DNA Analysis of Hair and Scat Collected Along Snow Tracks to Document the Presence of Canada Lynx

1. What they did

Snow tracking is a common way to find out which carnivores live in an area, but track identification depends on the judgement of whoever is in the field and gets harder as snow conditions get worse. The authors asked whether hairs and scats picked up along snow tracks could be used for DNA tests, so that the DNA, not the observer, confirms the species.

They worked in four study areas in Washington, Montana and Minnesota and set out to:

• Estimate encounter rates: how often hairs, scats and beds turn up per km of Canada lynx track.
• Test the samples: what share could be identified to species from their DNA, and what share were lynx.
• Compare with hair snares: in the Meadows area, compare winter snow-track samples with samples from baited summer hair snares.

Washington. Each study area was split into six zones of about 34 km². The team searched roads and trails in one zone each working day by snowmobile until they found a track they took to be lynx, then followed the freshest track for at least 2 km. They collected scats along the trail and hairs from beds: daybeds (oval, crusted hollows where the animal had curled up to rest), ambush beds (where the animal had sat with front legs stretched out, presumably waiting for prey) and, in Black Pine Basin, lookout beds (where the animal sat upright on their haunches). To find the few, hard-to-see hairs, an observer knelt or lay with their eyes about 30 cm from the bed and scanned it from two or three angles. In summer 2001, they also set 25 transects of five baited carpet-pad hair snares in the Meadows area.

Montana. They located 15 radio-collared lynx (6 females, 9 males) by radio telemetry and followed each animal’s tracks for 2 to 3 km, recording daybeds. No DNA was collected here.

Minnesota. They followed up 14 reports of lynx or lynx tracks from natural resource agency staff and in-house reports, 8 of them from one person who said they had regularly seen lynx tracks for about 30 years. They searched for sign near each location, followed any lynx tracks found for irregular distances, and collected all hairs and scats.

Samples were dried and stored, and the DNA was identified to species in the laboratory.

2. What they found

Overall

• Of 317 samples from all areas, 88% gave usable DNA and 247 were lynx.
• Scats were a better source of DNA than hairs: only 3 of 128 scats failed. About 81% of hair samples and 98% of scat samples from snow tracks could be identified.

Meadows, Washington

• Along 104 km of lynx track, the team collected 48 hair samples and 47 scats. Of these 95 samples, 84 were lynx, 6 were other species and 5 could not be identified: about 1 lynx sample per 1.2 km.
• Daybeds turned up about once every 1.8 km, and 88% (29 of 33) of daybeds held hair. Ambush beds were less productive (5 samples from 20 beds).
• Summer hair snares gave 37 samples; 31 had usable DNA, but only 6 were lynx. The rest were other felids, canids and other species.
• DNA success was similar for snow-track hairs and snare hairs (89.6% and 83.8%), but samples from snow tracks were far more likely to be lynx (88% against 16%).

Black Pine Basin, Washington

• Along 155 km of lynx track: 84 hair samples and 64 scats. Every one of the 128 samples with usable DNA was lynx, about 1 lynx sample per 1.2 km.
• Hair was found in 74% of daybeds, 22% of ambush beds and 1% of lookout sites. Daybeds turned up about once every 2.0 km.

Clearwater, Montana

• The team followed 240 sets of tracks from 15 collared lynx over 582 km and found 198 daybeds, about one every 2.9 km.

North Shore, Minnesota

• 20 hair samples and 17 scats. Of the 30 with usable DNA, 29 were lynx. One “lynx” hair sample was later shown to be from a lynx and bobcat hybrid.

How far to follow a track

• If samples are spread at random along the trail, each km of lynx track followed gives about a 55% chance of at least one lynx DNA sample. Following a track for 4 km gives more than a 95% chance.

Cost

• A species-level DNA test takes about 2 days, half an hour of technician time per sample, and about $5.00 in materials.

3. What they concluded

• Collecting hairs and scats along snow tracks is a workable way to confirm Canada lynx, and possibly other rare carnivores. In Washington, following a lynx track for 2 to 4 km gave a high chance of a usable DNA sample.
• Scats kept their DNA unusually well, possibly because they were fresh and cold.
• Encounter rates differ by place, so the decision about how far to follow tracks should start out cautious and can be cut back if samples turn up often.
• When DNA decides the species, a clear track matters less. Surveys can run in a wider range of snow conditions and for more days after snowfall, and a misidentified track costs extra work but does not produce false data.
• Snow tracking acts as a screen: it gathers mostly lynx samples, while summer hair snares collect many samples from other species. Each method has strengths and weaknesses that depend on snowfall, terrain, access and lynx density.
• Many other boreal carnivores travel shorter distances than lynx, so their beds and scats are likely closer together and the method may need even shorter tracking distances for them.
• The authors agree with Squires et al. (2004) in recommending hair and scat collection for DNA whenever reliable species identification is a high priority.

4. How tracking was or was not used

Snow tracking was the core of the study. The team found tracks they took to be lynx (from snowmobiles along roads and trails in Washington, by radio telemetry in Montana, and by searching near reported sightings in Minnesota), identified them in the field and then trailed them, usually for 2 km or more. While trailing, they read the tracks and signs for scats, daybeds, ambush beds and lookout beds, and searched each bed closely for hairs. Locations of tracks, beds and samples were recorded with GPS.

The authors point out that snow-track identification depends on field personnel, who vary in skill, and can be confounded by observer bias. DNA was used to check the species of each sample. In the Washington areas, nearly all usable samples from tracks taken to be lynx were confirmed as lynx (all 128 in Black Pine Basin).

5. Tracker or observer certification

Not reported. The paper does not say who followed the tracks or how skilled they were. In Minnesota, 8 of the 14 target locations came from one person who said they had regularly seen lynx tracks for about 30 years; their skill was not tested.

6. Other technologies, skills or innovations

• Differentially corrected Global Positioning System (GPS) for track, bed and sample locations, and Geographic Information System (GIS) software (ArcGIS) to measure distances along tracks.
• Radio telemetry to find collared lynx in Montana.
• Baited hair snares: carpet pads set with nails, baited with beaver castor, catnip oil and catnip, with an aluminium pie plate as a visual attractant.
• Hair stored in silica desiccant; DNA extracted with Qiagen kits; species identified from mitochondrial DNA (16S rRNA) and restriction enzyme digests.

7. Implications for conservation (from the paper)

• For rare carnivores of conservation concern, such as lynx and wolverine, verifiable evidence of presence and distribution is an essential first step for conservation plans and habitat management.
• Presence is often based on sightings or track descriptions, and anecdotal observations of forest carnivores are inherently unreliable. Where a report can be linked to a particular snow track, following it to collect DNA will likely be worth the time and money.
• DNA gives physical evidence that other researchers can check independently.