Environmental                       Studies




GPS collar products

GPS collars wildlife

GPS-GSM collars wildlife

GPS dog collar

GPS Hundehalsband


Link Manager

Remote GPS data download

Argos terminals

GPS tracking projects



Precision Farming




































































































































































































last update: November 1, 2005

   Tracking leopards with GPS collars in South Africa

   The leopard Panthera pardus in northern KwaZulu-Natal, South Africa:
     caught in a conservation blind-spot?


Luke Hunter Wildlife Conservation Society

Guy Balme Phinda Private Game Reserve


In 2002, we initiated a major study investigating the conservation issues surrounding leopards in northern KwaZulu-Natal, South Africa. Named the MunYaWana Leopard Project, the study grew from a concern that persecution of leopards in the region was reaching unsustainable levels.  Phinda Private Game Reserve forms the core of the study area and leopards are offered full protection within the reserve. They are also fully protected in the Mkhuze Conservancy abutting Phinda’s western boundary. The land to the south and east of Phinda, however, comprises a mosaic of local Zulu communities, livestock farms and private game farms where leopards are generally viewed as problem animals or as targets for legal trophy hunts. They are not constrained by the surrounding boundary fences and move freely between adjacent properties; the same individual may run the gamut of protected status within twenty four hours.  A dramatic increase in persecution levels (both legal and illegal) since 1997, raised concerns that the non-protected areas were not fostering resident populations of their own, but rather drawing off the surplus animals from the contiguous protected areas of Phinda and Mkhuze.



Taking measurements of male leopard M25 prior to placing on GPS collar 1883


Against this back-drop, the MunYaWana Leopard Project set out to investigate the following issues:

  • That protected areas in Zululand are providing a source population for hunting in adjacent areas,
  • That properties engaging in hunting are drawing entirely on ‘overflow’ from protected populations rather than fostering resident populations on those properties,
  • That illegal killing of leopards primarily by poisoning is widespread on private lands throughout the province, and 
  • That the levels of persecution from combined legal and illegal killing of leopards outside protected areas may be so high that even protected populations are unable to increase and may be diminishing.
  • The ultimate goal of the project is to establish a conservation management plan for leopards in KwaZulu-Natal which ensures that viable populations of the species are protected in the province while still allowing for regulated utilisation by landowners.  As the problems facing leopards throughout their range are similar, our results will also have relevance for leopard conservation throughout the rest of Africa.


The Future

Our research is entering a critical phase. We now have ample data on the leopards of Phinda to raise serious concerns about their persistence in the area. Despite the fully protected status of leopards on the reserve, Phinda may not be effectively conserving them. Geographically narrow and bounded by the private and communal lands to the south and east where leopards are intensely persecuted, Phinda might represent a ‘population sink’ for leopards. The continual removal of leopards from areas adjacent to the reserve sets up the possibility of a constant drain on the leopard population that, in the worst case, is not sustainable.

Accordingly, the next crucial stage of the project is to investigate what is happening in the larger Mkhuze Conservancy to the west of Phinda, as ultimately this area may be acting as a source population for the entire region.  Due to its larger size, the leopard population within Mkhuze may be ‘buffered’ to some extent and hence may represent a more natural situation.  If this area is acting as a source population, are recruitment levels sufficient to accommodate for the large number of cats being destroyed in adjacent non-protected areas?  Investigating the Mkhuze leopard population will give us the necessary control to compare with data collected from Phinda and the non-protected areas.   It will hopefully also give us an indication of the overall long-term sustainability of leopards in the region. 

The monitoring of radio-collared leopards in Mkhuze is, however, considerably more difficult than on the Phinda.  The Conservancy is three times the size of Phinda and yet there are significantly less roads, making regular radio-triangulation impossible.  Much of the area is also made up of extremely thick terrain, meaning that access on foot is exceedingly difficult.  It was for these reasons that we decided to use Vectronics GPS collars on the leopards in Mkhuze.  The region, however, has particularly poor cell phone coverage and hence downloading via GSM was not an option.  We therefore opted to use the Handheld Terminal and the UHF-downloading mechanism.  We have deployed two GPS collars to date and wish to deploy a further four in the near future (Figure 1). 



Figure 1. Camera trap photograph of male leopard M26 taken on the 23/7/05.














