You might think you know the roar of a from the iconic MGM logo that opens its films. However, the famous roar was actually voiced by a tiger. Compared to lion roars, tiger sounds are raspier and higher-pitched.
A team from the University of Exeter, led by researcher Jonathan Grocott from the Department of Ecological Research, claims that lion roars should be categorized into two distinct types: loud and intermediate, with the latter characterized by a more even and less varied sound.
Artificial intelligence played a crucial role in helping scientists reach this conclusion. They believe that identifying this feature could significantly impact lion conservation efforts. According to them, the findings from this new study could facilitate monitoring lion populations, which in turn would simplify the protection of these animals.
Currently, the total population of wild lions in Africa is estimated to be between 22,000 and 25,000 individuals, a figure that is half of what it was 25 years ago. The primary reason for this decline is habitat loss due to human activities and conflicts with local communities. According to the Red List of the International Union for Conservation of Nature, lions are now at risk of extinction.
What Did the Scientists Discover?
Roaring is a vital aspect of lion behavior. It not only signals the whereabouts of to relatives in the pride but also marks territorial boundaries for intruding lions.
For their study, the team focused on the lion population in Nyerere National Park in Tanzania. Researchers recorded the sounds of these animals using camera traps, and the results were analyzed by artificial intelligence.
The researchers employed supervised machine learning to classify the vocalizations that occur during lion roaring into different sound types, primarily focusing on loud and intermediate roars. On the spectrogram, the team observed that the loud roar was complex and arc-shaped, while the intermediate roar was flatter and less variable, always following the loud roar, as reported by Science Alert.
Using basic acoustic parameters (duration and maximum frequency of sound), the scientists were able to identify each type of vocalization with an accuracy of 95.4 percent.
Additionally, the researchers found they could identify individual lions with an accuracy of 94.3 percent. This means that through this roar recognition method, researchers will likely be able to more accurately assess lion population density and numbers.
Jonathan Grocott noted, “It was fascinating to learn that lion communication is more complex than previously thought. However, it remains unclear what the communicative differences are between the two types of roars. For now, we can celebrate that is helping us learn more about such iconic phenomena in the wild as the lion’s roar.”
Photo: Unsplash
