Every autumn, caribou gather along the shores of Victoria Island in the Canadian Arctic and wait.
Once the temperature drops and the ice gets thick enough, dozens of them cross together to mainland Canada. There, if they make it, they will spend the winter breeding before heading back to give birth and raise their young on the island.
The Inuit, who share an intimate connection with these caribou, have always said that a few lead and the rest follow along this treacherous Arctic journey. Now, new technology is confirming this centuries-old indigenous wisdom.
In November 2015, scientists led by Andrew Berdahl at the Santa Fe Institute in New Mexico and Colin Torney at the University of Glasgow in Scotland deployed drones to film this migration. With computer vision they analyzed the footage, identifying every individual and tracking their assorted behaviors as they moved together. They found that indeed, caribou paid close attention to the directional cues of their neighbors ahead of them. And despite earlier research models suggesting individuals in migrating groups all act the same, the researchers confirmed what the Inuit knew — some caribou led, while others followed along.
The results, published Monday in Philosophical Transactions of the Royal Society B, may help inform conservation efforts for island caribou, which already face other pressures that have been exacerbated by warming temperatures and increased ship traffic along their icy migratory routes.
Their methods may also contribute to the study of other animal migrations in the wild. There are even lessons for your average human: In scary situations, it's nice to have others around.
Scientists have known that social interactions of organisms from bacteria to humans influence the decisions they make as species. But it's difficult to study these interactions among migrating animals.
Dolphin and Union caribou, as this endangered population is called, provide an interesting case study because they're different from other caribou. They're less social — except during this sea ice migration. The benefits they gain from interacting with one another during their crossing have been a mystery.
Scientists have studied caribou migrations by counting passing animals, taking aerial photos and attaching GPS collars to track the movements of a few members. But drone videos offer a more complete view of the varying behaviors of migrating caribou.
The researchers in this study were able to observe how sex or age influenced interactions while traveling, looking at who stayed close to whom and which animals started movements and which copied. They discovered, for example, that calves rely on social cues and stay close to others, but mature bulls were more autonomous.
"If you want to understand the full picture you have to consider the social context, because in this group of moving species, the interactions between individuals can be just as important as the traits of individuals themselves," Berdahl said.
He thinks that when these caribou come together, they can better sense and respond to the perilous ice they cross. It's called collective sensing.
"What's interesting is that they can achieve collective behavior — migrating together — without any individual deciding that's what they need to do or making any global assessments," said Deborah Gordon, who studies collective behavior in ants at Stanford University and was not involved in this study. "That's how nature works at many scales — from the cells in our bodies to herds of caribou."
Understanding such collective behavior is important as the environments migratory animals live in change with the climate and human activity. And these new technologies can contribute to ongoing caribou conservation efforts, said Kim Poole, a wildlife biologist at Aurora Wildlife Research in British Columbia who has studied island caribou sea ice crossings.
In the future, combining drone footage of behavior with high resolution maps of terrain may offer clues to how these animals and others deal with danger.
"So many migratory species travel in groups," Berdahl said. "Nature's had countless years to discover and perfect these algorithms for controlling distributed systems," and now it's time we uncover and use them.