Switch to desktop

The AfriCat Predator Population Density Study in the Okonjima Nature Reserve


okonjima nature reserve map 2014

Research Proposal: Submitted by Jenny Noack

Title of proposed project: The assessment of leopard (Panthera pardus) density and population size via a capture – recapture framework in an island bound conservation area in Namibia.

Supervisor and Veterinarian: Dr. David G Roberts

Principal Investigator: Jenny Noack

Co-Worker: Louis Heyns, Wayne Hanssen, Donna Hanssen, Janek Hoth

Contact Details:
AfriCat Foundation Research Department
Contact Person: Jenny Noack
Tel: +264-67-304566 / Mobile: +264-81-3987594
Fax: +264-67-687051
P.O. Box 1889, Otjiwarongo, Namibia


Background and Introduction

Project Need

In the last decades human activities have led to the devastating destruction of large parts of natural habitats (Gaston, 2008) leading to the dramatic decrease and threatened status of many wildlife species worldwide. The leopard (Panthera pardus) is classified as "Near Threatened" according to the IUCN Red List of Threatened Species (IUCN, 2014.2.). Leopards occur across wide ranges of sub-Saharan Africa as well as inhabit parts of Northern Africa and tropical Asia (Friedmann & Holzer, 2008). Their adaptability and tolerance towards a wide range of various habitats as well as their secretive and elusive nature had let them survive in marginal areas from which other felid have disappeared completely (IUCN, 2014.2). Despite their wide distribution throughout sub-Saharan Africa, the felids are declining dramatically in numbers and have disappeared from approximately 36.7% of their historical range (Ray et al. 2005) due to habitat fragmentation as well as intense persecution by humans. While sub-populations in North Africa and Asia are on the verge of extinction, Namibia’s population maintains stable numbers (Stein, Andreas & Aschenborn, 2011). With a total of only 17% of protected areas in forms of national parks, game reserves and recreational areas in Namibia (Turpi et al 2010), the majority of leopards occurs on commercial and communal farmland where the sporadic predation of livestock induces an inevitable conflict between man and carnivore. The necessity for the development and expansion of protected areas as well as the implementation and execution of improved livestock farming techniques on farmland are therefore of utmost importance to secure the survival of the Namibian leopard population.


Protected areas maintain a higher density of predators than un-protected areas (Stein et al. 2011). Carnivore abundance is often correlated by the level of inter- and intraspecific competition as well as the availability and accessibility of resources such as water and prey. Human-caused mortalities and habitat loss are additional factors playing a crucial role in determining felid densities (Khorozyan et al, 2008). Reliable large carnivore estimations are essential for the development of sustainable and long-term management and conservation strategies (Hayward, Brien, Kerley, 2007).


The majority of large felids are nocturnal and characterized by a secretive and elusive nature, consequently most techniques used to estimate species richness and abundance provide unreliable results. The application of camera traps proved to be the most dependable method when it comes to the creation of reliable quantitive estimations of large carnivore species.


Camera trapping as a continuous, non-invasive monitoring method

leopard camera trapbrown hyaena in a tree at bait site

Photography as a method of studying and observing wildlife goes way back in the nineteenth century and finds their early beginning in the 1860’s. George Shiras, lawyer, conservationist and pioneer of remote wildlife photography, was the first one to capture animals on photograph without human presence by using a trip wire that triggered his magnesium flash-gun camera. The use of camera traps by the hunting community promoted their increased application and advanced technical improvements (Sollmann et al. 2012). Since then the methodology of camera traps experienced major enhancements - not only in a technical but also in a visual way: Trip wires and step treadles that used to trigger old models have been replaced by infrared triggered systems placed in compact, camouflaged and waterproof housings. Today camera trapping is an essential and widely used tool that has revolutionized wildlife research and is applied for a variety of different research questions including the detection of species and their distribution and activity patterns as well as abundance and density estimations.

2 leopards and spots


The total number of free ranging leopards that inhabit the 20 000 ha Okonjima Nature Reserve is yet unknown. Due to their secretive and elusive nature a sighting is rare, but sometimes, one is lucky enough to be at the right place at the right time to spot one of our un-collared leopards. To improve luck and to get a detailed insight of how many leopards roam the reserve we now are busy using camera traps in order to identify single individuals.


