With this technique being used in over 20 different projects around the world, our next big challenge will be to keep pace with the demand for FIT - to get it to where it's needed faster and more effectively.
Imagine a time long before modern technology, before the industrial revolution, before the wheel, even before agriculture. Humans had evolved a complex and powerful traditional ecological knowledge that enabled them to track and hunt effectively. This knowledge was arguably the origin of science. It required rigorous observation, deduction and repeatability. Those who acquired it lived to breed more, and prospered. But fast-forward a few thousand years and traditional ecological knowledge is vanishing with the decline in hunter-gatherer societies.
Some years ago, we spent a decade in Zimbabwe, researching black rhino population dynamics. We worked with local trackers, observing their extraordinary skills as we tracked through the bush for long hours every day in the blazing sun. The trackers who accompanied us, armed with old Russian AK47 rifles to ward off any threat from wild animals or poachers, used to laugh at our earnest attempts to track rhino using telemetry by by following erratic bleeps from their collars.
Why go to all the trouble of chasing a rhino, usually with a helicopter, in the heat of the midday sun, just to fit a VHF or GPS collar for tracking? All the evidence, they reasoned, was on the ground right under our nose in the shape of the animals' own footprints.
Truthfully, we were sceptical. Surely identifying animals from their footprints was bushcraft, not science? We settled into collecting data from the radio-collared rhino and after a few years had an unprecedented dataset, which revealed an alarming trend: Female rhino that were darted frequently for re-collaring (collar failure rate was high) had significantly lower fecundity, obviously a disaster for an endangered species.
We went back to the trackers. We already knew they could follow animal trails over thick grassland and rocky terrain, and amazingly, we found they really could identify which animal had made those footprints. And footprints, unlike the animals themselves, were ubiquitous.
Now this really grabbed our attention triggering the idea that footprint identification would be a less-invasive and more cost-effective way of identifying individual animals. We set about trying to distill just the very basics of the trackers' knowledge. Our first attempt was to trace rhino footprints and measure them. We spent three months with our noses in the deep red African soil, hunched over footprints. We did some preliminary biometrics. It was a total failure. Nothing matched, nothing made any sense.
With this technique being used in over 20 different projects around the world, our next big challenge will be to keep pace with the demand for FIT - to get it to where it's needed faster and more effectively.
Depressed, but undeterred, we adopted digital cameras, revised and standardized our data collection protocol and tried again. We purchased a copy of JMP data visualization software and with their help wrote a customized script for measuring footprints automatically. That was our true ‘AHA' moment becausewWhat the trackers had seen in their minds, we suddenly could see before us.
It was a revelation. The visual analytics allowed us to ‘see' differently. Not only did the software enable us to classify footprints by individual, we were also able to discriminate age-class and sex. We published our first paper on the footprint identification technique (FIT) in 2001. It caused quite a stir and before long we were inundated with requests from other researchers to use FIT for a range of species.
We now have algorithms for 15 species and more than 20 projects around the world but the icing on the cake has come with our most recent paper, published this month. It communicates FIT at a new level in the world's first peer-reviewed video-journal; the Journal of Visualized Experiments (JoVE). This a perfect forum for FIT as a technique rooted in visual analytics.
With our colleagues at the N∕a′an ku sê Wildlife Foundation in Namibia and Manchester Metropolitan University (UK), we have been able to demonstrate every step of the FIT software process in the cutting-edge software JMP from SAS and our next big challenge will be to keep pace with the demand for FIT - to get it to where it's needed faster and more effectively.
We have an exciting new initiative working with partners in Southern Africa, the UK and the USA, to develop the world's first open-access footprint database. A new partnership with the American Association of Zoos and Aquariums will aid the development of the database, significantly streamlining algorithm development making it possible to reduce development time and we will now start engaging fixed and mobile networks of Citizen Scientists to help us with the necessary stages of training datasets, algorithm development and field validation
Our footprint identification technique sprang from an unexpected source - traditional ecological knowledge. Yet it uses only a tiny sub-section of the skills we observed in trackers over the years. There is so much more to be re-awakened, revived and interpreted. Critically, integrating traditional ecological knowledge into conservation strategies also engages local communities - the key stakeholders in global conservation.
Technology has massive potential to transform the way we conserve endangered species, but our own research has pointed out that there are costs in not integrating this effectively into conservation strategy. If we are searching for techniques that work safely and sustainably within the natural world, what better resource and foundation than traditional ecological knowledge, which has been shaped, tested and evolved so rigorously by natural selection over millennia.
The Authors
Drs. Zoe Jewell and Sky Alibhai co-founded WildTrack (wildtrack.org) in 2004 to address a widespread need for less invasive and more cost-effective tools to monitor endangered species. This mission works hand-in-hand with a need for local conservation efforts to engage local people. WildTrack's award-winning footprint identification technique (FIT), based on traditional tracking skills, has now been developed for species ranging from black rhino to Polar bear and mountain lion. Jewell and Alibhai are Principal Research Associates in the JMP Division at SAS software (the world's largest private software company), Adjunct Faculty at Duke University's Nicholas School of the Environment, Adjunct Faculty at North Carolina State University's Department of Mechanical and Aerospace Engineering, and Associate Academics at the Centre for Compassionate Conservation of the University of Technology in Sydney, Australia.
.