杏吧原创

BETTY AND ABEL are two smart birds. Put a morsel of food just out of their reach and these New Caledonian crows will fish for it with a tool. And not just any tool. Given a choice, they鈥檒l pick the best one for the job. Smarter still, if they can鈥檛 find the right tool, they鈥檒l make one.

These two crows are in their third year at Oxford University, but they haven鈥檛 been taking an advanced course in toolmaking. Neither has received any training 鈥� their talent comes naturally and their wild relatives are equally skilled. When it comes to making tools, New Caledonian crows are experts. They show a keener understanding of form and function than even chimps. 鈥淭hey鈥檝e reached levels of toolmaking proficiency generally associated with an animal with a big brain, dextrous hands and symbolic language 鈥� in other words humans,鈥� says Gavin Hunt, a biologist at the University of Auckland.

True, crows aren鈥檛 likely to build computers or space stations any time soon, but their skills do challenge current ideas about how our ancestors became master toolmakers. Around 1.5 million years ago, ancient humans suddenly began to create symmetrical tools with carefully worked edges. This technological breakthrough has long been seen as evidence for the evolution of the sort of brain that is capable of planning and abstract thought. A second innovative leap 100,000 years ago, when early humans started experimenting with new materials and designs, has been linked with the evolution of symbolic language. If that鈥檚 right, then what鈥檚 a bird-brained New Caledonian crow doing with such advanced toolmaking skills? 鈥淢aybe when we鈥檝e got to the bottom of what makes crows such skilled toolmakers we鈥檒l have to think again about how toolmaking evolved in humans,鈥� says Russell Gray, an evolutionary biologist who works with Hunt.

Betty and Abel belong to a forest-dwelling species (Corvus moneduloides) confined to the islands of Grande Terre and Mar茅 in the South Pacific archipelago of New Caledonia. They are the only birds that habitually use a range of tools that they make themselves. 鈥淭he whole species relies on tools to get food,鈥� says Alex Kacelnik, an expert on animal behaviour at Oxford University. In the wild, they use the same tools repeatedly and carry them around from place to place. Hunt has been watching wild crows for the past decade, and he鈥檚 collected an entire tool kit of stick-like probes, nifty hooks and long, barbed tapers.

The simplest tool is a stick for poking about in cracks and crevices where insects and other small prey hide. Crows could just pick up sticks from the forest floor, but often they make their own. Hunt has found probes fashioned from a variety of materials: the central rib of a large leaf, a bamboo stem or a fern runner, even a sliver from the woody midrib of a palm frond. Crows wield these in different ways depending on their length and the job in hand. 鈥淭hey adjust their grip depending on whether precision or brute force is needed,鈥� says Jackie Chappell, a member of Kacelnik鈥檚 team at Oxford.

The crows mostly use these tools as beak extensions to flush out prey. But in the wooded, lowland parts of Grande Terre, a stick doubles as a grub-grabber. In this part of the island, wood-boring longhorn beetles (Cerambycidae) lay their eggs in the dead wood of candlenut trees. The grubs grow big and plump but are often beyond the reach of a crow鈥檚 beak. So the birds go fishing. 鈥淚t鈥檚 a very clever technique,鈥� says Hunt. 鈥淭hese grubs have big mandibles and if they are irritated around the head they grasp on to whatever鈥檚 annoying them.鈥� The crow pokes them with a stick, waits for a response, and then withdraws the stick with the grub clamped to the end.

The way crows manufacture and manipulate stick tools is clever, but biologists are more impressed by what they can do with hooks. The idea that a bird understands the function of a hook is highly controversial. 鈥淓ven our closest relatives the chimps don鈥檛 grasp it,鈥� says Gray. 鈥淭hey don鈥檛 have the same appreciation of the physical properties of objects as we do.鈥� Our ancestors cottoned on to the idea only recently, adding hooks to their tool kit around 80 000 years ago. And children don鈥檛 realise what a hook can do until they are two or three years old. But New Caledonian crows use them regularly and make them in a variety of ways from a range of raw materials.

