Scientists have discovered a new type of lion roar – an intermediate roar that is shorter and lower in pitch than the animal's iconic, full roar.
Researchers have found that this intermediate roar always follows a loud roar. The discovery shows that lions' vocalizations are more complex than previously thought, the study's lead author said. Jonathan Growcott, PhD student in Mathematics and Statistics at the University of Exeter in the UK.
The discovery was published Friday (November 21) in the journal. Ecology and evolutionThis challenges the long-held belief that there is only one species of roaring lion and could help researchers more accurately estimate lion population sizes, he said.
Lions roar to contact their pride and assert their territory. Previous research pointed out that the loud roars of lions are unique, individually identifiable signs that contain information about the sex, age and other characteristics of the animal. This suggests the sound recordings could help scientists count the number of lions in a landscape and thus estimate population density, Growcott said.
For the study, the researchers installed 50 custom-made microphones in Nyerere National Park in Tanzania and attached acoustic sensors to the collars of five lions in the Bubai Valley Game Reserve in Zimbabwe. In total, the team recorded 3,149 African lions (Panther Leo) vocalizations.
The team then used artificial intelligence (AI) analyze sounds in which an intermediate roar is detected.
Without artificial intelligence, understanding guttural roars and identifying lions by their voices has always been subjective, Growcott said. Lions also make other sounds—including grunts, growls, puffs, groans, and meows—but only the loudest roar has been shown to have identifiable signatures, so it's important to properly identify that roar and extract as much information from it as possible.
The researchers' artificial intelligence model allowed them to classify lion vocalization types with over 95% accuracy, largely eliminating the bias that occurs when researchers try to characterize lion calls manually. This approach also made it easier to identify individual lions by their loud roars and revealed a previously unknown intermediate roar, knowledge of which will help researchers isolate the full-throated roar in the future.
“This new method, which proves that the intermediate roar exists and is distinct from the full-throat roar, is an important first step in ensuring consistency in the selection of the full-throat roar,” Growcott said.
According to the organization, there are only about 23,000 African lions left in the wild, making the species vulnerable to extinction. IUCN Red List. Traditional methods for estimating lion populations include camera traps and track surveys, but these are resource-intensive and may be less accurate than acoustic surveys.
“My hope is that using data-driven roar forecasts will lead to more accurate acoustic population density estimates that can better inform critical conservation needs,” Growcott said.
