# Comparing Scat Detection Dogs, Cameras, and Hair Snares for Surveying Carnivores

> Which finds forest carnivores best: a trained dog’s nose, a remote camera, or a hair snare? Across Vermont, scat detection dogs found black bears, fishers and bobcats at more than three times as many sites as cameras, while hair snares never caught a fisher or bobcat hair. Dogs cost a little more per site, but for bears and fishers one dog visit did what took five camera surveys.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2007-long-donovan-comparing-scat-detection-dogs/

## Citation and access

- **Authors:** Robert A. Long, Therese M. Donovan, Paula MacKay, William J. Zielinski and Jeffrey S. Buzas
- **Year published:** 2007
- **Journal or source:** The Journal of Wildlife Management
- **Full citation:** Long, R. A., Donovan, T. M., MacKay, P., Zielinski, W. J. and Buzas, J. S. 2007. Comparing scat detection dogs, cameras, and hair snares for surveying carnivores. The Journal of Wildlife Management 71(6): 2018–2025.
- **DOI:** 10.2193/2006-292
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.2193/2006-292

## Tags

- **Species:** Mammals > American black bear; Mammals > Bobcat; Mammals > Fisher; Mammals > Grey fox
- **Region:** North America > United States
- **Tracking:** Hair; Scat; Sign surveys; Tracking; Tracks and signs identification
- **Methods and technology:** Camera traps; Detectability; DNA; Hair sampling; Occupancy modelling; Survey costs
- **Data type:** New field data
- **Observer qualification:** Not reported
- **Discipline:** Wildlife tracking
- **Year published:** 2007

## Detailed keywords

- **Species:** Studied: American black bear (Ursus americanus), fisher (Martes pennanti) and bobcat (Lynx rufus). Identified from hair samples but not target species: gray fox (Urocyon cinereoargenteus), and wolves (Canis lupus or C. rufus), domestic dogs (C. lupus familiaris) or eastern coyotes (C. latrans), which could not be told apart by mitochondrial DNA. Mentioned from other studies: Canada lynx. Scat detection dogs were the survey tool.
- **Location and habitat:** The whole state of Vermont (24,963 km²), plus several sites just west of central Vermont’s border with New York, USA. A mainly forested, topographically complex region of the northeastern United States.
- **Topics:** noninvasive survey methods; scat detection dogs; remote cameras (camera traps); hair snares; carnivore detection; detection probability given presence; unique detections; survey effort needed to reach 80% detection probability; latency to first detection; camera failure; attractants, bait and scent lures (and the biases they may bring); scat age and temporal mismatch between scat and animal presence; area sampled by search versus station-based methods; survey costs (leased versus purchased dog, cameras, hair snares); DNA species identification from scat and hair. Inconsistencies in the paper: the Methods give 74 camera sites as 50 in 2003 and 24 in 2004, while the Results give 49 in 2003 and 25 in 2004. In Table 2, the per-site figure for purchased-dog supplies and vet costs over both surveys is printed as 5.17, but 1,628 dollars over 120 sites is about 13.57 (and 13.57 is what makes the column add up to the 257 dollar total). The text says dogs detected each species at more than 3.5 times as many sites as cameras; for black bears it is exactly 3.5 (49 versus 14).
- **Statistics or models:** Single-season occupancy modelling in program MARK version 5.1 to estimate detection probability (p) per method, treating dog visits, camera and hair snare surveys as repeat surveys; cumulative detection probability P = 1 − ∏(1 − pᵢ)^nᵢ; Fisher’s exact tests between years; descriptive statistics (raw detection rates, means and standard errors); cost scenarios for two hypothetical 60-site surveys at 2005 prices and a 10 dollars per hour pay rate.
- **Data type:** New field data, May to August 2003 and 2004: 168 sites surveyed 220 times by detection dog teams, with cameras and hair snares at 74 of them. Cost estimates for hypothetical surveys. Detailed site selection, dog training, survey protocols and scat results are in the companion paper, Long et al. (2007), “Effectiveness of scat detection dogs for detecting forest carnivores”.

## Synopsis

### 1. What they did

Carnivores range widely, live at low densities and are hard to see, so researchers often survey them through what they leave behind. The authors compared three such “noninvasive” methods for detecting black bears, fishers and bobcats across Vermont: scat detection dogs, remote cameras and hair snares. They also compared what each method costs.

- Detection dog surveys. Teams of a dog, a handler and an orienteer searched a 2 km, diamond-shaped transect at each of 168 sites, roughly 5 km apart. Scats found inside the diamond or up to 100 m outside it counted towards whether a species was detected at a site. Some sites were visited two or three times.

- Camera surveys. At 74 randomly chosen dog sites, they set one camera for 14 days, with no revisits or rebaiting. They used a commercial trapping lure, a bag of fish food pellets and molasses for bears, and in 2004 raw chicken for fishers.

- Hair snare surveys. At each camera site they put up two carpet-pad snares studded with barbed nails, baited with a scent lure and catnip, with a hanging metal pie pan as a visual attractant. Bear hair was identified by its colour, shape and the large amounts left on the tree; other hair was sent for DNA species identification.

- Comparing the methods. They compared raw detection rates, “unique detections” (where only one method found the species at a site) and the probability of detecting each species in a single survey if it was present. From these, they worked out how many surveys each method would need to reach an 80% chance of detection.

- Comparing costs. They estimated the cost of two hypothetical 60-site surveys with a leased dog, a purchased dog, cameras or hair snares.

### 2. What they found

Detection dogs

- Dog teams surveyed 168 sites 220 times, and found 728 scats in 2003 and 868 in 2004. A survey took 4.1 hours on average.

