for all mortalities and mortalities of individuals with transmitting radio collars at time of death, Bakersfield, CA, 1998-2004. 12 Urban Roads and the Endangered San Joaquin Kit Fox
There was no overall interaction between vehicle strike as cause of death (Coded Yes/No) and gender (χ2
1 = 0.057, P = 0.812) or age (χ2
1 = 0.709, P = 0.400) for 78 foxes
that were actively being monitored at time of death. Of 47 adult mortalities, 38.3% were either confirmed or possible vehicle strikes. Confirmed or possible vehicle strike accounted for 29.0% of 31 juvenile mortalities. Mortalities were evenly distributed between the genders (male = 39, female = 39), and approximately one third of each sex died of confirmed or possible vehicle collision. While all mortalities were distributed throughout the year, vehicle strikes showed a distinct peak in December during the six- year period beginning 1 July 1998 (Figure 5). Further examination showed an association between gender, age, and vehicle strikes when the data were examined temporally. Of nine foxes that died of vehicle collision in December, seven were adult males. The winter (December/January) spike in male fox-vehicle collision corresponded with male rate of scent marking (Figure 6) and overall rate of interaction between focal foxes and conspecifics that were not within the social group (Figure 7). Both scent marking and interaction frequency may be indicators of reproductive behavior and territoriality (Murdoch 2004). There also appeared to be a smaller increase in vehicle strike from May to September. Many of these animals were pups that were at the age typical of increasing independence from the natal group, solitary foraging, and dispersal.
Vehicle traffic volume did not differ by month at six control stations spread across the urban area (Figure 8).
Figure 5. Fox-vehicle collisions by gender and month, Bakersfield, CA, 1998-2004. 13 Urban Roads and the Endangered San Joaquin Kit Fox
Figure 6. Fox-vehicle collisions (1998-2004) versus scent marking rates (Murdoch 2004) for male foxes, Bakersfield, CA.
Figure 7. Fox-vehicle collisions (1998-2004) versus interactions with foxes that were not within the immediate social group (Murdoch), Bakersfield, CA. 14 Urban Roads and the Endangered San Joaquin Kit Fox
We identified 23 segments of arterial or collector roads where strikes occurred. Number of strikes per segment ranged from 1 to 5 and accounted for 38 strikes in all (Figure 9).
Along these segments we identified 136 intersections. Kit foxes died more often than expected by chance at intersections with a linear right-of-way (χ2 2 = 15.75, P < 0.001).
Strikes within one road width (<=26m) of an intersection accounted for 47.3% of all roadkill. In contrast, only 22.5% of random points occurred within one road width of an intersection. Following a similar, but weaker, trend 21.1% of observed roadkill were between one and two road widths (26 < x <= 52m) of an intersection, versus 17.8% of random locations in the category. There did not appear to be a higher risk of strike at any one type of intersection. Across intersection types, frequency of vehicle strikes that occurred in the category nearest to intersections mirrored the frequency of random points (Table 4). For example, 38.9% of all strikes within one road width occurred at intersection with local roads, while 37.1% of corresponding random points occurred at local road intersections.
Figure 8. Mean (Std. Dev.) traffic volume by month at six control stations, Bakersfield, CA, 2003.
15 Urban Roads and the Endangered San Joaquin Kit Fox
Table 4. Observed versus random generated fox-vehicle collisions on arterial and collector roads that were within one road width (26m) of an intersection with a linear right- of-way, Bakersfield, CA, 1998-2004.
Observed Strikes Random Points Intersecting Linear Feature N % N %
Arterial road 1 5.6 99 11.6
Canal 4 22.2 191 22.4
Collector road 3 16.7 69 8.1
Golf course 1 5.6 59 6.9
Highway 1 5.6 58 6.8
Local road 7 38.9 317 37.1
Railroad 0 0.0 17 2.0
River 1 5.6 44 5.2
Total 18 100 854 100.0
Figure 9. Fox-vehicle collisions and road segments where mortality occurred, Bakersfield, CA, 1998-2004.
16 Urban Roads and the Endangered San Joaquin Kit Fox
DENS
Kit foxes did not appear to avoid using dens or building dens at sites that were adjacent to major roads. We calculated nearest major road (arterial, collector, highway, or highway ramp) to each of 327 fox dens used from 1 July 2001 to 30 June 2004. Of these, 9.5% occurred within 30.4m (100ft) of the nearest road centerline. A similar trend occurred for each subsequent 30.4m distance from road (Figure 10). Some foxes utilized features of a major road for den construction. One individual was known to den within a culvert beneath highway 99 (Figure 11). Other dens occurred in the embankments or underpasses of a raised portion of highway 99 (Figure 20b, Appendix), and at drainage basins (e.g., Figure 14c, Appendix) or canals directly adjacent to roads.
In some cases dens were directly adjacent to the roadway (Figure 12), but in other cases roadside dens were behind by a cement block wall or other barrier that increased isolation, decreased noise pollution, and may have made the site less likely to be impacted by human activity (Figure 13). Dens directly adjacent to roads are at risk of disturbance during road maintenance or construction. Four dens within a drainage basin at the intersection of two arterial roads (Figure 14c) were destroyed during earth moving associated with road widening and improvement. In fact, the entire drainage basin was removed in this instance. Four frequently occupied fox dens were destroyed when a local road, Kroll Way, was extended through a previously un-roaded area.
Figure 10. Number of known dens by distance to nearest centerline of major road, Bakersfield, CA, 2001-2004.
17 Urban Roads and the Endangered San Joaquin Kit Fox
Figure 11. Drainage culvert under Highway 99 in Bakersfield, CA that was used as a den by San Joaquin kit fox.
Figure 12. Kit fox dens at varying distances from collector and arterial roads in Bakersfield, CA, 2001-2004.
18 Urban Roads and the Endangered San Joaquin Kit Fox
Figure 13. Kit fox dens at varying distances from collector and arterial roads, Bakersfield, CA, 2001-2004. A barrier wall, indicated with dashed line, isolated dens from one roadway with relatively high traffic volume.
MOVEMENTS
We conducted 21 sessions of vehicle-based intensive monitoring on 15 different focal kit foxes, for a total of 4,633 minutes of observation. We collected an average of 16.2 unique locations per session. Mean interval between locations was 15.14 minutes. By connecting these locations sequentially, we were able to calculate minimum straight-line distance (MSLD) moved and minimum rate of travel. We recorded a total of 44,121m of MSLD, for an average of 2,101m per monitoring session. Using MSLD and observation time, we calculated that foxes traveled at an average rate of 0.5958 km/hr during monitoring. Of course, focal foxes were inactive during much of the focal period and minimum distances between points do not account for the more complex travel patterns that were likely to exist. Therefore, foxes traveled a greater distance in shorter time and actual rate of travel likely was considerably higher.
MSLD movements also allowed us to estimate number of road crossings during focal sessions. We directly observed foxes crossing local roads on 10 occasions and arterial and collector roads on one occasion, each. MSLD predicted that foxes made 108 road crossings during focal monitoring. The majority (79 crossings) occurred on local roads, while 12 and 17 crossings occurred on arterial and collector roads, respectively. On average, foxes made 2.43 road crossings per km MSLD and 1.46 crossings per hour during intensive monitoring. Crossing rate for arterial roads (the primary location of roadkill) dropped to just 0.32 and 0.16 crossings per km MSLD and hour, respectively.
Intersection of arterial roads with MSLD occurred in only six monitoring sessions (five individual foxes).
19 Urban Roads and the Endangered San Joaquin Kit Fox
DISCUSSION
MORTALITY
As the single largest source of mortality for the Bakersfield fox population, vehicle strike clearly is an important factor for urbanized foxes. These data are in marked contrast with all other studies on kit fox in which vehicle strike rarely exceeded 10% of mortalities and was not considered a significant influence on fox demographics or population ecology (Bjurlin and Cypher 2003). In the largest study to date on the San Joaquin kit fox in natural habitats, Cypher et al. (2000) found that predation accounted for 57.2% of 222 fox deaths while vehicle strike was identified as cause of death in only 9% of occurrences. In contrast, predation during this urban study accounted for 16.7% of mortality, while vehicle strike, at 26.9%, was a more important cause of death. The coyote is the primary predator and competitor of the San Joaquin kit fox (Ralls and White 1995, Cypher and Spencer 1998) and is not abundant in Bakersfield (Cypher unpubl. data). Coyote distribution appeared limited to the Kern River corridor in the urban area included in this study. Therefore, it is possible that vehicle strike simply replaces mortality that would otherwise have been caused by depredation, had predators been more abundant. Indeed, some urban fox populations have persisted for decades and appear relatively stable given the dynamic alteration of urban environments that occurs during development and land conversion. Analyses of urban fox population viability and the impact of vehicle strike on viability are pending.
Arterial roads accounted for most strikes, a relationship that was significantly different from expected based on availability of this road type in the urban environment. Arterial roads typically have multiple lanes, relatively higher traffic volumes, and relatively higher posted speed limits. The San Joaquin kit fox is capable of rapid travel over short distances. The swiftness of this species may enable it to safely navigate two-lane roads where vehicle speeds and traffic volumes are lower. Indeed, only 10% of strikes occurred on roads that had posted speed limits that were lower than 45 mph. Low volume and speeds on local roads make it less likely that a fox crossing the road will need to avoid more than one vehicle. In contrast, a fox crossing four- or six-lane roads may have to evade multiple vehicles, which may travel in different directions and at different rates. Average traffic volume for urban roads with fox-vehicle collisions was nearly three times the volume of two-lane highways that transverse the LoKern Natural Area, one of three core areas for the San Joaquin kit fox (Cypher et al. in prep).
The incidence of vehicle strike was not evenly distributed. Foxes died more frequently within one road width of intersections between major roads and other linear rights-of-way then expected by chance. In fact, 68.4% of strikes on major roads occurred within two road widths of an intersection. The association between intersections and vehicle strike likely was the result of kit fox movement patterns. Species of the family Canidae (dogs, wolves, coyotes, foxes) utilize linear features for travel (Trewhella and Harris 1990), and we observed urban kit foxes on linear corridors between habitat patches (C. Bjurlin, pers. obs.). Additionally, increased traffic at intersections and regular changes in vehicle speed or direction (particularly where two roads come together) could make navigating intersections confusing for foxes. Finally, roadways typically are wider at road/road 20 Urban Roads and the Endangered San Joaquin Kit Fox
intersections to accommodate turning lanes. Greater distance to traverse equals greater time on the road and greater exposure to fox-vehicle collision. The lack of association of fox-vehicle collision with a specific type of intersection suggests that this relationship is broadly applicable.
Fox gender also was linked with vehicle strike when time of year was added as a variable. Males appeared particularly vulnerable during the breeding season (Dec-Jan).
Increases in male strikes closely mirrored rate of scent marking and interaction with animals outside of the social group (Murdoch 2004). During the reproductive season, male foxes are more active, travel greater distances, are more aggressive, and spend more time chasing females in courtship and males in territory defense (C.Bjurlin, pers. obs).
Furthermore, male foxes travel outside of the home territory in search of mating opportunities during December and January (Cypher et al., unpubl. data). One male in particular made these movements nightly for a period of a week and was observed successfully mating with a female in these new territories. While this example stands out in its detail, many males were observed in extra-pair copulation and outside of the typical home range during this study. On one occasion we witnessed four foxes in rapid and aggressive chase make their way through traffic on a collector road where several mortalities were known to have occurred. As interloper, transient foxes have to contend with similar agonistic interactions with resident males. It is likely that animals occupied with mating, aggressive interactions, or in unfamiliar territory are less aware of roads and traffic.
Vehicle strike during winter months also may be affected by shorter days and daylight savings time. Peak traffic volume occurs during morning and evening commutes. In the winter, the evening commute occurs up to two hours after sunset when male foxes are likely active and seeking mates. Fox activity and peak traffic volume have the greatest likelihood of overlap on the winter solstice (the shortest day of the year). Male vehicle strikes also appeared to be centered on the solstice. In contrast, traffic volume was constant throughout the year, so it is unlikely that monthly changes in traffic (e.g., holiday travel) contributed to patterns of fox mortality.
DENS
Kit foxes did not appear to prefer or avoid dens with respect to roads in urban environments. Urban land use is non-random and areas of similar zoning often are clumped. Kit foxes have shown preference for den sites with respect to some land uses.
Of the 471 dens described from 1997-2004, 36.3% were located along the banks of canals or water detention basins. These areas, however, accounted for only a small percentage of available land area in Bakersfield. It appears that open spaces, or lightly and infrequently disturbed areas (such as canals and basins) are more influential on kit fox den site selection than the proximity or size of adjacent roadways. It is possible that a positive or negative association with roads may occur as a result of a relationship between roads and other, desirable den sites. For example, many of the water transport canals in Bakersfield are adjacent to major roadways. This pattern may reflect a planned strategy, simple convenience, or historical artifact of past development. There is no current strategy for land planning in urban areas with regards to the San Joaquin kit fox. 21 Urban Roads and the Endangered San Joaquin Kit Fox
Therefore, future development practices may increase or decrease the likelihood that foxes will select den sites near roadways.
Den proximity to major roads may elevate the risk of mortality in some cases. One family group along a collector road in Bakersfield lost two of nine juveniles to vehicle strike in 2004. The natal dens in this instance were less than 20m from the road. Kit fox pups are active and relatively naïve to external dangers, particularly in the first six months of life. Urban foxes also quickly habituate to human and vehicle presence.
While the process by which juvenile foxes learn how to successfully cross roads is unknown, there likely is a period of learning and perhaps elevated risk during youth, particularly when juveniles increase in independence and begin to venture farther from the natal site in search of prey and territory.
Fox dens are destroyed during road construction or modification. Because dens are an essential resource for the kit fox (see Appendix), development projects should avoid dens whenever possible, compensate for den destruction, and secure the necessary permissions from State and Federal authorities when altering lands upon which kit fox exist. In particular, kit foxes can be entombed within dens during earth moving. Road construction projects in Bakersfield have the potential to affect 9.5% or more of dens that occur within 100 ft of major road centerlines. The percentage of affected dens would have been far greater had local roads been included in this analysis. The Fish and Wildlife Service requires that an exclusion zone be created around these dens unless permission is given for den disturbance or destruction. The loss of den sites also decreases the general suitability of habitat, both for resident and for transient animals.
Whenever dens are potentially affected by transportation projects, we strongly encourage the supplemental protection and compensation strategies that are described in the appendix to this document.
MOVEMENTS
Our data on the movements of radio-collared animals show that kit foxes frequently cross roads. The majority of crossings occurred on local roads, mirroring their relative abundance in urban areas, but the majority of mortalities occured on arterial and collector roads. Clearly, the low traffic volumes and speeds associated with local roads decreases the danger to kit fox. While these road types, cover extensive area with the range of the fox (thereby decreasing habitat availability), loss of habitat in urban areas may be unavoidable. We do not proposal specific conservation measures for local road construction to decrease fragmentation of urban landscapes.
Arterial and collector roads offer a greater challenge. These roads accommodate an ever- increasing volume of traffic in Bakersfield and other southern San Joaquin Valley human population centers. Kit foxes have shown that they can successfully cross major roads, but many animals maintained territories that did not bring them into frequent contact. It is unclear how increases in traffic volume affect the ability of animals to disperse in urban environments, which is a critical function for healthy kit fox populations.
Crossings appear to be habitual in some cases – with individual foxes repeatedly using a few locations to navigate the roadway. That and the association between fox crossings, fox-vehicle collisions, and linear rights-of-way suggest that these intersections would be 22 Urban Roads and the Endangered San Joaquin Kit Fox
likely candidates for conservation measures. While reducing speed limits and road sizes would not be practical, it may be possible to reduce the likelihood of vehicle strike on arterial and collector roads by providing safe passage for foxes. Specifically, we propose that road modification projects include wildlife crossing structures where appropriate.
Criteria for selecting locations for crossing structures and instructions for building structures and associated infrastructure are provided in the appendix to this document.
CONCLUSIONS
Roads impact urban kit foxes. Fox-vehicle collision was the primary cause of mortality over a six-year intensive study of urban fox ecology. Because of the absence of predators in the urban environment, it is unclear whether vehicle strike was a limited factor for fox population growth, or simply a compensatory source of mortality for animals that would have been depredated had predator populations been robust. Analyses of survival and population viability are pending. Fox-vehicle collision spiked during the December reproductive season for male kit foxes. This spike in mortality may disrupt kit fox reproductive ecology by destabilizing pair bonds directly prior to whelping and rearing of young. Fox-vehicle collisions occurred more frequently than expected within one road width of an intersection with a linear right-of-way. Rights-of-way may be natural crossing locations for foxes due to habitual travel patterns. They also may be difficult to navigate safely due to wider roadways and irregular traffic speeds and directions. Kit foxes denned on roadsides at similar densities to more remote locations. Likewise, fox movements did not appear overtly affected by roads. Our study, however, was not specifically designed to test for an impact of roads on fox home range and dispersal patterns. An impact on either of these aspects of kit fox ecology is therefore unknown.
The effects of road construction and modification may be reduced by the installation of artificial dens and wildlife crossing structures as outlined in the appendix to this document. Efforts to conserve urban fox populations will contribute to the recovery of this endangered species. Because fox-vehicle collision is the primary source of mortality for urbanized foxes, changes to roads that decrease fox death should command a central role in future conservation programs.
23 Urban Roads and the Endangered San Joaquin Kit Fox
FUTURE RESEARCH
Based on the studies conducted to date by ESRP, further research and conservation efforts will benefit urban kit foxes. We have identified five topics for future investigation.
1) The factors that affect the suitability of urban habitats for kit fox are currently unknown. These factors may include road density, landuse (e.g., zoning), building density, and presence of refugia and/or corridors between habitat patches. A systematic census of kit fox in urban habitats (e.g., Bakersfield, Taft, Coalinga) using camera or hair traps is required to focus conservation and mitigation efforts in the areas that are of highest value. Previous urban fox research, while critical to identifying vital rates (births, deaths, migration), was not designed to rigorously test the variables that affect fox presence. Therefore, there is a significant data gap in this area.
2) Cause of death could not be determined for a number of foxes, and poisoning was strongly suspected in many of these cases. This question warrants further investigation to identify the types and sources of poisoning. 3) A recent pilot study has indicated that striped skunks, which are a host to rabies in the San Joaquin Valley, may present a future epidemiological threat to kit foxes in urban environments. Further research is required to quantify this threat and develop strategies to ameliorate it.
4) Conservation strategies have been identified that could potentially benefit urban kit foxes, such as road crossing structures, artificial dens, passages through walls and fences, and the construction of refugia and corridors in urban landscapes.
Further research on the efficacy of these strategies, and information packages detailing these strategies, would benefit urban kit foxes, city planners, and developers.
5) A recent survey indicated that Bakersfield residents harbored misconceptions regarding kit foxes and their population status (Bjurlin and Cypher, in press).
Resources directed toward developing and evaluating a coordinated outreach effort will benefit kit fox conservation in both urban and natural environments.
24 Urban Roads and the Endangered San Joaquin Kit Fox
LITERATURE CITED
Bjurlin, C.D., and B. L. Cypher. 2003. Effects of roads on San Joaquin Kit Foxes: A review and synthesis of existing data. Pages 397-406 in C.L. Irwin, P.Garrett, and K.P.McDermott, editors. 2003 Proceedings of the International Conference on Ecology and Transportation. Center for Transportation and the Environment, North Carolina State University, Raleigh, NC.
Bjurlin, C.D., and B. L. Cypher. In Press. Encounter frequency with the urbanized San Joaquin kit fox correlates with public beliefs and attitudes toward the species.
Endangered Species Update.
City of Bakersfield. 2000-2004. Geospatial data on roads and traffic volumes from Department of GIS and Traffic Engineering. http://www.bakersfieldcity.us/ 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. 2000. Effects of roads on San Joaquin kit foxes: a review and synthesis of existing data. Report for California Department of Transportation, Sacramento, California.
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. 43pp.
ESRI Inc. 1996. Arcview 3.2 for Windows. ESRI Inc., Redlands, CA.
Murdoch, J. D. 2004. Scent marking behavior of the San Joaquin kit fox (Vulpes macrotis mutica). M. S. Thesis. University of Denver, CO.
Ralls, K, and P. J. White. 1995. Predation on San Joaquin kit foxes by larger canids.
Journal of Mammalogy 76:723-729.
SPSS Inc. 2002. SPSS Base 10.0 for Windows User's Guide. SPSS Inc., Chicago IL.
Trewhella, W.J., and S. Harris. 1990. The effect of railway lines on urban fox (Vulpes vulpes) numbers and dispersal movements. J. Zool., Lond. 221:321-326 United States Fish and Wildlife Service. 1998. Recovery plan for upland species of the San Joaquin Valley, California. United States Fish and Wildlife Service, Region 1, Portland, Oregon. 319 pp.
White, P. J., W. H. Berry, J. J. Eliason, and M. T. Hanson. 2000. Catastrophic decrease in an isolated population of kit foxes. The Southwestern Naturalist 45:204-211. 25 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox APPENDIX. SUPPLEMENTAL RECOMMENDATIONS
FOR PROTECTION OF SAN JOAQUIN KIT FOX DURING
ROAD PROJECTS IN URBAN ENVIRONMENTS
Prepared by: Endangered Species Recovery Program
in cooperation with: U.S. Fish and Wildlife Service, Sacramento Field Office and the California Department of Transportation
July 2005
INTRODUCTION
The San Joaquin kit fox (Vulpes macrotis mutica) was listed as Federally Endangered in 1967 and California Threatened in 1973. Specific measures implemented to protect the kit fox for any given project shall be determined by the U.S. Fish and Wildlife Service (Service) based upon applicant consultation with the Service. The purpose of this document is to provide an introduction to measures that may reduce impacts of road construction or modification in urban environments. Specifically, we discuss the criteria for and construction of artificial dens and road crossing structures. The measures outlined in this document are supplemental to standard Service and Caltrans guidelines, which will not be repeated here. Furthermore, these procedures are subject to modification or revision at the discretion of Service. The recommendations made in this appendix are those of the authors and are based on the findings of this research project and other current knowledge of the San Joaquin kit fox. These procedures are applicable to all urban environments where kit foxes are present or will be present in the future. Kit fox presence has been confirmed in Bakersfield, Taft, Santa Nella, and Coalinga, but surveys have yet to be conducted elsewhere. Kit foxes are likely to be found now or in the future in other urban or urbanizing areas.
