Facebook
Twitter
Pocket
Reddit
Email
The leopard’s roar is a distinct sound. A repetitive pattern of hoarse and guttural calls, it is often compared to the rough music of a saw being pushed and pulled through a giant tree trunk. These “sawing” roars are also highly individualized, shaped by each leopard’s physiology and environment, differing in measurable ways from one leopard to the next.
Now researchers have figured out how to tell leopards in Tanzania apart by analyzing the differing temporal patterns in their roars. The researchers were able to use their analyses of leopard roars to accurately identify a leopard 93 percent of the time, a method that could help conservationists track these big cats over much larger areas than they do now. Understanding the population sizes of endangered species is critical for conservationists and policymakers when it comes to managing landscapes.
Leopards are classified as vulnerable to extinction—and in some places in Africa they are now considered “locally” extinct—but they are difficult to study because they are nocturnal, solitary, elusive, and roam widely over home ranges of up to 300 square miles. Approximately one third of Tanzania already comprises protected areas for the preservation of wildlife and ecosystems, but as human populations continue to grow, so does demand for land and resources. Habitat loss is among the greatest threats to leopards, which increases risk of human-wildlife conflict: Attacks on livestock, a target for the big cats, often drive these conflicts.
The project required sifting through 75,000 hours of acoustic data, but was something of a happy accident, says Jonathan Growcott, lead author and a Ph.D. student at the University of Exeter. “We got really lucky,” he says.
Growcott was working with another biologist named Matthew Wijers to track animals in Tanzania’s Nyerere National Park. They had placed acoustic monitors and camera traps together in the same locations so that they could capture not just video but also audio of creatures walking past. Growcott asked Wijers if he thought they would hear the footsteps of the leopards, given they are so stealthy in the wild.
To find out, they began listening to the audio paired with video of the leopards, and noticed that as leopards were walking past the cameras, they were also vocalizing. This meant they could sort through the audio data by noting time stamps on the camera traps when the leopards appeared. While audio data was captured continuously, the camera traps only recorded video when triggered by motion.
Growcott and his colleagues gathered data from 50 camera traps that were paired with audio recording units across 174 square miles of Tanzania’s Nyerere National Park. Over 62 days, researchers recorded 217 roars from seven members of one of Africa’s most elusive and solitary species. Once roars were isolated, audio analysis algorithms allowed Growcott to determine that the frequency contours of those roars could be used to identify individual leopards with incredible accuracy.
As leopards were walking past the cameras, they were also vocalizing.
Studying animal vocalizations, a field known as bioacoustics, has long been used to monitor birds, primates, and marine species, but researchers have only recently begun to use it on big cats in Africa. Beyond the challenges of finding leopards in the first place, Growcott says that suitable technology is still emerging—earlier attempts to study vocalizations of lions, wild dogs, or spotted hyenas happened mostly with handheld microphones.
“Those studies relied on going out and tracking the animals, spending an awful lot of time in the field, and having to know exactly which individual you’re recording. It takes a lot of resources. With the low density of these animals, acoustic studies just didn’t happen.” Putting monitoring collars on the cats, another alternative, is expensive, invasive, and risky for researchers.
Kirsten Young, a senior lecturer in ecology at the University of Exeter who was not involved in Growcott’s study, says there are limitations of acoustic monitoring for conservation purposes. “We need the animal to be making sounds, and we know that there are periods where they are silent,” she wrote in an email. “If we can’t hear them, we don’t know that they are there.”
But capturing species vocalizations is worth the often lengthy pursuit. Young is a marine scientist who studies the ecology of deep-diving whales. Much of her work relates to conservation. Beyond facilitating population estimates, vocalizations can be an important indicator of animal behavior.
“They help us understand what an animal is doing,” says Young. For instance, sperm whales make special “creaks” when they are foraging or capturing prey, which helps scientists identify feeding areas.
“Sound is such a good indicator of ecosystem health,” says Growcott. “We need to be measuring these sounds before it’s too late.” 
Lead image: Los t / Shutterstock
Get the Nautilus newsletter
Cutting-edge science, unraveled by the very brightest living thinkers.
