The Status of Snow Leopards (Panthera uncia) Conservation in Jigme Khesar Strict Nature Reserve at Haa, Bhutan

Background and Research Aims

Jigme Khesar Strict Nature Reserve (JKSNR) is the only Strict Nature Reserve in Bhutan, the rest of the protected areas in the country being either a National Park or Wildlife

Sanctuary. It has a total area of 609.5 Km2 and is composed of many high-value medicinal, globally endangered, rare, and endemic species of plants and animals plus the spectacular sceneries, and unique culture [1]. The Reserve has the strongest representation of pristine temperate and alpine eco-systems in the country [1]. It is the only protected area in Bhutan that is part of the trans-boundary conservation landscape, under Kanchenjunga, which is also the part of Sacred Himalayan Landscape of Eastern Himalayas [2] popularly known as “Snow Leopard Landscape”, forming part of Himalaya Biodiversity hotspots [3]. Further, the Reserve also is an involuntary part of the two of the WWF Global 200 ecoregions [4]. As the area name suggest, the Reserve is home to a good number of important flora and fauna including the elusive Panthera uncia.

The Snow Leopard (Panthera uncia) is now considered a Vulnerable species but until September 14, 2017 was considered a globally endangered wild cat by International Union for Conservation of Nature (IUCN) and is still listed under Appendix I of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). As such, trade and transport of any part of the body of the species is strictly forbidden [5]. In Bhutan it is listed in Schedule I of Forest and Nature Conservation Act of Bhutan 1995. Geographically, this magnificent cat is found in 12 Asian countries namely Afghanistan, Bhutan, China, India, Kazakhstan, Kyrgyzstan, Mongolia, Nepal, Pakistan, Russia, Turkmenistan and Uzbekistan [6]. In Bhutan it is found in Jigme Dorji National Park (JDNP), Jigme Khesar Strict Nature Reserve (JKSNR), Wangchuk Centennial National Park (WCNP) and Paro Territorial Forest Division. Quite recently newspaper sources [7] and personal communications confirmed its presence in Jigme Singye Wangchuk National Park and Bumdeling Wildlife Sanctuary (Tshering Dendup pers. com). However, other than the presence report, there are no reliable studies (abundance, population, etc.) conducted on the species throughout its range [8] and Bhutan is not an exception. As such, conservation of the species remained confined to the presence absence confirmation in its range thus far, compelling the conservationists in Bhutan to pursue further study of the species. The main problem faced managing the species is knowledge on abundance of the species in particular habitat or region. Bhutan does have information on the presence of the species but lack the exact points where they are located and on what they survive.

Therefore, the Reserve officially embarked on the Snow Leopard Survey along with the national level survey in the identified habitats to assess population and distribution of the species initially with the investigation of signs and evidences of prey species such as Blue Sheep (Pseudois nayaur), Himalayan Marmot (Marmota himalayana), pheasants and domestic preys including gathering of information from the residents of the Reserve (migratory cattle herders) with effect from December 2014. The actual camera trapping exercises began from October until December 2015 with camera installations, monitoring and retrieval exercises being carried out with the minimum of 60 trap nights within the given schedule in the study area. Today, the study is being continued to maintain the status quo along with an aim to further strengthen the conservation of the vulnerable species.

Thus, the main objective of this study is to confirm the number of species occurring in JKSNR using camera traps so that appropriate conservation management plan can be developed. This research is to also identify exact places where the snow leopard occurs and recommend appropriate management schemes.

Study Area

JKSNR covers two Districts Haa and Samtse and lies between 27034’ and 27011’ latitude and 88054’ and 89010’ longitude of which almost 98 percent of the area falls under Haa (Figure 1).

Click to enlarge

However, the study focused on the north and central part of the Reserve which has proven Snow Leopard presence. It

is the only protected area in the country which is virtually uninhabited and boasts of pristine temperate and alpine vegetation, home to several high-altitude wild animals besides Panthera uncia. The study was confined to areas which are mostly below 5200masl and above 3500masl. Most of the areas at 5200masl could not be covered because of topographic reasons and international border security issues. Therefore, most of the data obtained are from the grids laid over the map in Figure 2. While it is considered uninhabited permanently, the study includes the seasonal migratory yak herders from Haa valley who live in harmony with wildlife.