The individual identification of animals via the use of wildlife camera traps has become a vital tool in animal ecology and has been applied to a wide range of individually recognizable species: tiger Panthera tigris (Karanth 1995), jaguar Panthera onca (Wallace et al. 2003), puma Puma concolor (Kelly et al. 2008), Iberian lynx Lynx pardinus (Gil-Sanchez et al. 2011), striped hyena Hyena hyena (Harihar et al. 2010), tapir Tapirus terrestris (Noss et al. 2003).

Individual recognition relies on capture - recapture surveys (Foster & Harmsen, 2011) and is based on the assumption that an animal can be identified by a natural or artificial marking and that this marking clearly distinguishes it from other individuals (Heilbrunn et al. 2003). Natural markings that allow for individual identification include stripe and spot pattern (e.g. tigers P. tigris, leopards P. pardus, cheetahs Acinonyx jubatus) and whisker patterns in lions Panthera leo.

The application of camera traps to study faunal populations offer a number of advantages and benefits when compared to other field methods and techniques: In contrast to other methods photographic captures provide undeniable evidence of an individual, animal and/or species being present at a certain site at a certain time and therefore, makes review by other researchers possible.

leopard tree cameraleopard tree cameraleopard tree camera

Camera trap captures of (A) Jo Jo Farque and (B) Mafana (centre and right) at two different trap sites and while tracking him with radio telemetry. Distinctive spot patterns are visible. Besides individual patterns, additional features e.g. body size, colouring, sex, external marks (scars, ear cuts) can be used for identification purposes.


Camera trapping is a non-invasive method of observing wildlife with only a minimum of disturbance to the target species and therefore, provides ideal conditions to study rare, cryptic and elusive. They are able to illustrate a 24h-cycle providing day and night time data whereas many other methods are restricted to diurnal use only (e.g. line transect surveys, VHF-telemetry) which makes them ideal to study and observe nocturnal species. Remote cameras can be left in the field unattended for several weeks without the supervision or the continuous presence of a researcher or field assistant. Depending on the environmental conditions as well as technical factors such as battery operating time, cameras can operate in the field for several weeks and months.

Since camera traps are mainly independent from field conditions such as weather, vegetation, habitat and terrain their use can be applied for a huge variety of ecosystems (Silveira et al. 2003) in which the use of alternative methods can be limited or restricted. Thus, camera trap studies have been conducted in different ecosystems and vegetation zones ranging from humid and dry forests (Karanth, 1998; Cuellar et al. 2006) over grassland habitats (Harris, 1996) to alpine ecosystems (Jackson et al. 2006; Xu et al. 2008).

camera set-up

Fig. 1: (A) Exemplary camera set – up: The camouflaged white flash camera trap is fastened to a trunk of a tree targeting a piece of bait on the opposite site. (B) Diagram of a Cuddeback Capture: Cameras are equipped with a heat-and-motion sensor that triggers the camera when a difference between the target- and the surrounding background temperature is detected.


In order to successfully "camera-trap" leopards on Okonjima, the first thing that needs to be done is to lookout for a suitable tree that provides enough space and shelter for a leopard. In most of the times our choice involves whether the Camel Thorn Tree (Acacia erioloba) or the mighty Kalahari Apple Leaf Tree (Philenoptera nelsii) – tall trees with large spreading canopies found in dry open wood - and bushland. The camouflaged trap is fastened to a branch approximately 150 – 200 cm above the ground targeting a piece of bait - helping to attract the cats. Occasionally, we also place camera traps close to pathways or riverbeds without any bait – in areas that are known to exhibit a high rate of leopard occurrence. The cameras are adjusted in a way that allows us a view on the asymmetrical flanks of the leopard. Once a leopard is "captured" the distinctive pattern of the spots are going to be compared to photographs already existing in our data base in order to identify whether it involves an unknown individual or a cat already known to us.