- Dogs detected black bears at 57.1% of sites, fishers at 61.3% and bobcats at 12.5%.

Cameras

- Cameras or sensors failed in 18 of 74 surveys (24%): film shot all in one day, jammed cameras, flat batteries, chewed wiring and cameras knocked out of place.

- On sites with working cameras, they detected black bears at 23.7%, fishers at 19.6% and bobcats at 5.4%. Fishers were detected far more often in 2004 (36.4%) than in 2003 (8.8%).

- The first detection came after about 6 days on average (5.9 days for bears, 6.6 for fishers, 6.0 for bobcats).

Hair snares

- Of 29 hair samples, 7 were thought to come from contamination during setup or removal, and one was lost. Six were attributed to black bears by eye.

- Of 15 samples sent for DNA testing, 8 failed. The rest came from a black bear, a gray fox, and wolves, domestic dogs or coyotes (which the DNA test could not separate).

- Hair snares detected black bears at 9.5% of sites and never detected fishers or bobcats.

Comparing the methods (74 sites with all three)

- Dogs detected each species at at least 3.5 times as many sites as cameras, the next best method.

- Dogs gave the only detection of black bears at 65.3% of the sites where bears were detected, fishers at 74.5% and bobcats at 78.6%.

- The chance of detecting a species in a single survey, if it was present, was highest with dogs: 0.87 for black bears, 0.84 for fishers and 0.27 for bobcats. With cameras it was 0.33, 0.28 and 0.13; with hair snares, 0.08 for black bears.

- To reach an 80% chance of detection, one dog visit was enough for bears and fishers, and five for bobcats. Cameras would have needed five surveys for bears and fishers (three for fishers on 2004 results) and 12 for bobcats. Hair snares would have needed 20 surveys for black bears.

Costs per site

- Leased dog: 316 dollars. Purchased dog: 257 dollars. Cameras: 214 dollars. Hair snares: 153 dollars.

- DNA analysis was a large share of the cost for dogs and hair snares.

### 3. What they concluded

- Detection dogs were far more effective than cameras, which were in turn more effective than hair snares. Dogs cost somewhat more per site, but once the effort needed for a high chance of detection is taken into account, they were clearly the most cost-effective method.

- Dogs need only one site visit, which matters where sites are far away, hard to walk into or hard to get permission for.

- Dogs need no bait or lure, which may reduce bias (for example if males and females are drawn to lures differently), avoids feeding wildlife and avoids drawing animals to dangerous places such as road edges.

- Dogs find scat, not animals. Scat lasts days to months, so a scat shows that a species used the site in the recent past, not that it is there now. In this study only 1 of 39 black bear detections by dogs (2.6%) rested only on scats thought to be a month old or more. Fisher and bobcat scats could not be aged, but these species range less widely than bears and are strongly territorial, so the animals were likely still using the site.

- Dogs searched a large area, while cameras and snares worked from one point and relied on lures, so the area each method really sampled is hard to compare.

- The camera and hair snare protocols were deliberately minimal. More cameras per site, more frequent checks and rebaiting, and snare designs made for bears or mustelids might have done better, but would cost more.

- Cameras and hair snares keep real benefits: cameras give instant identification, show current use, record many species and produce photos that help public relations; hair snares are cheap and need little training.

- Costs could fall by buying a dog and using them over several seasons, by hiring experienced dog-and-handler teams, or by cutting DNA costs where scat or hair can be identified by eye (as for black bears here).

### 4. How tracking was or was not used

No human tracking or trailing was reported. The survey rested on signs: the dogs found scats, which were then identified to species (the methods are in the companion paper, Long et al. 2007), and black bear scats were aged from their appearance. On the hair snares, black bear hair was identified by eye from its colour, shape and the large amount left on the snare tree; other hair went for DNA testing.

### 5. Tracker or observer certification

Not reported. Five detection dog teams did the surveys, but this paper says nothing about the experience of the handlers or the people who identified hair and aged scats (dog training is described in the companion paper). The authors strongly encourage researchers to consult experienced professional trainers before starting a detection dog project.

### 6. Other technologies, skills or innovations

- Scat detection dogs working with a handler and an orienteer.

- Film cameras with Trailmaster active and passive sensors (2003) and CamTrakker camera-sensor units (2004), set for a 5-minute delay between photos.

- Carpet-pad hair snares made from coiled nail-gun nails clipped into small barbs.

- DNA species identification of hair and scat (Wildlife Genetics International).

- Occupancy modelling in program MARK to estimate detection probability, and a formula to combine detection probabilities across methods and repeat surveys.

### 7. Implications for conservation (from the paper)

- Researchers who need to detect carnivores and collect scat should consider detection dogs, especially when a high chance of detection and little bias matter, as in surveys for endangered species.

- A low chance of detection makes occupancy estimates less accurate and precise, so cost comparisons should account for how well each method actually detects the species.

- More studies are needed of how well dogs survey different species in different conditions, and how they compare with other methods.

- Survey goals, available people, climate and terrain will decide whether dogs are the best choice for a given study.

### 8. Implications for tracking (Original Wisdom's observations)

_These are Original Wisdom's observations, not the authors'._

- Scat surveys share a limit familiar to trackers: a sign tells you an animal was there, not that they are there now. The authors handled this by ageing bear scats from their appearance, a tracks and signs skill whose accuracy this paper does not test.

## Cite this summary

Lawrence, K. 2026. Summary of: Comparing Scat Detection Dogs, Cameras, and Hair Snares for Surveying Carnivores (Long and Donovan et al., 2007). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2007-long-donovan-comparing-scat-detection-dogs/