How Paul Allen’s $7 million and big data are combating Africa’s elephant crisis



Date Published

See link for photos and video.


Africa’s elephant population is in crisis.
Some 30,000 were killed for their tusks in 2012, according to the International Fund for Animal Welfare, and by some estimates, the population could be decimated in a decade if poaching continues at this rate.
That’s where Microsoft cofounder Paul Allen’s Great Elephant Census, the project he started with a $7.3 million grant and a partnership with the Botswana nonprofit Elephants Without Borders, comes in.
Nobody knows exactly how many elephants are left in Africa, which makes them extremely difficult to track and protect from poachers and other threats. The GEC seeks to solve that problem by counting the continent’s savanna elephants, living and dead, over two years. The idea is that, with accurate count data, researchers and policy makers not only will know how big the population is, but better understand the pachyderms’ dynamics as a species. That, in theory, will make the dwindling population easier to protect.
“The threat of local extinction feels very real,” Mike Chase, director and founder of Elephants Without Borders said when the project was announced last year. He recounted how he and his organization “flew a survey over a park where we had previously counted more than 2,000 elephants. We counted just 33 live elephants and 55 elephant carcasses. That is why this research is so important.”
(Population estimates cited by the Census vary from 410,000 to 650,000, though scientists at the Udzungwa Elephant Project in Tanzania estimate that the number of living elephants is closer to 250,000.)
Take Satao, one of the top 20 largest elephants left on the continent, Dr. George Wittemyer, an assistant professor at Colorado State University, said at the SXSW Eco conference in Austin, Texas, in October. Wounded in February, Satao was finally killed by poachers for his tusks on May 30 in Kenya’s Tsavo National Park. Born sometime in the late 1960s and weighing over 7 tons, the bull was one of an “estimated 100,000 [elephant] deaths in the last 3 years,” Wittemyer said.
That’s a tragic example of what scientists like Wittemyer call an acute stressor, though the species is also suffering from the “chronic stressor” of reduced habitat.
(The video below contains imagery of Satao after he was killed and stripped of his tusks that some viewers may find disturbing.)
The Allen-funded team of over 50 researchers started surveying 21 countries and 600,000 square kilometers of Africa in February. The two-year survey is comprised of “basic scientific work,” Wittemyer said in Austin, “to give numbers to policy makers, which they rely on to make decisions” to protect the species.
That means four people make observations from a Cessna flying at 300 feet. A constant altitude keeps count numbers consistent and each team employs one to three planes.
The task is not as straightforward as it may sound. “Elephants are being poached as we’re trying to count them,” said Ted Schmitt, senior program manager-conservation at Vulcan, Inc., the company Allen started in 1986 to manage his philanthropic projects. (The project will not cover forest-dwelling elephants, because they cannot be counted aerially.)
Researchers in the Cessnas are now sampling 3 to 6% of the targeted regions, and then using “analytical extrapolations to figure out total population from that area,” Schmitt said. To date, GEC team members have completed surveys in four countries, with twelve currently in process. They have flown over regions in Kenya, Tanzania and Zimbabwe, to name a few.
Surveyors have just finished counting the elephant population in Botswana, for example, after two months. Chase recently reported that there were “no fresh poached elephant carcasses seen on the entire survey.”
The aerial technique is a classic method of counting a wildlife population, because the threat to the species doesn’t allow for the luxury of time. Schmitt and his team, however, are “actively looking … to do this type of census in ways that use technology more effectively, taking cost down,” which would allow for more frequent counts.
For starters, the team at Vulcan helped researchers to automate some of the more tedious parts of the data collection process required for a quality survey.
“We built a tool for them we’ve … called our flight data logger,” Schmitt said in an interview. “We actually used a 3D printer to create the mount and we used a standard Android tablet,” he said, to create a device that would help reduce researcher fatigue and improve accuracy. “We wrote an app for that and basically equipped a device and application to record a lot of this survey data automatically,” he continued. That has “really taken the burden off the surveyors.”
The flight data logger, developed by Vulcan, in the cockpit of a Cessna, in the orange mount.
Surveyors were also provided with laser altimeters for their planes, which connect with the data loggers. That too allows for more accurate measurement of data. The Vulcan team has also built a secure website for researchers to submit data from the field. The secure site keeps data out of the wrong hands and allows for quality control measures to be implemented as it is collected.
Part of the GEC, too, is imagining and testing technologies that can streamline the survey process in the future, exploring methods that will allow more efficient surveys, tracking and protection of both the animals and their interaction with their dwindling habitats.
“We and others are looking at new survey collection methods, everything from drones to light aircraft … and the application of image recognition technology,” Schmitt said, indicating that the latter could potentially be used to identify and count the animals automatically.
“I suspect in the coming, say, three to five years we’ll see very new and hopefully more efficient ways to do these big surveys.”
As in the U.S., regulations on private drone use are still being tackled in African countries. Vulcan researchers, at least, figure they may as well be prepared with drone technology for use in future conservation scenarios, so they continue to study the possibilities.
Lucas Joppa, a Microsoft Research conservation scientist who also spoke on the panel in Austin, supported the idea that the current technological limitations to wildlife conservation are surmountable.
“There’s ways around all of this,” he said. “It just requires enough people to care.”