麻豆传媒

Eugene, Oregon

WHEN an octopus in a research laboratory in Naples learnt to choose a red ball instead of a white one by watching another octopus, students of animal learning were taken aback. Such 鈥渙bservational learning鈥� is supposed to be seen only in higher vertebrates-animals such as rats with sophisticated brains. Octopuses, on the other hand, are molluscs, a seemingly primitive animal group. True, octopus have huge brains. But they look nothing like the brains of the vertebrates that are so adept at learning.

For that matter, why would octopuses need to learn by example? They are short-lived, solitary creatures that usually meet only once, to copulate.

Jean Boal, who studies animal behaviour at the University of Texas in Galveston, epitomises the scepticism that greeted the announcement in 1992 of the educable octopus. 鈥淚f they really did show observational learning, it would be astonishing,鈥� she said recently. 鈥淲e have many mammals that aren鈥檛 doing that.鈥�

But not everyone is as doubtful. After all, the brains of animals like the octopus evolved entirely separately from the brains of the vertebrates, and they have an entirely different design-perhaps they also house a unique form of intelligence. 鈥淲e have two very different brains that can do some similar things-including perhaps observational learning,鈥� says Shelley Adamo at Dalhousie University in Halifax, Nova Scotia.

Others say that even if octopuses are incapable of something as sophisticated as observational learning, they excel at other types of behaviours that mainstream neurobiology has long since denied in so lowly an animal. 鈥淚t鈥檚 very clear from a cursory observation of cephalopods that they are extremely intelligent animals,鈥� says Nathan Tublitz at the University of Oregon in Eugene.

The debate about observational learning in octopuses is at the centre of one of the oddest challenges in neurobiology: the quest to understand the brain of the cephalopod, animals that include octopuses, squid, cuttlefish and the single-shelled nautilus. The weird construction of the cephalopod brain was pieced together in the 1960s by J. Z. Young鈥檚 classic studies of the common octopus, Octopus vulgaris. But getting to grips with how it works, and how well it works, has become fraught with disagreement.

Octopuses and squid are seductive creatures, so some critics suspect that their intelligence has been grossly exaggerated by anthropomorphising observers-鈥渢hey watch my every move, therefore they must be curious鈥�. On the other hand, because cephalopod behaviour and brain structure are so foreign, others argue that their greatest cognitive feats are probably still being overlooked.

Deep-sea brains

鈥淭hey鈥檙e very alien-an alien intelligence,鈥� says Adamo. 鈥淭he way they `think鈥� is very different. When you consider a cat, dog or rat, we pretend we can sort of understand some of their responses. With cuttlefish, it鈥檚 much more difficult.鈥�

But it is becoming easier. The advent of cheap, compact underwater video cameras makes it possible to chronicle the animals鈥� behaviour in their natural environment. Lab-bred squid and cuttlefish are also now available, making them a far more convenient animal to study. Boal has even developed a standardised aptitude test for octopus learning, which she plans to present at this month鈥檚 annual meeting of the Animal Behavior Society, to be held at the University of Maryland in College Park. Armed with these new techniques, researchers are laying the groundwork for what could be the first successful attempt to get to the bottom of what it means to be a cephalopod.

Cephalopod lives are unlike those of their clam and slug relatives. With the exception of the more primitive nautilus, these mighty molluscs dart about by jet propulsion, hunt down living prey, including fish, squirt ink and survive in a world filled with predators hungry for the taste of calamari. Some species grow barely larger than a thumbnail. Others may turn out to be the largest invertebrates ever to have lived. Take, for example, the mysterious giant squid Architeuthis dux, which is known from carcasses to be 18 metres long from tentacle tip to mantle, but has yet to be studied alive. All told, there are roughly 700 species, and cephalopods can be found in every ocean, from the tropics to the polar regions, from near-surface reefs to depths of 7000 metres.

Cephalopods have boneless bodies and keen senses. Their complex eyes, as large as car headlamps in some deep-water species, can distinguish detail as well as mammalian equivalents. Although cephalopods are thought to be colour-blind, they can see polarised light, which we cannot. They also have highly developed senses of touch, taste and smell, and can detect gravity, a sense which is used in the coordination of muscles during movement. And in the past few years, researchers have even discovered what can best be described as hearing: fine hairs along the head and arms that, in cuttlefish at least, can detect disturbances made by a metre-long fish up to 30 metres away.

