Some of the most intelligent animals can be found, not on land, but in the ocean. Cetacean mammals such as orcas and dolphins show strong signs of high intelligence, such as complex play behaviour, the ability to learn, the ability to plan and even regional dialects.
Although the sperm whale has the largest brain of any animal in the world, it is not generally considered one of the more intelligent Cetaceans. However, it has been known for some time that the patterns of clicking vocalisations (known as "codas") the mammals use to communicate vary from clan to clan.
In other words, each whale clan, or group consisting of several families, has its own specific accent, or dialect.
According to a study led by Mauricio Cantor, a Ph.D. candidate in biology at Dalhousie University in Canada, this is evidence of cultural learning; more specifically, that the whales learn the codas from each other, rather than other plausible methods, such as genetic inheritance.
Cantor and his colleagues used a dataset collected over 18 years of vocalisations made by sperm whales swimming around the Galapagos Islands. Although they all live in the same geographic area, these whale clans each had distinct codas. In the animal kingdom, groups of animals don't typically develop different accents in this way.
It does, however, happen in human populations and, Cantor believes, broadly results from cultural evolution. Humans learn from other humans that behave similarly to themselves.
To try to demonstrate his hypothesis, Cantor has spent time since 2013 studying the sperm whales of the Galapagos Islands. He and his team recorded whale conversations under water, studying their coda patterns.
They also photographed their tails, since the ridges on the edge of a sperm whale's tail are unique, like a fingerprint. This allowed the team to observe the behaviour of individual whales, and also match codas to individuals.
This data was combined with data from Dalhousie University's Whitehead Lab to model simulated interactions between individuals, demonstrating the most likely way in which the whales' accents emerged.
"We try to backtrack the patterns we observe in the wild to infer how the clan segregation could have evolved," Cantor said. "The computer will simulate the life of several sperm whale populations that acquire codas in different ways over thousands of years. At the end, we see which case could produce clans with different dialects."
Several methods of coda evolution were examined. In the genetic inheritance method, for instance, whales inherit the ability to know which sounds to produce. Another method tested was individual learning, in which individual whales develop codas on their own. A third method was pure social learning, in which young whales learn codas indiscriminately from older whales.
All of these methods produced a very low number of acoustic similarities between individual coda repertoires.
What the team found is that social learning with bias, as opposed to pure social learning, is the most likely way whales learn codas. This means that the whales are biased towards learning certain codas, based on specific codas from whales in their own clans, or the most commonly used codas. This is similar to how human dialects evolve.
"Our findings show that biased social learning is a required ingredient for the segregation of clans of sperm whales with different dialects," Cantor said.
"This gives us evidence that key features of human culture -- which we think makes us so different from everything else in nature -- might be at play in populations of other animals. Maybe we're not as different as we thought."
The full paper can be found online in the journal Nature Communications.