| In this dissertation, we used the endangered San Joaquin kit fox ( Vulpes macrons mutica) as a model species to examine the application of noninvasive fecal DNA monitoring in carnivore conservation and management. First, we explored the utility of using trained dogs to locate scats of kit foxes. Our results showed that most dogs could recover large samples of scats and were more efficient than human searchers, but rates of recovery can vary due to temperature and years of experience. In field trials, where red foxes (Vulpes vulpes) were not present, dogs were consistently accurate at distinguishing kit fox scats from scats of sympatric carnivores. Further discrimination trials showed that dogs were capable of selecting between scats of both fox species, yet were less reliable at ignoring red fox scats when scats of kit fox were not present. Finally, scat-detection dog surveys proved effective in multiple population areas, regardless of fox density or vegetation type. Second, we sought current information on both kit fox behavior and status. We demonstrated that, despite prior knowledge, the use of latrines is a common feature of kit fox marking behavior, and likely involved in chemical communication. Also, we found that the relative abundance of kit foxes in the San Joaquin Valley, CA is extremely variable. We provided recommendations for enhancing long-term persistence of kit foxes based on the result that greatest abundance occurred in specific population areas in the southern part of the range. Finally, we examined solutions for addressing errors in fecal genotypes, and provided suggestions for improving the recovery of genotypes in population studies. Using multiple genetic and GIS tests, we demonstrated an effective approach for both assessing genotype error and limiting the amount of additional laboratory work needed to prevent mis-scoring. We obtained information on sex ratio, relatedness, fox movement patterns, latrine use, and size of home range. Additionally, we showed that recovery of fecal genotypes is best maximized by using trained dogs, searching for scats on both unpaved roads and transects in vegetation, and surveying at the largest feasible geographic scale. |
