Fig 1.
Study area and sampling design.
A) Location of the study area in the Union Territory of Ladakh, India. B) Location of occupancy sampling grids (10 × 10 km) across the study area, where every grid had multiple sign surveys to record snow leopard signs. C) Locations of camera traps and cameras with snow leopard photo-captures.
Fig 2.
Identification of snow leopard individuals.
Process for identifying snow leopard individuals using ExtractCompare software. A) Strategically placed camera traps allow for a photo-captures of the forehead of snow leopards. B) Reference points are placed on the crown, eyes, and nose to permit fitting a 3D model on the forehead to correct for the varying angles at which the photo may have been taken. C) After correcting for the angle, the forehead pelage pattern is extracted as a 2D image. D-F) Reference cues are generated to digitize the pelage pattern and saved quantitatively as unique data per individual photograph, forming a library. G) The software finds the best possible matches of the forehead pattern from the library and ranks in order of the similarity index. Independent expert observers subsequently use the similarity index to verify the patterns and confirm same individuals (G) from other individuals (H).
Fig 3.
Relationships between snow leopard and environmental variables.
Snow leopard sign encounter rates significantly increased with A) increasing wild prey encounter rates, B) increasing livestock encounter rates, C) increasing terrain ruggedness, and D) decreasing human settlement density. These graphs were used to strengthen the choice of variables in distribution modelling.
Fig 4.
Snow leopard distribution and density.
A) Occupancy of snow leopard in the sampled region of Ladakh. B) Population density of snow leopard in Ladakh.
Table 1.
Snow leopard occupancy model. Model parameter estimates snow leopard occupancy (Ψ) and detection (P) in Ladakh The sign and magnitude of the ß estimate provides the relative influence of the ecological variables on snow leopard occupancy.
Fig 5.
Distribution of snow leopard and prey species.
Potential distribution derived using MaxEnt for bharal/blue sheep Pseudois nayaur (A), urial Ovis vignei (B), ibex Capra sibirica (C) and snow leopard Panthera uncia (D) in Ladakh.
Table 2.
Spatially explicit density model. Density parameter estimates used in modelling snow leopard population. is the modelled density, σ is the detection function, g0 is the activity centre.
Table 3.
Snow leopard population across different regions in Ladakh. Camera trap variables and population parameters of snow leopard in the protected areas and multi-use areas of Ladakh. (D: average density per 100 km2, g0: detection probability at the activity centre of the animal, σ: spatial scale of detection in km).