As the Crows Surmise

A fascinating new video shows the birds using a variety of tools to get their supper.

Confession: although I think this is a very cool study and an important clue to the evolution of intelligence, I might not be writing about it were it not for the fact that I can't resist this video of one remarkably smart crow. It shows Betty, a New Caledonian crow (Corvus moneduloides), figuring out on her own how to use up to three tools in the right sequence to achieve a goal (in this case, get food out of a tube). That's never before been observed in a nonhuman animal—not even in chimpanzees—that had not been explicitly trained to do it.

In the experimental setup, there is a morsel of food in one long tube (too far back to reach), a tool on a table, and other tools inside other tubes. The first notable thing about Betty's behavior is that she does not waste her time trying to get food out of the tube by brute force or by poking her too-short beak into the opening. She takes one look at the food sitting out of reach in the long tube and immediately grabs a short stick from the table with her beak. She then pokes the short stick into one of three "tool-tubes," dragging out a medium-length stick. She then looks into the food tube, but, remarkably, doesn't even try to use the medium stick to grab a snack: it won't reach. Instead, she uses the medium stick to pull out the long stick from the fourth tool-tube. Finally, she has what she needs: she uses the long stick to reach the food from the food-tube. Interestingly, Betty makes a point of disposing of each tool once she no longer needs it.

All of this is described in a fascinating paper by Jo Wimpenny of Sheffield University, Alex Kacelnik of Oxford University and their colleagues, published Wednesday in the journal PLoS One. The surprise isn't that New Caledonian crows used tools—in the wild, they use all sorts of sticks to scoop prey from holes and crevices, and in captivity they select tools of the right size to retrieve food. In earlier research, in fact, the scientists showed that Betty could invent new tools, bending strips of metal into hooks or straightening them so they reached farther. In all these cases, though, the tool was used directly to reach food.

The new study goes a step further, showing that Betty and her crew can use tools to retrieve other tools. That's considered a hallmark of human intelligence, and may have been a crucial step in our evolution. But it's extremely rare. Chimps have entire cultures of tool use (that is, troops living in different places use different tools for the same purpose), as Andrew Whiten of the University of St Andrews and colleagues have documented. But while chimps sometimes use two tools in sequence (a fat stick to puncture a termite mound, for instance, then a thin stick to fish out the insects), both tools are used to get food. Neither is used to procure another tool, which is what Betty and the other crows did.

This is called "secondary tool use," and although individual orangutans, gorillas, and some other primates have been reported to do it after training, it is a rarity in the animal world and appeared relatively late in human evolution. It's considered a marker of the ability to plan and exhibit foresight, which are crucial to higher-level cognition.

Betty performed the best, but three other crows successfully used tools sequentially (three on the first try), with Betty and two others solving the most demanding three-tool challenge. Even the "failures" offered fascinating glimpses of animal intelligence. Pierre, for instance, at first tried to use the short tool he grabbed from the table to extract a longer tool from one of the tubes, but it didn't reach. He left the room (I am tempted to say he stomped out in frustration, but that would be an anthropomorphic step too far) and returned moments later with a twig, longer than the one from the table, and used it to reach the longest tool, with which he then triumphantly pulled out the food. "By finding a suitable object outside the confines of the experimental setup," write the scientists, "he transformed the task from requiring three tools to two, and then proceeded to solve it appropriately."

You can't hope to figure out how high-level cognition evolved unless you have numerous examples of it; the crow experiments provide one more. (You can check out more videos here.)

For those of you with a classical bent, let me mention another study, published Thursday in Current Biology. Remember Aesop's fable of the crow and the pitcher? In it, a thirsty crow comes upon a pitcher, but the water level is so low he cannot reach it with his beak. The bird therefore drops pebbles in, one after another, until the water level rises enough for him to drink. (The moral is this: "Little by little does the trick.")

Back in real life, scientists led by Christopher Bird of the University of Cambridge presented rooks (relatives of crows) with clear plastic tubes, six inches tall, containing water with a worm floating on top. As in the fable, the water level was too low for the bird to reach the worm. Two of the four birds seized on the correct strategy on the first try, dropping in the precise number of equal-sized stones needed to raise the water to a height that let them grab the worm. That is, they did not check to see if they could reach the water after each stone, instead dropping in all the stones before testing the waters. In a second experiment, when the rooks had their choice of different sized stones, they picked larger rather than smaller ones, apparently realizing that bigger objects displaced more water. Watch it yourself here.