In order to cover and monitor major parts of the area and therefore increase the chances of "trapping" leopards, the position of the cameras will be alternated every four weeks.

camera on a tree trap locations

At the moment we are busy monitoring the southern part of the reserve. Current (red) and previous camera trap locations (grey). After successfully covering the area around Serenjima and the south-western part of the reserve, we are slowly moving further up north as well as into the eastern direction. Orange icons demonstrate location of box traps.

box trap set up

Carnivore box trap set up. The box traps are baited and thatched with grass on the in- and outside and additionally surrounded by Acacia-branches for camouflage purposes. The live camera is positioned 4 meters in front of the trap to ensure an optimal view. The trigger is remotely controlled and can be activated via an internet signal received from the radio aerial. Power is constantly supplied through the solar panel. To increase the capture probability and to cover and monitor major parts of the study area, positions of the cameras are alternated every four weeks.

box trap capture

mafana capture



Major aim of the study is to identify the leopard density (the exact number of leopards abundant) in the 20 000 ha Okonjima Nature reserve. Secondly, camera traps are used to locate suitable positions for the set up of carnivore box traps that are used for the capture and subsequent release of large carnivores. The steel mash box traps operate on a remotely triggered basis and can be closed via a live camera system which provides the advantage to specifically choose the targeted individual. Once an un-collared leopard is trapped, the animal gets darted and is fitted with a VHF – radio collar. Simultaneously body measurements (e.g. head – tail, leg, tail, canines) and temperature are taken. The animal is checked for injuries as well as other conspicuities and afterwards gets micro-chipped and vaccinated against rabies as well as treated against parasites.


leopard in box trap carrying caught leopard madiba transporting leopard
measuring leopard blood sample leopard collaring leopard
taking blood sample leopard measuring teeth leopard bwana 2014 collar
The radio collar allows us to precisely study the movement of the cats, create home ranges and analyze possible territory overlaps.
radio collars visitors with leopard jo jo farque leopard release
LeopardHomeRanges THIS map’s information is data collected from the 1st June 2014 – 1 Nov 2014: (min 24 positively ID'd images per cat, via direct observations and pictures taken of spot-patterns as well as camera trapping) Because leopards are easily identifiable via their spot patterns, camera trapping provides a reliable and practical tool to estimate abundance and density.


Project Description

Goal: To assess the density and population size of leopards (Panthera pardus) in the Okonjima Nature Reserve using photographic capture-recapture sampling and provide scientific data on their demography as well as spatial and temporal distribution patterns.


Study Objectives

  1. To determine leopard density and population size via a capture-recapture framework using remote camera traps
  2. To determine the demography of leopards within the Okonjima Nature Reserve
  3. To develop a dataset that can be applied as a baseline for comparisons to similar areas
  4. To develop a long-term population monitoring program


Study Area

study area map


Fig. 2: The Okonjima Nature Reserve located in central Namibia compromises a total area of 22 000 ha. Okonjima is home to The AfriCat Foundation whose mission is the long-term survival of Namibia’s carnivores in their natural habitat.


The Okonjima Nature Reserve (Lat/Lon: 20º49’19.36’’S, 16º38’21.25’’E) is located in central Namibia approximately 50 km south of Otjiwarongo and compromises a total area of 22 000 ha. The study area is semi-arid and characterized by a marked seasonality. The annual precipitation averages approximately 450 mm. The Okonjima farm boundary traces a central plateau, at average an altitude of 1 600 meters, surrounded by the Omboroko Mountains. The vegetation can be mainly described as tree- and scrub savannah, interspersed with Yellow wood (Terminalia sericea) and several Acacia-species. Artificially constructed water reservoirs ensure the perennial supply of surface water.

Okonjima was used intensively for the purpose of cattle farming from 1920 until 1993. Since then the private nature reserve has been used for carnivore rehabilitation and tourism purposes exclusively.

The reserve is surrounded by a 96 km electrified perimeter fence, completed in 2010, and is bordered entirely by commercial farmland. An additional fence is erected within the reserve and creates a 20 000 ha reserve for carnivore rehabilitation and a 2 000 ha “lodge area” that includes lodges and campsites as well as the AfriCat headquarters and the Environmental Education Centre.

Leopards as well as brown hyenas (Hyena brunnae) occur naturally within the borders while cheetahs (Acinonyx jubatus), African wild dogs (Lycaon pictus) and spotted hyenas (Crocuta crocuta) are part of AfriCat’s rehabilitation program that have been released into the area. Fifteen leopards are currently radio-collared in the reserve. Lions (Panthera leo) are absent from the study area.


