Using Remote Camera Traps to Better Estimate Snow Leopard Numbers
While SLIMS (Snow Leopard Information System) sign surveys can be improved through better training and use of sophisticated sampling regimes (see Occupancy Surveys), the relationship between the amount of sign present and the number of cats in the same area is just too variable and imprecise to produce consistent or sufficiently reliable results.
For these and other reasons, the Snow Leopard Conservancy pioneered the use of non-invasive, remotely-triggered camera “traps” for photographically “capturing” this seldom-seen carnivore. Ullas Karanth of the Wildlife Conservation Society was the first to use remote cameras to census tigers in India. But no-one had attempted this technique under the harsh conditions of Himalayan winters, or on a species whose elusiveness is so legendary.
Cameras are set where a snow leopard would be expected to visit; the camera automatically takes the image of any warm-bodied animal that happens to pass by. While some cameras use white or incandescent flash at night, others record pictures using only infra-red light which is invisible to many animals. The latest devices operate in sub-zero temperatures, and can be left unattended for weeks or even months at a time while the supervising scientist is miles away in the comfort or his/her home or office! Camera traps enable researchers and conservationists to accurately establish population size, identify resident cats or track specific individuals over extended periods of time within the camera-trapped area.
Initially, with older-model cameras that take only one or several photos, we deployed two cameras at each location, in order to photograph the snow leopard from both sides. As each snow leopard has its own unique pelage pattern (see below), views of the left and right sides are needed to identify a cat at the first “capture.” Now, with the latest digital cameras capable of recording an image every second or less, a single camera may be able to record both sides of a snow leopard as it moves about investigating a scrape or scent-mark.
Back in the office, the scientist carefully examines each image, matching the leopard’s spotting pattern against a photo library of animals whose identity is known from pictures taken earlier in the survey or from previous surveys. Identifying different snow leopards is most easily accomplished by comparing the rosette and spot patterns along each flank or side of the body; the more images one has to work with the better.
Example of identification of two separate individuals based on pelage patterns on the fore- and-hind limbs; HNP-1 (top row) and HNP-3 (bottom row). Red lines indicate the primary features and the blue lines indicate the secondary features used for identification. The number of identifiable features varies with body posture.
“Capture-mark-recapture” algorithms and computers are used to estimate the number of snow leopards present within the area surveyed. But since each snow leopard has its own distinctive pelage “fingerprint,” there is no need to physically capture or artificially mark animals for future recognition. This makes camera trapping highly non-invasive compared to radio-collaring.
There are two kinds of camera traps. The first uses active infrared technology, where a transmitter placed on one side of a trail sends out a beam of infrared light (which cannot be seen by people or snow leopards) to a receiver located several meters away on the other side of the trail. When this invisible beam is broken, the receiver electronically signals one (or more) cameras to take a picture. Like Dr. Karanth’s work with tigers, we used TrailMaster® TM cameras in India and Mongolia on snow leopards.
The other kind of detection system is known as passive infrared: a single camera is linked to a detector which is constantly scanning the immediately surroundings for sources of infrared energy which warm-bodied animals emit as their body heat radiates out into the environment. When the sensor detects both heat and movement, it instructs the camera to take a picture. Unfortunately, passive infrared cameras may also be triggered when heated rocks send out wafts of warm air!
Digital camera traps are rapidly replacing film cameras. New, faster and less invasive infra-red cameras like the Reconyx RapidFire©, Scoutguard SG 550V© or Bushnell Trophy Cam© are capable of capturing their first image within a second or less. They can also be set to record video or capture consecutive images seconds apart. Many hundreds, even thousands of pictures can be stored on a single memory card, making fast data collection routine. As noted, rapid image acquisition helps eliminate the need to deploy two cameras at each site, since snow leopards will often turn around while scraping or scent-marking, thus displaying both sides of their bodies to a single camera trap.
For more on camera trapping of snow leopards, download the following papers here:
(1) “Estimating snow leopard population abundance using photography and capture-recapture techniques” by Rodney Jackson, Jerry Roe, Rinchen Wangchuk and Don Hunter published in 2006 in the Wildlife Society Bulletin Volume 34, pages 772-781). You can go to the Wildlife Society Bulletin website for more information.
(2) Interested persons can consult Cat NEWS (www.catsg.org) for our recent camera trap survey undertaken in Mongolia’s South Gobi or request this paper from the authors: Jackson, Rodney M., Bariushaa Munkhtsog, David P. Mallon, Galsandorj Naranbaatar and Khurlee Gerelmaa. 2009. “Camera-trapping snow leopards in the Tost Uul region of Mongolia” Cat NEWS 51:18-21.
Camera Trap Handbook for Snow Leopards
Our definitive Surveying Snow Leopard Populations with Emphasis on Camera Trapping: A Handbook is based on four years of in-depth research undertaken in the Hemis National Park, Ladakh supplemented with training workshops held in Nepal and Pakistan. Download your own copy of the handbook below.
Surveying Snow Leopard Populations with
Emphasis on Camera Trapping: A Handbook
There are three versions available of increasing print quality:
Lower resolution “Screen Version” (pdf – 2.7 Mb).
Higher resolution “eBook” (pdf – 4.3 Mb).
If you would prefer to have this document on CD at nominal cost, please email us with your address and preferred postage option: Snow Leopard Conservancy
Handbook Now Available in Chinese
This definitive handbook on camera-trapping of snow leopards has now been translated into Chinese. This translation brings an important non-invasive survey technique to researchers and protected area managers working in the four provinces: Gansu, Qinghai, Sichuan and Yunnan, and two autonomous regions: Xinjiang Uygur, and Xizang (Tibet) harboring the endangered snow leopard.
Scientists estimate that China contains as much as 60% of all snow leopard habitat, with the primary threats being the growing trade in furs and bones (for traditional Asian medicine) and retaliation by shepherds when snow leopards kill their livestock.
Translation of the handbook is a collaborative effort between the Conservancy and the Xinjiang Conservation Fund, Irbis Mongolia, Fauna and Flora International-China, and the Japanese Wildlife Research Center in Tokyo.
For details contact the Conservancy or
Ge Yun at the Xinjiang Conservation Fund (XCF):
Liu Fang Bei Li No.9 Building, Unit 10, Room 102,
Chaoyang District, Beijing 100028, China
Watch for the Russian language version which will be released in May 2010.
We are focusing on training range-country biologists and village stewards in conducting sign transects and setting remote camera traps for deriving reliable population estimates or counts. Our work shows that wildlife guards and local villagers can be trained in camera trapping. The identification of individuals from their pelage pattern and the on-going cataloging of all images accrued over time will provide information on the minimum number of individuals present and the duration of their “residency” within the area surveyed. Knowing the individual snow leopards inhabiting a particular area helps promote stewardship of the species among interested households in the local community.