Click to enlarge

Camera Trap Installation Site Selection Process

Spatial Capture Recapture (SCR) method [9, 10] which is in vogue in recent times is expected to produce robust estimation of population and abundance [11] enabling conservationists to identify individual species of elusive large predators [12]. Based on the same principles postulated by experts above, potential ecological zone of the species was identified using the biological information of species (home range, prey base, breeding characters, etc.) in and around the Reserve. The results of signs and evidences surveys were used to identify the potential areas where the species could be found most probably. Having identified the habitats in the field, Google Earth Map was used to develop grids for installing camera traps. At least 17 square grids, each of size 4km by 4km covering 272 km2 area was laid on the identified habitat on Google map and subsequent field truthing (ridgeline crests and saddles, bases of rock cliffs, trails, and base of prominent isolated rock boulders) using GPS was done for the coordinates provided by Google by the surveyors. After locating the potential sites from here on referred as camera stations (Table 1), camera traps were installed physically keeping in mind the grids. Better alternative camera trap sites were chosen within the grid and cameras were set up ensuring that the distances between the cameras were not too close to each other. Specialized cameras, Scourt Guard (SG560C), Uway (VH2008) and RECONYX cameras fed with rechargeable batteries that can last more than three months with eight and four GB memory cards were used.

Stationing Camera Traps

In each selected survey grid, two cameras were placed facing in opposite directions to capture the hind (tail) and front (face and forelimbs) portions of Snow Leopard assigning camera codes such as Grid 10 Right (G10 (R)) and Grid 10 Left (G10 (L)) for instance (Figure 3). Using the clue for camera trap setting from the guideline [13], topographical features of the landscape were used to station the cameras. A total of 34 cameras were stationed in 17 grids using locally available stone slabs to protect the cameras from rain and sun (Figure 4). To ensure that the cameras could capture best images, team members mimicked Snow Leopards in front of the cameras to check if the cameras were placed at optimal heights and distances on the Snow Leopard trails (Figure 5). No camera lens faced the sunlight directly. Post installation, data sheets (Appendix I and II) were used to record way points, coordinates and altitudes of the trap sites (Table 1). Scats were collected on the way to supplement and confirm the species presence, beside camera traps.

Click to enlarge

Click to enlarge

Click to enlarge

Monitoring of Camera Traps

Camera traps once laid in the field were monitored on monthly basis as per the fixed schedule (Table 2). Every time the monitoring teams were on the camera spots, the cameras were switched off to replace the memory cards so that the earlier card could be checked for images of the animals.

Memory cards of the different camera of different grids were properly labelled and taken back to office for analysis. The status of batteries of each camera were checked during every visit and those needing replacement replaced. Cameras that did not record any animals were relocated to a new position within the grid and coordinates were recorded.

Installation of camera traps into 17 sampling grids were carried out in September-October 2015 and subsequent Phase I monitoring of camera traps undertaken from October- November 2015. The Phase II monitoring was carried out in between December and January 2016 considering a total of 60 traps nights of monitoring period. The standard 60 trap nights were adopted as suggested Jackson RM [14] who carried out camera trapping of Snow Leopards in Hemis National Park, Ladakh, India.

Encouraged by the images captured, the camera trap exercise was extended a month for another 30 trap nights which calculated to 90 trap nights. The extended trap nights produced an interesting result when a new member showed up in Grid 10.

The monitoring of camera traps, retrieval and adding of cameras along the identified trap sites continues and the Reserve expects better results.

Data Analysis-Identifying Individual Snow Leopards

After the retrieval of cameras, all memory sticks were scanned for the species photos. Of the many photos captured, clear images that prominently showed facial parts, forelimbs, flanks and tail parts was selected for individual identifications. Idea used and proposed, to identify individual Snow Leopards [6] were put into practice by comparing the rosette and spot patterns along each flank or side of the body. As such, individual Snow Leopards were identified based on unique pelage patterns the method for which is well prescribed [14]. Example from our study of identifying two separate individuals is as shown (Figure 6) where pelage patterns on the fore-and-hind limbs differ significantly.

