To protect our Wildlife we must conserve our Wilderness and for our Wilderness to be meaningful our Wildlife must be able to roam free within it.
The Sri Lankan leopard (Panthera pardus kotiya) in Yala National Park: A research summary

May 2012

Andrew M. Kittle
Anjali C. Watson

The Leopard Project

The Wilderness & Wildlife Conservation Trust



The following is a summary of the results of a 16-month long comprehensive field study conducted in Yala National Park (YNP) from 2001-2002. Earlier reports have been submitted to the DWC (2002, 2003 and 2004) but the current report includes updated results based upon additional analysis undertaken for a journal publication.

Currently an initiative titled Spotting the Spots is ongoing within the more frequented areas of Yala Block I. This initiative attempts to utilize photographed leopards by a pool of 5-6 volunteers who visit the park on a regular basis.  Standardized data sheets, UTM locations and the corresponding photograph of the leopard taken at a sighting are being catalogued so that individual identification of the leopards can be retro actively documented.  This allows for a basic monitoring of the current population trends and for a comparative data set to be established in relation to our 2001-2002 study. Preliminary results indicate similar population and spatial patterns to the earlier work.

This form of data, provided by park visitors, has been successfully used in a number of projects around the world. One of the best examples of such use of visitor records is from the Ngorongoro crater in Tanzania, where voluntarily provided tourist photographs have been extensively used to assist in estimating changes in lion population characteristics (Packer et al. 1991- he is now one of Andrew Kittles PhD supervisors). This was of particular importance to account for periods where researchers were absent from the park due to research commitments at other locations.


The Sri Lankan leopard (Panthera pardus kotiya) is the largest of four wild cat species recorded on the island of Sri Lanka. It is the island’s only large cat species and its top predator. This population has evolved geographically separated from the mainland Indian species (Panthera pardus fusca) and is now recognized as one of the nine subspecies of leopard currently extent in the world (Miththapala et al 1996; Uphrykina et al.2001). Furthermore this separation ensured that it has long been separated from any form of intra-guild competition. The long-term lack of intra-guild competition has some interesting potential ramifications in terms of the Sri Lankan leopard’s ecology and behaviour.

Study site and Methods

The study site (6°16’-24’ N, 81°23’-31’ E) was a 127 km2 section of YNP, Block I (Figure 1). The extreme north-western corner of the Block was not included as it was off limits due to security reasons during the period of the study. For details refer to Kittle & Watson 2004. For the population density analysis we used a 92 km2 portion of the larger study area.


Figure 1: Yala National Park, Block I showing roads and Water sources during study period (2001-02). The north-west corner of the Block was not included in study.

This study was conducted over 237 days from February 2001 to May 2002 with observations between 05.30-11.30 and 15.30-22.30 hrs to include perceived times of peak leopard visibility (Baily 1993). Focal routes were dependent on recent leopard activity, with the entire study area surveyed every 96 hours

Direct observations and track locations were used to determine adult male home ranges. Female “core areas” were based entirely on direct sightings and sighting locations of dependent cubs (Bailey 1993). These core areas are assumed to be smaller than true home ranges and were calculated for all females with ≥15 recorded locations. We defined male home ranges and female core areas using minimum convex polygons (MCP) of all recorded locations (Convex Hull extension for ArcView 3.2, ESRI systems).

Prey abundance was determined using a road strip census (Hirst 1969) with a modified fixed visibility profile (Norton-Griffiths 1978). The study site was divided into three sectors according to habitat type. The southeast sector (21.1% of the study area) is a coastal strip interspersed with open plains; the southwest sector (46.5%) is the forested interior, bordered on its western boundary by settlement areas; and the northeast sector (32.4%), bordered to the north by the Menik River, is sloping and well-drained, characterized by numerous small to medium tanks that hold water through most of the year. Road transects (15.6 – 18.4km) were delineated for each sector and a visibility index determined for leopard prey species in the dry, wet and intermediate seasons (Bailey 1993). The visibility distance was used to determine strip width per season which was multiplied by the transect length to establish the total seasonal area of each transect (Norton-Griffiths 1978). Open areas where visibility was greater than the road strip were measured on a 1:50 000 topographical map and added to the total transect area. Thirty transects were conducted, 10/sector, between 16:00 – 19:00 hrs at 5 – 10km/hr.