One type of hook, the 鈥渃rochet hook鈥�, is made by detaching a side twig from a larger one, leaving enough of the larger twig to shape into a hook. 鈥淭hey use a special technique to make it,鈥� says Hunt. 鈥淭hey can鈥檛 be doing it by chance.鈥� The 鈥渟tick-with-a-hook鈥� is equally ingenious. It is crafted from the leaf of a forest vine, which consists of a sturdy midrib with paired leaflets that each have a rose-like thorn at the base. The crows strip out a piece of midrib, removing the leaflets and all but one thorn at the tip. 鈥淭his comes with a ready-made hook,鈥� says Hunt. But it鈥檚 not easily come by. 鈥淭he vine isn鈥檛 very common and the birds must search for it,鈥� he says.

New Caledonian crows make a third kind of hook-bearing tool, one that is causing much head scratching among biologists. The 鈥減andanus stepped-cut tool鈥� is a tapered probe, pointed at the tip, broad at the base and bearing a row of tiny hooks along one edge. With its sharp point and backwards-facing barbs, it鈥檚 an ideal implement for winkling prey from tight spots. 鈥淚n engineering terms it鈥檚 beautifully designed,鈥� says Hunt. 鈥淭he tapered shape gives it uniform strength along its length.鈥� But what makes this tool special is that crows construct it to a standard design as if following a set of instructions. 鈥淭he shape has more to do with the crow鈥檚 behaviour than with the raw material. That suggests the birds are following some kind of plan when they make them,鈥� says Hunt.

The pandanus tree has strap-like leaves reinforced by tough parallel fibres that are hard to cut without a sharp blade, so the crow creates a tapered shape by cutting a series of steps. With precision beak work, it nips into the edge of the leaf then rips a short distance along the fibres. Next it makes another neat cut and rips again, repeating the process until it has a tool with two, three or four steps. 鈥淎 straight tapered edge would probably be more efficient but crows would have difficulty cutting such an edge with scissor-like precision,鈥� says Hunt. 鈥淭he steps are an efficient way for them to make a tapered shape.鈥�

It鈥檚 rare to catch a crow in the act of clipping out a pandanus tool. But for every tool, there鈥檚 a matching counterpart in a leaf. These provide a record of the birds鈥� industry and a faithful account of their methods. 鈥淚n one place we have counterparts going back more than 10 years,鈥� says Hunt. The amazing thing is, crows faithfully produce the same design every time. 鈥淵ou find no in-between or practice versions,鈥� says Hunt. It鈥檚 left the researchers wondering whether birds learn each step in the sequence separately and then put them together to produce the finished tool, or whether, like people, they have a mental image of the tool before they start and perform the actions they know are needed to make it.

Recent research has revealed that genetics plays a big part in the less sophisticated toolmaking skills of woodpecker finches in the Galapagos Islands. No one knows if that鈥檚 also the case for the New Caledonian crows, but it鈥檚 unlikely that their toolmaking skills are ardwired into the brain. Hunt and his colleagues suspect that young birds learn from older ones. 鈥淭he picture so far points to a combination of complex cultural transmission and individual resourcefulness,鈥� says Kacelnik. Just like us.

The right tool for the job

The crow鈥檚 toolmaking skills certainly suggest that its brain is rather better at complex mental processing than anyone imagined. But how much better? How much do these birds understand about the nature and purpose of a tool? And do they really form pictures in their mind before starting work? These questions are impossible to answer just by observing wild birds at work. But Betty and Abel can help.

If New Caledonian crows understand the function of a tool, then they should be able to pick the right one for a particular task. 鈥淭he diversity of the tools in the wild crows鈥� tool kit does suggest they might use different implements for different purposes,鈥� says Chappell. 鈥淏ut we can only test this in captivity by giving them new tasks, otherwise we don鈥檛 know if they鈥檝e just learned to associate a particular tool with a particular task.鈥�

At Oxford, she and Kacelnik offered the two captive crows a tool-rack containing ten wooden sticks of different lengths, then set them a challenge they had never encountered before. The birds had to retrieve a morsel of meat from inside a transparent plastic tube, well out of beak range. Both crows eyed up the meat through the side of the tube and through the hole at the end, then went straight to the tool rack and picked out a stick. They didn鈥檛 always pick a stick of exactly the right length but they almost invariably chose one that was either just long enough or longer than required to retrieve the food.

In a second test, Kacelnik, Chappell and their colleague Alex Weir offered the birds food in a bucket at the bottom of a 鈥渨ell鈥�. The only way to get the food was to hook the bucket out by its handle. Given a choice of two tools 鈥� a straight piece of wire and one with a hooked end 鈥� the birds picked the hook. 鈥淭heir ability to select an appropriate tool on their first exposure to a novel task is impressive,鈥� says Chappell. 鈥淚t tells us they understand something of the functional properties of the tool,鈥� adds Gray.