Click to enlarge

Results of the Photographic Analysis

Of the 215 images of Snow Leopards recorded by the camera traps laid in 10 grids over the period of 60 trap nights, 163 images (76%) could be used for individual identification while rest of the images (24%) were not usable for individual identification. We do not include the individual JKSNR 10 since it was captured outside the standard capture period of 60 days. However, we mention in this report about how important it is to extend the extension of trap nights from 60 to 90. More than/the number of capture nights, better the results.

JKSNR’s first Snow Leopard after the instalment of cameras traps was captured at grid number 2 (Nub Tshonapata Top) at 08.11pm on 1st October 2015 at an altitude of 4384masl. The camera trap stationed has alpine grassland with Rhododendron scrubs facing South East aspect at the geo-coordinates of 27°19’54.95”N and 89° 9’3.99”E.

Click to enlarge

Records of individual’s forelimbs, facial portion, flanks and dorsal surface of the tail, torso and other possible angles as used based on the guideline [14] enabled establishing individuals (Table 3). Many false images triggered by infrared light, snow fall, movement of clouds and movement of grasses by wind were captured in the cameras which had to be discarded. Non-target species like Blue Sheep (Pseudois nayaur), the main natural prey of Snow Leopard in Bhutan was also captured (41%) while few images of other carnivores like Wild Dog (Cuon alpinus) 0.8% and Red Foxes (Vulpes vulpes) made 5.3% of the total captures. The suspected Tibetan Wolf (Canis lupus chanku) and Woolly Hare (Lepus oiostolus) were not trapped in any cameras while images of Asiatic Golden Cat (Pardofelis temminckii), Himalayan Serow (Capricornis sumatraensis), Wild Pig (Sus scrofa) and the most unsuspected species Gaur (Bos gaurus) were trapped conspicuously. Musk Deer (Moschus chrysogaster), Himalayan Black Bear (Ursus thibetanus), Weasel (Mustela sp.) and Yellow Throated Marten (Martes flavigula) made their presence felt while birds like Tibetan Snow Cock (Tetraogallus tibetanus), Snow Partridge (Lerwa lerwa), Blood Pheasant (Ithaginis cruentus), Satyr Tragopan (Tragopan styra), Himalayan Monal (Lophophorus impejanus) and other bird species (18.4%) also showed up in most of the cameras traps as analyzed in Figure 17.

Of all the captured images besides Panthera uncia, Bos gaurus captured from 4500 masl above the legendary lake Nub-Tshonapata in grid 1 near Tshona beat other images as the species is classified as Vulnerable by IUCN and is protected in Bhutan as schedule I animal of FNCA 1995. The record of this species at such an altitude is first of its kind in Bhutan.

An interesting field knowledge is added with the capture of individual JKSNR 10 in 90 trap nights. The standard 60 trap nights may need to be reconsidered, because after the retrieval of cameras traps from December 2015-January 2016, some were monitored while others were retrieved. The final retrieval of all the remaining Cameras traps was done in between February and March 2016 during which an additional member which we named JKSNR 10 was photographed in Grid 10. It is noteworthy that no Snow Leopards were photographed from Grid 10 during the two earlier phases. Thus, this indicate that 60 trap nights is not enough to completely enumerate its population. To reconfirm the authenticity of the individual JKSNR 10 captured, repeated comparison with those already captured individuals was done. Scat and DNA collection survey was carried out two times in May and June 2016 and materials have been submitted to Nature Conservation Division (NCD) for analysis. While the result is awaited, JKSNR considers it to be a separate individual based on its rosette pattern and other features being different from others.

Key prey species for Snow Leopards in JKSNR is Blue Sheep unlike the species in Mongolia where Ibex, Markhor, Blue Sheep and Argali are equally important. With the good population of Blue Sheep in various places (Figure 18) inside the survey area viz. a herd each in Tshabjola (n=58), Tsang (n=20), Tshonapatra (n=30), Chuzomtse (n=30), Kampula (n=30), Tamzhing (n=30), Jazagortab (n=15), and Tshonathangkha (n=20) which is outside the Grid, the future of Snow Leopards in JKSNR is not insecure.

Click to enlarge

Click to enlarge

However, the diversity of prey species is lower than those in Mongolia which has at least four sympatric species on which the Snow Leopard can prey on. Though our study is ongoing, JKSNR can safely infer that the threat to Snow Leopard is not competition for food between the wild and the domestic prey species unlike the case in Mongolia. There are no threats of cattle population increasing since the numbers have dwindled with reducing number of herders over the last few years. However, a small threat of feral domestic dogs hunting Blue Sheep calves (Figure 19) are evident which JKSNR as a conservation entity may need to take care if the problem become worst in the near future.