To incorporate seasonal fluctuations in prey abundance, transects were spread across seasons, three in each of the dry and intermediate seasons and four in the wet season.  The age and sex class of individual axis deer (Axis axis), buffalo (Bubalus bubalis), sambhar deer (Cervus unicolor), grey langurs (Semnopithecus entellus), wild boar (Sus scrofa) and mouse deer (Tragulus memina) were recorded. Average seasonal abundance counts were divided by the associated seasonal transect areas to determine seasonal densities (individuals/km²). These were then amalgamated to determine a yearly average prey density. Each transect represented a different sized sector of the study area, so the relative contribution of each transect to the total study area prey data was weighted by the proportional size of the sector that it described.

The total biomass (kg/km²) of potential leopard prey in the study area was determined by multiplying the number of individuals in each age/sex class comprising the sector densities by their respective kg weights and weighting each sector’s contribution to the total as before. Age and sex classes considered potential leopard prey were determined from analysis of kills and consultations with YNP staff.

Leopard scat samples were collected both opportunistically and during regular trail monitoring.  To differentiate between leopard and fishing cat scat only samples with bolus width > 2.5 were retained (Henchel & Ray 2003).

Results and Discussion


The study population has a high density (Table 1) rivaling areas in South Africa considered to be prime leopard habitat (Bailey 1993). Other international studies (Norton & Henley 1987; Bothma & Le Riche 1984; Hamilton 1976; Schaller 1972; Kostyria A.V. 2004 (unpublished data)) conducted in a variety of habitat types show lower densities than this population.

At this point it appears that a combination of abundant prey and artificially maintained permanent water sources and the lack of con-specific competition as well as the natural life history attributes (social structure, reproductive strategy (see below)) of the leopard allows for such density.

Table 1:  The resident population size, demography and density in the YNP study area (92 km²).  The resident population included adult animals that were frequently observed or detected within the study area and displayed consistent territorial behaviour and/or had cubs.

Sex Number of individuals #/100 km² km²/ individual
4 4.3 23
12.5 13.5 7.4
♂♀ 16.5 17.9 5.6

Given these densities it is expected that the whole of YNP Block I (140 km²) holds 25 resident leopards (6 males and 19 females). A total of 45 leopards, including transients, sub-adults and cubs were observed within the 127 km² study area. The total number of leopards within YNP Block I is therefore expected to be ~50.

Reproduction and Dispersal:

There appears to be no birth season or peak for this population as recorded births (n = 17) were scattered across months (Figure 2).


Figure 2: Births in YNP, Block I observed during study period.

From the number of observed births, the reproductive rate of the study population appears healthy, however the cubs born during the study period appear to have a 45% chance of survival into full, reproductive adulthood (first year mortality estimated at 44.8% from 11 litters X 2.14 cubs/litter = 23.54 cubs).

On average young leopards did not leave their natal ranges until they were ~24 months old. A number of the study individuals were not observed again after this time. Given enough space, these young animals have the ability to spread to adjacent areas even though pressure upon them from human settlements, inferior habitat and existing resident animals exists. Block 1 is the southern terminus of a much larger conglomeration of protected forest habitat which is assumed to be large enough to allow a percentage to disperse successfully.

Spatial behaviour:

The study population’s land tenure system is very similar to what is accepted as typical for wild leopard populations whereby resident males occupy considerably larger areas than resident females and their boundaries generally overlap the entire or partial home ranges of 4-6 resident females (Bailey 1993). The mean home range size of resident males is 22.5 ± 1.7 km2 (n=3, range = 20.5 – 25.8 km²) (Figure 2). These ranges showed considerable overlap. A central 0.8 km2 was used by three resident males (Figure 2), all of which marked heavily and could be found here within a 24-hour period of one another. All male home ranges fully or partially encompassed the core areas of ≥ 5 female residents. Adult female core areas average 1.58 ± 0.33 km2 (n=4, range = 0.99 – 2.35 km²) (Fig.2). These are densely congregated but appear to be exclusive of other adult females. Sub-adult leopards shared their natal home range with their mothers.  Adult resident males appeared very tolerant of sub-adults, although after approximately 2 years most sub-adults disappeared or were seen with less frequency in their natal areas.  This could be associated with increased pressure to leave the area, exerted on the young animals by the residents.  Occasionally injuries were observed that would be consistent with con-specific rivalry.


