Impact Newswire

INTERVIEW: Machine Learning Cracks the Code on Lion Communication

INTERVIEW: Machine Learning Cracks the Code on Lion Communication
Credit: Panthera

A new era of AI-aided acoustic research has opened a window into the complex social world of the African lion, yielding a stunning discovery: the King of the Jungle has more than one roar, and its communication is nuanced by regional “dialects.” 

This finding, that lions possess a second type of roar and exhibit geographical variations in their vocalizations, is set to upend decades of established zoological understanding and offers a fresh angle on one of the great remaining mysteries of the wild: why lions are the only truly social cat.

The lion’s social structure, the pride, is a unique evolutionary anomaly in the solitary world of the cat family. For a comprehensive look at what this new acoustic data means for conservation and the ongoing debate over lion sociality, AI Editor Faustine Ngila spoke to Dr. Andy Loveridge, Director of the Lion Program for the wild cat conservation organization, Panthera. 

Born and raised in Zimbabwe, Dr. Loveridge received his Ph.D. in Zoology from Oxford University, and his research on lions has been published in 150 peer-reviewed papers. 

He is the author of Lion Hearted: The Life and Death of Cecil and the Future of Africa’s Iconic Cats – a book he wrote about Cecil the lion, whom Dr. Loveridge studied for eight years before the former was trophy hunted – leading to much media coverage – in 2015. He is also a contributor to IUCN Species Survival Commission Panthera Leo species conservation guidelines and IUCN Red List assessments.

Here is the interview, edited for clarity.

Ery of a second type of lion roar change our understanding of lion communication and social structure?  One thing in particular that this finding tells us is that we are still learning about the ecology of even the most social of wild cats that many from the scientific community may believe we fully understand. Whilst we knew from previous research that each lion’s roar is unique, it was not known that different components of the roar could be systematically identified using machine learning. This would further assist scientists in identifying the calls of individuals and hence monitoring lion populations using acoustic recorders. 

The study suggests lions have distinct “dialects” depending on geography. What does this imply about lion populations and their interactions across different regions? It seems likely that lions, as many other organisms, have regional ‘dialects’.  Ever-adaptable wild cats like the lion are known to change their behaviors, such as hunting preferences, depending on the topography of their landscape, availability of certain prey species, competitors and more. It’s not too far of a stretch to learn of these new findings regarding lions exhibiting varying dialects based on their locations. The authors of the paper in question suggest this is the case, but this would need to be verified in a dedicated study using a large sample of lion calls from multiple locations. 

How did AI help scientists identify this new type of roar, and what does this tell us about the potential of AI in wildlife research? AI has great potential to aid wildlife conservation and research, greatly enhancing and accelerating data analysis in ways that were unheard of even a few years ago. For instance we used AI to rapidly and accurately identify species in camera trap photos taken by Panthera’s lion program using the PantheraIDS data management platform. PantheraIDS uses artificial intelligence to automatically classify species and identify unique patterns of individual animals. Particularly considering camera trap surveys incorporate thousands of images, this tool makes arduous data processing projects quick and efficient. AI-aided acoustic monitoring is also helping to bring justice to the jungle and eavesdrop on coral reefs, for instance, and the impressive list of new tools goes on.  It is important, however, to remember that while AI is a highly valuable conservation tool, it will never replace the invaluable human scientists, law enforcements rangers and other wild cat advocates working on the frontlines to monitor and protect wild cats.  

Beyond this study, how might AI tools transform our understanding of lion behavior and other complex social patterns in the wild? We expect that developments in AI and machine vision will allow to accurately identify individual lions from photographs and videos, which will greatly enhance our ability to monitor and conserve populations. 

Lionesses often give birth around the same time and engage in alloparenting. How do these behaviors enhance cub survival and pride cohesion? True to their social nature, lionesses often get pregnant and give birth at the same time and also raise their cubs together in crèches, practicing alloparenting. Young lionesses will often practice their mothering skills by caring for cubs that aren’t their own and lionesses also suckle cubs that aren’t their own, including the occasional leopard.  Communal suckling and communal care makes cub-rearing (and life in general) easier for lionesses and safer for cubs. This may be one of the reasons that sociality evolved in lions. Rather than relying on one maternal source for food, training, rearing and care, lionesses band together as a team to fill much-needed gaps in raising young and protecting them from the very real human and wildlife threats that exist across lion landscapes. The more eyes and bodies dedicated to protecting cubs increases the likelihood of survival for such young.  

Male lions sometimes form coalitions, usually with brothers. How does this strategy impact pride defense, takeover dynamics, and genetic diversity? Just as lionesses band together, male lions form similar alliances with brothers and occasionally with strangers to defend their pride or take over a pride from a competing male. These brutal and often fatal takeovers lend the victor(s) the mating rights of a pride. At their core, male lions understand that bringing friends to a fight greatly increases the chances of coming out alive. The consequence of losing a fight for a pride is that the victorious male or coalition kills the vanquished male’s offspring. This allows the new males to sire their own cubs and pass on their genes. 

What role do spotted hyenas play in theories about why lions form prides, and how does predator pressure shape lion social strategies? Predators compete for food and evolve behaviors to protect their kills from competing species. Leopards, for instance, hoist kills into trees. Lions evolved alongside another social apex-predator – the spotted hyaena. One possible explanation for the evolution of group living is to protect kills and offspring from these potential competitors. Hyaenas will often steal kills from small groups of lionesses and never miss the opportunity to kill lion cubs. Power in numbers allows lions to ward off these attacks. 

From a conservation perspective, how can understanding lion communication, social behavior, and geographic dialects inform efforts to protect lion populations in Africa? While no single piece of information or technology will ever serve as a silver bullet to save lions and other wild cats, there is truth behind the ‘knowledge is power’ phrase. The more we learn about these species’ ecology, population sizes, use of their range, adaptability to threats, behaviors, reproductive capabilities and what is threatening their survival the better prepared we will be as a conservation community to adapt our strategies and better conserve wild cats.  

Get the latest news and insights that are shaping the world. Subscribe to Impact Newswire to stay informed and be part of the global conversation.

Got a story to share? Pitch it to us at info@impactnews-wire.com and reach the right audience worldwide


Discover more from Impact Newswire

Subscribe to get the latest posts sent to your email.

Scroll to Top

Discover more from Impact Newswire

Subscribe now to keep reading and get access to the full archive.

Continue reading