Spots aside: Researchers have discovered a new method to reliably distinguish between leopards (photo).

The Earth is home to an incredible number of species, but identifying individuals of wild, free-roaming animals is a complex and labor-intensive task. This is especially true for species that lead elusive lifestyles, such as solitary big cats that often are nocturnal and inhabit hard-to-reach areas, as reported by IFLScience.

Visual identification is often impossible, prompting scientists to explore alternative methods. This is particularly crucial given the need to protect and conserve species like leopards. Researchers have now developed a cheaper and more accurate way to identify leopards using their vocalizations.

During an experiment in Nyerere National Park in southern Tanzania, a team set up 64 camera traps to capture images of predators. Additionally, autonomous recording devices (CARACAL) were installed in the park, each containing four microphones to record leopard vocalizations. Recordings were made from September 20 to November 20, 2023.

In the first phase, scientists reviewed the photos to identify the animals and then listened to the audio files recorded simultaneously with the photographs. The results indicate that leopards have remarkably distinctive calls known as "sawing roars." The team also discovered that a leopard's roar consists of three stages, including an initial and final phase. The data were incorporated into their analysis when the leopard roared for 10 minutes after being photographed.

According to lead author of the study, PhD student Jonathan Grocott from the University of Exeter, their discovery is significant, though it is based on a fairly basic finding. Nevertheless, it highlights how little we actually know about leopards and large carnivores in general.

In total, the study identified 42 individuals from over 191 photos of leopards captured by the camera traps. There were 23 instances where a leopard also roared within 10 minutes of being photographed, involving 14 different leopards. Considering some technical challenges and exceptions during the analysis, the team ultimately identified seven leopards, 26 usable roar sequences, and 217 distinct roars.

The team then used the "fundamental frequency contour" of the second part of the leopard's roar, which helped distinguish individual animals by their vocalizations. The study authors utilized the data in their modeling system and were able to identify individual roars with an accuracy of 93.1%.

Scientists believe that leopards have vocalizations that can be recorded and assigned to individual animals, allowing for relatively easy and inexpensive identification with high accuracy. The team considers this the first example of detecting leopards using autonomous recording devices, demonstrating an increased detection range.

Researchers also suggest that in the future, this method, combined with others, could become a valuable tool for assessing population density. Furthermore, it could serve as an effective alternative to collars.