Use of track transects to measure the relative occurrence of some boreal mammals in uncut forest and regeneration stands

1. What they did

Large areas of Canada’s boreal forest have been logged, but little was known about how small carnivores and their prey use the forest as it grows back. The researchers wanted to know whether counting tracks in the snow could show which stand ages animals prefer, which winter months give the highest counts, and whether track counts can follow changes in numbers from year to year.

Near Manitouwadge, Ontario, they chose two sites in each of five stand types: uncut forest (150 to 200 years old) and areas logged less than 5, 10+, 20+ and 30+ years before. Each site had two permanent 1 km transects, laid out to cross upland and lowland in proportion to their occurrence.

From 1980 to 1985 they walked the transects two to eight times in each of three winter periods (late November to December, January to February, and March), 12 to 96 hours after fresh snow. They counted the tracks of marten, lynx, red fox, ermine, snowshoe hare and red squirrel that crossed each transect, and divided the count by the number of 12-hour periods since snowfall. Several red squirrel tracks together counted as one, and a hare trail was given an arbitrary value of four.

To test whether track counts reflect numbers, they compared them with live captures of marten, hare and red squirrels in spring and fall, and with local fur harvest records for lynx, fox and ermine.

2. What they found

Best time of winter

• Track counts for marten, red squirrel and ermine were generally highest in late November to December. Red squirrel tracks were fewest in January.
• Hare and lynx counts did not change significantly between the winter periods. Most species tended to have their lowest counts in January.

Stand age

• Marten: more tracks in uncut forest than in any logged stand, in all 5 years.
• Snowshoe hare: most tracks in stands logged 20+ years before, fewest in stands less than 5 years old. In uncut forest, hare tracks could also be high, reflecting openings in these old stands with shrubs hares eat.
• Red squirrel: generally most tracks in uncut forest, and fewest in stands less than 5 years old every year. In some years, 20- and 30-year-old stands had counts similar to uncut forest.
• Lynx: seldom recorded in uncut forest or stands under 20 years; most tracks in stands logged 20+ years before. On 113 km of transects in uncut forest they counted only four lynx tracks.
• Red fox: more tracks in logged stands than in uncut forest in 2 of the 5 years; most fox tracks were in 10- to 30-year-old stands.
• Ermine: no consistent stand age preference; their counts were the most erratic of all species.

Changes between years

• Marten, hare and fox tracks were significantly higher in 1980 and 1981 than later; red squirrel tracks were highest in 1980; lynx peaked in 1981 and 1982 in the 20+ year stands.
• Track counts were significantly correlated with live captures for marten (spring and fall), hare (spring and fall) and red squirrels (fall only). In 1984, squirrel tracks suggested a low population, yet that fall’s captures were the highest of the study, which the authors could not explain.
• Track counts were not significantly correlated with fur harvest for lynx or ermine, and only weakly for fox.

3. What they concluded

• Snow track counts can be used as an index of stand age (habitat) preferences in winter and of population trends among years, as long as each year is compared with the years before.
• Marten appear to need uncut forest: their use of logged stands stayed low for decades. Hare and lynx both favour stands about 20 years after logging, which fits lynx depending on hare.
• Track counts vary widely between transects and years, so non-parametric statistics were needed.
• Counts are best made in early winter (before mid-December), before trapping, cold weather and time spent under the snow reduce the number of tracks.
• Four or more permanent 1 km transects in at least two sites can index population trends for marten, ermine, hare and red squirrel. For fox and lynx, which live at low density, transects should be longer, probably 3 to 5 km, to avoid many zero counts.
• Fur harvest is a poor index of numbers, because the catch is affected by fur prices, snow depth and other work available to trappers.

4. How tracking was or was not used

Tracking was the main method. The team identified and counted the tracks of six species that crossed fixed transects in fresh snow, timed after snowfall so that old tracks were covered. No trailing was reported. Tracks were treated as an index of relative occurrence, not as counts of individual animals: squirrel tracks close together counted as one, and hare trails were given a fixed value.

5. Tracker or observer certification

Not reported. The paper does not say who identified the tracks or how experienced they were; L. Krysl and S. McGovern are thanked for helping with some of the transects.

6. Other technologies, skills or innovations

• A simple standardization: dividing track counts by the number of 12-hour periods since snowfall, so counts made at different times after snow can be compared.
• Live-trapping, with marten and hares marked for individual identification, to check track counts against numbers caught.
• Rough estimates of how many animals a 1 km transect could cross, from published home ranges and densities: 1 to 5 marten, 15 to 300 hares, 16 to 70 red squirrels and 0 to 5 lynx (more likely 0 to 1).

7. Implications for conservation (from the paper)

• Logging in upland boreal forest helped some species (such as hare and lynx in stands around 20 years old) while others, especially marten, stayed depressed in numbers for many years.
• Habitat needs can change with the season, especially in the breeding season, so winter track data alone cannot describe all the habitats these species need.