Analytics Helps Save an Endangered SpeciesBy Maria Behan | Posted 2011-07-28 Email Print
Marwell Wildlife, an international conservation charity and zoo, has added predictive analytics software as the latest tool in its 15-year effort to protect the Grevy’s zebra.
Predictive analytics technology, which maximizes the visibility of data and uses it to anticipate what will happen next, is increasingly coming into play—and in a wide range of contexts. Merchants use predictive analytics software to identify and target their most profitable customers, while insurance companies deploy it to help them decide whether to process a claim or flag it for investigation. Police departments enlist predictive analytics to focus resources on where and when crimes are most likely to happen, factoring in variables as diverse as crime statistics and weather patterns. But one of the most intriguing uses of predictive analytics isn’t focused on snaring consumers or criminals: It’s aimed at saving zebras.
Marwell Wildlife, an international conservation charity and zoo based in Hampshire, England, has added predictive analytics software as the latest tool in its 15-year effort to protect the Grevy’s zebra, an endangered species whose population in the wild is estimated to have fewer than 2,500 animals.
The organization is using the software to analyze an array of information from the field, including data from aerial surveys, camera traps and radio collars. That analysis gives the conservation group a detailed understanding of both the threats facing the zebras and the policies that would help the population bounce back.
“We’re using predictive analytics to focus our resources on areas that really matter,” says Guy Parker, head of bio-diversity management at Marwell Wildlife.
Some of the most illuminating data that the organization is analyzing is from a survey of nomadic herdsmen in northern Kenya, the area where most of the remaining Grevy’s zebras live. “The people in the local pastoral communities know a great deal about the landscape and its wildlife, including the zebras,” Parker explains.
To mine that wisdom, field workers from Marwell Wildlife and its partners in the survey—the Northern Rangelands Trust, Lewa Wildlife Conservancy, Grevy’s Zebra Trust, Kenya Wildlife Service, and both the St. Louis and Denver zoos—spent a month in remote areas conducting a questionnaire-based survey. “We gathered information about population distribution and threats to zebras,” Parker says, “but we also learned about people’s attitudes.”
The survey confirmed what conservationists have long recognized: The needs of wildlife and humans are often at odds. Some herdsmen were hunting the zebras, but there were also more subtle forms of conflict. “The zebras live in dry areas alongside people raising livestock, so there’s competition for water and pasture between the livestock and the zebras,” Parker explains.
To drill deeper into the survey data, Marwell Wildlife used IBM SPSS predictive analytics software, which it loaded on a single laptop, for insight into human practices and attitudes that may have an impact on the zebra population.
“With predictive analytics, we can now look at seven or eight variables—things such as education level and whether or not people had been previously exposed to conservation efforts—to better understand the reasons behind people’s attitudes,” Parker says. “That’s the real beauty behind predictive analytics: It lets you tease out complex patterns you otherwise couldn’t see.”
Some of the patterns related not to the zebras, but to the humans living alongside them. “We discovered that many people see benefits from living alongside the zebras,” Parker says. “For instance, zebras lead people to pasture in dry years, and they attract tourists.
“That kind of information is incredibly useful, because it helps us work out where there’s potential for conservation, and we can build on that. We’ve determined that the most promising areas for us to focus our efforts on are communities where there are already positive attitudes toward the zebras.”
Marwell Wildlife’s survey and subsequent analysis also unearthed another fact that might help save the Grevy’s zebra: They’re hunted less for food than for traditional medicine.
“Zebra fat is valued highly by pastoral communities for ailments ranging from headaches to tuberculosis,” Parker reports. “Many survey respondents said they’d be happy to move to conventional medicines if they were available. That’s a promising avenue that has the potential to benefit people and zebras.”
The organization’s use of predictive analytics is moving beyond its conservation efforts. “We don’t have many people on our fund-raising team, but they’re hoping to increase their effectiveness by using predictive analytics to analyze our donor base: who gives to Marwell, when they give and the factors that influence those choices,” Parker explains. “Since we first came to this from the scientific side of things, that’s an entirely new area for us.”
Marwell Wildlife has other plans to expand its use of predictive analytics. “We’ll definitely use it in the future,” Parker says. “For instance, we’re talking to IBM about integrating Landsat [satellite] imagery into our predictive analytics system. The more data we can put in, the more we can learn.”
Behan is a freelance journalist with a background in technology.