DEN DISTURBANCE
Summary
Dens are a vital resource to the kit fox in urban environments (Koopman et al. 1998).
Dens provide refuge from predators and climatic extremes, and a location to whelp and rear young. Urban kit foxes typically occupy dens diurnally, but may spend considerable time within dens during the night, especially immediately preceding parturition and in the months post-parturition when young are nursing and vulnerable to predation. In the following text we describe kit fox dens in urban environments and make recommendations for the treatment of dens when they are encountered at a work site. In addition, we provide criteria and instructions for the creation of fox-specific artificial 26 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox dens. The construction of artificial dens may in some cases offset the destruction or disturbance of dens or fox habitat during road projects.
Urban Kit Fox Den Description
Many Federal guidelines pertain to the “known den” (U.S. Fish and Wildlife Service 1999). A known den is defined as any existing earthen den or manmade structure that is used or has been used at any time in the past by a San Joaquin kit fox. Evidence of use may include historic records, past or current radio telemetry or spotlighting data, kit fox sign such as tracks, scat, and/or prey remains, or other reasonable proof that a den or structure is being or has been used by a kit fox. The U.S. Fish and Wildlife Service discourages use of the terms ”active” and “inactive” when referring to a kit fox den because occupied dens may show no evidence of use, and because kit foxes change dens often, with the result that the status of a given den may change frequently and abruptly.
Known dens in urban environments take a variety of forms. Most dens are earthen and have multiple entrances, occur in vacant or undeveloped lots, and appear similar to dens found in natural lands. Many earthen dens in urban environments, however, occur at atypical locations, such as dense landscaping, maintained lawns, building foundations, stacked materials, and canal, drainage basin, road or railroad embankments (Figure 14a-d).
Kit fox use of manmade structures presents a further complication when assessing fox presence in urban environments. Kit foxes regularly den within pipes, under storage containers, buildings with raised foundations, stacked materials, vehicles, or any other structure or object that offers shelter and is large enough for a fox to enter (Figure 15a-d).
Manmade structures may be permanent or temporary and may be occupied by kit fox for a single night or habitually. There may be little or no indication that a fox is present at the site. Because these features often are large and provide shelter without digging, typical sign of fox presence, such as dirt mounds, prey remains, or tracks, may be absent or obscured.
Detection and evaluation of kit fox dens in urban environments is challenging. Access to affected or adjacent properties may be restricted at the time of surveying. Irrigation associated with landscaping may degrade dirt mounds, prey remains, and tracks. In some cases dense thatch entirely obscures den entrances (Figure 16a-c). All potential earthen and manmade structures should be thoroughly inspected for sign of kit fox occupancy. In addition to standard procedures, the use of track plates and/or fiber-optic burrow probes may be necessary. Lastly, it is recommended that the surveying biologist interview local landowners, employees, and passersby about kit fox presence, den occupancy, and historic use of a potential project site. Local citizens often will have useful information on kit fox activity and habits that can focus survey efforts.
Natal dens, or dens where whelping and rearing of young occur, present a special circumstance that warrants increased protection. If a natal den is detected within or adjacent to the project site, all activities should immediately stop and the appropriate authorities should be notified. Natal dens typically have more entrances, tracks, scat, prey remains, and human garbage, and may have a broader apron of matted dirt and/or vegetation at one or more entrances (Figure 14b). As with non-natal dens, they may occur 27 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox in a wide range of circumstances, from earthen dens more typical of natural lands to manmade structures where presence of the family group is difficult to confirm without direct observation of young. If interviewed, local citizens often will provide detailed information on the whereabouts and historic use of natal dens.
Artificial Den Description
Every effort should be made to avoid destroying or damaging known dens during project activities, as per standardized recommendations by U.S. Fish and Wildlife Service (e.g., surveys, monitoring, and exclusion zones; U.S. Fish and Wildlife Service 1999). Den destruction may at times be unavoidable. Consultation must occur prior to disturbance or destruction of den sites. While careful excavation may reduce the potential for direct mortality, loss of den sites has long-term negative effects that reduce the quality and viability of urban environments for kit fox.
The construction of artificial dens is a recommended mitigation for the disturbance or destruction of earthen or manmade kit fox dens during project activities. Artificial dens are of two types: subterranean and escape. Subterranean dens consist of two entrance tunnels (3-m length of 20-cm diameter polyvinyl chloride or high-density polyethylene pipe) leading down at no more than a 30-degree angle to a subterranean chamber (0.5 m in diameter, 0.3 m tall). Entrance tunnels should be oriented such that water does not flood the chamber. It is also recommended that a longitudinal section of the tunnel floor be removed to increase drainage and provide better footing for foxes that are exiting the chamber. The floor of the chamber should be located approximately one meter below ground. The chamber may be either box- or dome-shaped, but must have an earthen floor that allows kit foxes to enlarge the den over time (Figure 17a-d).
Escape dens consist of 3- to 6-m lengths of 20-cm diameter pipe (polyvinyl chloride or high density polyethylene) placed on the surface of the ground and covered with 1-2 m of soil (Figure 18a-d).
Both subterranean and escape dens can be modified to exclude entrance by larger predators, particularly red foxes (Vulpes vulpes) and coyotes (Canis latrans). These predators could be excluded by narrowing the den entrances to a width of 10-15 cm. This can be accomplished in a variety of ways, such as driving stakes in front of the entrances, or installing prefabricated diameter-reduction couplings.
Artificial subterranean and escape dens are regularly used by kit foxes in urban environments for resting, predator avoidance, avoiding temperature extremes, moisture conservation, and rearing of young. From June 2001 to June 2004, 22 subterranean and 9 escape dens in urban Bakersfield, CA were monitored using tracking media, remotely- triggered cameras, and radio telemetry (Cypher et al., unpublished data). Kit foxes used 29 of the 31 dens during the study period. These dens were checked with tracking media 8548 times, and fox tracks were detected on 1198 occasions (14%). This rate of visitation was a minimum estimate because precipitation, irrigation, and visitation by other animals at times made detecting kit fox tracks difficult. In spring 2003, and again in spring 2004, two different family groups raised young in artificial subterranean dens (Figure 19a-b).
28 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox Artificial Den Installation Criteria
The following criteria should be met when selecting locations to construct artificial dens. 1) Artificial dens should be as close as possible to the destroyed or disturbed fox den. 2) The location should have a high likelihood of kit fox visitation. 3) Den entrances should be oriented to avoid flooding.
4) The den complex should have reduced likelihood of future disturbance, including development, regular maintenance, and human tampering.
Permanent cautionary devices that alert maintenance personnel to the presence of the artificial den may allow construction in regularly maintained areas. If a suitable site is unavailable along the road corridor it may be necessary to arrange for cooperative agreements with adjacent landowners to install artificial dens on their properties.
Suitable properties may include railway, canal, or utility corridors, drainage basins, golf courses, parks, and industrial sites. Throughout the site selection and construction process, the suitability for kit fox, the expected longevity, and the impacts on regular maintenance activities should be carefully evaluated.
KIT FOX ROAD CROSSING STRUCTURES
Summary
Vehicle strike is the leading cause of mortality for urbanized kit foxes. The likelihood of strike increases with increasing road width, speed limits, and traffic volume. Strikes often occur at the intersection of roads with linear rights-of-way. Linear rights-of-way include other roads (of all sizes), railroads, canals, utility corridors, golf course crossings, and riparian areas. Kit foxes appear to use linear corridors for movement, potentially accounting for the increase in strikes at these locations.
Major roads and associated features may restrict fox movements and consequently reduce gene flow, dispersal, colonization of habitat patches, and the resilience of fox populations to disturbance. This is especially likely in urban environments where road density, width, and traffic volume are increased and habitat is fragmented and susceptible to change.
Maintaining connectivity is identified as a primary recovery action for kit fox in section II.L.6.a. of the Recovery Plan for Upland Species of the San Joaquin Valley, California (U.S. Fish and Wildlife Service 1998).
The construction of fox-specific road crossing structures is a recommended conservation strategy for new high-volume roads, the widening of existing roads, or the rebuilding of existing structures that have the potential to facilitate kit fox road crossings. Potential crossing structures include underpasses, overpasses, and culverts. In the following text we describe crossing structures for urbanized kit foxes and provide criteria for site selection and instructions for building or modifying structures.
Road Crossing Structure Attributes
Little is known about the attributes that make crossing structures appealing to the San Joaquin kit fox. Clevenger (2005) conducted a comprehensive literature review on the 29 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox subject, but his conclusions suffered from a lack of data specific to kit fox. To fill this gap, Cypher et al. (Interagency Agreement No. 43A0068) have initiated a 2005-2006 study for California Department of Transportation to evaluate non-engineered crossing structures along 4-lane highways within the range of the San Joaquin kit fox and the ecologically similar desert kit fox of the Mojave eco-region. While these data will be useful when they become available, they will have limited applicability to kit fox crossing structures in urban environments.
Urban habitats are highly fragmented, and urban kit foxes develop habitual travel behaviors, often utilizing linear corridors for movement through the complex of human polygonal and linear development. In contrast, the Cypher et al. 2005-2006 study will examine crossing structures on roads that transect native habitat. There is no evidence that kit foxes in natural lands concentrate road crossings at specific features, such as drainage culverts, bridges, or road intersections when medians, fencing, or other impediments are absent.
In a recent study of the effects of two-lane highways on San Joaquin kit fox in natural lands (LoKern), Cypher et al (Interagency Agreement No. 06A0814, 55004872) found no evidence of fox avoidance of highways or of favored crossing locations. It should be noted, however, that traffic volumes on the highways in the LoKern study (approximately 8500 vehicles per day) were similar to volumes found on local roads in the city of Bakersfield (this study). Neither highways at LoKern nor local roads at Bakersfield significantly contributed to kit fox mortality. In other words, there may be little benefit to utilizing crossing structures or other habitual crossing locations when there is little risk of strike.
The following recommendations are based on knowledge of kit fox behavior, movements, territory use, and locations of mortality. We have combined these data with general recommendations on structure design and site selection for similar species.
Ideally, crossing structures in urban environments shall be institutionally linked with long-term monitoring programs and clear criteria for determining efficacy.
On the most basic level a crossing structure should satisfy two conditions. First, it should reduce mortality for the target species. Because vehicle strike is the primary cause of death for kit foxes in urban lands, and strikes occur disproportionately on major roads at identifiable geographic features (i.e., intersections with linear rights-of-way), we conclude that crossing structures have a high probability of reducing mortality.
Consequently crossing structures are an appropriate and necessary mitigation for urban road projects.
Second, crossing structures should increase habitat connectivity. This is a more difficult condition to evaluate. As noted, urban habitats are highly fragmented. To the extent that roads, and especially high-volume roads, contribute to fragmentation, we conclude that successful crossing structures will increase connectivity. As demonstrated (this study), kit foxes have seasonally influenced behaviors and susceptibility to vehicle strike.
During the month of December, male kit foxes increase scent marking frequency (an indicator of territory defense) and movements in search of mating opportunities. Not surprising, there is an increase in vehicle strike for male kit foxes during this period of instability in social dynamics. Furthermore, kit fox juveniles disperse during summer months, sometimes several miles from the maternal range, again encountering many 30 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox roads. Finally, urban kit foxes are displaced by new development and other human activities, resulting in range instability and dispersal to new territories. It is not known, however, whether the presence and types of roads in the urban environment influence the direction and length of movements and dispersal.
We take a hierarchical approach, from landscape to individual behavior, in describing requirements of fox-specific crossing structures. Criteria include: 1) road type; 2) kit fox presence and adjacent habitat suitability; 3) presence of natural crossing locations; 4) embankments, walls, or fencing that funnel individuals toward the crossing zone; and 5) passage design that meets kit fox behavioral requirements. We stress that these criteria are generated with expert opinion and data from ongoing observational studies of urban foxes. Hypothesis driven experiments will be necessary to evaluate the relative importance of these variables.
Criterion 1: Road type
We examined the distribution of fox-vehicle collisions on four types of urban roads - highway, arterial, collector, and local. Of these, kit foxes disproportionately died on arterial roads. Arterial roads typically were wider and had higher posted speed limits and traffic volumes than collector and local roads. Foxes also died on collector roads in Bakersfield from 1985-2004. In some cases, collector roads had similar traffic volumes and widths to arterial roads. Highways, which had the highest traffic volumes and speed limits, weren’t a marked source of fox mortality during this study. The absence of fox- vehicle collision on highways can be explained by a barrier effect –animals that do not attempt to cross, will not be struck. Dispersal of individuals between habitat fragments, however, is an essential function of kit fox populations (U.S. Fish and Wildlife Service 1998) and a reduction in dispersal due to roads has been shown to strongly impact the demographics and extinction probability of small populations (Jaeger and Fahrig 2004).
Therefore it is likely that highways posed a significant road effect (if not direct mortality) on urbanized foxes. Lastly, there was little evidence for an association between local roads and fox-vehicle collision or altered kit fox movements.
Based on these findings, we recommend that crossing structures be considered on all highway, arterial, and collector roads. We do not recommend crossing structures as a protection or compensation measure for local roads.
Criterion 2: Kit fox presence and adjacent habitat suitability Presence of kit foxes may be evaluated by spotlighting, surveillance or track stations, surveying for scat, tracks, and dens, or interviewing citizenry. Presence/absence data, however, may be fallible to some degree. Kit foxes may be absent not because the site is unsuitable, but because of localized extirpation. In fact, fox populations throughout the species’ range undergo cycles of extirpation and recolonization, known as metapopulation dynamics (U.S. Fish and Wildlife Service 1998). Metapopulation dynamics influence kit fox demographics and extinction probability, and increase in strength with increasing fragmentation. Furthermore, it has been shown that with growing isolation of habitat fragments, it is less likely that fragments will be recolonized in the event of local extinction. Nowhere is kit fox habitat more fragmented and more 31 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox frequently altered than in urban environments. Therefore, it is likely that at any given time a subset of suitable habitat in urban environments will be unoccupied by kit fox.
A conservative alternative to presence/absence surveys is to concentrate crossing mitigation at locations that have suitable adjacent habitats. As a general rule urban kit foxes occupy industrial and commercial lands, parks, golf courses, and schools. Within these lands, occupancy appears linked to the extent of contiguous habitat, the quantity of undeveloped or lightly developed lots, the presence of drainage basins (favored den sites), and the presence of linear features that provide den building and movement opportunities such as canals, utility corridors, railroads, and riparian areas. Kit foxes do not typically occupy residential zones unless a golf course or other large open space is present (in which case occupancy is highly likely).
Criterion 3: Areas of increased road crossing
Clevenger (2005) reports that animal use of a crossing structure is linked to placement near natural crossing points. This assessment is supported by crossing studies (e.g., Foster and Humphrey 1995, Yanes 1995). Urban kit foxes more often are killed in vehicle strikes at the intersection of major roads with linear rights-of-way including other roads (of all types), canals, railroads, utility corridors, and riparian areas (this study). The concentration of mortalities at these intersections increases the likelihood of successful mitigation. First it narrows the selection of sites for crossing structures to easily identifiable landscape features. Second, it decreases the need to inhibit crossing at other locations. Third, intersections with linear rights-of-way often have pre-existing crossing structures. For example, a bridge over a canal or river may only require a design modification to be suitable for kit fox. When implemented early in the planning process these modifications may have little additional cost.
Because of the concentration of road crossings and vehicle mortalities at linear rights-of- way, we recommend that mitigation efforts prioritize these locations. For practical reasons, it will not be possible to install crossing structures at most road-road intersections. Instead, we recommend placing highest priority on installing or modifying structures at intersections of roads with highways, railroads, canals, rivers, or utility corridors and other locations where an underpass, overpass, or culvert is required.
Criterion 4: Embankments, walls, and fences
Another important predictor of use of crossing structures and decrease in vehicle strikes is the construction of manmade barriers that funnel animals toward the passageway (e.g., Haas 2000, Clevenger 2005, Mata et al. 2005). Barriers may include embankments, walls, and fences.
Embankments appear to discourage road crossing. Cypher et al. (in prep.) observed regular use by kit foxes of a railway underpass on a raised portion of highway 99 in Bakersfield (Figure 20a). The embankment was approximately eight meters tall and the underpass span was approximately 50 meters wide. Radio-tagged foxes were never known to climb the embankment to travel across the highway surface. Kit foxes constructed dens in the earthen slopes within the underpass and even whelped and reared young at the site for multiple seasons (Figure 20b). Embankments on raised highways 32 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox may fragment the landscape in the absence of underpasses or other crossing opportunities. Crossing structures should always be considered when building or modifying raised roadways.
Most urban roads are at grade with the surrounding environment. Cement walls may border highways, arterial, and collector roads, but this is typical only of residential areas, which may not be suitable for kit fox. Other roads are bordered by vegetation, low walls, and metal or wood fencing, all of which may be permeable to kit foxes. Small, skilled at digging, and agile, the kit fox is adept at passing through barriers designed for humans or other large mammals.
We recommend the installation of walls at potential crossing sites that encourage use of the passage and discourage crossing the road above grade. Studies on other species suggest that barriers are more effective when they are shaped as a funnel that directs animals toward the structure. It also is recommended that there be a method of exit for animals that get onto the roadway within the barrier zone. The complexity of urban environments makes meeting these recommendations difficult. Nevertheless, barriers are an integral part of successful crossing structures and should be implemented with all crossing projects.
We make the following specific recommendations. Barriers should be approximately 1.5 meters in height and of a solid material, such as cement block, that discourages climbing.
Walls should be impermeable to kit foxes that approach from adjacent habitat, but permeable to foxes that are on the roadway. One method to achieve one-way permeability is to install a dirt embankment next to the wall on the side nearest the road that allows a fox to run to the top of the wall from roadside (Figure 21a-b). The wall should remain 1.5 meters above grade on the side that borders habitat to prevent entrance to the roadway. Based on behavioral observations of radio-collared kit foxes and location of kit fox mortalities (this study), barriers should lead directly from the passage and run parallel to the road for a distance of at least 50 meters. In some cases, barriers will best be implemented as part of and in addition to existing natural or manmade features, including embankments, walls, and buildings. Finally, landscaping is recommended between the barrier and roadside (i.e., on the embankment). Urban kit foxes are known to use landscaping for cover while traveling along arterial roadways that are bordered by cement block walls (C. Bjurlin, pers. obs.).
Criterion 5: Behaviorally suitable passage design
We have identified four characteristics that affect behavioral suitability of crossing structures for kit foxes. First, a clear line of sight should be maintained through the passage for the entire breadth of the roadway, with the exit visible from the entrance.
Second, approach to the entrances should be level or gently sloped and free from vegetation or obstacles. Third, the passage should be minimally one meter square, with passage dimensions proportional to the roadway (i.e., as road width increases, passage width and height increases). Fourth, one escape den (see above) should be installed within the passage and one escape den installed on the approach to each entrance to give kit foxes protection from predators. In the following text, we profile existing structures to illustrate the aforementioned criteria. We provide recommended modifications where appropriate.
33 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox Profiled Crossing Structures
Seven Oaks Golf Course underpass
The Seven Oaks Golf Course in Bakersfield, CA was bisected by Buena Vista road, a four-lane, divided arterial road (Criterion 1). The road segment of interest (980 m) was bounded to the north by intersection with Ming Avenue and to the south by intersection with Chamber Blvd. Traffic volumes were estimated in 2002 at 13,750 vehicles per day (vpd). To maintain connectivity for golfers, an underpass structure was installed. This underpass met nearly all of the criteria for a successful crossing structure. The golf course was occupied by kit foxes and extended on both sides of the passage (Figure 22, Criterion 2). The greens narrowed where they approached the crossing site, creating a natural transverse point (Criterion 3). Two-meter high walls bordered the road (Criterion 4). The passage was greater than two meters square and had flanged and unobstructed entrances (Figure 22b-c, Criterion 5). One confirmed vehicle strike occurred at the site after construction of the barrier wall (Figure 22a). There were gates in the walls at a maintenance yard directly southeast of the passage and at a road access point to the southwest of the passage. A kit fox may have used one of these gates to gain access to the roadway.
Analysis - This design was suitable for locations with pedestrian or bicycle routes, utility corridors, golf courses, or other linear rights-of-way that are not associated with water transport. It would be improved by the installation of escape dens and the modification of the nearby gates to prevent kit foxes from entering the roadway.
Truxtun Extension horse culvert
Truxtun Extension was a 4-lane, divided collector road in Bakersfield, CA (Criterion 1).
Traffic volumes in 2002 were estimated at 35,000 vpd. The road segment of interest (5.3 km) occured between the intersection with Wible Road to the east and Coffee Road to the west. The segment was bordered to the north by the Kern River and to the south by a canal and mixed residential and commercial zones. Kit foxes occupied the surrounding lands. Radio-tagged animals repeatedly crossed the road and five confirmed vehicle strikes occurred along the road segment (this study). An underpass culvert was built 270 m east of the junction with Coffee Road. The passage location had suitable habitat to the north (Criterion 2). A pedestrian trail connected the passage to a canal on the southern side of the roadway (Figure 23a-b, Criterion 3). There was a chain link fence that may have impeded access to the road on the southern side, but it was broken by a pedestrian trail 15 m to the west of the southern entrance. The northern entrance had a 20-m chain link fence directly over the culvert (Figure 23c). The road was easily accessible from the north along the remainder of the segment. The culvert was approximately 2.5 meters in diameter with an earthen bottom that was seasonally flooded (Figure 23d, Criterion 5). On one occasion a kit fox was seen running north across the road directly over the crossing structure.