Do they also have foresight and an ability to plan the construction of their tools? The pandanus stepped-cut tool hints that they might. But the only way to test this is to give the birds problems and materials they鈥檝e never come across before. 鈥淚f they make novel designs of tool exploiting the properties of new materials, this would provide evidence for a high level of creativity and planning,鈥� says Kacelnik.

And in a display of inventiveness that astonished the entire Oxford team, Betty has done precisely that. In this week鈥檚 issue of Science, the team describes how, during the fifth bucket-in-the-well test, Abel carried off the hook, leaving Betty with nothing but a straight wire. She had a go with this but failed. 鈥淲hat happened next was absolutely amazing,鈥� says Kacelnik. She wedged the tip of the wire into a crack in a plastic dish and pulled the other end to create her own hook. Wild crows don鈥檛 have access to pliable, bendable material that retains its shape, and Betty鈥檚 only similar experience was a brief encounter with some pipe cleaners a year earlier. What鈥檚 more, Betty鈥檚 invention was no fluke. In nine out of ten further tests, she again made hooks and retrieved the bucket. Abel wasn鈥檛 as innovative and had to resort to robbing Betty.

The question of what鈥檚 going on in a crow鈥檚 mind will take time and a lot more experiments to answer. What is clear is that they can construct sophisticated tools without large brains or symbolic language. There could be a lesson in this for understanding our own evolution. Maybe our toolmaking ancestors didn鈥檛 have some of the sophisticated mental abilities with which we credit them. Closer scrutiny of the brains of New Caledonian crows might provide a few pointers to the special attributes they would have needed. 鈥淚f we鈥檙e lucky we may find specific developments in the brain that set these animals apart,鈥� says Kacelnik.

One of those developments might be a very strong degree of laterality 鈥� the specialisation of one side of the brain to perform specific tasks. Lateralisation is thought to improve the efficiency of the brain. 鈥淚n the case of vertebrates, the left side seems to be specialised for processing complex, sequential tasks,鈥� says Hunt. And the more complex the task, the more lateralised the brain. In people, the left side of the brain controls language and speech, and one of the consequences is thought to be our right-handedness. No other species has such a strong bias towards right-handedness 鈥� but New Caledonian crows come a close second.

鈥淚 noticed early on that most of the pandanus tools were cut from the left side of the leaf,鈥� says Hunt. That means that the birds clip the leaves with the right side of their beaks 鈥� the crow equivalent of right-handedness. Last year, after examining almost 4000 pandanus counterparts, Hunt and his colleagues discovered that this was true of most groups of crows all over the island. 鈥淪uch a widespread bias is unique outside humans,鈥� says Hunt. The team thinks this reflects the fact that the left side of the crow鈥檚 brain is specialised to handle the mental processing required to make tools, or at least those tools that require most thought (Nature, vol 414, p 707). Right-handedness, then, probably stems from a general specialisation of the left side of the brain for sequential processing.

Finding other similarities between the brains of crows and people could prove tricky, but further studies of crow behaviour may be more revealing. For all the crow鈥檚 amazing skill, biologists know very little about this secretive forest bird. The Oxford group is investigating whether the New Caledonian crow鈥檚 superior brainpower extends to other aspects of its behaviour. In New Caledonia, Hunt is investigating the birds鈥� social behaviour and lifestyle, which in turn may give further clues about how and why our own ancestors became so good with tools. After all, such advanced toolmaking is very rare. It has evolved in just two species whose last common ancestor certainly didn鈥檛 have the brains for it.

Under what conditions might this extraordinary talent have emerged? Both humans and crows are social animals. Both have a complex brain. And when it comes to eating, both are generalists and supreme opportunists. All these were probably important but, ironically, it may have been their shortcomings that triggered the evolution of toolmaking.

Maybe the ancestors of crows and humans found themselves in a position where they couldn鈥檛 make the physical adaptations required for survival 鈥� so they had to change their behaviour instead. 鈥淭he stage was then set for the evolution of those rare cognitive skills that produce sophisticated tools,鈥� says Hunt. 鈥淣ew Caledonian crows may tell us what those crucial skills are.鈥