Back to Stories

People Beat Machines In Ability To Identify Wildlife

Although AI is frighteningly on the rise, study finds human eyesight and brains still hold advantage in recognizing subtle differences between pumas

Whether it is identifying pumas in Patagonia, wolves in Yellowstone's Lamar Valley, or the differences in appearance of members of Grizzly 399's extended clan in Jackson Hole, citizen wildlife watchers have demonstrated a remarkable adeptness in telling animals apart. Here, a puma races through Chile. Photo courtesy/copyrighted by Nicholas Lagos/Panthera
Whether it is identifying pumas in Patagonia, wolves in Yellowstone's Lamar Valley, or the differences in appearance of members of Grizzly 399's extended clan in Jackson Hole, citizen wildlife watchers have demonstrated a remarkable adeptness in telling animals apart. Here, a puma races through Chile. Photo courtesy/copyrighted by Nicholas Lagos/Panthera

by Todd Wilkinson

Is our species smarter than robots and other so-called “thinking machines” that fall under the umbrella of Artificial Intelligence or AI?

In the future—or so sayeth lyrics from the classic 1970s pop song, “In The Year 2525”—we’re not going to need truth, nor, for that matter, our teeth, eyes, arms and legs, because non-organic machines will be doing all the existential pondering and manual labor for us.

Yes, we’re racing headlong into uncharted territory with AI and gene modification, and no one can seem to tell us exactly where we’re headed. Are you worried? 

For now, however, here's some good news: humans are superior to digital computing firepower in at least one category. It's evident in our ability to better read feline forms and nuanced patters in their faces. 

According to the international wildcat conservation organization, Panthera, humans hold an advantage over facial recognition technology in detecting the difference between individual pumas. Indeed, a pair of much older technologies that our ancestors used to survive in Nature—our eyesight and brains—are, as Panthera note, still "giving AI a run for its money."

The findings grew out of a simple experiment, the results of which are published in the journal Science of the Total Environment. It pitted the acumen of human tour guides and “citizen scientists” on one side and AI on the other. The test was who was more adept at identifying individual physiological features of pumas in the Patagonia region of Chile’s Torres del Paine. (Note: in our part of the world, pumas are more commonly known as “mountain lions” and “cougars.”

To conduct the study, researchers created an online survey where participants were tasked with matching the faces of individual pumas. It found that responders achieved a 92.2 percent success rate in IDing pumas, a higher score than was achieved with automated facial recognition. 

The results mirrored what emerged in studies involving other species. For example, facial recognition technology, similar to what exists when we unlock our cell phones, achieved an 84 percent rate in recognizing brown bears and 88 percent success in telling harbor seals apart.

“Monitoring wildlife populations to determine changing abundance is the basis for conservation strategies and interventions. Monitoring, however, is expensive, and we lack baseline data for countless species and landscapes around the globe,” the study’s authors note. “One solution is to utilize methods that leverage observations collected by everyday people. Humans are not only excellent sensors for diverse data, but possess a remarkable ability to process data and differentiate patterns with minimal training.”

The study was led and co-authored by Panthera’s Puma Program Director Dr. Mark Elbroch who has spent years studying mountain lions in Greater Yellowstone. This winter, Elbroch has been in Patagonia where he says evidence indicates that the Torres del Plaine biosphere holds one of the large puma densities in the species’ range through the Americas. He is author of the critically-acclaimed book, Cougar Conundrum: Sharing the World with a Successful Predator.
Above is an example of the test given to participants in the Panthera study to determine if they could determine if the two images, presented side by side, were the same puma. Graphic courtesy Panthera
Above is an example of the test given to participants in the Panthera study to determine if they could determine if the two images, presented side by side, were the same puma. Graphic courtesy Panthera

This week, I had a remote interview with Elbroch who is still in South America. I asked him what the most important take-home of the study is and if there was anything which surprised him? “I would say it confirms that people are super computers,” he said. “That we under-utilize the skills of people working in the field in capacities other than traditional science—in this cast specifically the puma guides in Patagonia, but it would apply to diverse people in diverse settings.”