Figure 2: Home range sizes and structure of leopards in YNP, Block I.

Feeding ecology:

We collected over 250 fecal samples during the study, of which 214 were successfully analyzed (Table 2). Axis deer is the most common prey for leopards in YNP. This is no surprise given their abundance (see Table 3). Black-naped hare were also found frequently in leopard scat.

Table 2: Scat analysis results showing number of scat samples with different prey species.

Prey species # %
Axis deer 169 78.97
Black-naped hare 21 9.81
Sambhar 6 2.80
Water buffalo 6 2.80
Porcupine 5 2.34
Wild boar 4 1.87
Grey langur 2 0.93
Toque macaque 1 0.47
Peafowl 1 0.47
Snake 1 0.47

Prey abundance: 

Table 3: Total prey density and biomass and prey density and biomass available to leopards in YNP, Block I

Prey  Density  Potential Total Biomass Biomass available
Species  (#/km²) Prey (%)  (kg/km²)  (kg/km²)
Axis deera 121 100 6111.9 6111.9
Water buffalob 31.9 17 6641 502
Wild boarc 8 90 431.4 316.5
Sambhara 0.8 100 133 133
Grey langura 7 100 47.3 47.3
Mouse deera 0.1 100 0.6 0.6
TOTAL     13365.2 7111.3

a  all age and sex classes are potential leopard prey

b  only non-adults are potential leopard prey

c  females and young are potential leopard prey


Amerasinghe, F. P., Ekanayake, U. B. & Burge, R. D. A. 1990. Food habits of the leopard (Panthera pardus fusca) in Sri Lanka. Ceylon J. Sci. (Bio. Sci.) 21: 17-24.

Bailey, T.M 1993. The African Leopard: The Ecology and Behavior of a Solitary Felid Columbia     University Press, New York.

Bothma, J. duP. and E. A. N. Le riche.  1984.  Aspects of the ecology and behaviour of the leopard (Panthera pardus) in the Kalahari Desert. Koedoe (supplement) 27: 259- 279.

Hamilton, P.H.  1976.  The movements of leopards in Tsavo National Park, Kenya, as determined by radio-tracking.  M. Sc. thesis, University of Nairobi.

Henschel, P. & Ray, J. 2003. Leopards in African rainforests: survey and monitoring techniques. Wildlife Conservation Society global carnivore project.  

Hirst, S.M. 1969. Road-strip Census Techniques for wild ungulates in African woodland. Journal of Wildlife Management. 33: 40-48

Karanth, K.U. 1994. Estimating Tiger Panthera tigris populations from camera-trap data using capture-recapture models. Biological Conservation 71: 333-338.

Kittle, A.M. and A. Watson. 2004. Observations of a Rusty Spotted Cat population in an arid zone habitat. Cat News, Cat Specialist Group. London. No.40: 17-19.

Kittle, A.M. and A. Watson. 2002. Report to the Department of Wildlife Conservation and Ministry of Environment and Natural Resources, Biodiversity Unit. Sri Lanka.

Miththapala, S., J. Seidenstiker, and S.J. O’Brian. 1996 Phylogeographic subspecies recognition in leopards (Panthera pardus): molecular genetic variation. Conservation Biology 4:1115-1132

Miththapala, S., J. Seidenstiker, L.G. Phillips, S.B.U. Fernando and J.A. Smallwood. 1989. Identification of Individual Leopards (Panthera pardus kotiya) using spot pattern variation. Journal of Zoology, The Zoological Society of London, 218, 527-536.

Norton, P. M., and S. R. Henley.  1987.  Home range and movements of male leopards in the Cedarberg Wilderness Area, Cape Province.  South African Journal of Wildlife Research 17: 41-48.

Schaller, B.B. 1972.  The Serengeti lion.  University of Chicago Press, Chicago.

Singh, L.A.K. 1999. Tracking Tigers: Guidelines for estimating wildtiger populations using the pugmark technique (revised edition). WWF Tiger Conservation programme, New Delhi.

Uphyrkina O., Johnson W. E., Quigley, H., Miquelle, D, Marker, L., Bush, M and S.J.O’Brian.  2001. Phylogenetics, genome diversity and origin of modern leopard, Panthera pardus. Molecular Ecolgy 10, 2617-2633.