Search Dogs, Scats, and DNA: A Noninvasive Approach for Monitoring Endangered SpeciesEPA Grant Number: U915998
Title: Search Dogs, Scats, and DNA: A Noninvasive Approach for Monitoring Endangered Species
Investigators: Smith, Deborah A.
Institution: University of Washington - Seattle
EPA Project Officer: Michaud, Jayne
Project Period: January 1, 2001 through January 1, 2004
Project Amount: $90,906
RFA: STAR Graduate Fellowships (2001) RFA Text | Recipients Lists
Research Category: Fellowship - Terrestrial Ecology and Ecosystems , Academic Fellowships , Ecological Indicators/Assessment/Restoration
Endangered species often are rare and difficult to monitor with traditional methods (e.g., trapping, visual surveys) that become less effective as population density decreases. Recent advances in genetic analysis of DNA extracted from feces (scats) offer a potentially effective method for monitoring such species. With specific genetic markers, identification of species, individuals, and sex can be determined from scats, and potentially provide information on population size, sex ratio, home range, distribution, paternity, and kinship. However, for this technique to be successful, scats must be easily found and recovered. The detection of scats in a diversity of habitats may prove difficult. The objective of this research project is to incorporate an innovative and highly effective scat collection technique that uses dogs specifically trained to detect scats from my study species, the San Joaquin kit fox. I have found that these detection dogs were 100 percent accurate (n = 1,298 scats) at distinguishing kit fox scats under field conditions and were significantly more effective at locating scats than were human observers.
The purpose of my research is to validate the use and appropriateness of fecal DNA analysis, and offer a powerful survey method for rare species. My research addresses recovery tasks associated with the distribution, abundance, habitat management, and population genetics of endangered San Joaquin kit foxes. I used detection dogs and fecal DNA analysis to determine the current distribution and abundance of kit foxes in several parts of their range where trapping and spotlighting surveys have been unsuccessful. I found scat surveys using detection dogs to be an effective method for determining kit fox presence and relative abundance in both habitat types used by kit foxes: nonnative grassland, and saltbush scrub. Using the San Joaquin kit fox as a model species, I am directly comparing fecal DNA analyses and radiotelemetry techniques to determine if both methods provide similar data on kit fox population density, sex ratio, social groupings, survival, reproduction, and differential habitat use. Using detection dogs, I collected kit fox scats during the same period as telemetry observations. Preliminary comparison of individual foxes identified from fecal genotyping and direct capture from approximately 10 percent (95/963) of the scats collected in January and February 2002, found 22 unique genotypes. Sixteen of these genotypes match the genotypes of foxes already captured in the study area at all six microsatellite loci as well as sex. The remaining six genotypes found at least two times in 19 scats do not match genotypes for any of the foxes that were trapped in the study area. However, spotlighting in January and February 2002, indicated that there were still many foxes in the study area that had not been trapped and radio collared. This conclusion is supported by the fact that I detected the presence of four foxes by their unique scat genotypes from scats collected in January 2002, before their presence was verified by trapping 8 to 10 months later. I am evaluating whether kit foxes prefer grazed or ungrazed habitat by surveying for scats along appropriate transects on grazed and ungrazed plots. Currently, most of the remaining area inhabited by kit foxes is heavily grazed with unknown consequences. Results from this survey will provide information for habitat restoration and management prescriptions for kit foxes.