Masters Thesis

Urban Landscape Attributes and Intraguild Competition Affect San Joaquin Kit Fox Occupancy and Spatiotemporal Activity

Human population growth and rapid urbanization have created new, attractive environments for opportunistic animals including some species of wild canids. San Joaquin kit foxes (Vulpes macrotis mutica) are a federally listed endangered and California listed threatened canid that persists in the city of Bakersfield, California, where they form a unique ecological guild with three other canid competitors: coyotes (Canis latrans), red foxes (Vulpes vulpes), and gray foxes (Urocyon cinereoargenteus). These four canids typically exhibit avoidance and/or resource partitioning due to overlapping niches, with smaller fox species avoiding attacks from more dominant foxes and coyotes by selecting alternative resources, finding refuge, occupying different habitat types, or adjusting behavior. Recent carnivore sympatry in urban areas may be due to behavioral adjustments and adaptations to complex urban environments, including heterogeneous landscape matrices and new, abundant resources. I investigated carnivore sympatry in urban environments using 5 y of remote camera survey data collected throughout the city of Bakersfield to first determine how landscape attributes within heterogeneous urban landscapes influence San Joaquin kit fox occupancy patterns, and if canid competitors (i.e., coyotes, red foxes, and gray foxes) affect San Joaquin kit fox distributions with the use of occupancy modeling. Second, I investigated how the presence of canid competitors or domestic dogs (Canis familiaris) in the same 1-km2 area affects San Joaquin kit fox spatiotemporal activity with the use of Two-way Contingency Tables, One-way Analysis of Variance, and Kruskal-Wallis tests. I found the most supported occupancy model for San Joaquin kit foxes to be an additive effect of two urban landscape attributes, percentages of paved roads and campuses (e.g., schools, churches, and medical centers) in cells. The percentage of paved roads was a negative predictor of San Joaquin kit fox occupancy while the percentage of campuses was a positive predictor. The percentage of paved roads was ultimately the most supported covariate for predicting San Joaquin kit fox occupancy (or lack thereof) in my study system. Roads are the main source of mortality for urban San Joaquin kit foxes and have greater noise pollution, development, disturbance, and human activity, which may discourage San Joaquin kit foxes from incorporating roads into their urban home ranges. Conversely, campuses have landscaping, sports yards, quadrangles, and walkways that offer open space, which San Joaquin kit foxes select for in natural habitats. These sites also afford security from excess human disturbance and larger predators due to fences and other security measures employed by campuses, as well as anthropogenic food sources from cafeterias and people directly feeding San Joaquin kit foxes. I concluded that San Joaquin kit foxes were avoiding paved roads while selecting for campuses in the urban environment. The presence of coyotes, red foxes, and gray foxes was not a contributing factor of urban San Joaquin kit fox occupancy patterns, though this may have been a result of low sample sizes of other canids compared to San Joaquin kit foxes. Apart from one association between the number of days in which San Joaquin kit foxes occurred alone and the number of days in which they occur with other canids in 2018, I found no other associations between San Joaquin kit fox and other canid occurrences in cells or on given days. I also found differences between the number of days San Joaquin kit foxes occurred alone and the number of days they occurred with another canid for all years collectively and each year individually. I concluded that urban San Joaquin kit foxes rarely occur with coyotes, red foxes, gray foxes, and domestic dogs in the same 1-km2 area within the same day, same year, or 5-y span, suggesting spatiotemporal avoidance of canid competitors. In instances when San Joaquin kit foxes and other canids did occur on the same camera on the same survey night, I found San Joaquin kit foxes delay their time to appearance following sunset by about 3 h at camera stations where another canid species appeared. Furthermore, variances in mean consecutive min that San Joaquin kit foxes spent at stations showed that they had the least predictability in the potential window of time spent at the station if another canid visited the camera station on the same night but did not appear first. San Joaquin kit foxes had the most predictability in the potential window of time spent at the station if another canid appeared first. These results indicate that San Joaquin kit foxes may require a more immediate predator presence cue than scent to perceive imminent risk from nearby competitors. Finally, my results show that if multiple canid species did occur there were never more than three, though primarily only two canids occurred in any given cell or on any given day, with a majority of co-occurrences between kit foxes and domestic dogs. Because domestic dogs are abundant in urban areas, they may not be novel or threatening to kit foxes, allowing domestic dogs and kit foxes to co-occur at higher frequencies than kit foxes and other wild canids. Additionally, where coexistence does occur, canids may only be willing to exist with one other canid species at any given time. In both analyses, I confirm that San Joaquin kit foxes occur in higher abundances than any other wild canid species in Bakersfield. San Joaquin kit foxes may be more receptive and adaptive to highly developed urban areas than other canids and are frequently observed denning in inner city landscapes; whereas past studies show that coyotes require larger, connected ranges and natural habitat, that red foxes avoid coyotes in intermediate human-modified habitats (i.e., suburbs with house densities of < 20 houses/ha), and that gray foxes select for urban edges or more natural, tree covered areas. My results also demonstrated a sizeable decrease in kit fox abundance over the years, with a 69% decrease in San Joaquin kit fox abundance at camera stations and a 40% decrease in probability of San Joaquin kit fox occupancy from 2015 to 2019. This is explained by the recent outbreak of sarcoptic mange (Sarcoptes scabiei) skin disease in San Joaquin kit foxes in Bakersfield, which is highly infectious and 100% fatal in untreated kit foxes. Overall, I conclude that while San Joaquin kit foxes rarely occur with other canid species within a 1-km2 urban area, they may require immediate predator presence cues to perceive risk from competition, while avoidance of paved roads and selection for campuses as urban landscape characteristics may be of greater importance in explaining occupancy dynamics in urban San Joaquin kit foxes. Understanding how top predators adapt to developing landscapes provides insight towards species conservation and management in urban areas, which is particularly important for the San Joaquin kit fox. Conserving the unique urban population in Bakersfield may be significant for the overall health, survival, and recovery of this species as human development is projected to continue, and upcoming conservation efforts may be particularly critical considering the current mange epidemic within this population.


In Collection: