Ensuring the long-term persistence of snow leopards (Panthera uncia) in changing landscapes requires a deep understanding of their spatial ecology and movement behavior. To maintain viable metapopulations and ensure gene flow between populations, there is an urgent need to develop sound and effective conservation plans. This study presents findings from Nepal's first GPS telemetry study of snow leopards, shedding light on their home range size, habitat selection, and transboundary movements. GPS data were collected from four snow leopard individuals in the Kangchenjunga Conservation Area, eastern Nepal, over tracking periods ranging from 20 to 659 days, yielding a total of 4707 location points. We used three home range estimators for analysis: local convex hulls (LoCoH), fixed kernels (Kernel), and minimum convex polygons (MCP). Our results show that home range sizes were 6 to 97 times larger than previous estimates for Nepal, with LoCoH estimates of 310 and 102 km² (MCP = 730 and 211 km²) for two adult females and 312 km² (MCP = 1032 km²) for one adult male. Three snow leopards crossed international borders five to seven times, spending, on average, 10%–34% of their time in neighboring countries (China and India), with 28%–50% of their home ranges overlapping India. Our study demonstrates that snow leopards in Nepal have home ranges that are significantly larger than previously documented and frequently cross international borders. These extensive transboundary movements highlight the need for stronger coordination between Nepal, China, and India to ensure the long-term conservation of snow leopards in this key region of their distributional range.

摘要

要确保雪豹种群在不断变化的环境中的长期存续, 就必须深入了解其空间生态学特征及其运动行为。为维持雪豹区域种群的存续并确保种群间的基因流动, 迫切需要制定科学且有效的保护计划。本研究介绍了尼泊尔首次利用GPS遥测技术对雪豹进行追踪的研究结果, 揭示了雪豹的活动范围大小、栖息地选择及跨境迁徙行为。本研究在尼泊尔东部的Kangchenjunga保护区对四只雪豹进行了GPS追踪, 追踪时长从20天到659天不等, 共记录到4,707个位置点。研究使用了三种活动范围估算方法进行分析:局部凸壳法 (LoCoH) 、固定核密度法 (Kernel) 和最小凸多边形法 (MCP) 。研究结果显示, 雪豹的活动范围比尼泊尔之前的估算值大6到97倍。两只成年雌性雪豹的LoCoH估算活动范围分别为310 km²和102 km² (MCP法分别为730 km²和211 km²), 一只成年雄性雪豹的LoCoH估算活动范围为312 km² (MCP法为1032 km²) 。三只雪豹跨越国际边界5到7次, 在邻国 (中国和印度) 停留的时间平均占其总活动时间的10%到34%, 其活动范围有28%至50%位于印度境内。我们的研究表明, 尼泊尔雪豹的活动范围远超以往研究的估算, 并且经常跨越国际边界。雪豹的广泛跨境活动突出表明, 尼泊尔、中国和印度之间需要加强协调与合作, 以确保这一关键分布区域的雪豹能够长期存续。

通俗语言摘要

为了确保雪豹在不断变化的环境中的长期存续, 了解它们的移动模式和空间利用方式至关重要。本研究是首次在尼泊尔使用GPS追踪技术来研究雪豹的移动行为, 共追踪了尼泊尔东部Kangchenjunga保护区的四只雪豹, 监测时长从20到659天不等, 记录了4,707个位置点。研究结果显示, 尼泊尔雪豹的活动范围比以往研究估计的要大得多, 其中一只雄性雪豹的活动范围达到312 km², 两只雌性雪豹的活动范围分别为102 km²和310 km²。部分雪豹频繁跨越国际边界进入中国和印度, 最多有34%的时间在邻国活动, 并且有多达50%的活动区域位于邻国 (中国和印度) 。雪豹的这些跨境活动凸显了尼泊尔、中国和印度三国开展跨境合作的重要性, 以确保雪豹在其整个栖息地范围内都能得到有效保护。

Summary

To ensure the survival of snow leopards in changing environments, it is important to understand their movements and use of space. This study is the first to track snow leopards in Nepal using GPS collars, following four individuals in the Kangchenjunga Conservation Area in eastern Nepal for 20–659 days, recording 4707 locations. The findings show that snow leopards in Nepal have much larger home ranges than previously thought, with one male covering 312 km² and two females covering 102 and 310 km². Some of the snow leopards frequently crossed international borders into China and India, spending up to 34% of their time and overlapping up to 50% of their home range in neighboring countries. These cross-border movements emphasize the need for transboundary collaboration between Nepal, China, and India to ensure the effective conservation of snow leopards across their home ranges.

Practitioner Points
- ◦
  This study recorded the world's highest elevation for a snow leopard at 5,848 meters above sea level.
- ◦
  Home range estimates for snow leopards were 6 to 97 times larger than earlier estimates for Nepal: LoCoH estimates of 310 and 102 km² (MCP = 730 and 211 km²) for two adult females and 312 km² (MCP = 1032 km²) for one adult male.
- ◦
  Snow leopards frequently cross international borders between Nepal, India, and China, which highlights the need for transboundary collaboration to secure long-term conservation of the species in this key distribution region.

实践要点

本研究记录了雪豹活动的世界最高海拔, 达到海拔5,848米。
雪豹的活动范围估计值比尼泊尔之前的估计值大6到97倍:两只成年雌性雪豹的LoCoH估计值分别为310 km²和102 km² (MCP法分别为730 km²和211 km²), 一只成年雄性雪豹的LoCoH估计值为312 km² (MCP法为1032 km²) 。
雪豹频繁跨越尼泊尔、印度和中国之间的国际边界, 这说明为确保该物种在这一关键分布区域的长期保护, 各国有必要开展跨境合作。

1 Introduction

The globally threatened snow leopard (Panthera uncia) is one of the most elusive and least known of the large cats. Its range spans 1.2 to 1.6 million km² across the rugged and remote habitats of the central Asian mountains (McCarthy et al. 2017). These cats require large territories to meet their behavioral and ecological needs, as they reside in harsh, unproductive habitats (McCarthy et al. 2005; Johansson, Rauset, et al. 2016). However, their extensive habitat requirements present a significant conservation challenge, as they often overlap with human interests (Ikeda 2004). Protected areas alone are insufficient for ensuring the long-term viability of these felids unless they are connected, allowing safe genetic exchange within metapopulations (Johansson, Rauset, et al. 2016). Furthermore, securing connectivity is fundamental for adapting to the impacts of climate change, which is expected to fragment snow leopard habitats, particularly in lower-elevation areas of the Himalayas (Forrest et al. 2012).

In the Nepalese Himalayas, the snow leopard's range is generally considered to be separated by extensive, forested barriers or high mountain glaciers and peaks (MoFSC 2017). In 2013, Nepal, together with 11 other range countries, came together and committed to identifying and securing 23 snow leopard landscapes by 2020, ensuring community stewardship and strengthening transboundary cooperation (GSLEP 2013). Nepal identified three priority snow leopard landscapes and developed a 5-year national snow leopard conservation action plan (2017–2021) aligned with the goals of the Global Snow Leopard and Ecosystem Protection Program (GSLEP) (DNPWC 2017). Identifying and securing connectivity between these prime habitats is a prerequisite for landscape-level conservation and management.

The advent of advanced animal tracking systems, such as GPS telemetry, has provided new opportunities for studying the spatial ecology of elusive species like the snow leopard. This technology enables precise tracking that can provide vital information for conservation planning (Johansson, Simms, et al. 2016). Earlier tracking systems relied on Very High Frequency (VHF) telemetry, with findings often limited due to rugged and inaccessible mountainous terrain, which led to underestimated home range sizes (Jackson 1996; Oli 1997). VHF collars use unique high-frequency radio signals for animal tracking and identification within a limited range (Johansson, Simms, et al. 2016). Recent studies across the snow leopard distribution range (Afghanistan, Kyrgyzstan, Pakistan, and Mongolia), which have effectively employed Global Positioning System (GPS) telemetry that uses low-orbit satellite communications, indicate that VHF telemetry may not be the ideal method for giving precise and reliable home range estimates (Johansson, Simms, et al. 2016).

This study marks the first GPS-based telemetry research on snow leopards in Nepal. Our goal is to improve the understanding of their home range size, habitat preferences, and potential transboundary movements in northeastern Nepal. The findings from this study will be key for the development of national and regional management plans aimed at the long-term conservation of snow leopards.

2 Materials and Methods

2.1 Study Area

The Kangchenjunga Conservation Area (KCA) lies in the northeastern part of Nepal, within Taplejung District. It is bordered by two international protected areas: the Qomolangma National Nature Preserve (QNNP) in Xizang (Tibet, China) to the north and the Khangchendzonga National Park (KNP) in Sikkim (India) to the east (MoFSC 2017) (Figure 1). Established in 1998, KCA covers an area of 2035 km², with 65% of the area consisting of high mountain rock and ice. The remaining 35% is covered by forests (14.1%), shrublands (10.1%), grasslands (9.2%), and agricultural land (1.6%) (Thapa et al. 2021). KCA is also one of Nepal's three priority snow leopard landscapes (GSLEP 2015).

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Study area in the Kangchenjunga Conservation Area, Nepal, contiguous with protected areas in China (Qomolangma National Nature Preserve) and, India (Khangchendzonga National Park). The closed triangles denote the areas where the field work was conducted.

The study area covers three of the four habitat blocks within KCA—Yangma, Ghunsa, and Ramjer—since these blocks border China and India, allowing us to assess transboundary movement. The borders between Nepal, India, and China are relatively porous, with an open border policy between Nepal and India and no fencing along the Nepal-China border (MoFSC 2017). The altitudinal range of the areas extends from 1200 to 8586 m, resulting in diverse vegetation patterns. In the lower mid-hills, subtropical vegetation is found, while higher elevations are characterized by alpine grasslands. Forest types include mixed broadleaved forests between 1200 and 2800 m (comprising Larix griffithiana and Juniperus species), coniferous forests of Picea smithiana and Tsuga dumosa between 2800 and 3500 m, and Rhododendron (Rhododendron spp.) forests from approximately 3000 m to the tree line at around 3700 m (DNPWC 2018).

The alpine eco-region of KCA harbors large predators such as snow leopards, gray wolves (Canis lupus), and common leopards (Panthera pardus), with blue sheep (Pseudois nayaur) and a sparse distribution of marmots (Marmota himalayana) serving as their major prey. The conservation area is inhabited by approximately 500 households, primarily belonging to the largely Buddhist and Kirat communities of the Sherpa and Limbu ethnic groups. These Indigenous people raise mixed livestock herds (92% engaged in livestock rearing), with herds consisting of yak, cows, buffalo, dzo (a hybrid of yak and Tibetan cow), and mules (WWF 2017).

2.2 Capture and Collaring

We captured snow leopards during the spring (April and May) and autumn (October and November) seasons using modified Aldrich foot-hold snares (Johansson, Rauset, et al. 2016; Frank et al. 2003). Before the placement of the snares, we carried out camera trapping for 15 days to help identify potential sites for their placement (Jackson et al. 2006). Based on the camera trap data coupled with sign surveys, snares were strategically placed at snow leopard marking sites, movement trails, and kill sites (Jackson 1996). The capture periods lasted between 30 and 45 days, with kill sites used only when there was a fresh carcass and more than 25% of the carcass was undevoured. Each snare was connected to a trap transmitter (TBT 503-1, Telonics Inc., and TT3, Vectronic AEROSPACE), which transmitted radio signals to a communication tower (RA-6B VHF Antenna, Telonics Inc.) when triggered. Snares were monitored using a VHF receiver (R-1000, Telonics Inc.) every 2 h between 4 a.m. and 8 p.m. after being set. In addition, physical monitoring was carried out every 3 days to minimize disturbances at the snare sites.

Snow leopards were immobilized using a combination of medetomidine and telazol, administered via a Telinject air-pressured gun (McCarthy et al. 2010; Johansson et al. 2013). GPS collars (GPS Plus Globalstar, Vertex Survey Iridium, and Vertex Plus Iridium, all Vectronic AEROSPACE) were used. During immobilization, routine measurements such as weight, body dimensions, and vital rates were taken. The cats were revived by administering the antidote Antisedan. GPS collars were configured to transmit location fixes at 4-h intervals, yielding a total of six fixes per day. Each collared snow leopard was given a unique indigenous name representing the local habitat and culture.

The age of each collared snow leopard was estimated based on body weight, teeth coloration and wear, facial scars for males, and nipple size and pigmentation for females (McCarthy et al. 2010). Additionally, pelage texture and coloration were used to assist in aging the cats. Males with facial scars and females with darker and enlarged nipples were considered to be above 3 years old, whereas adult-sized cats without these features were considered to be 2 to 3 years old (Johansson, Rauset, et al. 2016).

2.3 Home Range Estimation

Home range estimates can be affected by the location frequency and duration of the GPS data collected (Borger et al. 2006). In our study, the GPS data for individual snow leopards varied in terms of both location frequency (fix rate < 100%) and duration (number of tracking days). To standardize the data, we used GPS fixes with a Dilution of Precision (DOP) value of less than 10 and validated 3D positions to remove inaccurate locations (Lewis et al. 2007). We also randomly selected four GPS fixes per individual per day to reduce autocorrelation. We removed data collected in the first 4 days after collaring to avoid potential bias related to capture (Farhadinia et al. 2019). Furthermore, we removed 897 locations for one cat (Lapchhemba) during dispersal to avoid including extra range movements (Johansson, Rauset, et al. 2016). We used net squared displacement (Bunnefeld et al. 2010) with the R package “adehabitatHR” (Calenge 2006) to identify temporal breakpoints for estimating home range instead of relying on preconceived annual home ranges (Johansson, Rauset, et al. 2016).

We applied three different home range estimators to calculate the home ranges: minimum convex polygon (MCP), fixed kernel (Kernel), and local convex hull (LoCoH). While MCP is the most commonly used estimator and Kernel is generally considered more accurate, both have been found to include areas that are not actually used by the animal (Getz and Wilmers 2004; Borger et al. 2006). LoCoH has been shown to provide more accurate delineations of the areas used by animals, especially when there are hard borders (Getz et al. 2007). We calculated home ranges using all three methods to allow for comparison with earlier studies that used these estimators. Moreover, we limited our home range calculations to snow leopards that were tracked for more than 3 months (n = 3 cats), following Johansson, Rauset, et al. (2016).

For the home range calculations, we used the “adehabitatHR” package in R (Calenge 2006) to calculate 95% MCPs, 95% kernel (using bivariate normal smoothing curve and h_REF *0.6) (Johansson, Rauset, et al. 2016), and the “TLoCoH” package in R (Lyons et al. 2013) to estimate adaptive local convex hulls (aLoCoH). According to Johansson, Rauset, et al. (2016), static home ranges were more focused than spatiotemporal dynamics (s = 0), as we were more interested in the overall space used by the animals. Since we were unsure about which range of “a” parameters should be applied, we followed heuristic guidelines and plotted the range of values for ‘a’ to examine different home range generations using this method (Getz et al. 2007). By assessing the extent of known holes generated by a range of “a” values (MSHC—Minimum Spurious Hole Covering) and considering both Type I and Type II errors, we selected the “a” value that best described the home range utilized by the individuals (Getz et al. 2007).

2.4 Transboundary Movements and Transboundary Home Range Overlap

We used the number of times the collared cats crossed the borders into China and India as our measure of transboundary movements, where we identified the number of border crossings by converting the collar data to travel lines using the “Points to Lines” tool in ArcGIS ver. 10.8 (ESRI 2019). We used all corrected GPS positions when calculating the movement paths to ensure that no potential crossings were missed. We also calculated how much the aLoCoH home ranges for the three cats overlapped with those in Nepal, China, and India. We chose aLoCoH estimates for our calculations because they have been shown to be more accurate in delineating the areas used by animals, especially in rugged terrain (Getz et al. 2007).

2.5 Habitat Selection

We examined the habitat selection of snow leopards by comparing the proportion of cat locations within different habitat types to the available proportion of each habitat type in the study area (Marcum and Loftsgarden 1980). Habitat types were defined based on land cover (alpine steppe and scrub, barren land, forest, snow/ice, and waterbody; Karra 2021) and terrain features, including slope, ruggedness, aspect, and elevation (Jackson and Hunter 1996). Data from all four GPS-collared snow leopards (n = 4011) were pooled for the analysis. Due to the small sample size and the fact that the cats followed were followed in different seasons, more detailed analyses were not feasible.

To determine habitat availability within the study area, we generated 8000 random points using ArcGIS ver. 10.8 (Marcum and Loftsgarden 1980; ESRI 2019). Habitat selection was analyzed using Ivlev (1961) electivity index, where positive values indicate habitat selection, negative values indicate avoidance, and values close to 0 indicate neutral selection. We tested for selection following Marcum and Loftsgarden (1980) by conducting χ² goodness-of-fit tests at a 0.05 significance level for each habitat type individually and applied Bonferroni corrections for multiple comparisons within habitat categories. We could not perform compositional analyses following Aebischer et al. (1993) due to the small sample size (n = 4 collared cats).

Land cover data were obtained from layers derived from Sentinel-2 imagery at a 10 m resolution (Karra 2021), and terrain features were calculated in ArcGIS ver. 10.8 (ESRI 2019) using a digital elevation model (DEM) layer of 12.5 m from JAXA/METI (2007). Ruggedness indices were calculated using code from Riley et al. (1999) and classified according to the method described in the Snow Leopard Information Management System (Jackson and Hunter 1996), with minor modifications for our study area.

3 Results

Four snow leopards were collared between November 2013 and May 2017 (Table 1), with a capture success rate of five captures in 3,004 trap nights. One adult male (Name: Ghangjenjwenga) was collared twice, while another adult male (Name: Omi Khangri) stopped transmitting locations after 20 days. One adult female (Name: Lapchhemba) exhibited dispersal behavior for 229 tracking days before settling in a new home range. The second adult female (Yalung) was followed for 167 days.

Table 1. Details of GPS collared snow leopards in Kangchenjunga Conservation Area (KCA), northeastern Nepal, from 2013 to 2017.

Snow leopard	Collar ID	Tracking period (days)	Sex	Estimated age	Body weight (kg)	Body measurement (cm)	Paw measurement (cm)	Capture date	Collar status
Yalung	21755	167	F	Adult 3 years	30	Total length: 165, Tail length: 85, Shoulder height: 57	Fore limbs: 13 × 12, Hind limbs: 9 × 6	May 8 2017	GPS stopped functioning on October 22, 2017
Lapchhemba	21245	392	F	Adult 3 years	30	Total length: 173, Tail length: 93, Shoulder height: 63	Fore limbs: 10 × 10, Hind limbs: 9 × 8	April 27 2016	Collar stopped functioning on May 24, 2017, with collar dropped
Omi Khangria	17103	20	M	Adult 5 years	41	Total length: 196, Tail length: 97, Shoulder Height: 70	Fore limbs: 10 × 7, Hind limbs: N/A	May 20 2015	GPS stopped functioning on June 10, 2015, with collar still intact
Ghangjenjwenga	13646 and 13647	659	M	Adult 5 years	40	Total length: 191, Tail length: 93, Shoulder height: 60	Fore limbs: 9 × 10 Hind limbs: 8.5 × 7.5	November 25, 2013 And May 15, 2014	GPS (ID 13647) stopped functioning on September 15, 2015, with collar dropped

a Satellite signal lost but inbuilt radio VHF performed for 549 days.

We obtained a total of 4707 locations over 1239 tracking days from the four collared snow leopards. After the removal of erroneous locations, the final data set included between 63 and 1834 locations per individual. For the home range estimates, we excluded potential capture bias (n = 22), dispersal movements by Lapchhemba (n = 1130), and the positions for Omi Khangri, who was followed for only 20 days (n = 63). Additionally, we randomly selected four locations per cat per day. The final data set for home range estimation included 2386 locations from 697 tracking days. The GPS fixation rates for the five collars ranged from 49% to 98%, representing the proportion of received locations to the total number of programmed locations.

3.1 Home Range Size

Home ranges estimated by MCP and Kernels were two to three times larger than those obtained using aLoCoH (Table 2, Figure 2). The MCP and Kernel estimates were relatively similar in size for the female Lapchhemba and the male Ghangjenjwenga, whereas the Kernel estimate for the female Yalung was considerably larger than the MCP estimate. The home range size estimated by aLoCoH ranged between 102 and 312.2 km², with an average of 241.4 km² (SD ± 120.7). There was no apparent difference in home range size between the sexes (Table 2).

Table 2. Home range sizes of three GPS collared snow leopards in Kangchenjunga Conservation Area, Nepal between 2013 to 2017.

Snow leopard	Collar ID	n	MCP (95%), km²	Kernel (95%), km²	aLoCoH (95%), km²	a*
Yalung (adult female)	21755	752	730.1	1043.9	309.9	45
Lapchhemba (adult female)	21245	651	210.7	235.5	102	39
Ghangjenjwenga (adult male)	13647	983	1031.6	1053.1	312.2	49

* adaptive distance threshold for selecting points in local convex hulls (Getz et al. 2007).

3.2 Transboundary Movements

Over tracking periods ranging from 167 to 659 days, the three snow leopards crossed international borders between five and seven times, each spending 10.3% to 34.3% of their time in China and India (Table 3, Figure 3). Transboundary movement of the female Lapchhemba into China was observed during her dispersal period, but once she settled in her home range in Nepal, no further crossings were observed. During her exploratory movements, Lapchhemba spent between two and 37 days in China (Qomolangma National Nature Preserve). Both male Ghangjenwenga and female Yalung crossed into India, entering Khangchendzonga National Park, where they spent between 0.5 and 21 days. Of the total home range (aLoCoH) for Ghangjenjwenga and Yalung, 28.2% and 49.7%, respectively, appeared to fall across the border in India (Table 3, Figure 3).

Table 3. Transboundary movements, proportion of time spent in neighboring countries, and home range (aLoCoH 95%) overlap with neighboring countries for the three snow leopards from 2013 to 2017.

Snow leopard	Neighboring country	Transboundary crossing	Date of crossing	Duration of length (days)	Time spent in neighboring country (%)	Home range overlap with neighboring country (%)
Yalung	India	1	5/28/2017	1/2	34.3	49.7
		2	6/8/2017	21
		3	7/22/2017	10
		4	8/11/2017	5
		5	9/30/2017	20
		6	10/21/2017	1
				Total: 57½
Lapchhemba	China	1	7/1/2015	31	23.7	0a
		2	8/13/2016	6
		3	9/15/2016	2
		4	9/30/2016	37
		5	11/14/2016	17
				Total: 93
Ghangjenjwenga	India	1	1/11/2014	20	10.3	28.2
		2	4/25/2014	7
		3	10/1/2014	11
		4	1/21/2015	14
		5	4/4/2015	4
		6	5/12/2015	6
		7	9/4/2015	6
				Total: 68

a No transboundary movements once Lapchhemba settled after initial dispersal movements.

3.3 Habitat Selection

The snow leopards exhibited selective habitat use, favoring certain habitat types disproportionately to their availability in the study area (Table 4). The habitat selectivity indices for the pooled locations showed that alpine steppe and scrub were the most preferred habitats, while forest and snow/ice habitats were the most avoided (Table 4). In terms of terrain, snow leopards favored rolling and very broken terrains on southwest- to east-facing slopes (15°–50°) more than anticipated, while they avoided north-facing cliffs and slopes greater than 50° (Table 4). Regarding elevation, snow leopards preferred areas between 4000 and 5000 m, strongly avoiding elevations below 3500 m and above 5500 m. The highest elevation recorded was 5858 m (the male Ghangjenjwenga), and the lowest elevation was 2040 m (the female Lapchhemba).

Table 4. Habitat selection for four snow leopards in Kangchenjunga Conservation Area (Nepal), Qomolangma Nature Preserve (China), and Khangchendzonga National Park (India).

Habitat variables		Index of electivity	X²	p	Bonferroni Simultaneous Confidence Interval (Z = 0.985)
Habitat variables		Index of electivity	X²	p	P RP	P SL	CI
Landcover	Alpine steppe and scrub	0.22	2344.4	< 0.05	0.29	0.63	−0.34	−0.34
	Barrenland	0.07			0.29	0.36	−0.07	−0.07
	Forest	−0.90			0.18	0.01	0.18	0.18
	Snow/Ice	−0.89			0.24	0.01	0.23	0.23
	Waterbody	−0.46			0.01	0.00	0.00	0.00
Ruggedness	Cliff	−0.10	61.6	< 0.05	0.10	0.07	0.02	0.02
	Very broken	0.02			0.42	0.44	−0.02	−0.02
	Moderately rugged	−0.09			0.15	0.12	0.03	0.03
	Rolling	0.04			0.33	0.37	−0.04	−0.04
	Slightly broken	−0.43			0.01	0.00	0.00	0.00
Aspect	Flat	−0.26	864.8	< 0.05	0.00	0.00	0.00	0.00
	North	−0.52			0.09	0.02	0.07	0.07
	Northeast	−0.34			0.11	0.04	0.07	0.07
	East	0.02			0.11	0.12	−0.01	−0.01
	Northwest	−0.37			0.11	0.04	0.07	0.07
	South	0.15			0.14	0.24	−0.09	−0.09
	Southeast	0.18			0.13	0.24	-0.11	−0.11
	Southwest	0.06			0.17	0.21	−0.04	−0.04
	West	−0.14			0.13	0.09	0.04	0.04
Slope (°)	0–15	−0.09	55.4	< 0.05	0.15	0.12	0.04	0.04
	15–30	0.04			0.33	0.37	−0.04	−0.04
	30–50	0.02			0.42	0.44	−0.02	−0.02
	50–90	−0.10			0.10	0.07	0.02	0.02
Elevation	< 3500	−0.86	2543.9	< 0.05	0.15	0.01	0.15	0.15
	3500–3750	−0.68			0.04	0.00	0.03	0.03
	3750–4000	−0.21			0.04	0.02	0.02	0.02
	4000–4250	0.23			0.08	0.19	-0.11	−0.11
	4250–4500	0.23			0.10	0.23	−0.13	−0.13
	4500–4750	0.25			0.10	0.26	−0.15	−0.15
	4750–5000	0.11			0.12	0.17	−0.05	−0.05
	5000–5250	0.00			0.09	0.09	0.00	0.00
	5250–5500	−0.48			0.08	0.02	0.06	0.06
	> 5500	−0.89			0.19	0.01	0.18	0.18

Note: P RP and P SL denote the proportions of the locations in the study area; CI is the Bonferroni confidence interval, where both negative values denote that snow leopards use more than in proportion to availability, and vice versa.

4 Discussion

Our study, the first GPS telemetry study to track snow leopards in Nepal, provides valuable data that complement earlier VHF-based studies conducted in the 1980s and 1990s in Mugu and Manang (Jackson 1996; Oli 1997). Although the small sample size necessitates cautious interpretation, our study offers improved insights into the spatial ecology of snow leopards in Nepal, with important implications for conservation management. Notably, our findings include: (1) home range estimates that are 6 to 97 times larger than previous VHF-based estimates for Nepal when comparing similar home range estimators (MCP, Jackson 1996; Oli 1997), (2) considerable transboundary movements, signifying the need for international collaboration, and (3) the world's highest recorded elevation for snow leopards at 5,848 meters.

4.1 Home Range Size

Using three different estimators (GPS-based: MCP, Kernel, and aLoCoH), we found that the home range of snow leopards was substantially larger than earlier estimates of 11 to 36 km² for Nepal (VHF-based: MCP, Jackson 1996; Oli 1997). Both Jackson (1996) and Oli (1997) acknowledged the limitations of VHF-based studies, and Johansson, Simms, et al. (2016) further supported this, noting that for low-density, wide-ranging species inhabiting inaccessible terrains, GPS-based methods prove to be much more effective.

Our study revealed differences in home range estimates based on the three methods. GPS-based studies using the aLoCoH estimator have been shown to provide more biologically accurate home range estimates than MCP or Kernel, as the latter often contain areas that were never actually visited by the collared animals (Getz et al. 2007; Johansson, Rauset, et al. 2016). In the Kangchenjunga Conservation Area, the habitat consists of valley gorges and high mountain ridges, where the ridges and low-lying forests act as barriers for snow leopards. Our results, particularly from the aLoCoH estimates, align more closely with the actual GPS locations of snow leopards, which tended to be concentrated in valley gorges and mountain passes used by our snow leopards. Based on these findings, we concur with Johansson, Rauset, et al. (2016) that aLoCoH estimates appear to be more biologically appropriate for snow leopards. However, recent advances in home range analyses using Autocorrelated Kernel Density Estimates (AKDE) have also demonstrated improved estimates that explicitly account for autocorrelation (Calabrese et al. 2021. The integration of AKDE with current methods could be considered for future studies of snow leopard home range.

Comparing the aLoCoH home range size from our study with those from other snow leopard range countries using GPS collars shows that our estimates are consistent with those from Mongolia (114–394 km² (n = 9) for adult males and 75–193 km² (n = 7) for adult females; Johansson, Rauset, et al. 2016) and China (174–261 km² (n = 2) for adult females; Yu et al. 2022).

4.2 Transboundary Movement

Our study demonstrates that snow leopard movements frequently extend across international borders, with individuals traveling into both China and India. Snow leopard habitats in the region stretch into India's Khangchendzonga National Park and China's Qomolangma National Nature Preserve (MoFSC 2017), both of which share similar geospatial features. This continuity suggests that frequent crossings and home range overlap are likely in such contiguous regions, which are not hindered by anthropogenic barriers.

For the female snow leopard, Lapchhemba, her border crossings appeared primarily driven by the search for suitable habitats for establishing her home range. Conversely, the territories of Ghangjenjwenga and Yalung spanned both sides of the international boundary, indicating that their crossings were motivated by the need to cover their entire home range. Ghangjenjwenga's lack of border crossing during the summer months could be attributed to a larger proportion of his home range within Nepal (71.8%). This season coincides with the parturition period of its wild and domestic prey (Schaller 1977), leading to higher prey densities, particularly blue sheep lambs and yak calves, which are preferred by snow leopards (Thapa et al. 2021). These increased prey densities may have reduced the need for Ghangjenjwenga to move into the smaller proportion of his home range in India.

While our findings provide insightful information into the motivations behind snow leopard border crossings, further investigation into the density and distribution of prey, the influence of competing predators, and herder activities across borders would help refine our understanding of the drivers behind these long-distance movements.

In recent decades, telemetry studies have generated crucial information on the spatial ecology of large-ranging carnivores like the snow leopard, information that is needed for landscape planning and the development of reliable and effective conservation plans (Woodroffe and Ginsberg 2000; Johansson, Rauset, et al. 2016). The conservation of these species must extend beyond a single protected area, as individual protected areas alone cannot sustain wide-ranging carnivores (Johansson, Rauset, et al. 2016). It is also necessary to account for transboundary movements to ensure the long-term viability of the snow leopard. For example, almost half of the protected areas within the snow leopard's global distribution range are too small to support even a single reproducing pair (Johansson, Rauset, et al. 2016). Our study revealed that snow leopard movements extend beyond international borders, with cats traveling into China and India. This evidence further reinforces the need to integrate transboundary conservation approaches into national action plans (DNPWC 2017; DNPWC and DoFSC 2024) and snow leopard landscape management plans (SLEMP; MoFSC 2017). Nepal's snow leopard conservation action plan (2024–2030) stresses the importance of strengthening transboundary coordination and regional and international cooperation for effective snow leopard conservation. The plan includes a 10% increase in the estimated 5-year budget allocation, up from 6% in the previous plan (DNPWC 2017; DNPWC and DoFSC 2024). Our study highlights the critical issue of transboundary conservation efforts among Nepal, China, and India and further validates the transboundary movement capacity of snow leopards, a species that requires large spaces to thrive.

4.3 Habitat Selection

Our study showed that forested areas and cliff terrains above 5500 m were strongly avoided by snow leopards, suggesting that these areas may act as natural barriers that limit their movement and distribution within the eastern snow leopard landscape of Nepal. Similarly, approximately half of the Kangchenjunga Conservation Area (KCA) has cliff terrains greater than 5000 m with steep slopes greater than 50°. As a result, snow leopards were confined to the narrow V-shaped valleys and river gorges, which are dominated by broken and rolling terrains of alpine steppe and alpine scrub lands.

We also report the highest elevation record for snow leopards to date at 5858 m. The preferred elevation range of 4000 to 5000 m observed in our study was higher than that reported in earlier studies from the Mugu district in the western landscape of Nepal (< 4000 m) (Jackson 1996). This difference in altitudinal use can probably be attributed to biophysical aspects such as variations in tree lines and mountain range elevations between the eastern and western landscapes. In the eastern landscape, the tree line and mountain ranges generally start at higher elevations compared to those in the western landscape.

5 Conclusions

Our study presents findings on the spatial ecology of snow leopards in eastern Nepal and highlights the large spatial requirements and transboundary movement capacity of these cats. Despite the small sample size, our study revealed larger home ranges for snow leopards than earlier estimates for Nepal. Similarly, we observed considerable transboundary movements, highlighting the urgency of snow leopard conservation in Nepal through transboundary collaboration among the neighboring countries of China and India. The Global Snow Leopard and Ecosystem Protection Program's (GSLEP) goal of securing 100 breeding adult snow leopards across 23 snow leopard landscapes, alongside Nepal's own commitment to the Snow Leopard Conservation Action Plan (SLCAP) 2024–2030 and the Snow Leopard Ecosystem Management Plan (SLEMP) 2017–2026, are important steps toward ensuring the long-term viability and continuity of snow leopard conservation. Ongoing research and conservation efforts within protected areas, community-based management, and transboundary coordination are imperative for the continued conservation of snow leopard populations in Nepal.

Author Contributions

Samundra Subba: conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, software, supervision, visualization, writing – original draft, writing – review and editing. Hem Raj Acharya: project administration, supervision, writing – review and editing. Sheren Shrestha: conceptualization, project administration, supervision, writing – review and editing. Saroj Koirala: formal analysis, software. Rinjan Shrestha: conceptualization, investigation, methodology, resources, writing – review and editing. Gokarna Jung Thapa: data curation, formal analysis, resources, software, writing – review and editing. Kamal Thapa: conceptualization, investigation, methodology, project administration, resources. Anil Shrestha: investigation, project administration, writing – review and editing. Sabita Malla: conceptualization, methodology, supervision, writing – review and editing. Gopal Prakash Bhattarai: conceptualization, project administration, resources, validation, writing – review and editing. Laxman Prasad Poudyal: project administration, supervision, writing – review and editing. Man Bahadur Khadka: conceptualization, project administration, resources, validation, writing – review and editing. Ghana Shyam Gurung: conceptualization, funding acquisition, writing – review and editing. Shiv Raj Bhatta: funding acquisition, project administration, writing – review and editing. Naresh Subedi: project administration, writing – review and editing. Maheshwar Dhakal: conceptualization, project administration, resources, validation, writing – review and editing. Narendra Man Babu Pradhan: conceptualization, funding acquisition, project administration. Ananta Ram Bhandari: conceptualization, funding acquisition; project administration. Shant Raj Jnawali: conceptualization, methodology, writing – review and editing. Khagendra Phembu Limbu: project administration, supervision. Bed Kumar Dhakal: conceptualization, project administration, resources, validation, writing – review and editing. Kanchan Thapa: conceptualization, funding acquisition, methodology, project administration, resources, supervision, validation, visualization, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Open Research

Data Availability Statement

The GPS datasets of the collared snow leopards from the current study are government owned and are restricted for public availability due to high sensitivity of threatened species and under Nepal's protected list of mammals under National Parks and Wildlife Conservation Act 1973. They can be made available from the corresponding author upon reasonable request and permission from the government of Nepal. The satellite layers used for habitat selection analyses (digital elevation model and Sentinel-2 imagery) are freely available at ALOS PALSAR-https://search.asf.alaska.edu/ [7] and ESRI-https://www.arcgis.com/apps/instant/media/index.html?appid=fc92d38533d440078f17678ebc20e8e2 [8] and repositories, respectively.

