Conservation and management of species at risk demands cost effective methods for rapidly monitoring change in animal abundance, distribution and physiologic health over time.
The value of our noninvasive approach was clearly demonstrated in our fist study employing detection dogs as part of a comprehensive monitoring program for large landscapes. Between 1999 and 2001, we examined impacts of human disturbance on grizzly and black bear populations on the east slope of the Canadian Rockies. These populations are exposed to a variety of human pressures including mining, forestry, oil and gas development and exploration, transportation corridors, trapping, hunting, and public recreational use. We also compared results from the dogs to data from hair-snag stations and radio-collared bears that other researchers gathered independently.
Detection dogs were trained to locate feces (scat) from grizzly and black bears, while sampling contiguous grid cells across the 5,200 km2 study area. Sample locations were recorded using a Global Positioning System (GPS). DNA extracted from feces indicated the species, gender, and identity of the individual that left each sample. Data on individual identifications were then applied to mark-recapture models to estimate population abundance. Stress and reproductive hormones extracted from these same fecal samples were used to indicate the physiologic health of these individuals. Data were layered onto a Geographic Information System (GIS) that also included location-specific data on human disturbances over the landscape.
DNA testing of samples showed that the dogs detected four times more individual grizzly bears per square mile than the hair-snag stations did. Statistical tests showed that the dog sampling was unbiased-all bears in the population had an equal probability of being detected. Telemetry provided massive amounts of data on the movements of nineteen radio-collared bears. In the end the telemetry showed the same bear distributions as the scat (Figures 1a-d), but at more than thirty-three times the cost (about $1,000,000 for telemetry versus $30,000 for the dog sampling). Moreover, two grizzly bears died and one was seriously injured as a result of the trapping-high stakes for a population of only 100threatened animals. That, more than anything else, made me appreciate the conservation value of our methods.
Grizzly bears, but not black bears, avoided areas of high tourist densities within the national park, while both species concentrated in the multiple human use areas outside the park. The bears appeared to be drawn to the ephemeral food sources created by the chronic disturbances in these multi-use areas. Grizzly bears also had lower levels of physiological stress, better body condition, and more successful reproduction in these high disturbance areas outside the park.
Unfortunately, these areas were also more accessible to poachers, creating a population sink. Five grizzly bears were poached in this area during the three years of our study. Moreover, no old female grizzly bears (>15 yrs) were found in these areas, despite being relatively common in the neighboring national park. Presumably, resident females were being killed before reaching old age in the multi-use areas.
We compared sampling by detection dogs to two other methods, radiocollaring and DNA collection from hair snag stations. Dogs detected three times the number of unique individuals per km2 than did hair snags. Dog detections were also far less biased as a survey method. Hair snag stations use scent to lure bears into them, biasing captures from individual differences in attractiveness of these lures.
By contrast, scat dogs detect their samples where subjects left them. These detections are relatively free of bias because dogs are driven by the reward they receive for sample detection. These occur independent of biasing factors that bias other methods such as gender, dominance, or territorial behavior.
Scat dog detections also provided comparable movement data to radio collaring and telemetry data, but at one tenth the cost. The telemetry data also proved highly invasive, resulting in tissue damage, loss of body condition, and in at least two cases, death. Sampling with dogs was entirely noninvasive. We never observed a single bear during three years of sampling and collection of over 1200 fecal samples.
Our integrative tools allowed comprehensive coverage of large remote areas and provided data on abundance, distribution and physiological health of grizzly and black bears. Based on these results, we recommended reducing off-road vehicle access to mitigate poaching impacts on grizzly bears. We also recommended tighter monitoring of tourist behavior in Japser National Park, including harsh penalties for tourists who leave their cars in the vicinity of bears.
Wassers et al. 2004. Scat detection dogs in wildlife research and management: Applications to grizzly and black bears in the Yellowhead Ecosystem, Alberta, Canada. Canadian Journal of Zoology (82): 475- 492.
Detection dog methods were developed in collaboration with the K-9 unit of the Washington Department of Corrections. This project was supported by National Fish and Wildlife Foundation and the Foothills Model Forest (Parks Canada, Alberta Fish and Game, Canadian timber, oil, and coal industries) and the Center for Conservation Biology.