Analysis – This was a suitable passage at an appropriate location that required fencing and escape dens to become fully operational.
34 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox Coffee Road bicycle path underpass
Coffee Road was a six-lane, divided, arterial road in Bakersfield, CA (Criterion 1).
Traffic volumes in 2002 were estimated at 45,100 vpd. The road segment of interest (1600 m) was bounded to the north by intersection with Brimhall Road and to the south by intersection with Stockdale Highway. Surrounding lands were a mixture of commercial and undeveloped properties and were suitable for kit fox (Figure 24a, Criterion 2). Several canals and the Kern River corridor intersected the road. A bridge that spaned the Kern River also provided passage for a bicycle path on the southern embankment. The river corridor, proximity to road and canal intersections, and presence of a bicycle path made the passage a natural crossing point (Criterion 3). The west side of Coffee Road was bounded by chain link fence, but the east side had open access to the roadway (Figure 24b-c, Criterion 4). The passage was greater than 3m in height and width, and free from obstruction (Figure 24d, Criterion 5). Five confirmed vehicle strikes occurred on the road segment, including one that was 60 m south of the passage (Figure 24a & d).
Analysis – This was a suitable passage under a road with high traffic volume that required fencing and escape dens to become fully functional.
Highway 58 canal bridge
Highway 58 was a four-lane, divided roadway (Criterion 1). The road segment (2200 m) was bounded to the east by intersection with Landco Drive and to the west by intersection with Patton Way. Traffic volumes in 2004 were estimated at 47,000 vpd. Surrounding lands were a mixture of commercial, industrial, and undeveloped properties. A canal and arterial road (Fruitvale Ave) formed a six-way intersection with Highway 58. Kit foxes and their habitat were abundant along the road segment (Figure 25a, Criterion 2). The confluence of canal, arterial road, and highway made it a natural kit fox crossing location (Criterion 3). Chain link fence occured sporadically along the canal, but gaps in the fence and at gates allowed foxes to gain access to the roadway. The road was not walled or fenced. This underpass complex was composed of two separate bridges, one on Fruitvale Ave, the other on Highway 58. The Fruitvale Ave. bridge was impassable at the time photos were taken. The canal was transporting water and there was no embankment along its margin beneath the bridge (Figure 25b). In contrast, the Highway 58 bridge remained passable while transporting water. There were dirt embankments on both margins of the canal beneath the bridge, along with unobstructed entrance and exit locations (Figure 25c-d, Criterion 5). No fox-vehicle collisions were recorded at the location, but kit foxes frequently crossed the road during a telemetry study (Figure 25a).
A radio-tagged kit fox was repeatedly tracked to the canal and twice tracked to the canal underpass when there was no water present.
Analysis - The confluence of multiple linear rights-of-way along with presence of kit foxes and frequent road crossings made this an appropriate location for a crossing structure. The fencing at the site was insufficient for keeping foxes off the roadway.
When transporting water, one of the bridges did not accommodate fox movement beneath the roadway. This location would require extensive modification to be fully functional. 35 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 14. Earthen kit fox dens in urban Bakersfield, CA. a) Single entrance den under building foundation. b) Natal den complex on canal bank. c) Multi-entrance den in drainage basin bordering intersection of two arterial roads. d) Single entrance den in vacant lot. 36 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 15. Non-earthen den locations for kit foxes in Bakersfield, CA. a) Fox emerging from stacked pipe. b) Narrow entrances to a kit fox den under a storage container. c) Fox resting beside stacked lumber at a construction site. d) Fox returning to a den within a modular housing unit.
37 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 16. Camouflaged earthen kit fox dens in urban Bakersfield, CA. a) Entrance to den in landscaping at parking lot. b) Thatch covered den entrance in a golf course fairway. c) Dirt mound from den in shrub-island at park. 38 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 17. Artificial subterranean dens for San Joaquin kit fox at Bakersfield, CA. a) Artificial den schematic. b) PVC two-entrance chamber den under construction. c) High-density polyethylene two-entrance den. d) PVC tunnel with floor removed longitudinally. 39 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 18. Artificial escape dens for San Joaquin kit fox at Bakersfield, CA. a) Escape den schematic. b) High-density polyethylene escape den under construction. c) Completed den. d) Kit fox entering escape den. 40 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 19. Kit fox pups at artificial subterranean dens at a golf course (a) and drainage basin (b) in Bakersfield, CA.
Figure 20. Railroad underpass on highway 99 in Bakersfield, CA. a) Habitat view. b) Kit fox natal den at the top of the underpass embankment.
41 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 21. Barrier wall to encourage use of road crossing structures for the San Joaquin kit fox in urban environments. The roadside embankment allows foxes that are trapped on the roadway to clear the barrier wall. a) Cross-section view. B. Three-dimensional rendering.
42 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 22. Buena Vista Road golf cart crossing structure, Bakersfield, CA. a) Aerial view of roadkill and radio-tagged kit fox locations at crossing structure. b) Habitat view from east entrance of crossing structure. c) Approach view to east entrance. d) Passage interior.
43 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 23. Truxtun Extension horse culvert, Bakersfield, CA. a) Aerial view of radio-tagged kit fox locations at horse culvert. b) View from pedestrian trail at south culvert entrance (canal in background). c) North culvert entrance. Yellow line depicts extent of fencing on north side of road. d) Culvert interior.
44 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 24. Coffee Road overpass at Kern River bicycle trail, Bakersfield, CA. a) Aerial view of kit fox roadkill and radio-tagged night locations at crossing structure. b) Habitat view from west entrance. c) East entrance to bicycle path on southern embankment of Kern River. d) View of roadway from East with location of kit fox roadkill in red. No fences prevent fox access to roadway on eastern side.
45 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
Figure 25. Highway 58/Fruitvale Avenue canal bridge, Bakersfield, CA. a) Aerial view of radio-tagged kit fox night locations at crossing structure. b) East entrance to Fruitvale Bridge obstructed by water. c) South entrance to highway 58 bridge. d) Dirt embankment under Highway 58 Bridge.
46 Appendix A. Supplemental Recommendations for Protection of San Joaquin Kit Fox
LITERATURE CITED
Clevenger, A.P. 2005. Design and performance of wildlife crossings for kit foxes and ecologically similar small carnivores: A literature review and semi-annotated bibliography. California Department of Transportation, San Luis Obispo, California.
Foster, M.L., and S.R. Humphrey. 1995. Use of highway underpasses by Florida panthers and other wildlife. Wildlife Society Bulletin 23, 95-100.
Haas, C.D. 2000. Distribution, relative abundance, and roadway underpass responses of carnivores throughout the Puente-Chino Hills. MSc thesis, California State Polytechnic University, Pomona.
Jaeger, J.A., and L. Fahrig. 2004. Effects of road fencing on population persistence. Conservation Biology 1651-1657.
Koopman, M. E., J. H. Scrivner, and T. T. Kato. 1998. Patterns of den use by San Joaquin kit foxes.
Journal of Wildlife Management 62:373-379.
Mata, C., I. Hervás, J. Herranz, F. Suárez, and J.E. Malo. 2005. Complementary use by vertebrates of crossing structures along a fenced Spanish motorway. Biological Conservation 124, 397-405.
U.S. Fish and Wildlife Service. 1998. Recovery plan for arid upland species of the San Joaquin Valley, California. U.S. Fish and Wildlife Service, Portland, Oregon.
U.S. Fish and Wildlife Service. 1999. Standardized recommendations for protection of the San Joaquin kit fox prior to or during ground disturbance. U.S. Fish and Wildlife Service, Sacramento, California.
Yanes, M., J.M. Velasco, and F. Suárez. 1995. Permeability of roads and railways to vertebrates: the importance of culverts. Biological Conservation 71:217-222. 47 View publication stats
SAN JOAQUIN KIT FOX HOME RANGE, HABITAT USE,
AND MOVEMENTS IN URBAN BAKERSFIELD
by
Nancy Frost
A Thesis
Presented to
The Faculty of Humboldt State University
In Partial Fulfillment
Of the Requirements for the Degree
Master of Science
In Natural Resources:
Wildlife
December 2005
iii
ABSTRACT
San Joaquin kit fox home range, habitat use, and movements in urban Bakersfield
Nancy Frost
Habitat destruction and fragmentation has led to the decline of the federally endangered and state threatened San Joaquin kit fox (
Vulpes macrotis mutica
). Kit foxes in the San Joaquin
Valley occur in and adjacent to several urban areas.
The urban kit fox population
in Bakersfield
has the potential to contribute to recovery efforts, however, little is known about habitat use in the
urban environment.
H
abitat use in the urban landscape
was examined in 28 radiocollared adult
kit foxes
between 1 May 1997 and 15 January 1998.
The area of the urban environment needed to meet kit foxes’ ecological requirements varied among the sexes and f
luctuated throughout the year. Mean home range size was 1.72 km 2 (100% minimum convex polygon estimate) and mean concentrated use area (core area ; 50% fixed
kernel estimate
) size was 0.16 km
2 . Females had a greater median number of core areas than did ma
les, but core area size did not differ between the sexes. Mean home range size was significantly larger during the dispersal season than during the pair formation and breeding seasons.
Home range size was greater for males than females during the breedin g season.
To assess the extent to which kit foxes were concentrated around limited resources, the percentage of overlap between adjacent kit foxes’ home ranges and core areas was determined
.
Mean home range overlap was greater than 71
%
for family group
members and less than 16%
for
unrelated individuals, which was l
ess than that reported for the nearby rural population. Overlap of adjacent kit foxes’ core areas indicated that kit foxes were concentrated around key urban habitat patches.
C
ore area overl
ap was greater for members of the same family
group
than
core
area overlap of neighboring males and females
or
neighboring males.
iv
To determine which attribute
s of the urban environment were
used by kit foxes and may
support the population over time, habita
ts used disproportionately
were compared
to what was
available in home ranges and the urban landscape. Kit foxes used open and sump (water catchment basins) habitats more than their availability in both the home range area and study area. There was more
open and sump habitat in home ranges than in the study area. Kit fox core areas contained a disproportionately high amount of sumps that had food and den sites and restricted public access.
Sump and open habitats
were more suitable for denning and consequ
ently, were used
more than their availability in the study area.
Kit foxes primarily used subterranean dens but also
used pipes and other man
-
made structures. Dens were located farther from roads than expected and in habitat patches that were larger than
the median habitat patch in the study area
.
To identify which urban habitats
kit foxes used for resting, foraging, and traveling,
kit
fox movements and travel paths through adjacent habitat patches were examined
. Kit foxes
moved an average speed of 1.3
3 km/h. Minimum speed for females was
greatest
from
approximately 0330
-
0730
h and for males was
great
est from
approximately 1830
-
2230 h
. Kit
foxes used sump and open habitats for resting. Kit foxes also foraged in these habitats, as well as in trans
ition and commercial habitats. Habitats used for traveling included manicured open and open habitats, in which total distance traveled per movement session was longest. Overall median speed was
great
est in residential habitat. Kit foxes were observed cr ossing roads and
using culverts and bridges to move under roads.
D
emographic characteristics
and weights
of the kit foxes that
were
captured
were used to
assess habitat quality
.
Two of the 28 radiocollared kit foxes died during
6,036 days of active
radi
otransmitters
.
Radiocollared kit foxes had 15 litters, comprised of 2.8 +
0.45 (range = 1
–
7)
pups.
Urban kit fox weight was similar to that for rural kit foxes.
v
The availability of sump and open habitats
appear
ed
to
shape the distribution and
abundan
ce of the urban population. Therefore, retaining sump and open habitats throughout Bakersfield may promote the continued existence of the urban kit fox population.
vi
ACKNOWLEDGEMENTS
I would like to thank all those who contributed to the successful comple tion of this
research project, especially my advisor, Dr. R. T. Golightly, for sharing with me both his advice and expertise. In addition, I thank my other committee members, Drs. T. L. George and D. W.
Kitchen for their support and constructive comments
that greatly improved the quality of this
thesis. My thanks go out to my professors and fellow students at Humboldt State University who contributed greatly to my educational growth. I would like to thank the San Joaquin chapter of The Wildlife Society f
or the research grant and the Eureka Rotary Club and the Stockton Sportsmen's Club for scholarships.
I would like to thank the California Department of Fish and Game, especially M. Potter and R. Reed, for assisting with this project by providing a vehicle and hiring three scientific aids,
T. Chilton, J. Storlie, and R. Siebdrath, whose hard work, dedication, and willingness to learn contributed immensely to this project. I appreciate the volunteer efforts of M. Otten, G. Warrick, B. Vanherweg, B. Cypher, J
. Hinshaw, and K. Brown
. I thank
the Department of Energy,
Chevron USA, and Enterprise Advisory Services, Inc., especially J. Killen and B. Cypher, for the loan of telemetry equipment, radio collars, kit fox handling equipment, and traps. I would also li
ke to thank P. Hoffman from the City of Bakersfield Water Resources Department and A. Abe from the City of Bakersfield Recreation and Parks Department for arranging access to city sumps and parks, respectively. Furthermore, I wish to thank J. Tobar, with the City of Bakersfield's
Management Information Services Department, for the City of Bakerfield geographic information system coverages. I thank Dr. C. Hammans of the Coffee Road Animal Hospital and D. Kroger from the California Living Museum for their g
enerous participation in the successful
rehabilitation of an injured kit fox pup. I wish to acknowledge the Bakersfield SPCA for
vii
"rescuing" one of my study animals that made its way to the third story of a parking garage and could not find the exit. My t
hanks go to A. Sullivan, J. Redmond, and M. Cantrell of the Sundale golf course, D. E. Keith Trucking, and K. S. Fabrication for allowing access to their property for implementation of the study. I wish to acknowledge Kelly Pipe Company for their cooperat ive
efforts in helping me retrieve a kit fox pup that went for an unexpected "ride" in one of their pipes that was delivered from Bakersfield to Santa Fe Springs, California.
Special thanks go to my husband, S. Telthorst, for his tremendous support, encour
agement, and participation in this research project. Finally, I dedicate this thesis to S.
Telthorst and my late mother, S. Castle, whose support and lifelong example of patience, perseverance, and persistence imparted me with the skills necessary to purs ue and fulfill my
dreams.
The
United States
Fish and Wildlife Service, California Department of Fish and Game, California Department of Transportation, and Humboldt State University Institutional Animal Care and Use Committee (protocol #96/97.W.62) issued
the necessary permits and memorandum
of understanding for this project.
viii
TABLE OF CONTENTS
Page
ABSTRACT
................................
................................
................................
.........................
iii
ACKNOWLEDGEMENTS
................................
................................
................................
....
vi
LIST OF TABLES
................................
................................
................................
.................
x
LIS
T OF FIGURES
................................
................................
................................
...............
xi
LIST OF APPENDICES
................................
................................
................................
.......
xii
INTRODUCTION
................................
................................
................................
..................
1 STUDY AREA
................................
................................
................................
......................
5 METHODS
................................
................................
................................
............................
9 Trapping
................................
................................
................................
....................
9 Locations
................................
................................
................................
...................
9 Home Range
................................
................................
................................
............
12 Habitat Use
................................
................................
................................
..............
15 Movements
................................
................................
................................
..............
17 Demographics
................................
................................
................................
..........
19 RESULTS
................................
................................
................................
............................
21 Trapping
................................
................................
................................
..................
21 Locations
................................
................................
................................
.................
21 Home Range
................................
................................
................................
............
21 Habitat Use
................................
................................
................................
..............
23 Movements
................................
................................
................................
..............
30 Demographics
................................
................................
................................
..........
42 ix
TABLE OF CONTENTS (CONTINUED)
DISCUSSION
................................
................................
................................
......................
43 Demographics
................................
................................
................................
..........
43 Locations
................................
................................
................................
.................
45 Home range
................................
................................
................................
..............
46 Habitat use
................................
................................
................................
...............
50 Movements
................................
................................
................................
..............
53 MANAGEMENT RECOMMENDATIONS
................................
................................
...........
57 REFERENCES
................................
................................
................................
.....................
59 x
LIST OF TABLES
Table
Page
1 Habitat types and their descriptions used
to typify the urban setting and kit fox
habitat use in Bakersfield, California from May 1997 to January 1998. .......................
7 2
Home range (100% minimum convex polygon) and core area (50% fixed kernel) size for radiocollared kit foxes in Bake
rsfield, California from May 1997 to
January 1998.
................................
................................
................................
.........
24 3
Seasonal home range (100% minimum convex polygon) size for radiocollared kit foxes in Bakersfield, California from May 1997 to January 1998. ..............................
25 4
Ranked radiocollared kit fox habitat use based on habitat availability at the study area and home range area scale, in Bakersfield, California between May 1997 to January 1998. Habitat types used similarly (p > 0.05) are grouped by column. ...........
28 5
Mean difference between the mean proportion of habitat available in the study area and kit fox home ranges (e.g., there was 16.7% less open habitat in the study area than within home ranges), and kit fox home ranges and core areas in Ba
kersfield, California between May 1997 to January 1998. ................................
......
29 6
Influence of sex, movement session time, and their interaction on minimum, maximum, and mean speed and total adjusted distance moved for 10 radiocollared kit foxes in
Bakersfield, California from May 1997 to January
1998.
................................
................................
................................
.....................
36 7
Movement session time, type of movement, and habitat associations for radiocollared kit foxes in Bakersfield, California from September 1997 to January 1998.
................................
................................
................................
.........
39 8
Rural kit fox and swift fox cause
-
specific mortality rates reported within their
geographic range for comparison to that for urban kit foxes in Bakersfield, California from May 1997 to January 1998.
................................
..............................
44 9
K
it fox home range size reported within various habitats within their geographic range for comparison to home range size of urban kit foxes in Bakersfield, California from May 1997 to January 1998.
................................
..............................
47 xi
LIST OF FIGURES
Figure
P
age
1 The study area in Bakersfield, Kern County, California for investigations of kit fox habitat use from May 1997 to January 1998.
................................
........................
6 2
Behavior exhibited by directly observed kit foxes at independent daily lo cations
(n = 2,087) in Bakersfield, California from May 1997 to January 1998. ......................
22 3
Percentage of home range overlap and core area overlap for pairs of radiocollared kit foxes in Bakersfield, California from May 1997 to January 1 998. Vertical
bars represent
+
1 SE.
................................
................................
.............................
26 4
Mean percentage of each habitat type in the study area, which bounded the available habitat, and radiocollared kit fox home ranges (n = 27) and core areas (n = 58) in Bakersfield, Calif
ornia from May 1997 to January 1998. Different
letters above the bars indicate a significant difference in habitat use versus availability.
................................
................................
................................
............
27 5
Mean number of passing vehicles, passing pedestrians, and people in the same habit
at patch as radiocollared kit foxes at all daily locations (n = 4,824) in Bakersfield, California from May 1997 to January 1998. Vertical bars represent +
1 SE.
................................
................................
................................
...................
31 6
Number of dens used by radiocollared kit foxes in each habita t type in
Bakersfield, California from May 1997 to January 1998.
................................
...........
32 7
Percentage of available habitat in the study area compared to percentage of habitat used by radiocollared kit foxes for denning in Bakersfield, California fro
m May 1997 to January 1998.
................................
................................
.............
33 8
Minimum speed per movement session for male and female radiocollared kit foxes in Bakersfield, California from September 1997 to January 1998. .....................
35 9
Mean distance radio
collared kit foxes traveled in each habitat type in Bakersfield, California from September 1997 to January 1998. Vertical bars represent +
1 SE.
......
37 10
Mean speed radiocollared kit foxes traveled in each habitat type in Bakersfie ld,
California from September 1997 to January 1998. Vertical bars represent +
1 SE.
......
38 xii
LIST OF APPENDICES
Appendix
Page
A
Sites of live
-
captured kit foxes, domestic
cats, and opossums between May 1997
and January 1998 in Bakersfield, California.
................................
.............................
69 B
Kit fox location data and estimates for home range size, core area size, and time to independence for 28 rad
iocollared kit foxes that were monitored between May
1997 and January 1998 in Bakersfield, California.
................................
....................
70 C
Weight, hind foot, and ear measurements for captured adult, subadult, and pup kit foxe
s in Bakersfield, California between May 1997 and January 1998. ..................
73 D
Area
-
observation curves using the 100% minimum convex polygon home range for 27 kit foxes in Bakersfield, California from May
1997 to January 1998.
................
74 E
The percentage of all radiocollared kit fox daily locations (n = 4,824) and passing vehicles documented in linear habitats during each hour in Bakersfield, California
from September 1997 to January 1998.
................................
.....................
75 1
INTRODUCTION
The kit fox (
Vulpes macrotis
) is the smallest North American member of the family
Canidae, with an average weight of 2.1 kg for adult females and 2.3 kg for adult male s in
California (Morrell 1972). Kit foxes range throughout the western United States in shortgrass prairies and other arid areas (Whitaker 1998). The San Joaquin kit fox ( Vulpes macrotis mutica
;
hereafter kit fox) was described by C. H. Merriam in 1902 ( McGrew 1979) and is the largest
California subspecies in skeletal measurements, body size, and weight (Grinnell et al. 1937). Kit foxes inhabit arid habitats with flat to sloping terrain (Morrell 1971).
The kit fox historically ranged in the San Joaquin
Valley of California, north of Tracy,
west to San Joaquin County, east to La Grange, Stanislaus County, and south to Kern County (Grinnell et al. 1937, Morrell 1975). In the early 1900’s, the kit fox population started to decline (
United States
Fish and
Wildlife Service
1983), and by 1930, its range had been reduced by more
than half, with the largest portion remaining in the southern and western parts of the San Joaquin Valley (Grinnell et al. 1937). In 1969, the kit fox was known to range from Los Bano s, Merced
County, east to White River, Tulare County, south to Kern County, and west to San Luis Obispo County (Morrell 1975). Currently, the largest remaining populations of kit foxes occur in western Kern County and in the Carrizo Plain Natural Area, Sa
n Luis Obispo County (
United States
Fish
and Wildlife Service 1998).
