Identifying Sex and Individuals of Pine Marten Using Snow Track Measurements

1. What they did

Counting pine martens is hard. Snow tracking only works for a day or two after fresh snow, and trapping and radio tracking can miss animals. Researchers working on mountain lions had identified every individual in their study area from track measurements, so the author asked whether the same could be done for pine martens: can snow track measurements tell males from females, and one marten from another?

From winter 1993 to winter 1996 he caught and radio-collared 8 pine martens (3 males and 5 females) in Białowieża National Park, Poland. He found the martens at their daytime resting sites, then followed their tracks in the snow for up to 10 m from the resting site. He measured only stretches where the marten moved in its usual 2-gait (paired print) pattern and was not jumping between logs, so he worked in places without logs or dense shrubs.

He took four measurements: the width and length of one print (called a “pit”), the width of the two prints side by side, and the mean of 10 leap lengths. Because the shortest leaps in this gait relate closely to body length, he also averaged the 5 shortest of the 10 leaps. He then used statistics (discriminant analysis) to see how well the measurements sorted the tracks by sex and by individual, and checked whether snow depth (1 to 60 cm) affected the results.

2. What they found

Telling males from females

• Males’ tracks were generally larger than females’, but the ranges overlapped (41 track samples: 15 from males, 26 from females).
• Single print measurements did not separate the sexes well. Pit length sorted only about half the tracks correctly.
• Leap lengths did much better. The average of the 5 shortest leaps grouped 86.7% of male and 92.3% of female tracks correctly (averages of 63.6 cm for males and 46.3 cm for females).
• Together, the average of the 5 shortest leaps and pit width gave a simple formula (a sex key) that classified 85.4% of tracks correctly (73.3% of males and 92.3% of females).

Telling individuals apart

• Differences between individuals were very small and measurements overlapped a great deal.
• Only the average of the 5 shortest leaps was useful, and it classified just 43.9% of track samples to the right individual.
• The largest male (number 3, 1525 g) was identified correctly every time; the other martens were identified correctly in few cases.
• Snow depth did not significantly affect the measurements.

Mapping uncollared animals

On a map of winter home ranges (Figure 2), track samples with similar leap lengths found in the same area were joined up as presumably one animal. This suggested four uncollared females (A to D) living around the collared martens.

3. What they concluded

• Snow track measurements cannot always identify individual pine martens, and probably not other medium-sized carnivores either. Population estimates based only on this method may be badly biased.
• Identifying individuals from tracks works in larger animals such as mountain lions, but some measurements used for them (such as the angle between toes) could not be taken from marten tracks, because hair around the foot pads blurs the prints.
• Track measurements are a more reliable way to tell male from female martens than the position of scat and urine relative to the hind footprints. Marten scats were on average 1.4 km apart along a trail, and martens often moved around after defecating, which made the position impossible to read.
• Track measurements, especially the average of the 5 shortest leaps, are recommended as a complementary method for studying where pine martens live. They can help estimate density or map uncollared animals, for example unmarked females inside a collared male’s home range. This works best when an animal is very big or when neighbours’ leap lengths do not overlap, and the map of females was possible only because one collared female’s leap lengths were known.

4. How tracking was or was not used

The author found radio-collared martens at their resting sites and followed their trails in the snow for up to 10 m, measuring prints and leaps where the animals moved in their usual 2-gait pattern. Because each track sample came from a known, collared animal, he could test how well the measurements matched sex and identity. Snow tracks of uncollared females were also mapped by their leap lengths (Figure 2).

5. Tracker or observer certification

Not reported. The author took all measurements himself; his tracking experience is not described or tested. As each track came from a known collared marten, track identification itself was not in doubt.

6. Other technologies, skills or innovations

• Radio collars on 8 martens, used to find them at resting sites and to map their winter home ranges.
• A simple sex key (discriminant function) from two measurements: the average of the 5 shortest leaps and pit width.

7. Implications for conservation (from the paper)

• Density estimates for pine martens based only on track measurements may be greatly biased.
• Trapping tends to catch more males than females, so track measurements can help find unmarked females living in or around a collared male’s home range, and the other way round.