Methodology and Data Analysis

The leopard population size will be sampled via the application of remote wildlife camera traps. Therefore a closed model capture – recapture sampling approach will be used that is estimating abundance based on the number of individuals captured and the frequency of their recaptures (Silver et al. 2004). Capture-recapture (C-R) models take into account that not all individuals in the population are necessarily detected (Rovero et al. 2013) and rely on the assumption that the target species can be individually identified via natural marker or other phenotypic features (Karanth & Nicholas 1998). C-R models are the commonly used strategy to estimate abundance of elusive terrestrial mammals with particular focus on felids (Foster & Harmsen, 2012). Because leopards are easily identifiable via their spot patterns, camera trapping provides a reliable and practical tool to estimate abundance and density.

Secondly, C-R assumes that the sampled population is geographically as well as demographically closed during the study period and that no births, deaths, immigrations or emigrations occur during the sampling period (Karanth & Nicholas 1998; Silver at al. 2004). It is crucial that the sampling period ensures demographic closure and therefore doesn’t exceed a certain threshold. Karanth et al. (1998, 2000) sampled a tiger population in India for 3 months while Silver at al. (2004) used a 2-months period to estimate abundance of a jaguar population in South America. The long intervals before a leopard returns to a defined starting point within its natural movement need to be considered additionally when determining the time frame of the study.

Finally, it is critical to optimise camera trap placement without leaving any gaps to ensure a complete coverage of the entire study area to maximize capture probability and the number of individuals caught on the camera (Karanth & Nicholas 1998). The determination of inter-camera trap distances is crucial so that no leopard has a zero capture probability during the sampling period. The smallest home range size of a female representative of the target species is commonly used to determine the area that should be covered by at least one trap site (Karanth & Nicholas 1998, 2000; Silver at al. 204; Soisalo & Cavalcanti, 2006, Marnewick et al. 2008). The set-up of more than one trap within the defined home range increases the chances that also individuals of the opposite sex as well as from different age classes will be exposed to the camera (Soisalo & Cavalcanti, 2006) and decreases the probability that individuals inhabit home ranges that fall between trap sites (Foster & Harmsen, 2012). Available VHF-telemetry data on the collared cats in the reserve as well as consulted literature will be used to determine the inter-camera trap distance based on average female leopard home range sizes.

Due to a limited number of available camera traps the study area will be divided into equal sized blocks that will be covered sequentially for the same amount of time (Karanth & Nicholas, 2002; O’Brien et al. 2003). A preliminary inspection will be implemented in order to identify suitable camera trap locations and Geographic Position System (GPS) points will be taken of every trap site and transferred onto a digital map. Camera trap studies aim to capture as many individuals as possible. To enhance accuracy of abundance estimates it is essential to chose trap sites that increase capture probability. Therefore camera traps should preferably be placed in areas that suggest an elevated frequency of leopard occurrence such as dry riverbeds, riverbanks and/or frequently used roads and pathways.

Camera traps will be placed in pairs on opposite sides to ensure that both flanks will be exposed to the camera and therefore increase the chances of identification. Cameras will be placed approximately 2 – 4 meters away from where the target is expected to walk. A direct facing of the cameras should be avoided in order to prevent overexposure. Trap sites will be checked every 2 – 3 days to check battery status, change SD cards and ensure the correct functionality of the traps. Collected data will be stored in a data base.

Photographs from each trap site will be evaluated. Each individually identified leopard will be assigned with a unique identification number. If applicable, sex and age class will be recorded. Identification will be performed by two independent observers to prevent bias. A capture history will be generated for each sub-sampled block and afterwards combined to form one data set for the entire study period (Soisalo & Calvalcanti 2006). Therefore a standard X-matrix format will be used where "1" indicates the presence of an individual during a sampling occasion and "0" the absence during that occasion (Karanth & Nicholas, 2004) (Tab. 1).


Tab. 1: Example of a capture history data set showing the number of positively identified leopards (Pp) and the number of days per sampling period (sampling occasion). 0 = no visit; 1 = visit detected. The capture history of Pp1 (00011000) indicates its capture during sampling occasion 4 and 5 and absence during 1, 2, 3, 6, 7 and 8.

  Sampling Occasions (n=8)
ID 1 2 3 4 5 6 7 8 ...
Pp1 0 0 0 1 1 0 0 0  
Pp2 1 0 0 0 1 1 0 0  
Pp3 ...                