In 1967, the San Joaquin kit fox was listed as a federally endangered subspecies ( United
States
Department of the Interior 1967)
, and in 1971, it was designated a State threatened species
(
United States
Fish and Wildlife Service 1998
). The destruction and fragmentation of its native
habitat has been a limiting factor that has led to the decline of the kit fox population (Knapp 1978, Spiegel 1996,
United States
Fish and Wildlife Service 19
98). This habitat fragmentation
2 has been due mainly to agriculture, urban development, and oil production (Morrell 1975, Spiegel 1996,
United States
Fish and Wildlife Service 1998). In 1993, the San Joaquin Valley Biological Technical Committee estimated
that only about 1
-
5% of the original habitat of the kit fox
remained.
Kit foxes in the San Joaquin Valley occur in and adjacent to several urban areas including the cities of Taft (Zoellick et al. 2002), Tulare, Visalia, and Porterville ( United States
F
ish and
Wildlife Service 1998), and Bakersfield, where kit foxes have been observed throughout the city (Morrell 1975, Murphy 1992, Cypher and Warrick 1993, Bjurlin and Cypher 2003). Kit foxes are known to use the Kern River Parkway in Bakersfield (Murphy 1992, Beedy et al. 1992) for
denning, foraging, and traveling, despite the fact that it is bounded on the north and south by urban and industrial development and receives intensive public recreation use (Beedy et al. 1992).
Additionally, urban kit foxes
have been found to den close to human activities and structures (Murphy 1992), including water sumps (catchment basins) in and adjacent to urban development in Bakersfield (Stafford and Murphy 1992).
Urban habitats may be marginal for kit foxes, because t he urban environment has a high
level of disturbance and consists of a patchy mosaic of habitats (Morrell 1975). While urban developments negatively impact kit fox habitat, kit foxes may still survive in or near them if there are adequate den sites and pr
ey base (
United States
Fish and Wildlife Service 1998). Positive
attributes of urban habitats may include a constant source of water from sumps, canals, and lawn sprinklers (Murphy 1992) that could lead to a more stable prey population. While fewer natur al
predators, such as coyotes (
Canis latrans
) and bobcats (
Lynx rufus
), are found in the urban
environment, domestic dogs (
Canis familiaris
) occur in greater abundance. Food may be
available on a more consistent basis due to the presence of refuse, which constitutes part of the kit
fox diet in Bakersfield (Murphy 1992, Cypher and Warrick 1993).
Such use of human
-
derived
foods suggests that kit foxes may be able to adapt to the urban setting.
3 Smaller canid species, such as coyotes and foxes, often do well in human
-
altered
environments because they scavenge and forage individually on a variety of small food items and are able to utilize new food sources (Kleiman and Brady 1978). If kit foxes are adaptable enough to persist in urban environments, as are coy
otes (Quinn 1995, Romsos 1998, Grinder and
Krausman 2001) and red foxes (
Vulpes vulpes
; Harris 1977, Kolb
1984,
1986, Harris and Rayner
1986, Doncaster et al. 1990, Lewis et al. 1993, Sallee 1998, Kamler and Ballard 2002), the urban kit fox population coul
d contribute to recovery efforts (
Cypher and
Warrick 1993). Urban kit fox
populations could protect
against extirpations in rural areas, and serve as a source of animals for reintroduction efforts (Cypher and Frost 1999)
. Favorable habitat features in ur
ban areas also
could serve as linkages or corridors between natural lands, which may minimize local extinction and genetic isolation of fox populations in fragmented habitats ( United States
Fish and Wildlife
Service 1998, Wandeler et al. 2003
).
It is impor
tant to understand how kit foxes use landscapes in all areas where they exist, including urban settings. Unfortunately, little is known about the urban kit fox population.
Morrell (1975) noted that a study should be conducted to gain a better understandi ng of kit fox
ecology in areas other than natural habitat. As part of the recovery strategy in the San Joaquin Valley Multispecies Recovery Plan (
United States
Fish and Wildlife Service 1998), it was
recommended that better demographic information be coll ected for kit foxes living in residential
and industrial areas. The San Joaquin Valley Biological Technical Committee (1993) expressed concern that urbanization threatens the remaining kit fox habitat and listed several kit fox research needs including fu
rther studies of mortality rates and causes, diet, distribution, home range requirements, and movements. Due to the continual loss of natural habitat, more information on urban kit fox space use, movements, survival, and reproduction may contribute to the
ir recovery (Cypher and Warrick 1993).
4 Canid home range size is influenced primarily by food abundance, and also food distribut
ion and
availability of adequate denning sites (Kleiman and Brady 1978). In order to understand the ecological factors that fac
ilitate habitat use, it is important to identify the areas of concentrated use (hereafter core areas) within the home range (Samuel et al. 1985). Optimal habitat conditions for a species may be identified by evaluating if and how core areas are used dispr
oportionately (Ables 1969). Core areas likely have dependable food sources and rest sites (Samuel et al. 1985) such as dens, the availability of which limits the presence of kit foxes (Goli
ghtly 1981, Spiegel et al. 1996
). If kit foxes are concentrated a
round limited resources, the
home ranges and core areas of adjacent kit foxes may have substantial overlap (Golightly 1981).
If sufficient food and den sites are available, mortality is minimal, and reproduction occurs at reasonable rates, then urban habit
at should support viable kit fox populations. The goal of my project was to determine which attributes of the urban environment were successfully used by kit foxes and may support the population over time. To identify the attributes of the urban environm
ent that kit foxes used most, areas and habitats used disproportionately were compared
to what was available in the urban landscape.
T
he quantity and type of urban habitat needed to
meet
the
ecological requirements
of urban kit foxes
(i.e., that
used for
denning, foraging, and
traveling
)
was identified
. To understand habitat use, kit fox home range size, core area size, spatial distribution of kit foxes in the urban landscape, habitat and den use, and movements were
evaluated
in relation to habitat type,
time of night, and human activity. Assessment of habitat quality should be based on the survival, reproduction, and density of the species occupying the habitat (Van Horne 1983). Therefore, I also reported the demographic characteristics of captured urba
n kit foxes.
5 STUDY AREA
The study was conducted within the city limits of Bakersfield (Figure 1), at the southern end of the San Joaquin Valley, California. The average elevation of Bakersfield was 124 m (U
nited States
Geological Survey 1981) and in 199
7, had a human population of 221,700 (State
of California,
Department of Finance
2002).
Bakersfield has an arid climate with an average annua
l precipitation of 14.5 cm (National
Oceanic and Atmospheric Administration
1997) that occurs mainly between Octo
ber and April.
Total precipitation during the study was 10.9 cm, 3.0 cm higher than normal (Meadows Field
Weather Station;
National Oceanic and Atmospheric Administration
1997,
National Oceanic and
Atmospheric Administration
1998). Summers are hot and dr
y, while winters are cool and moist.
During the study, average temperatures ranged from 7.9
°
C in December to 27.2
°
C in July and
annual mean temperature was 19.6
°
C, which was
-
14.3 °
C lower than normal
(Meadows Field
Weather Station;
National Oceanic an
d Atmospheric Administration
1997,
National Oceanic and
Atmospheric Administration
1998).
Vegetation in the study area varied by land use and was composed mostly of introduced and exotic species. Manicured open, commercial, and residential habitats were p lanted mainly
with non
-
native grasses and shrubs (Table 1). Conversely, open, industrial, and some linear habitats, such as the Kern River Parkway, had a mix of introduced and endemic grasses, forbs, and shrubs. Sumps were primarily barren on the sides,
often contained standing water, and had
scattered native and non
-
native vegetation near the upper edges. In addition, sumps were
distributed throughout the city, fenced, and closed to the public. Sumps had sloping sides and ranged from 0.01 to 0.03 km
2 i
n area.
ÊÚ
C
a
l
i
f
o
r
n
i
a
K
e
r
n
C
o
u
n
t
y
S
t
o
c
k
d
a
l
e
H
i
g
h
w
a
y
Interstate 5 Freeway
H
i
g
h
w
a
y
1 7
8 Freeway 99Gosford Road
F
r
e
e
w
a
y
5 8
S
T
U
D
Y
A
R
E
A
B
a
k
e
r
s
f
i
e
l
d
Kern River
k
m
3 0
3 6
N
Figure
1 . The study area in Bakersfield, Kern County, California for investigations of kit fox habitat use from May 1997 to January 1998.
6 7
Table
1 . Habitat types and their descriptions used to
typify the urban setting and kit fox habitat
use in Bakersfield, California from May 1997 to January 1998.
Habitat type
Description
Commercial
Office buildings, hotels, medical facilities, and their associated parking areas.
Transition
Areas in which res
idential, commercial, or industrial construction is in
progress or has recently been completed.
Industrial
Areas of industrial use not generally open to the public, such as manufacturing plants and shipping yards.
Linear
Kern River Parkway, canals, roads
, railroads, and highway and power
line right
-
of
-
ways.
Manicured Open
Parks, schools, and golf courses.
Open
Vacant lots and fallow agricultural fields.
Residential
Apartments and single
-
family housing developments.
Sumps
Water catchment basins used to
contain urban run
-
off and rainwater.
8 Kit fox predators, including bobcat, coyote, and domestic dog (Cypher et al. 2000), were observed at the study site. Bobcats and coyotes were only observed in the Kern River
Parkway
(personal observation
). Califor
nia ground squirrel (
Spermophilus beecheyi
), Botta's pocket
gopher (
Thomomys bottae
), and desert cottontail (
Sylvilagus audubonii
) (all urban kit fox prey;
Cypher and Warrick 1993), were observed at the study site. Domestic cats ( Felis catus
)
and
opossums
(
Didelphis virginiana
),
which
utilize some of th
e same food resources as
urban kit
foxes
and
red foxes
, were observed in the study area
(
Catling 1988,
Golightly et al. 1994
,
Read
and Bowen 2001
) (Appendix A).
9 METHODS
Trapping
Between 3 May and 18 July 1
997, wire
-
mesh live
-
traps (38 x 38 x 107 cm; Model 208,
Tomahawk Live Trap Company, Tomahawk, WI) were set in sumps, parks, the Kern River Parkway, and privately
-
owned land in Bakersfield. Traps were covered with canvas tarps to protect captured animals f
rom sun and inclement weather, and baited with canned mackerel.
Trapping frequency and trap placement was opportunistic. The trapping was limited to areas in southwest Bakersfield where kit foxes or their sign (i.e., tracks, scat, and dens) were observed .
Captured kit foxes were transferred from traps to heavy cloth bags for handling. Each animal was sexed, eartagged, weighed (
+
3%), measured (
+
1 mm; right ear and hind foot length),
and released at the capture location. Uniquely
-
numbered metal eartags
(Model 1005
-
4, National
Band and Tag Company, Newport, KY) were affixed to the right ear for females and left ear for males. Kit foxes were aged based tooth eruption and wear (Harris 1978). Individuals with partially erupted teeth were classified as juve
nile, and those with permanent teeth were classified
as adults. Young of the year were classified as puppies until 14 July, and subadults from 15 July to 30 November, after which
they were considered
to be adults.
Locations
I attached 50
-
g collar
-
mounted
radiotransmitters (Model 80, Telonics, Mesa, AZ) to
captured adults. Radiocollars were equipped with mortality sensors that pulsed approximately twice the normal rate if there was no movement for four hours. Between 14 December 1997 and 15 January 1998,
all but four study animals were
captur
ed and had their radiocollars removed
.
10 Radiocollared kit foxes were located using portable receivers (Model TR -
4, Telonics,
Mesa, AZ), an omni
-
directional antenna affixed to a car, and a hand
-
held 2
-
element yagi anten
na.
Kit foxes were
located between 1630 and 0900 h
because they were
primarily active at night
(Egoscue 1956, Golightly 1981, White
et al. 2000
)
.
At the onset of the study in May 1997, nine
study animals were randomly monitored to verify that they were n ot active during daylight hours.
With the exception of delineating the study area, only locations obtained between one hour before sunset and one hour after sunrise were used in analyses. Each k it fox was
systematically located
approximately once per nig
ht (hereafter referred to as daily locations) but rotated in the schedule each night. Although sampling effort for each study animal was equal at study end, every kit fox could
not
be found during each sampling period. When
a kit fox could not be found,
the area
where it was previously located
was searched
for at least
30 min and then, if not located,
it was
searched for
elsewhere in the study area concurrent with searches for other kit foxes.
The
transmitted signal
was followed
until the kit fox was obs
erved (sometimes with the
aid of a flashlight) or tracked to a den, or the habitat patch from which the signal was emitted was
encircled
(White and Garrott 1990). This technique ensured that the location of and habitat used by each kit fox was classified
without error (Samuel and Kenow 1992). I tried to stay at least 100 m from the study animals, so as to not alter their behavior (i.e., cause them to move from one habitat patch to another). To assess if this technique created a sampling bias by altering their
behavior, each animal was classified as
resting/
walking,
trott
ing, or running at the time of
observation. When this technique was not feasible (e.g., the habitat patch was not accessible), kit fox locations were determined with triangulation, based on three bearings taken as close as
possible (
<
300 m) from the loudest signal source (Springer 1979). Immediately after a kit fox was located, the observer plotted the location on a 1996 City of Bakersfield street map (scale
1:29,000, Compass
Maps, Inc.,
Modesto, CA)
and recorded the number of people
that
were
visible (
i.e., not in buildings or
large habitat patches)
in the same habitat patch as the kit fox.
11 M
apped kit fox locations
were digitized
into a geographic information system (GIS;
ArcView 3.2 wit
h Spatial Analyst 2.0, ESRI, Redlands, CA), and
d
isplayed
over
the City of
Bakersfield's 1997 GIS file titled “Land Status Land Condition Land -
use,
”
as a background for
guidance during digitizing. This GIS file was developed through the input of bearing a nd
distance measurements (taken by land surveyors) to define parcel boundaries (i.e., the coordinate geometry method;
Amos, R. 2005. P
ers
onal
com
munication
.
City of Bakersfield, GIS
Department, 1501 Truxtun Avenue, Bakersfield, CA 93301).
Using
methods in
Romsos (1998), I
characterized the urban landscape and described urban kit fox habitat use by collapsing similar urban landscapes into eight categories (Table 1).
To verify the accuracy of the habitat types and boundaries depicted in the GIS, maps generat
ed from the GIS
were compared
to the actual location of habitat patches throughout the study area. When habitat was not correctly represented in the GIS, a global positioning system (GPS Pathfinder Pro XR, Trimble, Sunnyvale, CA) was used to obtain the ge ographic coordinates
of the edges of each habitat patch. These coordinate data were differentially corrected using the GPS Pathfinder Office program to improve accuracy to within +
1 m, imported into the GIS, and
used to digitize polygons representing the
correct habitat patches.
The accuracy of the habitat types and boundaries depicted in the GIS was further verified by comparison to aerial photographs (Farm Services Agency, USDA, Bakersfield, CA) of the study area that were photographed in June 1997, at
the beginning of the study, and in June 1998,
after the study concluded. This comparison was done primarily to identify habitats that were classified as open in 1997 but were in transition or residential in 1998. In these instances, the date each house w
as built
was determined
by consulting Kern County public records data. Any
habitat patch that was open in the June 1997 aerial photograph, on which a house was built prior to the end of the study, was classified as transition habitat.
12 Home Range
Home rang
e estimates were obtained using
the GIS
to generate the 100% minimum
convex polygon, which defined the bounds of the area used (Mohr 1947 , Hooge and Eichenlaub
1997
). Use of kernel
-
based estimators may be important to identify disjunct areas of activity i n
urban landscapes where habitat may be fragmented (Grinder and Krausman 2001). For comparison to other urban canid studies (Romsos 1998, Grinder and Krausman 2001), home range
was also estimated
using the 95% fixed kernel, a probabilistic model that iden tified the
most
probable area of utilization (W
orton 1989, Seaman and Powell 1996). To represent the core area for each study animal, I used the 50% fixed kernel, which identified the areas that encompassed 50% of the animal’s locations (Calhoun and Casby 1958, Jennrich and Turner 1969,
Anderson 1982).
L
arge bodies of water (e.g., Lake Truxtun)
were removed
from the estimates of
home range area and available habitat (White and Garrott 1990).
Probabilistic home range techniques (e.g., fixed kernel) are sen sitive to serial
autocorrelation (Anderson 1982, Hooge et al. 1999). The minimum convex polygon home range estimator assumes that location data are not time correlated (Anderson 1982, White and Garrott
1990), but are
robust to violation of this assumption
.
O
nly
statistically
independent data (i.e., the
data point provided no indication of the category for the next data point) were used
to estimate
home range and core area size.
T
ime to statistical independence between successive locations was based
on th
e time it took each kit fox to cross its home range (White and Garrott 1990, Lucherini et al. 1995,
Harrison 1997). To derive this estimate, I divided the mean speed for each of 10 kit foxes that were located approximately every 15 -
30 min during two to
six 3
-
5 h
movement sessions (hereafter referred to as sequential locations), by the longest axis of their 100% minimum convex polygon home range (Jenness 2003).
A
rea
-
observation curves
were used
to determine the minimum number of kit fox
locations necess
ary for home range estimation (Odum and Kuenzler 1955). To do so, I iteratively 13 estimated the 100% minimum convex polygon home range for each kit fox, beginning with the first 10 locations and adding one consecutive location at a time.
The
area of each 1
00%
minimum convex polygon
was plotted
against the number of locations from which it was
estimated. The point on the curve at which each 10 additional locations did not increase the home range estimate by 10% was the minimum number of locations needed to estimate home range.
Because home range size may vary with sex (Burt 1943), a 2 -
sample t
-
test
was used
to
determine if home range size differed between males and females (Hintze 1999). For comparison to rural kit fox studies in Arizona (Golightly 1981) an d California
(Zoellick et al. 1987
, White
and Ralls 1993, Koopman 1995), the 100% minimum convex polygon home range estimate was
used
for this analysis. All data were checked for normality and equal variance, and data that could not be normalized by trans
formation were analyzed with non
-
parametric tests. Significance
level (α)
was 0.05.
Home range size may also vary by season (Burt 1943). To assess if the area of the urban environment needed to meet kit foxes’ ecological requirements fluctuated through out the year, I
determined if home range size varied with biological season for urban kit foxes, as it did for rural kit foxes (Zoellick et al. 1987
). Because kernel
-
based estimates from small sample sizes (e.g.,
from subsetting the data by season) tend t
o overestimate home range sizes (Seaman and Powell
1996), seasonal home ranges
were estimated
using 100% minimum convex polygons, which is
more robust than other techniques when sample size is small (Harris et al. 1990). Data were normalized with a square
-
root transformation and analyzed with a 2
-
way analysis of variance
(Winer et al. 1991, Hintze 1999) to compare mean home range size for male and female kit foxes during three biological seasons: dispersal (1 June to 30 September); pair formation (1 Octobe r to
30 November); and breeding (1 December to 15
February;
Zoellick et al. 1987
, Reese et al. 1992,
Koopman et al. 1998). A Fisher's least significant difference multiple comparison test (Hintze 1999) was used to determine what factors (i.e., season and sex) in the analysis of variance were
14 significantly different. Study animals were located too seldom to estimate home range during the pup rearing season (16 February to 31 May).
The presence of food
-
rich habitat patches within territories may allow addit ional group
members. Macdonald (1983) predicted that the home range of a mated pair of red foxes may overlap and support other foxes if there were sufficient resources. However, if food is evenly -
dispersed among habitat patches, a larger home range would be required and the tolerance of
additional adults in the territory of a pair would be predicted to be less. I predicted that if kit foxes were concentrate
d around limited resources,
home ranges and core areas of adjacent kit
foxes would have a high perc
entage of overlap.
The GIS was used to identify
the percentage of
overlap between a) adjacent 100% minimum convex polygon home ranges and b) adjacent core area polygons between kit fox pairs comprised of a) mated pairs or family groups, b) adjacent female
s, c) adjacent males, and d) adjacent but unpaired males and females.
I
t was
assumed that habitats that were used disproportionately more than available contained key resources. To assess if habitats with key resources (i.e.,
denning and foraging
sites
)
w
ere equally
-
dispersed or conversely, concentrated, the mean percentage of area of each 100% minimum convex polygon that the core area represented was determined
. I
f areas of
habitat use were concentrated, the
n the
50% fixed kernel
sh
ould encompass less th
an 50% of the
area used by each kit fox. If areas of use were uniform rather than concentrated, the 50% fixed kernel would comprise half of the area within the 100% minimum convex polygon. To assess if use o
f key resources
differed among the sexes,
the m
edian number of core areas for females and
males
was compared
with a Mann
-
Whitney U
-
test, and
the
mean core area size for males and
females
was compared
with a 2
-
sample t
-
test.
15 Habitat Use
For those kit fox locations determined from triangulation, the like lihood for habitat
misclassification increased with animal proximity to habitat edge and decreased as habitat -
patch
size increased (Samuel and Kenow 1992).
To assess the
probability
of habitat misclassification,
the GIS
was used
to determine the distance
between each triangulated location and the nearest
habitat
-
patch edge
(Jenness 2004a)
.
I reasoned that triangulated kit fox locations
less than
100 m
from the nearest habitat
-
patch edge may have
had
misclassification
of habitat type
. The 100 m
value was
selected because
u
rban coyote studies that
used
similar
methods
to
study
habitat
use,
had
average telemetry error
s
of 46 m (Grinder and Krausman 2001) and 102 m (Romsos 1998) .
Use of habitat is considered selective if it is exploited disproportional to its availability
(Johnson 1980). Habitat selection can occur at a variety of scales (Wiens 1973, Orians and Wittenberger 1991, Litvaitis et al. 1996, Apps et al. 2001). I analyzed habitat use
within the
individual kit fox home range and core area and for th
e kit fox population in the study area. To
determine if kit foxes used urban habitat types (Table 1) with a frequency proportional to their availability, the independent locations that fell within a) the study area and b) each home range
were selected
. T
o delineate the study area, a 100% minimum convex polygon was generated
using all daily, sequential, and trapping locations for all study animals (Litvaitis et al. 1996, Romsos 1998, Kamler et al. 2003). For comparison to other urban canid habitat use stu dies
(Romsos 1998, Grinder and Krausman 2001), the 95% fixed kernel was used
to represent the
home range of each kit fox.