Repeated capture of Individual Snow Leopard

The most frequently captured Snow Leopard is individual JKSNR 6 which appeared at least 7 times in Grid number 2, 6, 8 and 12 followed by JKSNR 4 who appeared 5 times in grid number 9 and 13 (Table 4). JKSNR 2, 5 and 8 appeared only once each while JKSNR 1, 7 and 9 appeared two times each in various cameras traps. Similar analysis shows that the maximum number of Snow Leopard images (n=12) was captured in grid 9 Kampula while no images were captured at Grids 1, 3, 5, 7, 15, 16 and 17 (Figure 20). The highest captured and recaptured individual was JKSNR 6 which was recorded 7 times in Grid Numbers 2, 6, 8 and 12 (Figure 21). Individuals were captured more than once in at least 10 grids as explained in Table 4. During the extended period (January 2016), a new individual was captured in Grid 10. JKSNR, 1, 4 and 6 also appeared once each in the grid during the extended period (Table 5).

Click to enlarge

Click to enlarge

Identification of Snow Leopard Using the Captured Images

Cameras traps were set to photograph Snow Leopard

directly approaching or departing trap sites to primarily document facial features, forelimbs, flanks and dorsal surface of tail. Distinctive shapes of rosette groups were mostly used for distinguishing individuals because most of the images showed these body parts (Figure 22).

Click to enlarge

Only nine individuals from ten grids could be identified from 215 images of Snow Leopard. Many images could not be used for identification as some were blurred while others contained body parts that could not be analyzed. About 26 percent of Snow Leopard images were captured facing the camera traps directly while 5 percent of images taken were photographed from behind. At least 29 percent of the photographs were taken sideways resulting into 59 percent of all images captured showing only three-quarters of the animal’s torso while 41 percent of images showed up full torso. Limb visibility was restricted to 32 percent with one limb images making up 9 percent, two limbs 20 percent, and three limbs making up 25 percent while all four limbs visibility made up 14 percent. About 29 percent of images captured showed no tails while dorsal surface visibility made up 20 percent, and lateral surface visibility made up 21 percent. At least, 50 percent of the images captured showed

30 percent of tails. All these images made the laboratory work easier to identify individuals which are all named after the Reserve.

Activity Pattern and Habitat Use

The distribution of the Snow Leopard activity patterns on a 24 hours scale, based on the number of images of individuals photographed by different camera traps showed bimodal (two peaked) distribution, one in early morning and one in the late evening (Figure 23). Such a pattern suggests crepuscular habit of Snow Leopards in conformity to various other studies conducted on the species. Very few images were also taken from 8 am to noon. Similarly, a study on Snow Leopard conducted by Thinley [18] in Jigme Dorji National Park, Bhutan have also reported similar observation.

Click to enlarge

Habitat use inferred from camera-trap images indicate that Snow Leopards spent most of their time in the rugged rocky outcrops, a pattern which was generally observed in other range countries too [17]. A similar observation was made on estimating the species abundance and distribution in Jigme Dorji National Park using camera traps [18].

Distribution Pattern

Analysis of the capture history data showed that the three individuals JKSNR 1, 6 and 7 were captured in camera trap-station G2 (Tshonatop), G6 (Tshabjola/Peak Base) and G8 (Tshabjola), while individuals JKSNR 3, 4, 8 and 9 were captured only in camera trap-station G9 (Kampula), indicating high concentration and overlap of habitat among the individuals. Thus, the species could be mostly found in Kampula and Tshebjola ridges backed by the number of individuals photographed by camera-traps around these areas (Figure 24). The highest concentration of the species is in Grid Number 9 (Kampula) where as many as 4 individuals JKSNR 3, 4, 8 and 9 were captured followed by Grid Number 2 where at least two individuals JKSNR 1 and 6 roam around (Table 4). Therefore, JKSNR can safely conclude that most of its mountains above 3500masl has Panthera uncia as a primary predator. The Reserve, therefore, may focus more of their patrolling and monitoring activities in these high- density areas.

Click to enlarge