Just like humans, wolf species apparently have more than 2,000 distinct howling "dialects," each corresponding to a particular kind and subspecies of the canine.
In the largest study of its kind, scientists distinguished and classified various distinct howls from canids such as wolves, coyotes and jackals.
The distinct howls were further divided into 21 types based on sound patterns and pitch fluctuations.
Howls can match certain species and subspecies. For instance, the timber wolves have low, flat howls, while the endangered red wolves have a high, looping vocalization.
Zoologist Arik Kershenbaum, author of the study, said the distinct howls resemble vocal dialects. Each species has its own identifiable use of the howl.
The findings can help track and manage wild wolf populations better, as well as help mitigate conflict with farmers, Kershenbaum said.
Researchers wanted to study the vocalizations of non-primates in order to understand how our own systems of language evolved. But why did they choose wolves?
Kershenbaum said wolves are not taxonomically close to humans, but ecologically, their behavior in a social structure is similar to ours.
"That's why we domesticated dogs - they are very similar to us," said Kershenbaum, who is a Post-doctoral research fellow at the University of Cambridge.
Kershenbaum and his colleagues examined howls that were recorded from both wild and captive animals from Australia, India, Europe and the United States. They created a database containing 6,000 howls, a number that was shrunk down to 2,000 howls for the study. They also combed YouTube for domestic dog howls.
Data were then fed to machine-learning algorithms to classify the howls into separate types.
Previous studies on howling have had to depend on subjective human comparisons by looking at soundwave patterns. The machine-learning algorithms allowed the howl types to be studied objectively, thus revealing that various wolf species have characteristically different assortments of howl usage.
The howling dialects of most of the 13 wolf species were very distinguishable, but some bore close similarities. This may be the result of interbreeding, and may even threaten the survival of a species, researchers said.
Meanwhile, Kershenbaum and his team are currently working on research in the Yellowstone National Park, and they're using triangulation technology and recording devices to pick up the location of howl sounds. This would help them tell whether certain calls are linked to distance communication or pack warnings.
The howl study, which was conducted by scientists from the United Kingdom, U.S., India and Spain, is featured in the journal Behavioral Processes.