Habitat Conditions Associated With Lynx Hunting Behavior During Winter in Northern Washington

1. What they did

Canada lynx are listed as threatened in Washington and across the contiguous United States. Northern Washington lies at the southern edge of their range, where snowshoe hares, their main prey, are scarcer than further north. An earlier study in the same area found that lynx preferred Engelmann spruce and subalpine fir forests and avoided Douglas-fir and ponderosa pine forests and openings. This study asked why, by looking at how lynx hunt inside each forest stand.

The researchers expected lynx to travel in fairly straight lines through poor habitat, saving energy on the way to better places, and to wind back and forth through good hunting habitat as they searched cover to find or flush hares. So they tested whether trail sinuosity (how winding a trail is) could serve as an index of hunting behaviour and habitat quality.

• Finding trails. They divided the study area into six search zones, each about the size of a female lynx home range, and searched by snowmobile more than 12 hours after fresh snow. Once they found lynx tracks, they followed them on snowshoes for as long as possible each day.
• Following trails. To avoid disturbing the lynx, they back-tracked trails less than 24 hours old and forward-tracked older ones. A high-precision Global Positioning System (GPS) receiver logged the trail about every 3 m.
• Recording behaviour. Along the trail they recorded bounding gaits and the species chased, which they read as predation attempts, and identified prey from tracks or carcass remains. They collected scats for DNA to identify individuals and their sex.
• Recording habitat. Every 200 m they recorded vegetation type, tree stems, and canopy and understory cover. Hare densities for each forest stand came from a concurrent pellet-count study.
• Analysing. For each trail segment within a forest stand, they calculated sinuosity and compared it between segments with and without predation attempts, and between hare density classes, forest types and cover classes. They also compared where kills were found with what would be expected from the amount of each habitat.

2. What they found

The trails

• They snow-tracked lynx for 149 km along 51 trails (19 in the first winter, 32 in the second). Trails averaged about 2.6 km and 3.1 km in the two winters.
• 17 trails (3 with kill sites) were left out of the analyses because of poor tracking conditions or because they were shorter than 600 m.
• 11 of the 51 trails were family groups. Where several trails ran together (presumably a mother with kittens), they recorded only the one that seemed to be the mother’s: the most direct route, with no signs of play.
• DNA from 57 scats showed they had tracked at least 9 lynx (6 males and 3 females). Some may have been kittens.

Kills

• Prey remains were found at 21 places: 17 snowshoe hares, 3 red squirrels and 1 unidentified cricetid.
• Lynx made more kills than expected in stands with high hare densities (13 observed against 7.1 expected) and fewer in low-density stands (6 against 11.2).
• Lynx made more kills than expected in spruce and fir forests (17 against 11.6), and fewer in Douglas-fir and ponderosa pine forests (4 against 7.7) and openings (none against 1.7).

Sinuosity

• Trail segments with predation attempts were more winding (sinuosity 2.65) than those without (1.63).
• Trails were more winding in stands with high hare densities (1.99) than low (1.46), in spruce and fir forests (1.80) than in openings (1.26), with 11 to 100% canopy cover (1.76 to 1.89) than under 10% (1.49), and with 11 to 39% understory cover (1.82) than under 10% (1.27).
• Smoothing the trails to allow for GPS error under the canopy did not change these results.

3. What they concluded

• Trail sinuosity was a useful index of lynx hunting behaviour in the Black Pine Basin.
• Lynx hunted more, and more successfully, in the spruce and fir forests where hares were densest. These were the same habitats lynx selected at the landscape scale in the earlier study, so the hunting results give a behavioural reason for that selection.
• Canopy cover itself probably mattered little: both young regenerating stands (open canopy) and mature multistoried stands (closed canopy) had enough understory to support many hares.
• Habitat that suits hunting may not meet other needs, such as denning or resting.

4. How tracking was or was not used

Snow tracking was the core of the study. The team searched for fresh lynx tracks after snowfall, then followed the trails on snowshoes, back-tracking fresh trails and forward-tracking older ones so as not to disturb the animals. While trailing, they mapped the route with GPS and recorded gaits (bounding) and the species chased as signs of predation attempts, identified prey from tracks and carcass remains at kill sites, and collected scats for DNA. They separated a mother’s trail from her kittens’ by its more direct route and lack of play. Trails broken by poor snow conditions were excluded.

5. Tracker or observer certification

Not reported. Nothing is said about who did the tracking or how experienced they were, although the predation attempts, prey identifications and the choice of the mother’s trail all rest on reading tracks and signs. The authors thank D. Harrington, J. Tigner, J. Porter and M. Crabb for “enduring harsh conditions during winter field seasons.”

6. Other technologies, skills or innovations

• Submeter-precision Trimble Pathfinder ProXL and ProXR GPS receivers, logging the trail every 2 seconds (about every 3 m), with a datalogger to record behaviours along the trail.
• Geographic Information System (GIS) mapping of 330 habitat polygons in ArcGIS 9.2, and trail smoothing to check for GPS error under canopy.
• DNA analysis of scats collected along snow tracks to identify individuals and sex.

7. Implications for conservation (from the paper)

• To keep forest stands valuable as lynx hunting habitat, the authors encourage forest managers to maintain or create enough understory cover to support high densities of snowshoe hares.
• Ways to do this include deferring or avoiding precommercial thinning, thinning with reserves, limiting brush and understory removal, and planting regeneration stock at high densities.
• Lynx also need habitat for denning and resting, which hunting habitat may not provide.