Kasim Rafiq had spent all day in the woodlands of Botswana looking for a one-eared leopard named Pavarotti—with no luck—when his Jeep sank grill-first into an abandoned warthog burrow. Two hours later, frustrated and exhausted after extricating his vehicle, he passed some tourists on safari. He told the guides about what had happened, and they laughed. They’d seen Pavarotti earlier that morning.
The encounter got Rafiq, a wildlife researcher at Liverpool John Moores University in the United Kingdom, thinking: Wildlife tourists encounter animals like Pavarotti every day and take hundreds of photos. But once the tourists leave, they take those data with them. Rafiq wondered whether there was a way to use the photos they took for research. Now, a monthslong trial with two dozen different groups of tourists suggests not only is it possible, but it’s also far cheaper than traditional methods of tracking.
To count large animals like lions and hyenas in a given area, researchers typically use one of three methods. They use motion-triggered camera traps to photograph passing animals. They look for tracks by driving along predetermined routes. Finally, they bring the animals to them by playing sounds at specific stations. (If you’re interested in lions, Rafiq says, “you play the sound of a dying wildebeest.”)
But all these methods have their drawbacks. Camera traps, for instance, are expensive and easily knocked down by curious creatures. When it’s too dry or too windy, tracks don’t show up in the dust. And researchers often have difficulty obtaining permits for call-in stations, especially in areas with lots of tourists.
To test the method of using tourist photos, Rafiq and his colleagues surveyed the populations of several large carnivores in Botswana’s Okavango delta, including lions, hyenas, leopards, wild dogs, and cheetahs using traditional methods. They then had more than 50 people in 26 tour groups provide safari photos over 3 months.
Before they set out, tourists were outfitted with GPS devices to record their location at 1-minute intervals. When they returned, they uploaded more than 25,000 photos to Rafiq’s computer. Rafiq and his team tagged the photos with times and locations and identified individual animals based on patterns on their faces or bodies. They then estimated the number of animals in the area using computer models.
It worked: The researchers’ estimates using the photos were similar to those using traditional methods, at least for lions, wild dogs, and leopards, they report today in Current Biology. The tourists found far fewer hyenas than the researchers, but their photos were the only surveying method to identify cheetahs at all. The method was also cheap—a 3-month survey cost researchers less than $300 including labor and equipment, while a camera trap survey of the same length cost more than $9000.
“The paper looks solid,” says Greg Newman, a research scientist at Colorado State University’s Natural Resource Ecology Laboratory in Fort Collins, who has worked on citizen science projects in the past. “To me, this demonstrates that volunteer collected data can, when done in a very purposeful approach, generate high-quality scientific information—equal to, or perhaps in some cases even a little bit better than professional monitoring.”
There are some downsides, though. First, the photo survey method only works in areas where there are lots of tourists. It also does little to capture smaller or harder-to-photograph animals, as most tourists focus on charismatic megafauna such as lions and elephants. “It’s a tool in the tool bag,” Newman says. “You need to say, ‘OK, what’s the right use for this tool, in which context?’”
In the future, Rafiq says, the painstaking work of tagging thousands of photos could be outsourced to artificial intelligence. That, and streamlining the data collection process, will be his research focus for the next year while he is a Fulbright fellow at the University of California, Santa Cruz. Rafiq also hopes to partner with multiple tourism agencies to make such data collection a regular part of their routine. “With all the species currently at risk in the era we’re in, there is real potential to use this data for good,” he says. “You can really change people’s perspectives on certain issues by giving them a firsthand opportunity to experience conservation.”