The use of drones to fight elephant and rhino poaching could prove vital to the endangered elephants in sub-Saharan Africa.
Tests conducted last year (2015) by Air Shepherd Technology, a US-based non-profit organisation, in partnership with the Lindbergh Foundation, found that drones could be highly effective in spotting poachers and alerting nearby rangers.
According to the new Elephant Status Report (ESR) released at the end of September (2016) by the International Union for Conservation of Nature (IUCN) African Elephant Specialist Group, the number of forest and savannah elephants in Africa decreased by more than 100,000 elephants since 2007. The report covers all 37 nations in the elephant’s range and includes data from more than 180 surveys with estimates suggesting there are now only 415,000 elephants in Africa. A further 117,000 to 135,000 elephants may also exist in areas not yet systematically surveyed.
The ESR includes the survey data collected by the Great Elephant Census (GEC), whose results were released earlier in September, but adds data from the forests and remote savannah populations. The GEC conducted aerial counts of savannah elephant populations in most of the range states and found a decline of 30% in 7 years. The IUCN Status Report compiled aerial surveys from savannahs and ground dung counts in the forests and reported a 22% decline across the entire range.
Patterns of poaching
In an article for National Geographic magazine, Oliver Payne writes that drones can complement the efforts of park rangers and others to provide comprehensive poaching data. “New data is acquired daily from drones, tour operators, rangers on patrol, and GPS collars on individual animals. In aggregate, the analytics reveal patterns of poaching attacks and can predict with 90% accuracy where poachers will strike.”
While experts agree drones are not yet the ‘silver bullet’ in the fight against poaching, mostly because of the high cost of drones as well as limitations in flight time due to short battery life, Zimbabwe announced in August that it is about to deploy drones in its biggest wildlife sanctuary, Hwange National Park in the west of the county bordering Botswana, to combat the poaching of elephants.
The New Zimbabwe newspaper quotes a park official as saying at least two drones have been purchased with the aim to test the effectiveness of utilising data from drones to inform park rangers of movements on the ground. Tourism contributes 11 percent to Zimbabwe’s $14 billion economy with the country’s wildlife parks popular with overseas visitors.
Poachers have in the last two years killed dozens of elephants in the park by lacing watering holes with cyanide, a toxic substance that kills within hours. Hwange holds two thirds of Zimbabwe’s 80,000 elephants.
Drones in the African night sky
Thomas Snitch, a visiting professor at the University of Maryland’s Institute for Advanced Computer Studies, says the use of drones should be combined with high resolution satellite imagery, mathematics and algorithms to determine where animals are likely to be on any given night. “With this knowledge, we can fly drones in the African night sky with infrared cameras to alert us to where the poachers are coming from to attack the prey.”
Snitch did tests over 15 months at the Amakhala Game Reserve in the Eastern Cape of South Africa in efforts to curb rhino poachers. He says “you must know where to fly and this is where modelling comes in – Africa is so big that simply launching drones into the sky will have no impact. You must learn the patterns of how animals and poachers move and then use drones to be your eyes in the sky.
“We use very high resolution satellite imagery as a base and then layer on data covering everything from animal collar data, GPS routes of rangers, weather, ground intelligence, previous poaching incidents, moon phases, everything we can get our hands on. This allows us to statistically recreate the environment that was present during previous poaching events… We are constantly adding new data to the model – we get data dumps every week – and the model is continually learning. The model looks for changes in the data and teaches itself to monitor new patterns. This is critical because as our work successfully catches poachers, we must believe that their behaviour will change. The key is early recognition of these changes so that we can stay one step ahead.”