And yet despite the wiring required to process the input from such a sensory smorgasbord, the brain itself promises so much more. At a fundamental level-cells communicating by chemical signals and so forth-the brain of the cephalopod is essentially the same as any vertebrate. Indeed, research on the squid 鈥済iant axon鈥� has been instrumental in showing how nerves work throughout nature.

But not surprisingly, given that the common ancestor of the vertebrates and the molluscs probably had no brain at all, there are also some dramatic differences. The vertebrate central nervous system comprises one main nerve cord that has swollen at one end to create a brain. Most molluscs, on the other hand, have dual nerve cords running like a set of railway tracks along the length of the body. These are dotted with five or six pairs of ganglia, clumps of neurons that are capable of only the most primitive form of information processing. In the cephalopods alone among the molluscs, evolution has also constructed a brain. It has greatly expanded the forwardmost pairs of ganglia and moved them closer together to create a tightly packed mass of lobes that lies between the eyes and encircles the oesophagus. This is an awkward arrangement in some ways-researchers have discovered spines lodged in octopus brains, the result of a meal going down the wrong way.

鈥淭he brain is anatomically complex,鈥� says neuroscientist Ted Bullock of the University of California in San Diego. 鈥淚t is very highly differentiated. It has a lot of texture, it isn鈥檛 smooth or monotonous. It looks like a complicated brain, histologically and microscopically.鈥�

Make no mistake, cephalopod brains cannot compare with the complexity of the brains of mammals or even birds. In brain size to body mass ratios, however, they outrank those of lower vertebrates, like reptiles and most fish.

Educating octopussy

鈥淭here is still this burning question out there about why this animal group has such large brains,鈥� says Roger Hanlon, a cephalopod expert at the Woods Hole Marine Biological Laboratory in Massachusetts. 鈥淲e can鈥檛 really explain it-or the unbelievable centralisation and processing that you see in the part of the central nervous system that we call the brain.鈥�

One answer could be that nature has been guilty of poor engineering: perhaps all those neurons are merely the result of having a less efficient way to handle the information needed to direct simple reflexes and process sensory information. Another answer might be that the cephalopod is capable of extremely complex behaviour that requires a lot of neurons for processing-a view to which Adamo and Tublitz subscribe.

Among the most striking examples of what might constitute complex behaviour is the way cephalopods alter their appearance. To avoid being eaten, many species blend in with their surroundings by 鈥減osturing鈥�, positioning tentacles to mimic floating sea grass, for instance, or by flexing skin muscles that can change the texture and the colour of their skin. Indeed, cephalopods are unrivalled when it comes to adopting new looks. They can adopt patterns that range from full-body speckles to dramatic black-on-white tiger stripes. The more distinctive patterns are used as signals during courtship, hunting, male-to-male aggressive encounters and in response to a threat.

Many animals change colour, of course, but usually by releasing hormones that trigger the migration of pigment molecules within each cell. In cephalopods, the ability is due to thousands, sometimes millions, of chromatophores-multicelled organs that consist of pigment sacs of various colours and muscle fibres that stretch the sacs when contracted. Because the brain controls this through nervous impulses to the muscles, it takes less than a second for cephalopods to adopt a new look. Hanlon recorded an Octopus cyanea off the coast of Hawaii changing patterns nearly 1000 times during 7 hours of foraging.

The flashiest species are like swimming slide shows. One, the Caribbean reef squid, Sepioteuthis sepioidea, has at least 35 patterns in addition to its almost magical ability to blend with its background. It can flash a different display on each side of its body when positioned between a potential mate, which sees a uniform light grey, and a rival male, which sees tiger striping called the 鈥渋ntense zebra display鈥�. If the positions change, so do the patterns. Young cuttlefish sneaking away from predatory fish will lie camouflaged against the bottom until algae and sediments are disturbed by a breaking wave. Then, they ride the surging water, simultaneously adopting a new pattern to match the swirling debris.

In the 1980s, the late Martin Moynihan, who worked at the Smithsonian Tropical Research Institute on Barro Colorado Island in Panama, went as far as to speculate that the Caribbean reef squid use their patterning as a visual language-with different signals for nouns, modifiers and verbs.

Unlike octopuses, the reef squid live in schools and, Moynihan reasoned, might need to communicate about, say, approaching predators. Moynihan was struck by the richness of the squids鈥� signalling repertoire. The patterns even broke down into component parts that the squid used interchangeably, providing an almost infinite number of signals-far more than any other nonhuman animal, he claimed. The big hole in Moynihan鈥檚 theory, however, was that nobody knew what the signals meant.