T
he area of each habitat type in the study area and each home range
was summed
, and use of habitat types in each
was compared
with a
Friedman's test (Alldredge
and Ratti 1986, Hintze 1999; individual kit fox was the blocking factor, habitat type was the treatment factor, and the ranked numerical difference between the proportion of habitat used and the proportion of habitat available wa
s the response variable). To determine which habitat types 16 were used significantly more than their availability, a Fisher's least significant difference procedure (Alldredge and Ratti 1986, Hintze 1999)
was used
.
H
ow kit foxes used urban habitats
was dete
rmined
at two scales: within each home range
in proportion to what was available in the study area; and within each core area in proportion to what was available in the corresponding home range. For each habitat type, I compared the proportion a) within e
ach home range and the study area and b) within each core area and corresponding home range. Prior to analysis, the proportion data were arcsine transformed (Winer et al. 1991) and analyzed with a matched
-
pair t
-
test, or a Wilcoxon signed
-
rank test if dat
a
could not be normalized (Hintze 1999).
Fox habitat use may also be influenced by human presence (Lucherini et al. 1995) ;
therefore, I treated human activity as an attribute of habitat . After each kit fox was located, an
index of human activity
was gaine
d
by counting the
number of vehicles and pedestrians that
passed by an imaginary line that was perpendicular to the nearest road in a 15 second period. The number of people that were visible (i.e., not in buildings or
large habitat patches
) in the same
h
a
bitat patch as the kit fox was
also recorded. To determine if human activity (i.e., the number of passing vehicles, passing pedestrians, and people in the same habitat patch as the kit fox) va ried
among the habitat types,
a Kruskal
-
Wallis 1
-
way analysis o
f variance
was used
, followed by a
Kruskal
-
Wallis multiple comparison test to determine what factors in the analysis of variance were significantly different.
Relative to some rural landscapes that have uniformly
-
distributed resources (Koopman
1995, Kamler
et al. 2004), urban landscapes are highly fragmented, have isolated, clumped resources (Grinder and Krausman 2001), and
consist of a patchy mosaic of habitats (Morrell
1975). Some urban habitats may be more suitable for denning and consequently, are used more
than their availability in the study area. If certain habitats were selected for denning, the spatial 17 arrangement of home ranges would be influenced by the availability of the selected habitat types (Golightly 1981, Hardenbrook 1987).
The geograph
ic coordinates of each den (i.e., refugia receiving repeated use by the study animals) were determined using a GPS
and imported into the GIS
. Dens were categorized as
subterranean, culvert (or pipe), or man
-
made structure.
To understand if den si
tes infl
uenced
habitat use,
a 1
-
proportion z
-
test (Winer et al. 1991, Hintze 1999)
was used
to determine if
habitats were used for denning in proportion to their availability in the study area.
Using the GIS, I generated random points (with a uniform distribution) equivalent to the
number of dens in the study area
(Hooge and Eichenlaub 1997)
. To assess if road proximity
influenced den locations,
the GI
S
was used
to determine the distance of each den and random
point to the centerli
ne of the nearest road
(Jenness 2
004a)
. A
Mann
-
Whitney U
-
test
was used
to
determine if
den sites and random points differed with regard to distance to the nearest road.
A
Wilcoxon signed
-
rank test
was used
to determine if the median size of
habitat patches within
which dens were located
differed from the median habitat
-
patch size in the study area.
Movements
Canid movement behavior may indicate use of habitat for resting, foraging, and traveling (Ozoga and Harger 1966,
Laundré
and Keller 1981, Romsos 1998).
Between 8 September 1997
and
7 January 1998, a randomly
-
selected subsample (five females and five males) of
radiocollared kit foxes was examined for movements and travel paths through adjacent habitat patches in order to identify which
behaviors
were associated with
each
habitat
type
.
Based on
behavior
-
specific movement types for canids in
Ozoga and Harger (1966) and
Laundré and Keller
(1981), I categorized kit fox movement as resting/walking (i.e., inactivity at low speeds between 0 -
0.8 km/h), trotting (i.e., potentially searching
for food at intermediate speeds between 0.8
-
1.8 km/h), and running (i.e., traveling from one point to another at high speeds greater than 1.8 18 km/h).
Movements were assessed during three nocturnal sampling per iods, approximately 1830
-
2230
h, 2230
-
033
0 h, and 0330
-
0730 h. Each kit fox was followed two to six times during each of the three sampling periods.
While following kit foxes, sequential locations were determined ap proximately every 15
-
30 min during
the sampling period. Movement sessions wi
th more than one hour between any
two successive locations were deemed incomplete and repeated on another day. The locations identified during each movement session were plotted on a 1996 City of Bakersfield street map (scale 1:29,000, Compass Maps, Inc.,
Modesto, CA). Time, habitat type, kit fox activity (e.g., resting/walking, trotting, running
), and number of passing vehicles, passing pedestrians, and people in the same habitat patch as the kit fox were recorded.
GIS was used to
digitize
kit fox loca
tions and generate
a line from the set of sequential
location points for each movement session.
Each line was
then intersected with the habitat types
through which it passed (Jenness 2004b).
K
it fox speed
(the distance between two sequential
locations di
vided by the time elapsed)
was regressed
against the number of passing vehicles,
passing pedestrians, and people in the same habitat patch to examine the influence of humans on
kit fox habitat use
.
A
multivariate analysis of variance
was used
to determine
if sex and
movement session time influenced minimum
travel
speed,
maximum
speed,
mean speed
(
using
all speed estimates except
when the kit fox was at a den or resting)
,
and total distance moved
per
movement session
(Woodruff and Keller 1982). Because mov
ement session length varied, total
distance moved
was standardized
by multiplying the distance measure by the ratio of 1) the length of time elapsed during the shortest movement session to 2) the length of time elapsed during the
movement session with whic
h the distance measure was associated.
To determine if total distance moved per movement session varied by habitat type, I used
a Kruskal
-
Wallis 1
-
way analysis of variance (Zoellick et al. 1989)
. A
Kruskal
-
Wallis 1
-
way
analysis of variance
was used
to de
termine if overall median speed was influenced by habitat
19 type.
Kruskal
-
Wallis multiple comparison tests were used to determine what factors in these analysis of variance tests were significantly different.
A
loglinear model
was used
to examine
correlat
ions between type of movement,
frequency of relocations per
habitat type, and time of movement session. To identify the habitat type and movement session (i.e., evening, middle of the night, and early morning) within which kit foxes
rested/
walked, trotted
, and ran,
standardized residuals
were used
to identify the greatest
deviations from expected for kit fox movements.
Demographics
R
epr
oduction and mortality in
radiocollared kit foxes
were measured
to determine if
demographic parameters of
urban kit fox po
pulation
s were similar to those
for rural kit fox
populations, and assess habitat quality in the urban landscape (Van Horne 1983). Kit foxes with a radiocollar in a mortality mode were immediately located and necropsied to assess cause of death.
This inf
ormation was used to determine sources and rates of kit fox mortality, which were compared to data for rural kit fox populations. I determined the number of radio -
days since each
kit fox was radiocollared
and
the survival rate for that time interval
,
desc
ribed by
Heisey and
Fuller
(
1985)
as the
interval survival rate
. Paired tooth puncture wounds associated with
subcutaneous hematomas on the carcass were indicative of a predator kill, and if the carcass exhibited massive trauma and was located near a road
, the cause of death was determined to be
vehicular
impact (Spiegel and Disney 1996
).
T
he number of pups incidentally trapped or observed with radiocollared adult kit foxes and their mates
were identified
.
Additionally, as part of a concurrent study
(St
orlie
1998)
of
juvenile kit fox dispersal and mortality, trapping effort was conducted in areas where pups were captured or observed.
C
apture data from both studies
were
used
to estimate litter size for kit
20 foxes in the study area.
The a
nnual reproductiv
e success rate
was based on
the percent
age
of
radiocollared
females that produced pups
.
21 RESULTS
Trapping
Between 1 May 1997 and 15 January 1998,
107
(30 adult males, 31 adult females, and 46
juveniles)
kit foxes
were captured in all habitat types in which
traps were placed (Appendix A).
Mean adult kit fox weight was 2.47
+
0.05 (
x
¯
+
1 SE) kg for males and 2.16
+
0.03 kg for
females.
Locations
R
adiocollars
were attached
to 28 adults (13 males, 15 females). Each radiocollared kit fox was locate
d 5.1
+
0.3 days each week. Of the
35 k
it foxes located
between 0900 and 1600 h
,
34 were
stationary
(e.g., resting)
or in a den. In 252 study days, a total of 1,882 sequential locations (during 111 movement sessions) and 4,824 daily locations were determ ined for the 28
study animals. Of these locations, 4,249 were independent, of which 2,898 (68%) were considered to be exact and the habitat types in which they occurred were correctly classified, based on being tracked to a den (n = 811) or directly obser ved (n = 2,087). Directly observed kit
foxes were
classified as stationary for 52% of the observations (Figure 2).
Of the 1,351
triangulated locations, 66 were
>
100 m from the nearest habitat
-
patch edge. The remaining
locations were on average 31.3
+
0 .6 m from the nearest habitat
-
patch edge.
Home Range
Based on the area
-
observation curves (Odum and Kuenzler 1955; Appendix D), the average number of locations needed to estimate a 100% minimum convex polygon home range was 27
+
1.7. All sequentially
-
loca
ted kit foxes could cross the longest axis of their 100%
22 0
10 20
30 40
50 60
Resting/Walking
Trotting
Running
Other
Behavior Type
Observations (%)
Figure
2 .
Behavior exhibited by directly observed kit foxes at independent daily locations (n = 2,087) in Bakersfield, California from May 1997 to January 1998.
23 minimum c
onvex polygon home range in less than four hours (range = 1.23 -
3.89 h; Appendix
B).
Therefore, f
our hours
was used as the
time to independence between kit fox locations.
The
100% minimum convex polygon
home range size for all study animals combined
wa
s 1.72
+
0.19 km
2 and was larger for males than females (Table 2). Average 95% fixed kernel
home range size for all study animals combined (excluding one outlier) was 1.2 +
0.19 km
2 (range = 0.04
–
2.78). Excluding one outlier, core areas comprised 9.1 +
1.2% (range = 1.9
–
19.6) of the
100% minimum convex polygon
home range. Kit foxes had an average of 2.15
+
0.23 distinct core areas, and f
emales had more than males. Mean
100% minimum convex
polygon home range size was significantly larger
during the d
ispersal season
than during the pair
formation
and breeding seasons (
F = 7.79, df = 2, p < 0.001)
and
for males than females
during
the breeding season (F = 4.47, df = 1, p = 0.04), and t here was no interaction between season and
sex (F = 2.38, df = 2, p =
0.1)
(Table 3)
. The percentage of home range and core area overlap
between kit foxes was highest for family group members (Figure 3).
Habitat Use
The study area was 76 km
2 , and had 3,908 habitat patches that averaged 0.01
+
0.001 km
2 (range was < 0.01
-
2.03, except for a 17.88 km
2 patch that comprised all inte
rconnecting linear
features that
was excluded from this calculation). Kit foxes used urban habitats disproportionate to their availability in the study area (Q = 103.5, df = 7, p < 0.001) and withi n home ranges (Q =
91.36, df = 7, p < 0.001; Figure 4). Open and sump habitats were used more than their availability in both the study area and in home ranges (Table 4). There was less area of residential habitat but more area of open and sump habitats in home ranges than in the study area.
Area of transition, commercial, manicured open, linear, and residential habitats were less, but area of sump habitat was greater, in core areas than in home ranges (Table 5).
Table
2 . Home
range (100% minimum convex polygon) and core area (50% fixed kernel) size for radiocollared kit foxes in Bakersfield, California from May 1997 to January 1998.
All
Males
Females
Males vs. Females
Measurement
x
¯
SE
Range
x
¯
SE
Rang
e
x
¯
SE
Range
Statistic
df
p
Home range size (km
2 )
1.72 0.19
0.31 -
4.49 2.12
0.31 0.31
-
4.49 1.36
0.21 0.36
-
2.71 t =
-
2.08 25
0.02 Core area size (km
2 )
0.16 a
0.03 0.01
-
0.5 0.18
0.05 0.01
-
0.5 0.15
0.04 0.02
-
0.43 t =
-
0.57 24
0.57 Co
re areas per kit fox
2.15 0.23
1 –
6 1.85
0.37 1
-
6 2.43
0.27 1
-
4 z =
-
1.98 25
0.02 a
Excluding one outlier (the male with two distinct home ranges).
24 25
Table
3 . Seasonal home range (100% minimum convex polygo
n) size for radiocollared kit foxes
in Bakersfield, California from May 1997 to January 1998.
Home range (km
2 )
All
Males
Females
Season
x
¯
SE
x
¯
SE
x
¯
SE
Dispersal
1.24 0.11
1.32 0.16
1.17 0.15
Pair formation
0.77 0.11
0.79 0.16
0.74 0.16
Breeding
0.66 0.12
0.95 0.16
0.38 0.16
26 0
20 40
60 80
Family Group
Male-Female
Male-Male
Female-Female
Pair Type
Overlap (%)
Home Range
Core Area
Figure
3 . Percent
age of
h
ome range overlap and core area overlap for pairs of
radiocollared
kit
foxes in Bakersfield, California from
May 1997 to January 1998
. Vertical bars represent
+
1 SE.
0 5
10 15
20 25
30 35 40
Residential
Linear
Commercial
Open
Manicured
Open
Industrial
Sump
Transition
Habitat Type
Area (%)
Study Area
Home Range
Core Area
a
c
b
b
a
a
b
a
b
c
a
b
a
b
a
b
Figure
4 . Mean percent
age
of each habitat type in the study area, which bounded the available habitat, and radiocollared kit fox home ranges (n =
27) and core areas (n
= 58) in Bakersfield, California from May 1997 to January 1998 . Different letters above the bars indicate a significant difference in habitat use versus availability.
27 Table
4 . Ranked radiocollared kit fox habitat use based on h abitat availability at the study area and home range area scale, in Bakersfield, California between May 1997 to January 1998. Habitat types used similarly (p > 0.05) are grouped by column.
Study Area
Home Range
Habitat Types
Statistically Similar in U
se
Mean
Rank
Habitat Types
Statistically Similar in Use
Mean
Rank
Rank
of
Use
Open
2.1 Sump
2.3 Most
Sump
2.6 Open
2.5 Transition
4.0 Transition
3.5 Industrial
4.0 Industrial
3.9 Manicured Open
4.4 Manicured Op
en
5.3 Commercial
5.4 Commercial
Commercial
5.8 Linear
5.6 Linear
Linear
6.2 Residential
7.8 Residential
6.6 Least
28 29
Table
5 . Mean difference between the mean proportion of habitat availabl e in the study area and
kit fox home ranges (e.g., there was 16.7% less open habitat in the study area than within home ranges), and kit fox home ranges and core areas in Bakersfield, California between May 1997 to January 1998.
Study Area and Home Range
Home Range and Core Area
Habitat Type
Percent
Difference
Statistic
p
a
Percent
Difference
Statistic
p
b
Commercial
-
1.3 t = 0.89
0.38 2.8
t =
-
2.79 <0.01
Industrial
-
1.2 z = 1.43
0.15 -
2.2 z = 0.05
0.47 Linear
2.5 t = 1.47
0.15 4.3
z =
-
2.25 0.01
Manic
ured Open
-
6.2 t =
-
0.59 0.56
3.7 z =
-
1.77 0.02
Open
-
16.7 t =
-
3.83 <0.01
-
7.5 t = 1.46
0.16 Residential
25.8 z = 4.52
<0.01
6.4 t =
-
5.5 <0.01
Sump
-
1.9 z =
-
2.88 <0.01
-
7.6 z = 2.19
0.01 Transition
-
1.0 t =
-
0.66 0.51
0.2 t =
-
2.30 0.01
a,b
Becaus
e the analysis used multiple t
-
tests, the probability of at least one false
rejection of the null hypothesis was 1
–
0.95 8
= 0.34
(Winer et al. 1991).
30 N
umber of people in the same habitat patch as the kit fox varied be tween habitat types
and was great
est
in manicured open habitat (H = 285.7, df = 7, p < 0.001; Figure 5). Industrial and linear habitats had a higher incidence of passing vehicles than all other habitats (H = 355.1, df = 7, p < 0.001; Figure 5). A greater number of pedestrians were observed along roads near
sumps than all other habitats (H = 136.8, df = 7, p < 0.001; Figure 5).
The study animals used 190 dens, comprised of 159 subterranean dens, 17 culvert or pipe dens, and 14 man
-
made structure dens (e.g., metal storage containers, modular office buildings,
junkyard debris, a shed, and a garbage dumpster), most of which were located in open, sump, and linear habitats (Figure 6). Sump (z = 45.74, df = 1, p < 0.001) and open (z = 11.47, df = 1, p <
0.001) habitats were used for denning more t
han their availability in the study area (Figure 7).
Residential (z =
-
10.5, df = 1, p < 0.001) and commercial (z =
-
3.67, df = 1, p < 0.001) habitats
were used for denning less than their availability in the study area. Linear (z = -
1.24, df = 1, p =
0. 19), manicured open (z =
-
0.53, df = 1, p = 0.59), industrial (z =
-
0.3, df = 1, p = 0.76), and
transition (z =
-
0.45, df = 1, p = 1.0) habitats were used for denning in proportion to their availability in the study area.
Dens were located farther from roa
ds (0.10 + 0.01 km, range = 0.01
-
0.5) than randomly
-
generated points (0.06 + 0.01 km, range = 0.0001
-
0.55; z = 8.17, df = 378, p < 0.001). The
median size of habitat patches within which dens were located (0.03 km2) was significantly larger than the m
edian habitat
-
patch size in the study area (0.004 km2; z = 10.56, df = 1, p <
0.001).
Movements
Movement sessions (n = 100) lasted from 184 to 315 min long. Kit foxes moved at an average rate of 1.33
+
0.02 km/h. Minimum speed during each movement sessi
on was influenced
29 34
Vehicles Passing on Nearest Road
0 0.5
1 1.5 2 2.5
1 2
3 4
5 6
7 8
Vehicles/15 s
Pedestrians Passing Along Nearest Road
0 0.1
0.2 0.3 0.4
1 2
3 4
5 6
7 8
Pedestrians/15 s
People in Habitat Patch
0 0.2 0.4 0.6 0.8
Linear
Industrial
Residential
Sump
Commercial
Open
Manicured
Open
Transition
Habitat Type
People/15 s
Figure
5 . Mean number of passing vehicles, passing pedestrians, and people in the same habitat patch as radiocollared kit foxes at all daily locations
(n = 4,824)
in Bakersfield, California from May 199
7 to January 1998.
Vertical bars represent
+
1 SE.
31 0
10 20
30 40
50 60 70
Open
Sump
Linear
Manicured
Open
Industrial
Commercial
Residential
Transition
Habitat Type
Number of Dens
Man-made Structures
Culvert/Pipe
Subterranean
Figure
6 . Number of dens used by radiocollared kit foxes in each habitat type in Bakersfield, California from May 1997 to January 1998 .
32 0
5 10
15 20
25 30
35 40
Residential
Linear
Commercial
Open
Manicured
Open
Industrial
Sump
Transition
Habitat Type
Habitat (%)
Study Area Available
Used for Denning
Figure
7 .
Percent
age of available
habitat in the study area compared to percent
age of habitat
use
d
by radiocollared kit foxes for denning in
Bakersfield, California from May 1997 to January 1998.
33 34
by the
interaction of sex and time of night (Figure 8, Table 6).
The regressions of kit fox
speed
and the
number of passing vehicles (r
2 = 0.0077), passing pedestrians (r
2 = 0.0016), and people in
the same habitat patch as the kit fox (r
2 = 0.0002), respectively, indicated that these variables (all of which could not b
e normalized through transformation) were not related to kit fox speed. The distance traveled during each movement session was longer in manicured open and open habitats than all other habitats (H = 125, df = 7, p < 0.001; Fi gure 9). Overall speed was fa
st
er in
residential habitat than all other habitats (H = 111.6, df = 7, p < 0.001; Figure 10).
There was a significant interaction between kit fox movements, movement session time, and habitat type (
c
2 = 61.48, df = 28, p < 0.001). Kit fox
resting/
walking
movements deviated
most from expected from
1830
–
2230 h,
during which linear habitat was used more than
expected and from
2230
–
0330 h,
during which residential habitat was used less than expected (Table 7). Kit fox trotting movements deviated most fro m expected from
2230
–
0330 h
during
which open habitat was used more than expected
,
and from
2230
–
0730 h
during which sump
habitat was used less than expected. The greatest deviations from expected kit fox running movements were from
2230
–
0330 h
duri
ng which open habitat was used more than expected,
and from
0330
–
0730 h
during which industrial habitat was used less than expected.
Between 27 September and 4 October 1997, a male kit fox moved >
5.9 km from his previously
-
established home range to anot
her part of the study area, where he remained until the end of the study. Kit foxes were observed crossing roads (up to four lanes), but also used culverts to move under roads. One study animal was observed to use a 0.46 -
m in diameter culvert to cross un
der a
State Route 99 offramp and a 4.26 m wide by 3.66 m high vehicle access road to cross under
State Route 99. Other kit foxes were observed to cross under the State Route 99 bridge over the Kern River and use a foot bridge to cross a canal.
35 35
0 0.05
0.1 0.15
0.2 0.25
0.3 0.35 0.4
1830 - 2230
2230 - 0330
0330 - 0730
Time Period
Minimum Speed of Travel (km/hr)
Males
Females
Figure
8 . Minimum speed per movement session for male and female radiocollared kit foxes in Bakersfield, California from September 1997 to January 1998.
Table
6 . Influence of sex, movement session time, and t
heir interaction on minimum, maximum, and mean speed and total adjusted distance moved for 10 radiocollared kit foxes in
Bakersfield, California from May 1997 to January 1998
.