This is my speculation, but odds are that this finding would also be paralleled in Greater Yellowstone with the high level acuity that  exists among professional scientists and avid wildlife watchers in detecting the identifies of individual animals. For example, wolves in a place like Yellowstone’s Lamar Valley or grizzlies in Yellowstone and Jackson Hole, the latter being the domain of Grizzly 399 and several different broods of cubs.

Perhaps an extrapolation could also be extended to avid followers of the late cougar P-22 and others in southern California where Beth Pratt with the National Wildlife Federation and colleagues closely tracked big cats trying to survive in that massive human megalopolis. (Before her high profile work in California, Pratt spent several years working for the hotel concessionaire Xanterra in Yellowstone).

With Panthera being a leading global cat conservation organization, I wondered what Elbroch thought the best ways were to educate citizens who have a huge interest in protecting felids, be they cougars, bobcats, Canada lynx, jaguars, African lions, cheetahs, leopards, and tigers, to name a few. And I further asked if there are lessons for people involved in U.S. conservation to win over public support?

“Big questions and likely complex answers,” he replied. “Education of course—but what kind? What are the impacts of a film versus written word? A book versus a science article? Social science can provide some guidance. Moral-ethical arguments are better at changing beliefs—showcasing the positive roles of carnivores does improve beliefs, so does tourism, and so does education about tools to mitigate conflicts with carnivores. But certainly we lack strong guidelines for best practices. Maybe people are too complex for such guidelines to exist, and they will always need to be tailored to a specific culture and community.”

Panthera’s Greater Yellowstone work has, by design, worked to expand public awareness and invite the public to play  more meaningful roles in protecting mountain lions and other carnivores. Indeed, over the last few years, Elbroch and crew have completed intriguing research and taken innovative approaches to better understanding not just the biology of mountain lions but how they fit into the larger web of life and interact with other species.

One memorable assessment published in the journal Biodiversity and Conservation in 2017, involved Elbroch, Panthera’s Jenny Fitzgerald and Jackson Hole conservationists Lisa Robertson and Kristin Combs assessing the eco-tourism value of live bobcats in Yellowstone. 

For a single bobcat that attracted photographers and wildlife watchers—and based upon what they were willing to pay to have the experience—they calculated its non-consumptive economic value to conservatively be $308,105 in just a single winter tourist season. That single bobcat had a financial value 1,000 times greater than the exploitive value of $315.17 if the bobcat had been trapped or hunted in Wyoming during the same season. 

Much of Panthera’s other field work in Greater Yellowstone and the West have taken deep dives into the behavior and home ranges of wildcats.

“Our GYE-derived insights into the secret social lives of cats and their positive contributions to natural systems remain model studies,” Elbroch notes. “ The work on competition with wolves also remains some of the strongest work on the impacts of one carnivore on another in a system with human hunting. Lots to celebrate from that work—and it’s still coming.” 

He said to stay tuned. Analysis is now underway to better understand the interactions between mountain lions, wolves, grizzly and black bears—four key members of Greater Yellowstone’s fully intact “predator guild.”  

Not only is it advantageous, but forerunning ecologist Aldo Leopold pointed out how ecosystem health depends on saving all of the biological parts. It's also true that having smart, knowledgeable people out every day keeping a watchful eye on wildlife is priceless. 


Todd Wilkinson
About Todd Wilkinson

Todd Wilkinson, founder of Mountain Journal,  is author of the  book Ripple Effects: How to Save Yellowstone and American's Most Iconic Wildlife Ecosystem.  Wilkinson has been writing about Greater Yellowstone for 35 years and is a correspondent to publications ranging from National Geographic to The Guardian. He is author of several books on topics as diverse as scientific whistleblowers and Ted Turner, and a book about the harrowing story of Jackson Hole grizzly mother 399, the most famous bear in the world which features photographs by Thomas Mangelsen. For more information on Wilkinson, click here. (Photo by David J Swift).
Increase our impact by sharing this story.
GET OUR FREE NEWSLETTER
Defending Nature

Defend Truth &
Wild Places

SUPPORT US
SUPPORT US