Happy hunting

If not processing a language, the cephalopod brain clearly exercises fairly high-level control over the patterning process. Adamo is now trying to work out whether this control is merely governed by reflexes to simple chemical or visual cues, or involves some decision-making. When hunting, some species of cuttlefish flash what is known as the 鈥減assing cloud display鈥�, a moving pattern in which a patchwork of light browns is broken by waves of darker brown that radiate along the body and down the outstretched tentacles.

The performance, which resembles light dancing on stones, is thought to have a mesmerising effect on prey. It is also highly conspicuous. So how do cuttlefish avoid becoming someone else鈥檚 meal? To find out, Adamo set up an aquarium in which common cuttlefish, Sepia officinalis, were free to hunt fish. When she passed a fake bird over the tank, the animals stopped their hunting display. 鈥淭hey still hunt,鈥� she says. 鈥淏ut they don鈥檛 show it.鈥�

What took her by surprise was that not every animal reacted the same way. 鈥淎ll of them stop doing the `passing cloud鈥� when the bird is present,鈥� she adds, 鈥渂ut one does the coolest thing-he inks and then hunts under the ink. The first time he didn鈥檛 do it, but the second, third and fourth times when the bird flies by, he secretes a bunch of ink and goes down into the substrate and continues hunting.鈥� As Adamo sees it, such behavioural variability-both between individuals and by the same individual at different times-hints at the presence of a sophisticated brain.

Ivy League octopuses

And it is not only when it comes to patterning that the cephalopods show a startling flexibility in their behaviour. Jennifer Mather, a psychologist at the University of Lethbridge in Alberta, has studied how octopuses feed on shellfish. In one strategy, the octopus uses its beak to bore a hole through the shell near the abductor muscle. Next, it injects poison to weaken its victim and then pulls open the shell. On other occasions, the same octopus will simply yank the shells open, or smash them, or chip a little off the shell鈥檚 edge and inject the poison there.

鈥淭hey鈥檙e really being very flexible not just in terms of finding their prey,鈥� says Mather, 鈥渂ut also in terms of how they get at the food once they鈥檝e got the prey-which we found very intriguing.鈥�

Young鈥檚 early description of the octopus brain had certainly suggested a capacity for complex behaviour-a theory that was tested in a series of experiments at the Naples Zoological Station, culminating in the observational learning experiment published in the journal Science in 1992.

By giving octopuses a reward or a punishment when they attacked different objects, Young, Martin Wells from the University of Cambridge, and others discovered in the 1950s and 1960s that the animals can learn to distinguish between different shapes, orientations, sizes and degrees of brightness.

In one experiment, Young trained octopuses to select between large and small squares, horizontal and vertical stripes, and black and white circles. He found that the animals could retain all three preferences at once. In other experiments, blinded octopuses learnt to distinguish between differently shaped objects using only their highly sensitive suckers. One octopus remembered the differences for four months.

鈥淭hey learn these things rather faster than a vertebrate will-a pigeon or a rat,鈥� says John Messenger, a neurobiologist at the University of Sheffield. 鈥淭hat is quite impressive.鈥� He also points out, however, that although cephalopods learn faster at first, their skills level off. A trained octopus will always make more mistakes than a trained rodent, he says.

Next, Young and Wells removed various parts of the octopuses鈥� brains and repeated the tests. That showed that learning is dependent on the bits of the brain that lie above the oesophagus, particularly the vertical and superior frontal lobes. In the 1970s, Messenger discovered that the speed with which cuttlefish learn not to strike at prawns that are encased in a glass tube increases during the first four months of life, correlating precisely with the development of the vertical and superior frontal lobes.

But the 1992 鈥渓ook-and-learn鈥� study, by neuroscientists Graziano Fiorito and Pietro Scotto at Naples, is the most controversial of all the attempts to understand learning in cephalopods. To test if O. vulgaris could learn a skill by observing the activities of other octopuses, the researchers trained one group to choose a red ball or a white one.

When the trained animals reliably approached one or the other ball, untrained octopuses were allowed to watch. When later presented with a choice of their own, these animals not only selected the same ball more often throughout the five days of the trial, but also learnt more quickly through observing than the original subjects had under classical Pavlovian conditioning.