Males
Females
Males vs. Females
Movement Session
Time
Interaction
Measurement
x
¯
SE
Range
x
¯
SE
Range
F
-
Ratio
df
p
F
-
Ratio
df
p
F
-
Ratio
df
p
Minimum
Speed (km/hr
2 )
0.26 0.03
0.02 -
0.75 0.27
0.03 <0.01
-
1.16 0.05
1 0.82
0.58 2
0.56 3.17
2 0.047
Maximum
Speed (km/hr
2 )
2.19 0.13
0.26 -
4.65 2.17
0.22 0.32
-
9.06 0.1
1 0.75
2.5 2
0.09 0.21
2 0.81
Mean Speed
(km/hr
2 )
1.39 0.03
0.02 -
4.65 1.25
0.03 <0.01
-
9.06 1.32
1 0.25
1.08 2
0.34 0.15
2 0.86
Total Adjusted
Distance
Moved (km)
1.76 0.11
0.12 -
3.34 1.62
0.15 0.13
-
4.77 0.88
2 0.35
1.81 2
0.17 1.16
2 0.32
36 0
0.2 0.4
0.6 0.8
1 1.2
Manicured
Open
Open
Residential
Linear
Industrial
Commercial
Transition
Sump
Habitat Type
Mean Distance Traveled (km)
Figure
9 . Mean distance radiocollared kit foxes traveled in each habitat type in Bakersfield, California from September 1997 to January 1998.
Vertical bars represent
+
1 SE.
37 0
0.2 0.4
0.6 0.8
1 1.2
1.4 1.6 1.8 2
Residential
Manicured
Open
Commercial
Sump
Industrial
Open
Linear
Transition
Habitat Type
Mean Speed (km/hr)
Figure
10 . Mean speed
radiocollared kit foxes traveled in each habitat type in Bakersfield, California from September 1997 to January 1998.
Vertical bars represent
+
1 SE.
38 Table
7 . Movement session time, type of movement, and habitat associations for radiocollared kit foxes in Bakersfield, California from September 1997 to January 1998.
Movement Type
Movement Session Time (h)
Habitat Type
Observed Frequency
a
Standardized Residual
b
Resting/W
alking
1830
-
2230
Sump
11 -
5.19 Transition
18 -
4.16 Residential
19 -
4.02 Commercial
30 -
2.40 Industrial
58 1.71
Manicured Open
107
8.91 Open
185
20.37 Linear
205
23.30 2230
-
0330
Residential
6 -
5.93 Transition
12 -
5.05 Sump
13 -
4.90 Commercial
28 -
2.69 Manicured Open
76 4.36
Industrial
78 4.65
Linear
89 6.27
Open
194
21.69 0330
-
0730
Commercial
15 -
4.60 Residential
15 -
4.60 Transition
26 -
2.99 Sump
35 -
1.67 Industrial
38 -
1.23 Manicured Open
57 1.56
Open
164
17.28 Linear
174
18.75 39
Table 7.
Movement session time, type of movement, and habitat associations for radiocollared
kit foxes in Bakersfield, California from September
1997 to January 1998 (continued).
Movement Type
Movement Session Time (h)
Habitat Type
Observed Frequency
a
Standardized
Residual
b
Trotting
1830
-
2230
Sump
12 -
5.05 Transition
12 -
5.05 Residential
25 -
3.14 Industrial
29 -
2.55 Commercial
33 -
1.96 Manicured Open
45 -
0.20 Open
96 7.29
Linear
105
8.62 2230
-
0330
Sump
6 -
5.93 Transition
13 -
4.90 Re
sidential
14 -
4.75 Industrial
25 -
3.14 Manicured Open
30 -
2.40 Commercial
41 -
0.79 Linear
94 7.00
Open
115
10.08 0330
-
0730
Sump
6 -
5.93 Transition
20 -
3.87 Industrial
25 -
3.14 Commercial
28 -
2.69 Manicured Open
29 -
2.55 Re
sidential
46 -
0.05 Linear
72 3.77
Open
97 7.44
40 Table 7. Movement session time, type of movement, and habitat associations for radiocollared
kit foxes in Bakersfield, California from September
1997 to January 1998 (continued).
Movement Type
Movem
ent Session Time (h)
Habitat Type
Observed Frequency
a
Standardized Residual
b
Running
1830
-
2230
Sump
6 -
5.93 Industrial
14 -
4.75 Transition
16 -
4.46 Commercial
19 -
4.02 Linear
50 0.54
Manicured Open
54 1.12
Residential
61 2.15
Open
6 4
2.59 2230
-
0330
Transition
6 -
5.93 Sump
7 -
5.78 Industrial
25 -
3.14 Residential
27 -
2.84 Commercial
35 -
1.67 Manicured Open
38 -
1.23 Linear
53 0.98
Open
66 2.89
0330
-
0730
Industrial
2 -
6.51 Sump
4 -
6.22 Transition
6 -
5. 93
Commercial
8 -
5.63 Linear
21 -
3.72 Manicured Open
24 -
3.28 Open
30 -
2.40 Residential
30 -
2.40 a
Number of times a kit fox movement path intersected each habitat type, separated by movement type and movement session time.
b
Negative value
s indicate that a habitat was used less than expected and positive values indicate that a habitat was used more than expected.
If the value is >
+
2.00 (2 standard deviations), it is significant.
41 42
Demographics
D
uring
6,036 day
s of active radiotransmitter
s,
two of the 28 radiocollared kit foxes died
and
the interval survival rate was 0.92.
O
ne female
was hit by a vehicle and
another female was
likely killed by a coyote or domestic dog, based on 30 to 35 -
mm paired puncture wounds on the
torso of the intact
, unburied carcass (Spiegel and Disney 199
6 ).
It was unknown if another
female that could not be located after 3 September 1997, died, dispersed, or had a malfunctioning radiocollar.
Pups were detected
with 15 family groups, comprised of 22 radiocollared
adult kit foxes.
Eleven of the 15 radiocollared females
reproduced during
one breeding season
. Four of these 11
females had radiocollared mates. An additional four radiocollared males had uncollared mates that
also
reproduced
in the same breeding seaso
n
. Kit fox litters averaged 2.8
+
0.45 (range = 1
–
7) pups. The litter of 7 pups was observed denning with two radiocollared females, of which one exhibited reproductive condition and one did not.
36 43
DISCUSSION
Demographics
The urban kit foxes had lower m
ortality and similar reproduction and body weight
compared to rural populations (Morrell 1972). Conversely, Cypher and Frost (1999) found that kit foxes in Bakersfield (i.e., females in winter and juveniles) were heavier than their counterparts in a nearb
y rural population. Urban kit foxes’ normal weights may be attributed to good nutrition. The
constant source of water from sumps, canals, and lawn sprinklers in Bakersfield (Murphy 1992) may provide a more stable prey population.
Like urban red foxes in
Orange
County, California (Lewis et al. 1993) and Toronto, Canada (Rosatte et al. 1991), urban kit foxes were supplementally fed (i.e., anthropogenic sources of food, including bread and burritos) by people and observed opportunistically foraging on both
anthropogen
ic and natural food items
(personal observation
).
Urban kit foxes in Bakersfield may have lower mortality rates than rural foxes (Table 8).
Of 22 adult kit foxes from a nearby rural population that were radiocollared prior to or during late N
ovember and December 1995 through September 1996, 15 (68.2%) died and two were lost (Enterprise Advisory Services, Inc. 1997).
Of 35 adult kit fox mortalities (including the two from this study) in the Bakersfield area from 1997 to 2003, 23% were due to v ehicular impact (Bjurlin
and Cypher 2003).
T
he interval survival rate for urban kit foxes
(0.92)
was much great
er than
that
for rural kit foxes (
0.58 ; Ralls and White 1995;
0.44 ; Cypher et al. 2000).
Coyote predation may be more likely if a fox is far fr
om its den (Kitchen et al. 1999) or
in the peripheral areas of its home range (Sovada et al. 1998, Olson and Lindzey 2002). The
majority of swift fox
(
Vulpes velox
)
deaths were due to coyotes and occurred outside the 85%
Table
8 .
Rural kit fox and swift fox cause
-
specific mortality rates reported within their geographic range for comparison to that for urban kit foxes in Bakersfield, California from May 1997 to January 1998.
Cause
-
Specific Mortality Rate (%)
Reference
Species
L
ocation
Predation
Vehicular Impact
Cypher et al. 2000
Kit Fox
Kern County, CA
48.5 5.0
Ralls and White 1995
Kit Fox
Carrizo Plain, CA
78.0 4.0
Hardenbrook 1987
Kit Fox
Churchill County, NV
60.0 10.0
Spiegel and Disney 1996
Kit Fox
Kern County, CA
66.7 1.7
Standley et al. 1992
Kit Fox
Monterey and San Luis Obispo Counties
, CA
75.0 6.3
Kamler et al. 2003
Swift Fox
Dallam County, TX
33.0 42.0
Matlack et al. 2000
Swift Fox
Sherman and Wallace Counties, KS
45.0 36.0
44 45
(Kitchen et al. 1999) or 95% (Sova
da et al. 1998, Olson and Lindzey 2002) isopleth of the fox’s home range and more than
1 km from the den that the fox most recently occupied. Likewise, the urban kit fox that died as a result of coyote or dog predation was located outside the 85% isopleth
of her range.
Production of young exceeded the number of kit foxes that died, which may have indicated that the urban kit fox population was demographically healthy during the period of study and habitat was of sufficient quality to support a source popul ation (Van Horne 1983,
Pulliam 1988, Noss et al. 1997
). Seventy
-
three percent of
radiocollared fem
ales produced pups,
more than
that
observed
for rural kit foxes
in southwestern Kern County (64%; Spiegel and Tom
1996; 61%; Cypher et al. 2000). The mean n
umber of pups per litter was within the range of that
for rural populations: 2 (White and Ralls 1993) to 4 (Morrell 1971, Cypher et al. 2000).
Kit fox reproduction has been positively correlated with prey availability (White and Ralls 1993).
Raccoons an
d skunks in Toronto, Canada weighed more and had higher reproductive success than their rural counterparts, which Rosatte et al. (1991) attributed t o
abundant and varied natural and human
-
derived food sources in the urban setting. The results of a supplem
ental feeding study indicated that kit fox survival and reproductive success were influenced by food availability (EG&G Energy Measurements 1993).
Supplemental feeding by
humans
and the availability of
refuse
may contribute to the demographic health of ur
ban kit
foxes.
Locations
During daylight hours, urban kit foxes were found to be resting or in dens on all but one occasion. Like rural kit foxes in California (Morrell 1971, Morrell 197 2, White et al. 2000), Utah
(Egosc
ue 1956, O’Neal et al. 1987, Arjo e
t al. 2003), and Arizona (Golightly 1981, Golightly and Ohmart 1983) and urban red foxes in Orange County, California (Sallee 1998), urban kit foxes 46 were almost entirely nocturnal. Similarly, urban coyotes in Orange County, California (Romsos 1998), Seatt
le, Washington (Quinn 1995), and Tucson, Arizona (Grinder and Krausman 2001) were less active during daylight than at night. Urban kit foxes were primarily nocturnal and limited their daytime activity to the immediate vicinity of the den, similar to rural kit foxes
(Morrell 1971, Morrell 1972), rural swift foxes (Kilgore 1969, Kitchen et al. 1999, Kamler et al. 2004), and urban red foxes (Voight and Macdonald 1984). Urban canids may benefit from being active only at night when interactions with humans are
less likely.
Home range
Knowledge of home range size and spatial distribution can be used to manage urban habitat for the benefit of kit foxes
.
Home range estimates were much smaller than those for rural kit foxes (Table 9), but consistent with observati
ons that foxes (Voight and Macdonald 1984,
Murphy 1992) and raccoons and skunks in urban areas have smaller ranges than their rural counterparts (Rosatte et al. 1991).
Macdonald (1983) and Doncaster et al. (1990) attributed the tendency for red fox home ra
nges to be smaller in cities (e.g., Oxford) than
in
rural areas, to the spatial and temporal
patchiness of food dispersion. Red foxes in Orange County, California had smaller home ranges where they were supplementally fed (Sallee 1998). Ables (1969) note d that the smaller home
ranges for red foxes in south
-
central Wisconsin may have been due to habitat richness (i.e., a high interspersion of habitat types and abundant wild fruits and berries). A rtificial food sources
associated with urban and oilfield de
velopment may have led to decreases in the home range size of kit foxes in southwestern Kern County
(
Zoellick et al. 1987,
Spiegel and Bradbury 1992)
.
Conversely,
kit fox home ranges increased in size during times of low prey availability (White and Ralls
1993) and were negatively correlated with prey density (Morrell 1971, White and Garrott 1997). Thus,
the small size of urban kit fox home ranges indicated that abundant food
Table
9 . Kit fox home range size reported within vari
ous habitats within their geographic range for comparison to home range size of urban kit foxes in Bakersfield, California from May 1997 to January 1998.
Reference
Location
Habitat
Estimator
Range Type
Area km
2 Knapp 1978
Kern County, CA
Agriculture
100%
Minimum convex polygon
a
Home Range
2.52 Koopman 1995
Kern County, CA
Saltbush scrub
100% Minimum convex polygon
Home Range
3.52 Zoellick et al. 1987
Kern County, CA
Saltbush scrub
100% Minimum convex polygon
Home Range
4.63 White and Ralls 1993
Ca
rrizo Plain, CA
Grassland
100% Minimum convex polygon
Home Range
11.60 O’Neal et al. 1987
Great Basin, UT
Saltbush scrub
100% Minimum convex polygon
Home Range
3.11 Spiegel and Bradbury 1992
Kern County, CA
Saltbush scrub
97% Minimum convex polygon
b
H
ome Range
6.13 Spiegel and Bradbury 1992
Kern County, CA
Saltbush scrub
95% Minimum convex polygon
b
Nocturnal Range
5.82 Hardenbrook 1987
Churchill County, NV
Greasewood
Modified Minimum Area Method
Home Range
9.74 Arjo et al. 2003
Tooele County,
UT
Salt playa
95% Adaptive Kernel
Home Range
43.09 Arjo et al. 2003
Tooele County, UT
Greasewood
95% Adaptive Kernel
Home Range
13.29 Arjo et al. 2003
Tooele County, UT
Sagebrush
95% Adaptive Kernel
Home Range
11.51 Koopman 1995
Kern County, CA
Salt
bush scrub
75% Harmonic Mean
Core area
0.73 Spiegel and Bradbury 1992
Kern County, CA
Saltbush scrub
50% Hamonic Mean
b
Core area
1.18 Zoellick et al. (2002)
Kern County, CA
Saltbush scrub
50% Hamonic Mean
Core area
1.2 a
May have been underestimated
due to small sample size.
b
Estimates using non
-
independent data may have been inflated, as demonstrated in Koopman (1995).
47 48
resources were
concentrated in key urban habitats around which home ranges were configured.
Further evidence
of this was that ur
ban kit foxes’ core area size constituted only 9% of the home range size, whereas that for rural kit foxes constituted 17% of the home range (Spiegel and Bradbury 1992). This may be because suitable denning habitat for urban kit foxes was concentrated aro
und sumps and undeveloped lots.
Female urban kit foxes had a greater median number of core areas than did males, indicating that females had a higher number of distinct denning sites. Female kit foxes in Arizona (Golightly 1981) and Nevada (Hardenbrook 19
87) spent more time near their dens than
males. Fenced sumps in core areas provide denning habitat that may ensure safe reproduction in the urban population.
The home ranges and core areas of adjacent kit fox
es may have substantial overlap
if kit
foxes ar
e concentrated around limited resources (Golightly 1981), such as food -
rich habitat
patches (Macdonald 1983).
Mean home range overlap was greater than 71
%
for family group
members and less than 16%
for unrelated individuals, which was l
ess than that for
the nearby
rural population
(Zoellick et al. 2002, Spiegel and Bradbury 1992)
.
C
ore area overlap was greater for members of t
he same family group (61%
) than
core
area overlap of
neighbori
ng males and females (29%
) or
neighboring males (11%
), similar to t
hat
for rural kit foxes
(Spiegel and Bradbury 1992). Unpaired male and female urban kit foxes had a higher percentage of core area overlap than home range overlap. Because core areas were centered around denning sites, m
ale and female kit foxes’ home ran
ges may have overlapped in a
central denning area, as was the case for kit foxes in Arizona (Golightly 1981) and Nevada (Hardenbrook 1987). Kitchen et al.’s (2005) findings that male swift foxes spent more time on the boundary of the home range, while fem
ales spent more time in the group core area, may
explain why female urban kit foxes had the least home range overlap and no core area overlap.
47 49
Overlap of
home range
s
and core area
s
indicated that
kit foxes were concentrated around
key urban habitat patch
es
and food resources were not evenly
-
distributed
. It wa
s likely that urban
kit foxes were able to find foraging and denning habitat in close proximity because
of uniformly
-
distributed sumps, which may be important to the success of the urban kit fox popu lation.
D
ue to refuse and human
-
derived foods, urban kit foxes may be able to meet their resource needs in a smaller area than rural kit foxes. The presence of these food resources distributed among disjunct habitat patches may explain why u
rban kit fox
es had smaller and less
overlapping home ranges than their rural counterparts. Given the constant and widespread availability of human
-
derived food in the urban landscape, suitable denning habitat may be a limiting factor for the urban kit fox population.
This is consistent with Golightly’s (1981)
conclusion that kit fox den sites in rural Arizona were limited by soils suitable for denning.
Estimates of 100% minimum convex polygon home range size were greater for males than
females, perhaps due to fewe
r seasonal movements by females and increased foraging
requirements for males during pup
-
rearing (Spiegel and Bradbury 1992). While Golightly (1981) found that male kit foxes in Arizona had much larger home ranges than females, home range size for other r
ura
l kit fox (Zoellick et al. 1987
, Spiegel and Bradbury 1992, White and Ralls 1993,
Koopman 1995), swift fox (Kitchen et al. 1999, Kamler et al. 2003), and Blanford’s fox ( Vulpes
cana
; Geffen and Macdonald 1992) populations did not differ among the sexes.
Because home range size may vary with season (Burt 1943), the quantity of urban habitat needed to meet kit foxes’ ecological requirements may fluctuate over time. Home range size was greater for males than females during the breeding season, but did not differ between the seasons
for kit foxes in Utah (Arjo et al. 2003), swift foxes (Kitchen et al. 1999), Blanford’s foxes (Geffen and Macdonald 1992), and culpeo foxes (
Pseudalopex culpaeus
; Salvatori et al. 1999).
In order to search for a mate during the
breeding season, male foxes may increase their number of exploratory movements, which may be especially pronounced in the urban environment where 53 50
patches of habitat suitable to support kit foxes are small and disjunct. Home range size may have been under
estimated during the pair formation and breeding seasons, because fewer kit foxes were located during this time. This may explain why mean home range size was significantly larger during the dispersal season than the pair formation and breeding seasons.
Habitat use
Habitats were known to be correctly classified for 68% of all independent locations,
which were obtained by
following the transmitted signal until the animal was observed or tracked to a den
. I
t was
unlikely that visually observed kit foxes w
ere behaviorally pushed from one
habitat to another, because kit foxes were resting or walking 67% of the time .
However, distance
between the animal and the radio
-
receiver, as well as broken topography, can increase telemetry error (Harris et al. 1990).
Telemetry error was a concern because correct assignment of habitat use required ac
curate locations. Based on
average telemetry error in urban coyote studies
(Romsos 1998, Grinder and Krausman 2001),
triangulated kit fox locations less than
100 m from
the
nearest habitat
-
patch edge may have been misclassified.
The
remaining 32% of locations
were determined by triangulation or encircling the habitat patch . These locations
average
d
31 m
from the nearest habitat
-
patch edge
and
may have
included some misclas
sifi
cation
.
Like urban kit foxes, rural kit foxes do not use habitat within their home range in a uniform manner (Golightly 1981,
Hardenbrook 1987
) and have distinct core areas of concentrated
activity (Koopman 1995). Kit foxes used open and sump habitat s more than their availability in
both the study area and in home ranges, and there was significantly more open and sump habitat in home ranges than in the study area. Likewise, urban red foxes (Sallee 1998) and coyotes (Romsos 1998) in Orange County, Cal
ifornia used vacant field habitats more than expected. The higher vegetative cover in vacant fields may facilitate predator avoidance and support a more abundant prey base (Adams and Dove 1989) than the other less used urban habitats. Likewise, 49 51
sumps had
marsh vegetation, which provides food and cover for wildlife and facilitates use by terrestrial wildlife (Adams and Dove 1989). Therefore, sumps may serve as artificial urban wetlands that provide denning habitat with their sloping sides, foraging opport unities based on the
abundance of ac
tive small mammal burrows (personal observation
), and a constant source of
water, which I observed an urban kit fox drinking. In hot environments, such as Arizona, rural kit foxes meet their water needs through prey con
sumption (Golightly and Ohmart 1984). With the
availability of free
-
standing water in the urban environment, kit foxes may require less food.
The number of people in the same habitat patch as the kit fox was lowest in open and sump habitats. While the
greatest number of pedestrians observed along roads were near sumps, the fencing around sumps provided a physical and visual barrier between kit foxes and passing pedestrians. The shelter and security that the fenced perimeters provided may also explain w hy
Stafford and Murphy (1992) found kit fox dens in 10 of the 95 sumps that they surveyed in Bakersfield. Likewise, urban red foxes disproportionate use of industrial habitats may be attributed to security personnel and perimeter fencing, which limited in trusion by humans and
their pets (Sallee 1998).
There was significantly less residential habitat in urban kit fox hom e ranges than in the
study area;
the opposite
was true
for urban coyotes in Tucson, Arizona (Grinder and Krausman 2001). Harrison (1997)
found that gray foxes (
Urocyon cinereoargenteus
) in New Mexico
avoided high
-
density residential subdivisions. Urban coyotes (Romsos 1998) and red foxes (Sallee 1998) in Orange County, California avoided high -
density residential and commercial
habitats, pe
rhaps due to a greater possibility for human and domestic animal interactions.
