Orcas, or killer whales, live in family groups. Each group has its own way of calling, like a dialect. Young orcas learn these calls from their families. This is a kind of culture.

In the North Pacific, some groups eat fish, but other groups hunt seals in the same sea. Older females often lead the group when there is not much food. They know where to find fish.

In biology, culture means behavior that is passed on by learning from others, not through genes. Killer whales seem to have it. In 1991, a Canadian researcher found that each family group, or pod, off British Columbia had its own set of calls, a kind of dialect, and closely related pods had similar dialects.

In the eastern North Pacific, some killer whales eat mainly fish, while others hunt mammals such as seals, even in the same waters. In 1987, a strange fad of carrying dead salmon on their heads spread between groups but disappeared the next year. A 2015 study found that older females often led their groups when salmon were scarce.

In 1987, a female killer whale began carrying a dead salmon on her head. Other pods picked up the habit, and within a year it had largely vanished. Biologists count fads like this as evidence of culture, meaning behavior passed on through social learning instead of genes, and killer whales are among the strongest examples outside humans.

Their calls are another sign. In 1991, John Ford reported that resident pods off British Columbia each have a distinct repertoire of calls. More closely related pods share more similar dialects, which points to calves learning calls from their families. Diet follows traditions too.

In the eastern North Pacific, fish-eating and mammal-eating killer whales share the same waters but rarely interbreed. The fish-eaters often use echolocation, while the mammal-hunters stay quiet so their sharp-eared prey do not hear them. In 2015, Brent and colleagues found that females past reproductive age were more likely to lead their groups in years when salmon were scarce.

Their ecological knowledge may help explain why menopause evolved. Proving that all this is culture is hard. The two ecotypes also differ genetically, and dialect learning is inferred largely from correlations, so genes and environment are difficult to rule out.