Evaluation of Road Track Surveys for Cougars (Felis concolor)

1. What they did

Wildlife managers often use tracks and other signs to judge how many animals live in an area, because it is cheap and needs little equipment. But this only works if the link between the signs and the population is understood. An earlier study (Van Dyke and others, 1986) found only a weak link between how often cougar tracks were found on roads and how many cougars lived there. The authors set out to test this again in an area where cougar numbers were known, and to see whether the way cougars are spread across the land, measured by their home ranges, affects how many tracks are found.

• Knowing the cougars. Cougars were tracked on horseback, treed with trained hounds, and fitted with radio collars. They were located from the ground and the air, aiming for at least once a week. The area was searched regularly for signs of new cougars (tracks, scats and scratches), and new residents and transients were caught and collared too.
• Systematic surveys (May to June 1988). One 11.3 km dirt road was chosen in each of three separate areas, crossed by 2 to 3, 4 to 5, and 6 to 7 cougar home ranges. Each road was dragged with a conifer tree pulled behind a pickup truck. The next day both sides of the road were searched for cougar track sets from the truck at 8 to 12 km per hour. Each road was dragged and surveyed again 3 to 10 days later.
• Random-systematic surveys (July and August 1988). Four separate areas with different cougar densities (0.017 to 0.057 cougars per km²) each had a randomly chosen 16 km stretch of dirt road, surveyed in the same way but from an all-terrain vehicle and dragging only one shoulder. New routes were chosen each round until all dirt roads in an area had been used.
• Measuring the tracking surface. Every km, before and after dragging, the observer took 10 steps (5 on each shoulder) and scored each shoe print from 1 to 4 (a “dust rating”). They also recorded rain and traffic.

A track set was one continuous set of tracks left by one cougar on one occasion.

2. What they found

Track finds

• The systematic surveys covered 407 km and found only two track sets, both on the road crossed by 4 to 5 home ranges. This was too few to analyse.
• The random-systematic surveys covered 684 km and found seven track sets.

Tracks and cougar numbers

• There was no relationship between track-finding frequency and cougar density in cougars per km² (r² = 0.00, P = 0.886, 4 data points).
• There was a positive relationship between track-finding frequency and the number of cougars whose home ranges overlapped the road (r² = 0.73, P = 0.066, 5 data points). The point for 7 overlapping home ranges was dropped because less than 20 km of road was surveyed there.

The road surface

• Dragging did little to improve the tracking surface, or the dust ratings were not sensitive enough to show a change.
• Moderate rain had little effect on the surface, but heavy rain that deformed the road surface made it worse. Traffic had no significant effect.
• Too few track sets were found to test whether surface conditions or dragging changed how many tracks were found.

3. What they concluded

• Road track surveys were a poor index of cougar density in southern Utah. Both this study (based on a small sample) and Van Dyke and others (1986) found a weak link between density and track-finding frequency.
• The main problem is sampling. Cougars are rarely spread evenly, and the roads available are seldom many enough, or in the right places, to sample them properly. Roads might miss all home ranges, or run only where cougars live.
• Because road location matters so much, using track counts to compare cougar density between areas is doubtful. Using them to show that cougars are present is feasible, but still depends on roads crossing a home range.
• The link with the number of home ranges crossing the road suggests the method could help monitor cougars where roads are plentiful and placed so that every cougar’s home range is crossed. Even then, with 27% of the variation unexplained, it may only show fairly large changes in cougar numbers.

4. How tracking was or was not used

• Road track surveys: observers searched freshly dragged dirt roads from a slowly moving vehicle for cougar track sets, and counted them per km.
• Sign searches: the study area was searched for tracks, scats and scratches to find cougars not yet collared.
• Trailing for capture: cougars were tracked on horseback and treed with trained hounds before being collared.
• Judging the tracking surface: the observer’s own shoe prints were scored to rate how well the road would hold tracks.

5. Tracker or observer certification

Not reported. Nothing is said about who searched the roads for tracks or how their skill was judged. C. S. Mecham and M. C. Mecham are thanked for field help and for working as houndsmen. With so few track sets found, observer skill could matter to the results, but the paper does not discuss it.

6. Other technologies, skills or innovations

• Motion-sensitive radio collars (Telonics) and radio telemetry from the ground and the air, with locations recorded to the nearest 100 m.
• Dragging roads with a conifer tree behind a truck to make a fresh tracking surface.
• Dust ratings from the observer’s shoe prints to measure road surface condition (following Van Dyke and others, 1986).

7. Implications for conservation (from the paper)

• Road track surveys appeal to agencies managing cougars because they are cheap, but they gave a poor index of cougar density in this study.
• They may be useful for monitoring cougars only where roads with good tracking surfaces are plentiful and well placed, for example in parts of the Northwest where logging is common, and even then mainly for detecting fairly large changes in numbers.