Likewise, Lucherini et al. (1995) found that red foxes used habitats with less human access and disturbance.
Industrial and linear habitats had a higher incidence of passing ve hicles than all other
habitats. This may be because industrial habitats tend to occur along major roads. In linear 52 habitat, from 1800 to 1859 h and 0700 to 0759 h, both the number of kit fox locations and number of passing vehicles peaked (Appendix E).
However, during these times, 84% of the kit
foxes using the linear habitat were located in dens that sheltered them from the disturbance of the passing vehicles.
The kit fox is one of the few canids that utilizes an underground den throughout the year (B
erry et al. 1987, Koopman et al. 1998). Because dens are used for whelping and rearing pups, for escape cover from predators (Grinnell et al. 1937), and to minimize water loss and heat stress in summer and metabolic costs in winter (Golightly 1981, Goligh tly and Ohmart 1983), kit fox
presence is limited by the avail
ability of (Spiegel et al. 1996
) or their ability to build dens. Like
rural kit foxes in California (Knapp 1978, Berry et al. 1987, Reese et al. 1992) and Utah (Egosc
ue
1956, Arjo et al. 2003)
and swift foxes (Kilgore 1969), urban kit foxes primarily used subterranean dens. However, they also used alternative dens, including pipes. Use of pipes that are known to interfere with radio signals, may have caused an underestimation of the number of alternative dens used in this study (Berry e
t al. 1987, Spiegel et al. 1996
, Koopman et al. 1998).
Like Golightly (1981), I found that kit foxes did not randomly select den sites. Because kit foxes selected sump habitats
for
den
ning
, there was more area o
f sump habitat in core areas
than in home ranges. Sumps may also be used for denning because they are fenced and not accessible to the public. Likewise, most urban red fox dens occurred within fenced a reas that
may have isolated
dens from human disturban
ce (Sallee 1998). Stafford and Murphy (1992)
found that the majority of sumps used by urban kit foxes for denning were adjacent to undisturbed habitat, which likely has a low human presence.
Furthermore, I found that urban kit foxes denned in open habit ats more than their
availability in the study area. Likewise, rural kit foxes in California (Morrell 1971, Morrell 1972), Utah (E
g
o
sc
ue 1956,
Egoscue
1962), and Arizona (Zoellick et al. 1989) and urban red 53 foxes (Lewis et al. 1993) frequently denned in fl
at, open areas. Kit foxes may be less vulnerable
to coyote and bobcat predation on flat or rolling grasslands (Warrick and Cypher 1998).
Linear habitats (i.e., along canals, railroads, freeway embankments, and powerline right -
of
-
ways) were used in proport
ion to their availability for denning. Like swift foxes, urban kit foxes utilized dens in road right
-
of
-
ways (Matlack et al. 2000). However, urban kit foxes denned near roads less than expected and
in larger
habitat patches
than expected
. Similarly, red
foxes
used habitats with less human access and disturbance for resting (Lucherini et al. 1995).
Movements
Kit foxes moved at an average rate of 1.33 km/h, nearly identical to that for kit foxes in southwestern Kern County (1.35 km/h; Koopman 1995) but
s
lo
wer than that for kit foxes in
Arizona (males: 2.73 km/h; females: 1.4
9 km/h; Golightly 1981) and fast
er than that for swift
foxes (males: 0.68 km/h; females: 0.59 km/h; Kitchen et al. 1999). Between dusk and dawn (i.e., approximately 10 to 14 h, dependin
g on time of year), rural kit foxes in Arizona moved an average distance of 14.3 km for males and 11.8 km for females (Zoellick et al. 1989) and in southwestern Kern County, moved an average distance of 9.8 km for males and 10.9 km fo r
females (Zoellick et
al. 1987
), yielding average movement rates similar to that for urban kit foxes.
The average distance that captured urban red foxes, raccoons, and skunks in Toronto, Canada moved from the point of release was less than that for their rural counterparts, w hich indicated
that food and habitat in the urban environment were not limited in availability (Rosatte et al. 1991).
Males, which had larger 100% minimum convex polygon home ranges than females, may travel at higher rates of speed during the early part of the evening to cover their larger territory.
In Arizona, kit fox activity peaked in the early evening and a few hours before sunset, and males spent twice as much time out of the den, traveled 3.6 times farther, and maintained home ranges 54 2.6 times large
r than females (Golightly 1981). However, in the urban setting, I did not detect a
difference between sexes,
movement session times, or their interaction with regard to maximum speed, mean speed, or total adjusted distance moved during each movement sessi on. Similarly,
lengths of nightly mov
ements did not differ among
sexes for rural kit foxes (Zoellick et al. 2002)
and Blanford’s foxes (Geffen and Macdonald 1992). Compared to Arizona, where male kit foxes spent much time by the boundaries of their terri
tory to defend a non
-
uniformly distributed
resource base (Golightly 1981), urban kit foxes were able to meet their resource needs within small home ranges and did not need to travel far due to the concentration of resources within distinct habitat patches
in the urban landscape.
Kit fox movement speed was not correlated with the number of passing vehicles, passing pedestrians, or people in the same habitat patch. This is consistent with the findings in Lewis et al. (1993), which indicate
d
that urban red fo
xes were tolerant of humans. Nevertheless, kit foxes
may have attempted to avoid human interactions by traveling at a high rate through residential, manicured open, and commercial habitats, all which had a high incidence of people.
Sump and transition h
abitats were typically small patches that were free of people at night when kit foxes were active. Kit foxes were observed foraging in and caching food scavenged from construction sites. Furthermore, the remains of several species that live in sumps, inc
luding bullfrog (
Rana catesbeiana
), burrowing owl (
Athene cunicularia
), duck
(unidentified species), pocket gopher, squirrel, and opossum, were observed on kit fox dens in sumps (pers
onal observation
). Use of sump and transition habitats for hunting or sc avenging
refuse may explain why kit foxes traveled short distances at low speeds in these habitats.
Kit foxes used linear habitats (i.e., the Kern River Parkway, canals, and right
-
of
-
ways for
highway
s
, railroad
s
, and powerline
s
) as corridors to move throug
hout their home range. Because
highway, railroad, and powerline right
-
of
-
ways were fenced to prohibit public access, kit foxes
may not have been compelled to move quickly through this habitat type. Kit foxes moved long 55 distances quickly through manicured
open habitats, perhaps because this habitat had a high
occurrence of humans and was used mostly as a corridor for traveling, similar to urban red foxes’ use of golf courses in Toronto, Canada (Rosatte et al. 1991).
Kit foxes in urban areas may have shorte
r nightly movements due to anthropogenic foo
d
sources (Zoellick et al. 1987
). Consequently, movement distance may be influenced by foraging efficiency. Movement distance was less in commercial habitats (where kit foxes were observed foraging in parking l
ots and trash dumpsters) and transition habitats, both which tend to have an abundanc
e of discarded food items (personal observation
).
Conversely, kit foxes moved longer
distances in open, grassland habitat. This may have been because small, insectivorou s prey of kit
foxes and swift foxes tend to be evenly distributed in grassland habitats (Kamler et al. 2004).
Similarly, Koopman (1995) noted that food resources were uniformly distributed in rural kit fox home ranges.
M
ovements
of kit foxes
differed wi
th regard to habitat type and time of night. From
1830
–
2230 h,
linear habitat was used more than expected for kit fox
resting/
walking
movements. This was probably
because 23% of kit fox dens were in linear habitat. From 2230
–
0330 h
, kit fox trotting
and running movements were higher than expected in open habitat, which was likely used for foraging and traveling. From
2230
–
0330 h,
kit fox
resting/
walking
movements were less than expected in residential habitat, in which only one den was located. K it
foxes’ high rate of speed through residential habitat indicates that they used this habitat type as a means to get to a more used habitat. From
2230
–
0730 h,
kit fox trotting movements were less
than expected in sump habitat, which may be used primari ly for denning rather than foraging.
From
0330
–
0730 h
, industrial habitat was used less than expected for kit fox running movements. Kit fox speed may have been lower in industrial habitats during these hours due to the absence of humans or the presenc
e of refuse.
56 U
rban habitat was of sufficiently high quality to support kit foxes with low mortality and normal weight and offspring production. This was likely due to the widespread availability of natural and human
-
derived food sources and protected (i.e
., fenced) areas with cover. Kit fox
core areas contained a disproportionately high amount of sumps, which had food and den sites and restricted public access. The availability of sump and open habitats, which kit foxes selected for denning, may shape th
e distribution and abundance of the urban population. Portions of the urban landscape without these prime denning habitats may not be able to support kit foxes.
57 MANAGEMENT RECOMMENDATIONS
The kit foxes in Bakersfield had low mortality, similar weight and mean litter size, and
high reproductive success, compared to that for rural kit foxes. Urban kit foxes appeared to have adapted to urban conditions and resources, and may contribute to the San Joaquin Valley metapopulation.
Urban kit foxes were observ
ed crossing roads (up to four lanes), but also used culverts and a bridge to move under them, which may help explain the relatively low mortality. Vehicular mortality should be monitored around sump and open habitats. If increased mortality is observed, additional culverts may need to be placed and maintained under roads, along with animal -
proof
fencing to direct kit foxes to the culverts (Ng et al. 2004). Mid -
sized mammals’ (i.e., raccoons
and skunks) use of culverts is positively correlated to culvert length and negatively correlated to
culvert area (Ng et al. 2004). Small diameter culverts (i.e., 0.5 m) may serve a dual purpose by providing a place for urban kit foxes to hide from larger predators and facilitating kit fox movement in a fragmented land
scape.
To promote the continued existence of kit foxes in urban settings, such as Bakersfield, habitat will need to be retained and managed to meet their needs. Urban kit foxes denned farther from roads than expected
in habitat patches that were larger th
an expected,
and thus, may benefit
from the retention of larger open and sump habitat patches for denning. Urban kit foxes had less spatial overlap and smaller home ranges and core areas than their rural counterparts. The greater presence of anthropogeni
c food sources in cities may enable urban kit foxes to meet their energetic needs in a smaller area than kit foxes in rural, grassland habitats, where food resources may be more dispersed (Kamler et al. 2004). Kit foxes utilized open lots and sumps more t han all
other urban habitat types.
58 Golightly (1981) concluded that dens were important to kit fox energetics and a consideration in the management of this species. If kit foxes are to be retained in Bakersfield, open lots, to provide foraging and dennin
g habitat, and uniformly
-
distributed, sumps that are
fenced and closed to the public, to provide kit fox denning habitat, will need to be maintained throughout the city. Sump maintenance (e.g., grading or paving) may be a potential hazard to kit foxes (Mu
rphy 1992) and should be avoided. With sensitivity to kit fox presence in the urban environment, much protection can be afforded at a low cost.
The kit fox recovery strategy identified that “connections need to be established, maintained, and promoted b
etween populations to counteract negative consequences of inbreeding, random catastrophic events (e.g., droughts) and demographic factors” ( United States
Fish and Wildlife Service 1998). Urban kit foxes used the Kern River Parkway to move throughout their
home range. The Kern River Parkway should be maintained and managed to facilitate kit fox movement. In addition, dog management may be required because dogs may pose a threat to urban kit foxes as they have for rural kit foxes (Ralls and White 1995, Cyp her et
al. 2000).
Like urban r
ed foxes in Great Britain (Kolb 1984) and Toronto, Canada (Rosatte et al. 1991) that use railroad corridors to move throughout their home range, urban coyotes (Romsos 1998), red foxes (Lewis et al. 1993, Sallee 1998), and ki t foxes used canals (including foot
bridges over canals) and railroad, highway, and powerline right -
of
-
ways for moving and denning.
The continued existence of these right
-
of
-
ways, which are fenced to prohibit public access, will
facilitate kit fox movemen
t throughout the city. Kit foxes can persist in urban areas if an adequate mosaic of habitat is conserved and managed appropriately.
59 REFERENCES
Ables, E. D. 1969. Home
-
range studies of red fox (
Vulpes vulpes
). Journal of Mammalogy
50:108
-
119.
Adams,
L. W
.
and L. E. Dove. 1989. Wildlife reserves and corridors in the urban environment: a guide to ecological landscape planning and resource conservation. National Institute for Urban Wildlife. Columbia
, Maryland
. 91 pp.
Allredge, J. R.
and J. T. Ratti. 1986.
Comparison of some statistical techniques for analysis of resource selection. Journal of Wildlife Management 50:157 -
165.
Anderson, D. J. 1982. The home range: a new nonparametric estimation technique. Ecology 63:103
-
112.
Apps, C. D., B. N. McLellan, T. A.
Kinley, and J. P. Flaa. 2001. Scale
-
dependent habitat
selection by mountain caribou, Columbia Mountains, British Columbia. Journal of Wildlife Management 65:65
-
77. Arjo, W. M., T. J. Bennett, and A. J. Kozlowski. 2003. Characteristics of current and histor ical kit
fox (
Vulpes macrotis
) dens in the Great Basin Desert. Canadian Journal of Zoology 81:96
-
102.
Beedy, E. C., V. K. Getz, and D. A. Airola. 1992. Status of the San Joaquin kit fox ( Vulpes
macrotis mutica
) in the urban Kern River Parkway, Bakersfield,
California. Pages 47
-
54 in
D. F. Williams, S. Byrne, and T. A. Rado, editors. Endangered and sensitive s pecies of
the San J
oaquin Valley, California.
California Energy Commission, Sacramento
,
California
.
Berry, W. H., T. P. O'Farrell, T. T. Kato, and P. M
. McCue. 1987. Characteristics of dens used by
radiocollared San Joaquin kit fox,
Vulpes macrotis mutica
, Naval Petroleum Reserve #1,
Kern County, California.
United States
Dep
artmen
t of Energy
,
Topical Report,
EG&G/EM Santa Barbara Operations Report No. E
GG 10282
-
21 77, Santa Barbara,
California.
32 pp.
Bjurlin, C. D.
and B. L. Cypher. 2003. Effects of roads on San Joaquin kit foxes: a review and synthesis of existing data.
Pages 397
-
406
in
C. L. Irwin, P.
Garrett, and K.
P. McDermott,
editors.
2003 Proceed
ings of the International Conferenc
e on E
cology and
Transportation.
Center for Transportation and the Environment, North Carolina State University, Raleigh
, North Carolina
.
60 Burt, W. H. 1943. Territoriality and home range concepts as applied to mammals. Jou rnal of
Mammalogy 24:346
-
352.
Calhoun, J. B.
and J. U. Casby. 1958. Calculation of home range and density of small mammals.
United States Public Health Monograph 55, United States Government Printing Office, Washington
,
D.C.
Catling, P. C. 1988. Similariti
es and contrasts in the diet of foxes,
Vulpes vulpes
, and cats,
Felis
catus
, relative to fluctuating prey populations and drought. Australian Wildlife Research 15:307
-
317.
Cypher, B. L. and N. Frost. 1999. Condition of San Joaquin kit foxes in urban and ex urban
habitats. Journal of Wildlife Management 63:930
-
938.
Cypher, B. L.
and G. D. Warrick. 1993. Use of human
-
derived food items by urban kit foxes.
Transactions of the Western Section of the Wildlife Society 29:34 -
37. Cypher, B. L
., G. D. Warrick, M. R.
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. 43 pp.
Doncaster, C.
P., C. R. Dickman, and D. W. Macdonald. 1990. Feeding ecology of red foxes (
Vulpes vulpes
) in the city of Oxford, England. Journal of Mammalogy 71:188 -
194.
EG&G Energy Measurements. 1993. Effects of supplemental feeding on survivorship, reproduction, and d
ispersal in San Joaquin kit foxes.
United States
Department
of
Energy
,
Topical Report, EG&G/EM Las Vegas Operat
ions Report No. EGG 111265
-
2059, Las Vegas, Nevada.
28 pp.
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 Tooele County, Utah. Ecology 43:481
-
497.
Enterprise Advisory Services, Inc. 1997. Endangered species and cultural resources program, Naval Petroleum Reserves in California
, Annual
R
eport FY96. Report 97
-
1,
Enterprise
Advisory Services, Inc., Tupman, California
. 56 pp.
Geffen, E.
and D. W. Macdonald. 1992. Small size and monogamy: spatial organization of Blanford’s foxes,
Vulpes cana
. Animal Behaviour 44:1123
-
1130.
61 Golightly, Jr., R.
T. 1981. The comparative energetics of two desert canids: the coyote and the kit fox
. Ph.D. Dissertation
.
Department of
Zoology,
Arizona State University, Tempe
,
Arizona
. 174 pp.
Golightly, Jr., R. T., Faulhaber, M. R., K. L. Sallee, and J. C. Lewis. 1994 . Food habits and
management of introduced red fox in southern California . Pages 15
-
20 in
W. S.
Halverson and
A.
C. Crabb,
editors
.
Proceedings of the 16
th
Vertebrate Pest Conference.
University of California
,
Davis
, California.
Golightly, Jr., R. T.
and R
. D. Ohmart. 1983. Metabolism and body temperature of two desert canids: coyote and kit foxes.
Journal of Mammalogy
64:624
-
635.
Golightly, Jr., R. T. and R. D. Ohmart.
1984. Water economy of two desert canids: coyote and kit fox.
Journal of Mammalogy
65:51
-
58. Grinder, M. I.
and P. R. Krausman. 2001. Home range, habitat use, and nocturnal activity of coyotes in an urban environment. Journal of Wildlife Management 65:887 -
898.
Grinnell, J., J. Dixon, and J. Linsdale. 1937. Kit foxes. Pages 399 -
420
in
J. Grinn
ell, J. Dixon,
and J. Linsdale, editors.
Fur bearing anim
als of California. Vol. 2.
University o
f
California Press, Berkeley, California
.
Hardenbrook, D. B. 1987. Habitat usage and den sites of kit foxes in west -
central Nevada.
Master’s
Thesis.
Department
of Wildlife,
Humboldt State University,
Arcata, California
.
83 pp.
Harris, S. 1977. Distribution, habitat utilization and age structure of a suburban fox ( Vulpes
vulpes
) population. Mammal Review 7:25
-
39. Harris, S.
1978. Age determination in the red fox (
Vulpes vulpes
)
–
an evaluation of technique
efficiency as applied to a sample of suburban foxes. Journal of Zoology 184:91 -
117.
Harris, S., W. J. Cresswell, P. G. Forde, W. J. Trewhella, T. Woollard, and S. Wray. 1990. Home -
range analysis using radio
-
track
ing data
-
a review of problems and techniques
particularly as applied to the study of mammals. Mammal Review 20:97 -
123.
Harris, S.
and J. M. V. Rayner. 1986. Urban fox (
Vulpes vulpes
) population estimates and habitat
requirements in several British cities
. Journal of Animal Ecology 55:575
-
591.
Harrison, R. L. 1997. A comparison of gray fox ecology between residential and undeveloped rural landscapes. Journal of Wildlife Management 61:112 -
121.
Heisey, D. M
.
and T. K. Fuller. 1985. Evaluation of survival and
cause
-
specific mortality rates
using telemetry data. Journal of Wildli
fe Management 49:
668
-
674.
62 Hintze, J. L. 1999. User's guide
-
NCSS statistical system for windows. Number Cruncher
Statistical Systems, Kaysville
, Utah
. 777 pp.
Hooge, P. N. and B. Eich
enlaub. 1997
. Animal movement extension to ArcV
iew. Version 2.0.
Alaska Biological Science Center,
United States
Geological Survey, Anchorage
, Alaska
.
Hooge, P. N., B. Eichenlaub
, and E. K. Solomon.
1999. Using GIS to
analyze m
ovements i
n the
marine e
nviro
nment.
United States
Geological Survey and National Park Service,
Gustavus
, Alaska
. 20 pp.
Jenness, J. 2003. Longest straight line extension for ArcView 3.x. Version 1.3. Jenness Enterprises, Flagstaff
, Arizona
.
Jenness, J.
2004a.
Nearest features
extensio
n for ArcView 3.x. Version 3.8a. Jenness Enterprises,
Flagstaff
, Arizona
.
Jenness, J.
2004b. Alternate animal movement routes extension for ArcView 3.x. Version 2.
Jenness Enterprises, Flagstaff
, Arizona
.
Jennrich, R. I. and F. B. Turner. 1969. Measurement
of non
-
circular home range. Journal of
Theoretical Biology 22:227
-
237.
Johnson, D. H.
1980.
The comparison of usage and availability measurements for evaluating resource preference. Ecology 61: 65
-
71. Kamler, J. F.
and W. B. Ballard. 2002. A review of nat
ive and nonnative red foxes in North
America. Wildlife Society Bulletin 30:370
-
379.
Kamler, J. F., W. B. Ballard
, E. B. Fish, P. R. Lemons, K. Mote, and C. Perchellet. 2003. Habitat use, home ranges, and survival of swift foxes in a fragmented landscape: c onservation
implications. Journal of Mammalogy 84:989
-
995.
Kamler, J. F., W. B. Ballard
, E. M. Gese, R. L. Harrison, S. Karki, and K. Mote. 2004. Adult male emigration and a female
-
based social organization in swift foxes,
Vulpes velox
.
Animal Behavio
u
r 67
: 699
-
702.
Kilgore, Jr., D. L. 1969. An ecological study of the swift fox ( Vulpes velox
) in the Oklahoma
panhandle. American Midland Naturalist 81:512
-
534.
Kitchen, A. M., E. M. Gese, and E. R. Schauster. 1999. Resource partitioning between coyotes and swi
ft foxes: space, time, and diet. Canadian Journal of Zoology 77:1645 -
1656.
63 Kitchen, A. M., E. M. Gese
, S. M. Karki, and E. R. Schauster. 2005. Spatial ecology of swift fox social groups: from group formation to mate loss. Journal of Mammalogy 86:547 -
554.
K
leiman, D. G.
and C. A. Brady. 1978. Coyote behavior in the context of recent canid research:
problems and perspectives. Pages 163
-
188
in
M. Bekoff, editor.
Coyotes: Biology
,
behavior, and management.
