Poachers in Africa Have a New Enemy: Math

Researchers have devised a mapping method to predict where wildlife poachers will operate, making it easier to catch them.
(Photo: Wolfgang Kaehler/LightRocket via Getty Images)
May 28, 2015· 2 MIN READ
John R. Platt covers the environment, wildlife, and technology and for TakePart, Scientific American, Audubon, and other publications.

How can a team of just 70 rangers more effectively protect the wildlife in an African national park the size of Rhode Island?

One word: statistics.

That’s the theory, anyway, behind a new paper published this week in the journal Conservation Biology. The study used a new statistical model to look at 12 years of illegal activity within Uganda’s Queen Elizabeth National Park as a way to predict the possibility of future crimes.

It’s not quite Minority Report, but it may help foretell behavior and therefore protect the park’s elephants, hippos, buffalo, and other heavily poached wildlife.

The problem with existing antipoaching strategy is that rangers tend to concentrate their efforts on areas where they have previously encountered illegal activity. They may continue to find poachers there but fail to visit other parts of the park where crimes are being committed, said Andrew Plumptre, a coauthor of the paper and the scientist at the Wildlife Conservation Society who devised the idea.

“Over the 12 years of data we examined, there were many areas of the park where the patrols have never been,” he said by phone from Uganda. Because there is no data about poaching in those areas, the rangers have little incentive to visit them.

The new model may change that. It shows that poachers are creatures of habit: They come back to the same locations over and over again. More than that, though, they come back to the same kinds of locations time and time again. For example, Plumptre and his coauthors found that noncommercial poaching for the bushmeat trade was concentrated in wet areas and near rivers, which may provide the overgrowth to conceal snares and create a “funnel” to guide wildlife into those snares.

RELATED: Supercomputer-Powered Drones Shut Down Rhino Poaching in This Park

By looking for areas in the park that match those criteria, Plumptre said, rangers can target their patrols and hopefully catch or stop more poachers.

“This approach is highlighting areas that rangers haven’t been to very much and trying to target patrols to these areas, and perhaps not wasting as much time in areas where they haven’t been finding very much but they’ve been patrolling quite regularly,” Plumptre said.

The model also maps out different types of crime, such as commercial poaching of elephants, fishing, logging, and encroachment of livestock onto national park land. The approach, the authors said, will help the park to budget its resources and manage different kinds of threats.

“If you focus exclusively on elephant, hippo, and buffalo poaching, you could be much more effective by deploying rangers to the central areas of the park, but you could lose a lot of antelope and other animals to snares at the edge,” Plumptre said. “The optimum trade-off is something we are working on at the moment.”

Will the theory work? WCS is already working with rangers to test out the models by visiting areas pinpointed by the new mapping method. If the theory proves useful, they hope to roll out new software that will allow them to update their models on a monthly basis.

They also hope to expand the strategy into other national parks across the world, all of which, Plumptre said, have a need to patrol more wisely and effectively. “This new research is allowing us to do that scientifically,” he said.