References

Aebischer, N. J., P. A. Robertson, and R. E. Kenward. 1993. “Compositional Analyses of Habitat Use From Animal Radio-Tracking Data.” Ecology 75: 1313–1325. https://doi.org/10.2307/1940062.
10.2307/1940062
Web of Science® Google Scholar
Borger, L., N. Franconi, G. De Michele, et al. 2006. “Effects of Sampling Regime on the Mean and Variance of Home Range Size Estimates.” Journal of Animal Ecology 75: 1393–1405. https://doi.org/10.1111/j.1365-2656.2006.01164.x.
10.1111/j.1365-2656.2006.01164.x
CAS PubMed Web of Science® Google Scholar
Bunnefeld, N., L. Borger, B. Van Moorter, et al. 2010. “A Model-Driven Approach to Quantify Migration Patterns: Individual, Regional and Yearly Differences.” Journal of Animal Ecology 80: 466–476. https://doi.org/10.1111/j.1365-2656.2010.01776.x.
10.1111/j.1365-2656.2010.01776.x
PubMed Web of Science® Google Scholar
Calabrese, J. M., C. H. Fleming, M. J. Noonan, and X. Dong. 2021. “Ctmmweb: A Graphical User Interface for Autocorrelation-Informed Home Range Estimation.” Wildlife Society Bulletin 45, no. 1: 162–169. https://doi.org/10.1002/wsb.1154.
10.1002/wsb.1154
Web of Science® Google Scholar
Calenge, C. 2006. “The Package “adehabitat” for the R Software: A Tool for the Analysis of Space and Habitat Use by Animals.” Ecological Modelling 197: 516–519. https://doi.org/10.1016/j.ecolmodel.2006.03.017.
10.1016/j.ecolmodel.2006.03.017
Web of Science® Google Scholar
DNPWC. 2017. Snow Leopard Conservation Action Plan (2017-2021). Department of National Parks and Wildlife Conservation.
Google Scholar
DNPWC. 2018. “ Kanchenjunga Conservation Area.” http://dnpwcgovnp/protected_areas/details/kanchenjungaconservationarea.
Google Scholar
DNPWC and DoFSC. Snow Leopard Conservation Action Plan for Nepal (2024-2030). 2024. Department of National Parks and Wildlife Conservation and Department of Forests and Soil Conservation.
Google Scholar
Environmental Systems Research Institute (Esri), Inc., 2019. “ArcGIS Online.” Accessed July 15, 2019. https://www.esri.com/en-us/arcgis/products/arcgis-online/overview.
Google Scholar
Farhadinia, M. S., D. R. Heit, R. A. Montgomery, et al. 2019. “Vertical Relief Facilitates Spatial Segregation of a High Density Large Carnivore Population.” Oikos 129: 346–355. https://doi.org/10.1111/oik.06724.
10.1111/oik.06724
Web of Science® Google Scholar
Forrest, J. L., E. Wikramanayake, R. Shrestha, et al. 2012. “Conservation and Climate Change: Assessing the Vulnerability of Snow Leopard Habitat to Treeline Shift in the Himalaya.” Biological Conservation 150: 129–135. https://doi.org/10.1016/j.biocon.2012.03.001.
10.1016/j.biocon.2012.03.001
Web of Science® Google Scholar
Frank, L., D. Simpson, and R. Woodroffe. 2003. “Foot Snares: An Effective Method for Capturing African Lions.” Wildlife Society Bulletin 31, no. 1: 309–314.
Web of Science® Google Scholar
Getz, W. M., S. Fortmann-Roe, P. C. Cross, A. J. Lyons, S. J. Ryan, and C. C. Wilmers. 2007. “Locoh: Nonparametric Kernel Methods for Constructing Home Ranges and Utilization Distributions.” PLoS ONE 2: e207. https://doi.org/10.1371/journal.pone.0000207.
10.1371/journal.pone.0000207
PubMed Web of Science® Google Scholar
Getz, W. M., and C. C. Wilmers. 2004. “A Local Nearest-Neighbor Convex-Hull Construction of Home Ranges and Utilization Distributions.” Ecography 27: 489–505. https://doi.org/10.1111/j.0906-7590.2004.03835.x.
10.1111/j.0906-7590.2004.03835.x
Web of Science® Google Scholar
GSLEP. 2013. Global Snow Leopard & Ecosystem Protection Program: A New International Effort to Save the Snow Leopard and Conserve High-Mountain Ecosystems. Global Tiger Initiative Secretariat.
Google Scholar
GSLEP. 2015. Global Snow Leopard & Ecosystem Protection Program (GSLEP): First Steering Committee Meeting (SCM). Global Snow Leopard Conservation Forum. https://globalsnowleopard.org/wp-content/uploads/2020/09/SL_Report_2015_EN_July_03.pdf.
Google Scholar
Ikeda, N. 2004. “Economic Impacts of Livestock Depredation by Snow Leopard Uncia Uncia in the Kanchenjunga Conservation Area, Nepal Himalaya.” Environmental Conservation 31: 322–330. https://doi.org/10.1017/S0376892904001778.
10.1017/S0376892904001778
Web of Science® Google Scholar
Ivlev, V. S. 1961. Experimental Ecology of Feeding Fishes. Yale University Press.
Google Scholar
Jackson, R., and D. O. Hunter. 1996. Snow Leopard Information Management Handbook. International Snow Leopard Trust.
Google Scholar
Jackson, R. M. 1996. “Home Range, Movements and Habitat Use of Snow Leopard (Uncia uncia) in Nepal.” PhD thesis, University of London.
Google Scholar
Jackson, R. M., J. D. Roe, R. Wangchuk, and D. O. Hunter. 2006. “Estimating Snow Leopard Population Abundance Using Photography and Capture-Recapture Techniques.” Wildlife Society Bulletin 34, no. 3: 772–781.
10.2193/0091-7648(2006)34[772:ESLPAU]2.0.CO;2
Web of Science® Google Scholar