Academic Press, New York, New York.
Knapp, D. K. 1978.
Effects of agricultural development in Kern County, California, on the San Joaquin kit fox in 1977. Final Report, Project E
-
1 -
1, Job V
-
1.21, Non
-
game Wildlife
Investigations. California Department of Fish and Game, Sacramento , California
. 48 pp.
Kolb, H. H
. 1984. Factors affecting the movements of dog foxes in Edinburgh. Journal of Applied Ecology 21:161
-
173.
Kolb, H. H. 1986. Some observations on the home ranges of vixens ( Vulpes vulpes
) in the
suburbs of Edinburgh. Journal of Zoology 53:636
-
639.
Koopman,
M. E. 1995. Food habits, space use, and movements of the San Joaquin kit fox on the Elk Hills Naval Petroleum Reserves in California
. Master’s
Thesis.
Department of
Forestry and Resource Management,
University of California, Berkeley
, California
. 41
pp.
Ko
opman, M. E.
, J. H. Scrivner, and T. T. Kato. 1998. Patterns of den use by San Joaquin kit foxes. Journal of Wildlife Management 62:373
-
379.
Laundré
, J. W.
and B. L. Keller. 1981. Home
-
range use by coyotes in Idaho. Animal Behaviour
29: 449
-
461.
Lewis, J.
C., K. L. Sallee, and R. T. Golightly. 1993. Introduced red fox in California. California Department of Fish and Game
,
Nongame Bird and Mammal Sect
ion Report 93
-
10,
Sacramento, California
. 70 pp.
Litvaitis, J. A., K. Titus, and E. M. Anderson. 1996. Measur ing vertebrate use of terrestrial
habitats and foods in Research and management techniques for wildlife and habitats.
Pages 254
-
274
in
H. S. Mosby, R. H. Giles, Jr., S. D. Schemnitz, and T. A. Bookhout, editors. Research and m
anag
ement techniques for wildl
ife and habitats.
The Wildlife
Society, Bethesda
, Maryland
.
Lucherini, M., S. Lovari, and G. Crema. 1995. Habitat use and ranging behaviour of the red fox (
Vulpes vulpes
) in a Mediterranean rural area: is shelter availability a key factor? Journal of Zoolo
gy 237:577
-
591.
Macdonald, D. W.
1983. The ecology of carnivore social behavior. Nature 301:379 -
384.
64 Matlack, R. S., P. S. Gipson, and D. W. Kaufman. 2000. The swift fox in rangeland and cropland in western Kansas: relative abundance, mortality, and body s ize. Southwestern Naturalist
45:221
-
225.
McGrew, J. C. 1979.
Vulpes macrotis
. Pages 1
-
6 in
Virginia Hayssen, editor. Mammalian
s
pecies. American Society of Mammalogists
, Northampton, Massachusetts
.
Mohr, C. O. 1947. Table of equivalent populations of North American small mammals. American
Midland Naturalist 37:233
-
249.
Morrell, S. 1971. The life history of the San Joaquin
kit fox. Master’s
Thesis.
Department of
Biological Sciences
,
University of California, Santa Barbara
, California
. 67 pp.
Morrell, S.
1972
. Life history of the San Joaquin kit fox. California Fish and Game 58:162 -
174.
Morrell, S.
1975. San Joaquin kit fox distribution and abundance in 1975. California Department of Fish and Game, Wildlife Management Branch Administration Report Number 75 -
3. 27
pp.
Murphy, T. D. 1992. Kit fox distribution and activity in Bakersfield, California. Page 380 in
D. F.
Williams, S. Byrne, and T. A. Rado, editors. Endangered and sensitive s pecies of the San
J
oaquin Valley, California.
California Energy Commission, S
acramento, C
alifornia
.
(abstract only).
Ng, S. J., J. W. Dole, R. M. Sauvajot, S. P. D. Riley, and T. J. Valone. 2004. Use of highway undercrossings by wildlife in southern California. Biological Conservation 115:499 -
507.
National Oceanic and Atmospheric A
dministration
/ National Weather Service. 1997. Local
climatological data
-
Meadows Field We
ather Station.
National Oceanic and Atmospheric
Administration,
National
Cli
matic Data Center, Asheville, North Carolina
.
National Oceanic and Atmospheric Administ
ration
/ National Weather Service
. 1998. Local
climatological data
-
Meadows Field Weather Station.
National Oceanic and Atmospheric
Administration,
National
Climatic Data Center, Asheville
, North Carolina
.
Noss, R. F., M. A. O’Connell, and D. D. Murphy.
1997. Species and habitats. Pages 1
-
18 in
R. F.
Noss, M. A. O’Connell, and D. D. Murphy, editors. The science of conservation planning: habitat conservation u
nder the Endangered Species Act. Island Press,
Washington, D. C.
Odum, E. P. and E. J. Kuenzler. 1
955. Measurement of territory and home range size in birds.
Auk 72:128
-
137.
65 Olson, T. L.
and F. G. Lindzey. 2002. Swift fox survival and production in southeastern Wyoming. Journal of Mammalogy 83:199
-
206.
O’Neal, G. T., J. T. Flinders, and W. P. Clary. 1
987. Behavioral ecology of the Nevada kit fox
(
Vulpes macrotis nevadensis
) on a managed desert rangeland. Pages 443
-
481
in
H. H.
Genoways, editor. Current mammalogy. Volume 1.
Plenum Press, New York
, New York
.
Orians, G. H. and J. F. Wittenberger. 1991. Sp
atial and temporal scales in habitat selection.
American Naturalist 137:S29
-
S49.
Ozoga, J. J.
and E. M. Harger. 1966. Winter activities and feeding habits of northern Michigan coyotes. Journal of Wildlife Management 30:809
-
818.
Pulliam, H. R. 1988. Sources
, sinks, and population regulation. American Naturalist 132:652 -
661.
Quinn, T. 1995. Using public sighting information to investigate coyote use of urban habitat.
Journal of Wildlife Management 59:238
-
245.
Ralls, K. and P. J. White. 1995. Predation on San
Joaquin kit foxes by larger canids. Journal of
Mammalogy 76:723
-
729.
Read, J. and Z. Bowen. 2001. Population dynamics, diet and aspects of the biology of feral cats and foxes in arid South Australia. Wildlife Research 28:195 -
203.
Reese, E. A., W. G. Standl
ey, and W. H. Berry. 1992.
Habitat, soils, and den use by San Joaquin
kit fox (
Vulpes velox macrotis
) at Camp Roberts Army National Guard Training Site,
California.
United States
Dep
artmen
t of Energy
,
Topical Report, EG&G/EM San
ta
Barbara Operations Report No. EGG 10617
-
2156
, Santa Barbara, California
. 36 pp.
Romsos, J. S. 1998. Home range, habitat use, and movements of coyotes in a southern California
urban environment. Master’s
Thesis.
Department of Wildlife,
Humboldt Sta
te Uni
versity,
Arcata, California
. 60 pp.
Rosatte, R. C., M. J. Power, and C. D. MacInnes. 1991. Ecology of urban skunks, raccoons, and foxes in metropolitan Toronto. Pages 31
-
38 in
L. W. Adams and D. L. Leedy, editors.
Wildlife c
onservation in
metropolitan envi
ronments.
National Institute for Urban
Wildlife
,
Symposium Series 2, Columbia, Maryland.
Sallee, K. L. 1998. Spatial
land use of urban red fox. Master’s
Thesis.
Department of Wildlife,
Humbo
ldt State University, Arcata, California
. 70 pp.
Salvatori, V., G.
Vaglio
-
Laurin, P. L. Meserve, L. Boitani, and A. Campanella. 1999. Spatial organization, activity, and social interactions of culpeo foxes ( Pseudalopex culpaeus
) in
north
-
central Chile. Journal of Mammalogy 80:980
-
985.
66 Samuel, M. D., D. J. Pierce, and E.
O. Garton. 1985. Identifying areas of concentrated use within the home range. Journal of Animal Ecology 54:711
-
719.
Samuel, M. D.
and K. P. Kenow. 1992. Evaluating habitat selection with radio -
telemetry
triangulation error. Journal of Wildlife Management 5
6:725
-
734.
San Joaquin Valley Biological Technical Committee. 1993. A biological framework for natural lands and endangered species in the southern San Joaquin Valley.
San Joaquin Valley
Biological Technical Committee,
Turlock,
California.
26 pp.
Seaman,
D. E. and R. A. Powell. 1996. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology
77:2075
-
2085.
Sovada, M. A., C. C. Roy, J. B. Bright, and J. R. Gillis. 1998. Causes and rates of mortality of swift foxes in western
Kansas. Journal of Wildlife Management 62:1300
-
1306.
Spiegel, L. K. 1996. Studies of San Joaquin kit fox in undeveloped and oil -
developed areas: an
overview. Pages 1
-
14 in
L. K. Spiegel, editor. Studies of the San Joaquin kit fox in undeveloped and oil
-
dev
eloped areas. California Energy Commission, Environmental Protection Office, Sacramento, California.
Spiegel, L. K.
and M. Bradbury. 1992. Home range characteristics of the San Joaquin kit fox in western Kern County, California. Transactions of the Western Section Wildlife Society
28:83
-
92. Spiegel, L. K.
, T. C. Dao, and M. Tom. 1996
. Characteristics of San Joaquin kit fox dens at oil
-
developed and undeveloped sites in southwestern Kern County, California. Pages 15 -
38 in
L. K. Spiegel, editor.
Studies of th
e San Joaquin kit fox in undeveloped and oil
-
developed areas, California Energy Commission, Environmental Protection Office, Sacramento
, California
.
Spiegel, L. K.
and M. Disney.
1996. Mortality sources and survival rates of San Joaquin kit fox in oil
-
deve
loped and undeveloped lands of southwestern Kern County, California. Pages 71 -
91 in
L. K. Spiegel, editor.
Studies of the San Joaquin kit fox in undeveloped and oil -
developed areas, California Energy Commission, Environmental Protection Office, Sacramento
,
California
.
Spiegel, L. K.
and J. Tom. 1996. Reproduction of San Joaquin kit fox in undeveloped and oil -
developed habitats of Kern County, California. Pages 53 -
69 in
L. K. Spiegel, editor.
Studies of the San Joaquin kit fox in undeveloped and oil -
develope
d areas, California
Energy Commission, Environmental Protection Office, Sacramento , California
.
Springer, J. T. 1979. Some sources of bias and sampling error in radio triangulation. Journal of Wildlife Management 43:926
-
935.
67 Stafford
, R. W.
and T. D. Murph
y. 1992. Use of water sumps as denning sites for San Joaquin kit fox. Page 383
in
D. F. Williams, S. Byrne, and T. A. Rado
, editors.
Endangered and
sensitive s
pecies of the San J
oaquin Valley, California.
California Energy Commission,
Sacramento
, Californi
a
. (abstract only).
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
Depa
rtment
of Energy
,
Topical Report, EG&G/EM Santa
Barbara Operations Report No. EGG 10617
-
2157
, Santa Barbara, California
. 19 pp.
State of California, Department of Finance. 2002. Revised Historical City, County, and State Population Estimates, 1991
-
2000, wi
th 1990 and 2000 Census Counts.
State of
California, Department of Finance, Sacramento, California .
Storlie, J. T. 1998. Sources of mortality, survivorship, and dispersal of juvenile San Joaquin kit foxes in an urban environment. Senior Thesis.
Department
of Wildlife,
Humbo
ldt State
University, Arcata, California
. 21 pp.
United States
Department of the Interior. 1967. Federal Register 32, 4001.
United States
Fish and Wildlife Service. 1983. San Joaquin kit fox recovery plan.
United States
Fish and Wildlife
Service, California/Nevada Operations Office
, Portland
, Oregon
. 84 pp.
United States Fish and Wildlife Service
. 1998. Recovery plan for upland species of the San
Joaquin Valley, California.
United States Fish and Wildlife Service, California/Nevada Operati
ons Office, Portland, Oregon
. 319 pp.
United States
Geological Survey. 1981. Geographic Names Information System (GNIS).
United
States Geological Survey,
Reston
, Virginia
.
Van Horne, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife
Management 47:893
-
901.
Voight, D. R. and D. W. Macdonald. 1984. Variation in the spatial and social behaviour of the red fox,
Vulpes vulpes
. Acta Zoologica Fennica 171:261
-
265.
Wandeler, P., S. M. Funk, C. R. Largiade’r, S. Gloor, and U. Breitenm oser. 2003. The city
-
fox
phenomenon: genetic consequences of a recent colonization of urban habitat. Molecular Ecology 12: 647
-
656.
Warrick, G. D.
and B. L. Cypher. 1998. Factors affecting the spatial distribution of a kit fox population. Journal of Wildli
fe Management 62: 707
-
717.
68 Whitaker, Jr., J. O.
1998. National Audubon Society f
ield
guide to North American mammals.
2 nd
edition.
Alfred A. Knopf, Inc. New York
, New York
.
White, G. C.
and R.
A. Garrott. 1990. Analysis of wildlife radio
-
tracking d
ata. Ac
ademic Press,
Inc.
San Diego
, California
.
White, G. C. 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 ch
anging prey availability. Journal of Wildlife Management 57:861 -
867.
White, P. J.
, K. Ralls, and D. B. Siniff. 2000. Nocturnal encounters between kit foxes. Journal of Mammalogy 81:456
-
461.
Wiens, J. A. 1973. Pattern and process in grassland bird communiti es. Ecological Monographs
43:237
-
270.
Winer, B. J., D. R. Brown, and K. M. Michels. 1991. Statistical Principles in Experimental Design, 3
rd
edition.
McGraw
-
Hill, Inc. New York
, New York
.
Woodruff, R. A. and B. L. Keller. 1982. Dispersal, daily activity, and home range of coyotes in
southeastern Idaho. Northwest Science 56:199
-
207.
Worton, B. J. 1989. Kernel methods for estimating the utilization distribution in home -
range
studies. Ecology
70:164
-
168.
Zoellick, B. W., C. E. Harris, B. T. Kelly, T. P. O'Far rell, T. T. Kato, and M. E. Koopman. 2002.
Movements and home range of San Joaquin kit foxes (
Vulpes macrotis mutica
) relative to
oil
-
field development. Western North American Naturalist 62:151 -
159.
Zoellick, B. W., T. P.
O'Farrell, and T. T. Kato. 1987
. M
ovements and home range of San
Joaquin kit foxes on the Naval Petroleum Reserves, Kern County, California.
United
States
Department
of Energy
,
Topical Report, EG&G/EM Santa Barbara Operations
Report No. EGG 10282
-
2184
, Santa Barbara, California
. 38 pp.
Zoe
llick, B. W.
, N. S. Smith, and R. S. Henry. 1989. Habitat use and movements of desert kit foxes in western Arizona. Journal of Wildlife Management 53:955 -
961.
69 Appendix
A
.
Sites of live
-
captured kit foxes, domestic cats, and
opossums between May 1997 and
January 1998 in Bakersfield, California.
Species captured
Habitat type
Trap
-
nights
Kit foxes
Domestic cats
Opossums
Residential
5 1
0 0
Commercial
11 6
2 0
Industrial
41 13
5 0
Linear
84 19
3 1
Manic
ured open
53 21
0 0
Open
54 23
3 0
Sump
327
128
33 2
Total
575
211
a
46 3
a
Includes recaptures but
e
xcludes 32 captures
which
occurred when
the
number of traps
in use
was not
record
ed.
76 Appendix
B
.
Kit fox location data and estimates for home range size, core area size, and time to independence for 28 radiocollared kit foxes that were monitored between May 1997 and January 1998 in Bakersfield, California.
Kit Fox Locations
Home Range
(km
2 )
Eartag
Sex
Sequential
Daily
Independent
100% MCP
a
95% Fixed Kernel
50% Fixed Kernel
Core Area (km
2 )
Time to Independence (h)
3468
F
-
166
149
1.17 0.41
0.05 -
5468
F
190
193
166
1.36 0.94
0.08 1.69
5702
F
190
196
184
0.72 0.40
0.06 2.29
5703
b
F
-
8 7
-
-
-
-
5705
F
-
183
163
2.71 2.70
0.20 -
5708
F
133
164
151
1.89 1.67
0.32 1.70
5713
F
-
162
131
2.19 1.59
0.28 -
5723
F
-
190
161
0.93 0.41
0.02 -
5741
F
166
181
174
1.32 0.66
0.08 3.89
5742
F
-
213
177
2.69 2.78
0.43 -
5747
F
165
187
1 73
0.38 0.28
0.04 1.23
5752
F
-
182
155
0.59 0.52
0.03 -
70 Appendix
B.
Kit fox location data and estimates for home range size, core area size, and time to independence for 28 radiocollared kit foxes that were monitored between May 1997 and January 19
98 in Bakersfield, California (continued).
Kit Fox Locations
Home Range (km
2 )
Eartag
Sex
Sequential
Daily
Independent
100% MCP
a
95% Fixed Kernel
50% Fixed Kernel
Core Area (km
2 )
Time to Independence (h)
5753
F
-
198
162
1.07 0.67
0.06 -
5754
F
-
79 64
1.63 2.75
0.32 -
5756
F
-
189
164
0.36 0.35
0.07 -
5674
M
167
181
168
3.33 2.69
0.46 2.61
5701
M
-
206
178
0.86 0.22
0.02 -
5704
M
230
203
193
2.16 2.40
0.39 1.80
5707
M
-
179
160
2.70 2.50
0.50 -
5710
M
187
153
145
1.21 0.37
0.05 2.67
5711
M
-
180
164
2.60 0.15
0.05 -
5712
M
-
167
137
1.90 1.30
0.11 -
5721
M
-
185
158
2.02 1.70
0.27 -
71 Appendix B
.
Kit fox location data and estimates for home range size, core area size, and time to independence for 28 radiocollared kit foxes that
were monitored between May 1997 and January 1998 in Bakersfield, California (continued).
Kit Fox Locations
Home Range (km
2 )
Eartag
Sex
Sequential
Daily
Independent
100% MCP
a
95% Fixed Ker
nel
50% Fixed Kernel
Core Area (km
2 )
Time to Independence (h)
5743
M
219
186
165
2.13 1.14
0.09 1.90
5744
M
-
190
156
2.72 2.27
0.22 -
5746
M
-
176
147
0.31 0.04
0.01 -
5755
M
163
168
154
1.11 0.30
0.03 1.74
5757
M
-
159
143
4.49 c
15.58 d
4.00 d
-
Total
1810
4824
4249
-
-
-
-
Mean
181
172.3
151.75
1.72 1.2
0.16 2.15
SE
9.04 7.62
6.89 0.19
0.19 0.03
0.24 a
Minimum convex polygon.
b
Located too few times to be included in the home range and habitat use analyses.
c
Sum of
two
distinct home ranges in the southeastern (
3.19 km
2 ) and northwestern (
1.31 km
2 ) parts of the study area, approximately 3.1 km
apart.
d
Outlier that was excluded from calculation of
x
¯
+
1 SE.
72 Appendix
C
.
Weight, hind foot, and ear measurements for captured adult, subadult, and p up kit foxes in Bakersfield, California between May
1997 and January 1998.
Male
Female
Body Measurement
Age
n
x
¯
SE
n
x
¯
SE
Statistic
a
p
Weight (kg)
Adult
29 2.47
0.05 32
2.16 0.03
t =
-
5.53 <0.001
Subadult
7 2.09
0.11 6
1.87 0.14
t =
-
1.29 0.22
Pup
20 1.73
0.04 13
1.44 0.05
t =
-
4.25 <0.001
Hind foot (mm)
Adult
27 125.5
1.0 31
119.8
0.8 t =
-
4.4 <0.001
Subadult
6 121
1.2 5
118.4
0.9 t =
-
1.67 0.13
Pup
18 121
1.2 9
112.2
1.4 t =
-
4.36 <0.001
Ear from notch
b
(mm)
Adult
28 81.4
0.6 31
78.6 0.7
z = 3.28
<0.001
Subadult
6 82.5
0.6 5
78.4 1.3
t =
-
3.04 0.007
Pup
19 80.1
0.8 10
75.3 1.1
t =
-
3.65 <0.001
Ear from rear
c
(mm)
Adult
28 84.4
0.7 31
81.5 0.7
t =
-
2.81 0.002
Subadult
6 84.8
1.1 5
81.2 2.2
t =
-
1.56 0
.15
Pup
19 80.8
0.8 10
77.1 1.1
t =
-
2.6 0.007
a
I used a 2
-
sample t
-
test to determine if means differed between the sexes.
For data that was not normally
distributed, I used a Mann
-
Whitney U
-
test to
determine if medians differed between the sexes
.
b
Tip of ear to notch at anterior base.
c
Tip of ear to posterior base.
73 74
Appendix
D
.
Area
-
observation curves using the 100% minimum convex polygon home range for 27 kit foxes in Bakersfield, California from May 1997 to Ja nuary 1998.
Females
0 0.5
1 1.5
2 2.5
3 3.5 4
0 20
40 60
80 100
120
140
160
180
Home Range Area (km2)
3468
5468
5702
5705
5708
5713
5723
5741
5742
5747
5752
5753
5754
5756
Males
0 2
4 6
8 10
12 14 16
0 20
40 60
80 100
120
140
160
180
200
Number of Locations
Home Range Area (km2)
5674
5701
5704
5707
5710
5711
5712
5721
5743
5744
5746
5755
5757
Appendix
E
.
The percent
age
of all radiocollared kit fox daily locations (n = 4,824) and passing vehicles documented in linear habitats during each hour in Bakersfield, California from September
1997 to January 1998.
0 2
4 6
8 10
12 14 16
1500-15591600-16591700-17591800-18591900-19592000-20592100-21592200-22592300-23590000-00590100-01590200-02590300-03590400-04590500-05590600-06590700-07590800-08590900-0959 Time
Percent
Kit Fox Locations in Linear Habitat
Vehicles Passing on Nearest Road
75 EXHIBIT E