Astounding thought

鈥淭he rapid acquisition and the stability indicate that observational learning in Octopus vulgarisis a powerful mechanism of learning,鈥� the researchers concluded (Science,vol 256, p 545). The finding was astounding not least because observational learning is considered by some to be a preliminary step to conceptual thought.

To critics such as Boal, however, the question isn鈥檛 what kind of complex learning cephalopods are capable of, but whether they鈥檙e capable of it at all. First, she says, there鈥檚 the problem of why the solitary octopus has evolved the capacity to learn by observing other octopuses. Supporters of the idea of smart octopuses argue that other cephalopods form loose social groups, so the octopus may have inherited its penchant for observational learning from a common evolutionary ancestor.

Boal is not convinced. She has attempted to duplicate many of the early learning experiments on octopuses, as well as perform similar tests on lab-bred cuttlefish, but with little success. 鈥淎s I added more controls, I got less and less evidence for learning,鈥� she says. 鈥淚 found there were a lot of places where people could have provided inadvertent cues to their performance.鈥� For example, she says, sometimes the animals could see the human researchers during presentation of objects. At other times, the researchers failed to take into account the fact that the animals often have their own unpredictable preferences for certain objects.

Late last year at Woods Hole, Boal, Hanlon and graduate student Kim Wittenberg allowed animals to observe trained cuttlefish attack and eat a crab, and then compared their performance in the same situation with a naive animal. The observers did learn more quickly how to hunt down a crab. But they also hunted better if they had previously seen only a crab without a predation event, or even if they had simply smelt a crab that was kept hidden behind a partition.

鈥淚f smelling a crab means you perform better than if you hadn鈥檛 smelled one before, and watching a predation event is no better than simply smelling a crab,鈥� says Boal, 鈥渢hen we鈥檙e not talking about complex learning. We鈥檙e talking [about] some kind of releaser of an innate behaviour.鈥�

Most researchers agree that understanding cephalopod psychology is going to a require a thoroughly rigorous approach. For a start, the studies should analyse types of learning and behaviour that might reasonably be expected of cephalopods, even if they seem odd by comparison with vertebrates like rats, the animals in which most learning studies have been done to date.

With that in mind, Boal is steering clear of tests of observational learning. Instead, she has developed an apparatus to study cephalopod spatial learning, the means by which animals find their way by remembering landmarks or the distance they have travelled.

Cephalopods are certainly adept at navigation. Mather and a team of volunteers have mapped the travels of fist-sized O. vulgaris as they forage off the coast of Bermuda. The animals venture from their dens on complicated trips lasting up to three hours, and return by different, more direct routes. Although O. vulgaris usually ends up no more than 9 metres from home, other species of octopus can find their dens after journeys of up to 120 metres-over a landscape that easily disorients human scuba divers.

Octopus brainteaser

Boal鈥檚 new apparatus comprises a well-lit circular tank with six den-like holes, five of which are fake. The apparatus is based on the successful Morris water maze for rats in which animals鈥� abilities to remember the location of a platform that lies just below the surface of coloured water is tested. And just as rats in a Morris water maze will always try to get out of the water onto land, octopuses will always try to escape from open spaces. In her first test of the maze, Boal was able to demonstrate that the California mud-flat octopus, O. bimaculoides, learnt which hole was the correct one within 20 trials, and sometimes as few as five, and that they remembered its position for a week. 鈥淲e don鈥檛 know yet if it was learning a landmark or spatial relationship. All we know is they did find their way back to the same place quickly and efficiently,鈥� says Boal, who will present her findings at the University of Maryland meeting.

Boal hopes that ultimately her octopus maze will become a standardised test for working out how the cephalopod brain operates. For example, neurophysiologists could use it to analyse the effect of removing different parts of the animal鈥檚 brain or chemically blocking different neurotransmitters, and so get a far better idea of the similarities and differences between the brain of a cephalopod and that of a vertebrate. For the first time, researchers will also be able to compare directly the same type of learning in the rat and the octopus.

But until fresh results are in, the debate about the intelligence of the cephalopod will remain deadlocked. On the one side, researchers like Boal predict that the outlandish cephalopod will prove nowhere near as intelligent as the earlier studies claimed. On the other side are those who firmly believe it鈥檚 only a matter of time before the depth and breadth of cephalopod psychology is finally revealed. 鈥淲e have not come up with the right set of experiments to illuminate the intelligence these animals possess,鈥� says Tublitz. 鈥淭he problem is the limitation of humans, rather than the limitation of the cephalopods.鈥