The number of captured and re-captured animals will be analysed by using a statistical model (e.g. CAPTURE) and abundance and population size (N) generated. The calculated abundance will be used to derive an estimation on the leopard density (D): D = N/A whereas A is defined as the actively sampled area.

Trap success between trap stations will be evaluated and compared in order to reflect spatial distribution patterns. Distances moved between traps during the sampling period will be used to calculate minimum home range estimates for each cat.

After the effective sampling period camera traps will be used as part of a long-term monitoring program for the sampled leopard population.



Cuellar, E., Maffei, L., Arispe, R. & Noss, A. (2006) Geoffroy’s cat at the northern limit of their range: activity pattern and density estimates from camera trapping in Bolivian dry forests. Studies on Neoptropical Fauna and environment, 41, 169 - 177.

Foster, R. & Harmsen, B. J. (2012) A critique of density estimation from camera-trap data. The Journal of Wildlife Management, 9999, 1 - 13.

Gaston, Kevin J., et al. The ecological performance of protected areas. Annual Review of Ecology, Evolution, and Systematics 39 (2008): 93-113.

Gil-Sanchez, J. M., et al. (2011) The use of camera trapping for estimating Iberian lynx (Lynx pardinus) home ranges. European Journal of Wildlife Research, 57, 1203 - 1211.

Hayward, M. W., O'Brien, J., and Kerley, G. I. H. (2007) Carrying capacity of large African predators: predictions and tests. Biol. Conserv. 139: 219-229.

Harihar, A., Gosh, M., Fernandes, M., Pandav, B. & Goyal, S. P. (2010) Use of photographic capture-recapture sampling to estimate desity of striped hyena (Hyaena hyaena): implications for conservation. Mammalia, 74, 471 - 479.

Harris, R. B. (1996) Wild ungulate surveys in grassland habitats: Satisfying methodological assumptions. Chinese Journal of Zoology, 31, 16 - 21.

Heilbrunn, R. D., Silvy, N. J., Tewes, M. E. & Peterson, M. J. (2003) Using automatically triggered cameras to individually identify bobcats. Wildlife Society Bulletin, 748 - 755.

Henschel, P., Hunter, L., Breitenmoser, U., Purchase, N., Packer, C., Khorozyan, I., Bauer, H., Marker, L., Sogbohossou, E. & Breitenmoser-Wursten, C. 2008. Panthera pardus. The IUCN Red List of Threatened Species. Version 2014.2. ( Downloaded on 13 November 2014.

Jackson, R. M., Roe, J. D., Wangchuk, R., & Hunter, D. O. (2006). Estimating Snow Leopard Population Abundance Using Photography and Capture‐Recapture Techniques. Wildlife Society Bulletin, 34(3), 772-781.

Khorozyan, I. G., Alexander G. M., and Alexei V. A. (2008) Presence–absence surveys of prey and their use in predicting leopard (Panthera pardus) densities: a case study from Armenia." Integrative Zoology 3.4 322-332.

Karanth, K. U (1995): Estimating tiger (Panthera tigris) populations from camera trap data using capture-recapture models. Biological Conservations, 71, 333 - 338.

Kelly, M. J. (2008) Design, evaluate, refine: camera trap studies for elusive species. Animal Conservation, 11, 182 -184.

Marnewick, K., Funston, P. J., & Karanth, K. U. (2008). Evaluating camera trapping as a method for estimating cheetah abundance in ranching areas. South African Journal of Wildlife Research, 38(1), 59-65.

Noss, A. J., Cuellar, R. L., Barrientos, J., Maffei, L., Cuellar, E., Arispe, R., Rumiz, D. & Rivero, K. (2003) A camera trapping and radio telemetry study of lowland tapir (Tapirus terrestris) in Bolivian dry forests. Newsletter of the IUCN/SSC Tapir Specialist Group. Tapir Conservation, 12, 24 - 32.

O'Brien, T. G., & Kinnaird, M. F. (2008). A picture is worth a thousand words: the application of camera trapping to the study of birds. Bird Conservation International, 18, 144.

Ray, J. C., Hunter, L.T.B & Zigouris, J. (2005) Setting conservation and research priorities for larger African carnivores. Working paper 24. Wildlife Conservation Society, New York.

Rovero, F., Zimmermann, F., Berzi, D., & Meek, P. (2013). " Which camera trap type and how many do I need?" A review of camera features and study designs for a range of wildlife research applications. Hystrix, the Italian Journal of Mammalogy, 24(2), 148-156.