The collars are set to take a position five times a day and I attempt to locate the animals once a month to download all the stored locations.

Both collared leopards are males.  The first, M26, is approximately three-years-old, weighs 55 kg, and was captured on the 24/6/05. At this age he would almost certainly still be a transient individual; ranging over a large area, avoiding adult resident males and looking for a territory in which to possibly settle.  We have received 197 fixes (Figure 2) from the first two months of data.  Home range size using a 95% Fixed Kernel Analysis is 58km2 (Figure 3). Interestingly, his home range appears to be relatively linear in shape, stretching from southern boundary all the way up through the central portion of the reserve.  He has not yet moved out of the Conservancy.



Figure 2. GPS locations downloaded from Vectronics collars 1883 and 1884 deployed on male leopards M25 (red) and M26 (blue) in Mkhuze Game Reserve from the 24/6/05 to the 3/9/05


















Figure 3. 95% Kernel Home Ranges of male leopards M25 (red) and M26 (blue) in Mkhuze Game Reserve from the 24/6/05 to the 3/9/05.









The second male, M25, was initially captured on the 2/6/05 and a VHF radio-collar placed on him. We recaptured him on the 5/7/05 and the VHF collar was exchanged for a GPS collar.  Judging from tooth wear, he is older than M26 (estimated between four and six years), although he only weighed one kilogram more.  We have downloaded 134 locations (Figure 1) from the collar and he is using a home range of approximately 57km2 (Figure 2).  This area is predominantly made up of the mountainous regions in the south and the west of the park, though he has crossed out of the Conservancy on at least one occasion into a neighbouring Zulu community. 


The Research to Date

We have captured and radio-collared 32 leopards since the inception of the study in April 2002. We attempt to locate every radio-tagged individual at least once daily and have so far logged over 10 000 ‘leopard radio-transmitter days’ (the total number of days combined for which all radio-tagged leopards are monitored) and over 2 000 direct observations of leopards. This is more than any previously published and is giving rise to one of the most comprehensive datasets that exists on the species as a whole.

The most important facet of our monitoring has been establishing the survival and reproductive patterns in the population . Ultimately, these parameters will determine if the population is increasing, stable or diminishing. Accordingly, we record the numbers of cubs born to radio-collared females and closely monitor their survival. We also assess the dispersal rates and success of these youngsters once they reach independence and leave the mother. Finally, we note all deaths of leopards and where possible, investigate their causes.

The results to date paint a bleak picture. During the three year study period we have recorded the deaths of 20 leopards on Phinda and adjacent properties. The average annual mortality rate (AMR) of the population between April 2002 and August 2004 was 44.7%; in other words, almost half the population is dying every year. In comparison, a study conducted on the protected leopard population in the Kruger National Park revealed an AMR under 25% or a quarter of the population. The worst-affected segment of our population is males; 82% of independent males (resident adults and dispersing sub-adults combined) are killed each year. Males are more desirable to trophy hunters due to their larger size and that they utilise larger home ranges and cover greater daily distances than females, increasing their chances of moving off the reserve into areas where they can be hunted.


Our results show that removing an excessively high number of males produces a cascade of deleterious effects on the population. Unnaturally high turnover among adult males can result in increased levels of infanticide, in which new males entering the population kill cubs sired by the former males. Although infanticide is a natural event in leopards, very high levels of male turnover promotes a situation in which females cannot successfully raise cubs because of constant incursions by new males. We think the consequence of this in the MunYaWana leopard population is very low survival rates of cubs. During the entire study period to date, only three litters have been produced, although we have recorded females mating with various males on many occasions. Of these three litters, only two cubs survived long enough to reach independence. Only one of those survivors is alive today, a young female. She is still too young to reproduce at the time of writing and she may not necessarily survive long enough to reproduce.

The conservation prospects represented by these figures are not positive. Excluding immigration, the tally stands at a minimum of 20 leopards that have been removed from the population and only one that has been recruited into it. Although these are preliminary totals, this is clearly not sustainable. If these patterns apply to the entire population, the leopards of the region are in trouble.




Author's address:


Dr Luke Hunter: Wildlife Conservation Society;








Guy Balme: Phinda Private Game Reserve;





More information:







Top of leopard GPS tracking page