Data from: From shadows to data: First robust population assessment of snow leopards in Pakistan
Data files
Oct 22, 2025 version files 27.47 KB
-
README.md
2.94 KB
-
SCR_model_fitting.R
14.36 KB
-
sensitivity_all.R
10.17 KB
Abstract
The snow leopard (Panthera uncia) is a flagship species of the greater Himalayan region and symbolizes the integrity of this ecological system. Within the greater Himalayas, Pakistan holds special significance as the north of the country represents a confluence of four major mountain ranges (Hindu Kush, Pamir, Karakoram, and Himalaya). However, robustly surveying and monitoring elusive, low-density species such as snow leopards has historically been difficult in the region. As a result, our understanding of the spatial patterns in density and overall population size of snow leopards has remained conjectural in Pakistan. This lack of objective information is an obstacle to realizing effective conservation planning for the species in Pakistan, as well as the broader ecosystem within which it plays a key role. This study aimed to empirically derive population estimates for snow leopards in Pakistan, based on extensive camera trapping conducted over a decade (2010 to 2019), covering about 39 % of the species' range across four major mountain ranges in northern Pakistan. A total of 828 cameras were placed over 26,540 trap days, resulting in 4,712 photos of snow leopards obtained from 65 different locations. Among the 53 unique individuals identified, the majority (53 %) were detected only once, with an overall recapture frequency of 2.28 times per individual. Spatial capture-recapture (SCR) was employed for population and density estimation. Model selection strongly favored a model in which density was negatively associated with distance to the closest glacier and positively associated with elevation, and baseline encounter rates were higher in the Karakoam-Pamir region and with Reconyx cameras than in other regions and types of cameras. The estimated population size for snow leopards in Pakistan was 155 (95 % CI 100-239), with a mean density of 0.16 (95 % CI 0.10-0.24) animals per 100 km². This research provides the first robust population estimate for snow leopards in this region, establishing a foundation for long-term population monitoring and assessing the effectiveness of conservation measures. We recommend the integration of complementary approaches, such as non-invasive genetic methods, to validate and refine population estimates.
README: Snow Leopard Population Estimation: R Code
We have submitted the R scripts used in our current study on snow leopard population and density estimation. Two main scripts were utilized: SCR_model_fitting.R and Sensitivity_all.R.
· SCR_model_fitting.R: This script was used to estimate snow leopard population size and density across the study area in Pakistan. The Spatially Explicit Capture–Recapture (SCR) modeling was implemented using the secrv4.6.5 package in R.
· sensitivity_all.R: This script was applied to evaluate the robustness and stability of the SCR model outputs.
The following dataset is required for SCR modeling:
· SL_Capture.csv: This CSV file contains the detection history of uniquely identified snow leopard individuals. A sample format of this capture file is provided below.
| Session | IndivID | Occasion | DetectorID |
|---|---|---|---|
| 1 | 1 | 1 | SL_KNP01 |
| 1 | 2 | 2 | SL_KVO06 |
| 2 | 1 | 1 | SL_KVO09 |
Session: each session represents a defined camera trapping period, e.g., 1, 2.
IndivID: Uniquely identified snow leopard individual ID
Occasion: Survey/sampling occasion within a session
DetectorID: Installed camera ID in which animal was trapped.
SL_Trap.csv: This CSV file provides information about all installed camera traps.
| DetectorID | X | Y |
|---|---|---|
| SL_KNP01 | 484619.3 | 4066813 |
| SL_KVO06 | 484247.7 | 4066528 |
| SL_KVO09 | 484072.6 | 4066937 |
DetectorID: Installed camera ID.
Detector_Location:
X coordinates: Easting coordinate of the installed camera trap location in UTM meters.
Y coordinates: Northing coordinate of the installed camera trap location in UTM meters.
Survey_Variable: Habitat types, camera types, etc.
Environmental/topographic/anthropogenic variables
A range of environmental, topographic, and anthropogenic variables—including elevation, slope, terrain ruggedness, distance to the nearest glacier, protected area status, human population density, road network density, and settlements—were used in raster (.tif) format in the current analysis. All raster layers were standardized to the same coordinate reference system (UTM, meters) with identical spatial extent, projection, and resolution to ensure spatial consistency and compatibility in subsequent modeling.
Data Availability Statement:
Due to regulatory compliance requirements, we are unable to provide the full datasets referenced in this script. However, data formats and structures are provided to enable users to set up their own datasets. Researchers interested in accessing additional information or seeking assistance with the code may contact the corresponding author at nawazma@gmail.com.