Silveira, L., Jacomo, A. T. A. Diniz-Filho, J. A. F. (2003) Camera trap, line transect census and track surveys: a comparative evaluation. Biological Conservation, 114, 351 - 355.

Silver, S. C., Ostro, L. E., Marsh, L. K., Maffei, L., Noss, A. J., Kelly, M. J., ... & Ayala, G. (2004). The use of camera traps for estimating jaguar Panthera onca abundance and density using capture/recapture analysis. Oryx, 38(02), 148-154.

Soisalo, M. K., & Cavalcanti, S. (2006). Estimating the density of a jaguar population in the Brazilian Pantanal using camera-traps and capture–recapture sampling in combination with GPS radio-telemetry. Biological Conservation, 129(4), 487-496.

Sollmann, R., Mohamed, A., Samejima, H. & Wilting, A. (2012) Risky business or simple solution - Relative abundance indices from camera trapping. Biological conservation, 159, 405 - 412.

Stein, A., Andreas, A. & Aschenborn, O (2011) Namibian National Leopard Survey – 2011. Final Report. Ministry of Environment and Tourism.

Turpi, J., Barns, J., De Longcamp, M. & Paxton, M. (2010) Sustainable Financing Plan for Namibia’s Protected Area System: Ministry of Environment and Tourism. Directorate of Parks and Wildlife Management.

Wallace, R. B., Gomez, H., Ayala, G. & Espinoza, F. (2003) Camera trapping for jaguar (Panthera onca) in theTuichi Valley, Bolivia. Journal of Neotropical Mammals, 10, 133 - 139.

Xu, A. et al. "Status and conservation of the snow leopard Panthera uncia in the Gouli Region, Kunlun Mountains, China." Oryx 42.03 (2008): 460-463.


Leopard Behaviour in the Okonjima Nature Reserve

1 female 2 cubs
1 female and 2 cubs.
1 female scent marking
Female leopard scent marking.
leopard hunting warthog
Leopard hunting warthog.
leopard cleaning prey before eating it
Leopard 'cleaning' prey before eating it.
female and male during mating
Mating leopards.
female and male during mating
Mating leopards.
female and male during mating
Mating leopards.
female and male during mating
Mating leopards.
female and male during mating
Mating leopards.

"Both leopards and lions have exactly the same mating rituals which, when averaged out, has them mating every 15 minutes for up to 5 days. This means that if they last a full 5 days, they can mate more than 250 times. This may seem a little excessive, but there is a good reason for this. In humans, females produce an egg every 28 days. If it is not fertilised, the egg and the lining of the uterus will be discarded. This does not happen with leopards though. The act of producing something that is not used is a waste of energy, so the female leopard requires a stimulus to start ovulation.


The female leopard will therefore go into oestrus. This is a state in which her hormones are at a level at which she is able to produce eggs. She will also leave a scent in her urine which will indicate to the male that she is ready to mate. As she enters oestrus the female will begin to mark her territory more often than usual and will call to attract the dominant male in the area. He will latch onto her scent using a gland on his palate called the organ of Jacobson, which is able to measure hormone levels and determine whether or not the female is ready to mate.


Once the mating begins it is a non-stop affair, filled with uneasiness and violence. In order to stimulate the female to ovulate, the male has barbs on his penis which dig into the female. When the penis is retracted it hurts the female causing her to lash out at the male. As the mating ritual continues she will produce eggs. Due to his weak sperm, a male leopard has to mate with a female often enough to ensure that fertilisation takes place.


It is interesting to note that current research suggests female leopards can tell whether or not a male is capable of being the territorial male and remain dominant for a long period of time. If she determines that he is not, she has the ability to make herself less fertile. This is because if she conceives and the male is ousted from his territory, a new male will kill her existing cubs. If this happens she would have completely wasted her energy.


It has also been established that females will mate with more than one male within a very short period, thus lowering the chances of infanticide. This is particularly important for female leopards whose territories overlap those of two dominant males. Infanticide in leopards accounts for 40% of cub deaths and is therefore one of the major considerations for females when choosing a mate.” Sabi Sabi

Copyright AfriCat All rights reserved Copying of images is prohibited

Top Desktop version