|
WILDTRACK VISITS OKONJIMA AND AFRICAT
Spotting cheetah footprints in the sand....
WildTrack at Okonjima - update September 29th 2011
WildTrack (www.wildtrack.org) is working with Okonjima to help develop a new approach to cheetah monitoring in Namibia - using footprints.
The footprint identification technique (FIT), developed by WildTrack in JMP statistical software from the SAS Institute, uses complex statistical modelling of footprint geometrical variables which can then provide animal identification at the species, individuals, age-classes and sex levels.
Thanks to the ambassador cheetah at Okonjima, and some willing volunteers, we have now started collecting footprints from Okonjima to help establish a database for this species.
Because each species has a different foot geometry, the FIT algorithm must be determined for each species before it can be used for monitoring unknown animals. That's why we're collecting prints from known individuals first, to figure out what it is in the geometry of the foot which makes each cheetah unique.
Recently a volunteer, Kristina Killian, began the process of collecting prints at Okonjima. Five different individual cheetah were persuaded to walk along a sandy trail, one at a time, and their prints were photographed using a simple digital camera and sent electronically to our research base in Portugal.
Processing the images in FIT involves optimising and orientating the images, taking more than 100 measurements and then using a statistical model to output not only predicted identity, but also relationships between individuals. This is perhaps the most exciting part of the analysis, and although the database is far from complete, the model has already been able to identify sibling groups.
Take a look at the five footprints below. There's one from each cheetah - Gomes, Jago, Morticia, Pugsley and Wednesday. They are all from the left hind foot of the individual. Can you see something unusual for a large cat footprint? Cheetah prints leave claw marks because they have only semi-retractable claws. Can you also see the characteristically sharp lines at the back of the heel, produced by the sharp indentations in their tarsal pads? These may play a role in stabilising the foot during very high acceleration.
Now....looking at these five images could you tell they come from five different animals? That's tough....and where FIT comes in.....
We'll be updating regularly on how the analysis pans out as AJ and Kristina at Okonjima, and our other cheetah partners in Africa collect more prints, and send news from JMP software in the USA as FIT v.1 nears completion. Stay tuned....
|
|
 |
  
|
Gomes |
Jago |
Morticia |
Pugsley |
Wednesday |
|
September 2011
Earlier this year, we had the great pleasure of a visit to Okonjima – home of the AfriCat Foundation, at the kind invitation of Wayne Hanssen and his family. Our mission was rather different from other visitors - we were not there to look for cheetah, or leopard, or hyena but rather we were searching for their footprints.
WildTrack (www.wildtrack.org) is our endangered species monitoring organisation. We were in Namibia co-hosting a carnivore monitoring workshop at N/a'an ku se Wildlife Sanctuary, which Wayne Hanssen – representing AfriCat, attended. Our primary interest is in using footprints to monitor endangered and elusive species. Using our Footprint Identification Technique (FIT) we can identify individuals, and sometimes also their age and sex.
Monitoring endangered species is central to successful conservation. Take the cheetah as an example. Unless we have accurate figures for the numbers of cheetah in Namibia, and where they are, it's very difficult to implement a successful conservation strategy. Yet current estimates range from around 2,500 animals, up to more than 9,000.
How can we get more accurate figures? In the past 30 years radio-telemetry (fitting collars, or transmitters) has been widely adopted, but as anyone using this technique will tell you, it is very expensive, collars are often unreliable, and the animal has to be routinely immobilised for the procedure. That's not to say that collaring doesn't have a role - it can be invaluable for monitoring a few animals if funding is not an issue and welfare is carefully maintained - but it's just not practical for population censusing.
 Some years ago we worked in Hwange National Park in Zimbabwe, monitoring black rhino. At that time the Govt. of Zimbabwe had decided to radio-collar every rhino in the hope that this would help protect them from poachers. Over a 10 year period our research found that not only did radio-collars fail at a very high rate, but that female rhino fertility declined dramatically as the animals were repeatedly immobilized for re-collaring. During these years we'd been out in the bush tracking with expert local trackers who were able to identify individuals from footprints. That's when we thought about trying to translate their ancient tracking skills into a format accessible for non-experts, with the benefit of modern technology.
After many years of trial and error, and working with world-experts from such diverse fields as image-recognition, software engineering, mathematics, statistics and forensic science, we published our findings. These attracted attention from wildlife biologists all over the world who were interested in using footprints as a new approach to monitoring their study species. We now have a footprint identification system for several species, including black and white rhino, the Bengal tiger, the Polar bear, mountain lion (cougar), striped hyena, Baird's and lowland tapir and even small mammals at the species level. Development is in the pipeline already for the cheetah, African elephant, brown hyena and leopard.
|
  
|
|
FIT is able to identify individuals from measurements of their footprints. Each species has a unique geometric profile of the measurements which allow us to identify them at the required level. For most species FIT has provided an accuracy of more than 95% in individual identification. All that's needed to collect images is a digital camera, a GPS, and a ruler! The data processing can be done back at base on a laptop computer using JMP data visualisation software from the SAS Institute.
As well as the science, FIT has considerable local community benefits. Because it is based on traditional tracking techniques, local scouts, rangers and trackers find it a very accessible tool. For little or no extra cost they can carry simple digital cameras in the field and collect footprint images. Their participation in conservation will also help make it sustainable - all conservationists agree that conservation can only be sustainable with the support of local people.
Our next step is to bring this technique into the field as a stand-alone package for wider use by wildlife managers and custodians, scientists and researchers. We receive support from JMP software from the SAS Institute, in North Carolina, USA, to streamline the imaging and statistical components, and introduce new features to allow animal distributions to be mapped - all in one package. We hope to develop this for mobile field devices, such as the iPad or Android tablets, so that data can be gathered, processed and relayed direct from the field.
  
We're working together with a group of Namibian and Botswanan wildlife organisations to get FIT into the field where it is needed.
Wayne Hanssen and the Okonjima, AfriCat & PAWS Teams will be collecting footprints from cheetah, leopard and brown hyena in the first instance, from which we will form a reference database for these species. FIT will then be able to monitor these species on site, using footprints.
We're very excited to be working with the AfriCat family to bring this vision to reality for Namibian wildlife!
Keep watching this space,
Zoe Jewell and Sky Alibhai
Sky Alibhai, D.Phil
Zoe Jewell, M.A., M.Sc., Vet. M.B., M.R.C.V.S
Apartado 210, 8550-909 Monchique, Portugal
& The SAS Institute, SAS Campus Drive, Cary, NC 27513
info@wildtrack.org www.wildtrack.org |
|