JAXA/METI. 2007. “ALOS PALSAR L 10, Datasets.” https://search.asf.alaska.edu.
Google Scholar
Johansson, O., J. Malmsten, C. Mishra, P. Lkhagvajav, and T. McCarthy. 2013. “Reversible Immobilization of Free-Ranging Snow Leopards (Panthera uncia) With a Combination of Medetomidine and Tiletamine-Zolazepam.” Journal of Wildlife Diseases 49: 338–346. https://doi.org/10.7589/2012-02-049.
10.7589/2012-02-049
PubMed Web of Science® Google Scholar
Johansson, O., G. R. Rauset, G. Samelius, et al. 2016. “Land Sharing Is Essential for Snow Leopard Conservation.” Biological Conservation 203: 1–7. https://doi.org/10.1016/j.biocon.2016.08.034.
10.1016/j.biocon.2016.08.034
Web of Science® Google Scholar
Johansson, O., A. Simms, and T. McCarthy. 2016. “ From VHF to Satellite GPS Collars—Advancements in Snow Leopard Telemetry.” In Snow Leopards: Biodiversity of the World: Conservation From Genes to Landscapes, edited by T. McCarthy and D. Mallon, 355–365. Elsevier, Academic Press. https://doi.org/10.1016/B978-0-12-802213-9.00026-2.
Google Scholar
Karra, K., C. Kontgis, Z. Statman-Weil, J. C. Mazzariello, M. Mathis, and S. P. Brumby. 2021. “Global Land Use/Land Cover With Sentinel-2 and Deep Learning.” In IGARSS 2021-2021 IEEE International Geoscience and Remote Sensing Symposium [Data Sets]. http://wwwarcgiscom/apps/instant/media/indexhtml?appid=fc92d38533d440078f17678ebc20e8e2.
Google Scholar
Lewis, J. S., J. L. Rachlow, E. O. Garton, and L. A. Vierling. 2007. “Effects of Habitat on GPS Collar Performance: Using Data Screening to Reduce Location Error.” Journal of Applied Ecology 44: 663–671. https://doi.org/10.1111/j.1365-2664.2007.01286.x.
10.1111/j.1365-2664.2007.01286.x
Web of Science® Google Scholar
Lyons, A. J., W. C. Turner, and W. M. Getz. 2013. “Home Range Plus: A Space-Time Characterization of Movement Over Real Landscapes.” Movement Ecology 1: 2. https://doi.org/10.1186/2051-3933-1-2.
10.1186/2051-3933-1-2
PubMed Google Scholar
Marcum, C. L., and D. O. Loftsgarden. 1980. “A Nonmapping Technique for Studying Habitat Preferences.” Journal of Wildlife Management 44: 963–968. https://doi.org/10.2307/3808336.
10.2307/3808336
Web of Science® Google Scholar
McCarthy, T., D. Mallon, R. Jackson, P. Zahler, and K. McCarthy. 2017. “Panthera uncia The IUCN Red List of Threatened Species.” e. T22732A50664030. https://dxdoiorg/102305/IUCNUK2017-2RLTST22732A50664030en.
Google Scholar
McCarthy, T., K. Murray, K. Sharma, and O. Johansson. 2010. “Preliminary Results of a Long-Term Study of Snow Leopards in South Gobi, Mongolia.” Cat News 53: 15–19.
Google Scholar
McCarthy, T. M., T. K. Fuller, and B. Munkhtsog. 2005. “Movement and Activities of Snow Leopards in Southwestern Mongolia.” Biological Conservation 124: 527–537. https://doi.org/10.1016/j.biocon.2005.03.003.
10.1016/j.biocon.2005.03.003
Google Scholar
MoFSC. 2017. Snow Leopard and Ecosystem Management Plan (2017-2026). Ministry of Forests and Soil Conservation.
Google Scholar
Oli, M. K. 1997. “Winter Home Range of Snow Leopards in Népal.” Mammalia 61: 355–360.
10.1515/mammalia-1997-610303
PubMed Web of Science® Google Scholar
Riley, S. J., S. D. DeGloria, and R. Elliot. 1999. “A Terrain Ruggedness Index That Quantifies Topographic Heterogeneity.” Intermountain Journal of Sciences 5, no. 1–4: 23–27.
Google Scholar
Schaller, G. B. 1977. Mountain Monarchs. Wild Sheep and Goats of the Himalaya. The University of Chicago Press.
Google Scholar
Thapa, K., N. Schmitt, N. M. B. Pradhan, H. R. Acharya, and S. Rayamajhi. 2021. “No Silver Bullet? Snow Leopard Prey Selection in Mt. Kangchenjunga, Nepal.” Ecology and Evolution 11, no. 23: 16413–16425. https://doi.org/10.1002/ece3.8279.
10.1002/ece3.8279
PubMed Web of Science® Google Scholar
Woodroffe, R., and J. R. Ginsberg. 2000. “ Ranging Behaviour and Vulnerability to Extinction in Carnivores.” In Behaviour and Conservation, edited by L. M. Gosling and W. J. Sutherland, 125–140. Cambridge University Press.
Google Scholar
WWF. 2017. Socio-Economic Outcomes of Conservation Interventions in Kangchenjunga Conservation Area, Nepal. WWF Nepal. https://wwfasia.awsassets.panda.org/downloads/socio_economic_outcomes_kca_2017.pdf.
Google Scholar
Yu, C., N. Ding, Y. Li, et al. 2022. “Preliminary Results From Applying Satellite-Tracking on Snow Leopards for the First Time in China.” Global Ecology and Conservation 40: e02346. https://doi.org/10.1016/j.gecco.2022.e02346.
10.1016/j.gecco.2022.e02346
Web of Science® Google Scholar

Volume4, Issue1

March 2025

Pages 45-56

This article also appears in:

Snow leopard – a large cat with high potential for conservation success

Lessons for Transboundary Snow Leopard Conservation: Findings From a GPS Telemetry Study in Kangchenjunga Conservation Area, Nepal

跨境雪豹保护经验:来自尼泊尔Kangchenjunga保护区的GPS追踪研究结果

ABSTRACT

摘要

通俗语言摘要

Summary

实践要点

1 Introduction