Avian intelligence

In the English language, the term “bird brain” is often used in reference to intellectually challenged individuals. This is, of course, based on the notion that birds are dim-witted creatures whose behaviour is largely based on instinct. The main assumption is that a six-layered neocortex, like that of humans, is a prerequisite for anything that might be classed as intelligent, and even ornithologists have generally believed that, because they have a “smooth” brain, birds aren’t too clever.

However, it has in recent years become clear that we have grossly underestimated the cognitive abilities of birds. Some of the behaviours observed in birds are just as complex, if not more so, than those seen in non-human primates – and “birdbrain” no longer seems so much of an insult. In humans, there is no correlation between brain size and intelligence. In birds, however, it would appear that there is some link between brain size and cognitive abilities. The most remarkable examples of avian cognition are seen in birds of the Corvidae family, which includes jays, ravens, rooks, and crows.

Corvids live in large, complex hierarchical societies, consisting of permanent flocks containing up to several thousand individuals.Species of birds belonging to this family also have the largest brains in relation to body size all other types of birds. The crow’s brain, for example is the same relative size as a chimpanzee’s, and crows, along with jays, score best on an IQ test for birds devised by Louis Lefebvre, a professor of biology at MgCill University in Montreal.

Recent research shows that African cichlids use logic to infer their position within the social hierarchy. By observing other pairs of fish fighting, male cichlids can make an informed guess about their position in a dominance hierarchy by using a simple method of logical reasoning. For example, if fish A is always seen to beat fish B, and fish B is always seen to beat fish C, then it can easily be inferred that fish A will always beat fish C, even though A and C have never been observed fighting.

However, the first demonstration of transitive inference in animals under controlled conditions was in pinyon jays (Gymnorhinus cyanocephalus). The work was carried out by researchers at the University of Nebraska’s Center for Avian Cognition, who captured 16 adult male pinyon jays in Northern Arizona. The birds were divided them into three groups, and 5-minute encounters were staged between closely-matched individuals within each of the 36 possible pairs in each group, so that dominance relationships were established. Each of the encounters was filmed and, for every bird, the frequency of dominance displays (staring) and subordination displays (looking away, crouching down, pushing the chin up or begging) was determined for each bird.

The ability of the jays to draw inferences about their position within the social order was then investigated. Individuals were made to observe encounters between a bird from another group – the ‘demonstrator’ – and two opponents. It was found that the observers could use their prior knowledge, based on their observations of the staged encounters, to determine their chances of winning a confrontation with the demonstrators, even though they had never encountered each other directly.

Earlier this year, members of the Comparative Cognition Lab, at Cambridge University’s Department of Experimental Psychology, published a study which shows that rooks use social interactions as a means of stress management. Like pinyon jays, rooks live in societies containing large numbers of individuals. Rooks are monogamous, and male-female pairs typically mate for life. Pairs do not fight with each other, but with neighbours, usually over food or nest-building materials.

Amanda Seed and her colleagues filmed and analyzed the post-conflict behaviour of a group of 10 rooks, which included 4 monogamous pairs. They found that, although birds that come into conflict never reconciled with each other, they often sought solace in their partner afterwards, by sharing food, preening each other, and twining bills. Prior to the study by the Cambridge group, this behaviour had been observed in monkeys, but never in a non-primate.

Corvids are also capable of remarkable feats memory. Pinyon jays, for example, live in flocks of up to 500 individuals, and will determine their status within a dominance hierarchy by inference, rather than by direct interaction with other individuals, which is both time-consuming, and dangerous. The pinyon jay is able to recognize large numbers of individuals within their group, and to track social relationships over long periods of time.

The Clark’s nutcracker – another member of the Corvid family – displays an extremely accurate spatial memory. It is known to store around 100,000 pinyon seeds in 20,000-30,000 separate caches, which can be recovered up to 9 months later. Although exactly how the bird acheives this is unclear, the use of visual cues, particularly large landmarks, is known to be involved.

Astonishingly, it was recently found that Western scrub jays use counterespionage methods to prevent their stored food caches from being stolen. If they are aware that they have been observed by a competitor while storing their food, they re-hide the cache so that the other bird loses track of where it is. This is evidence that they are able to infer the mental state of other birds and that, maybe, birds have a “theory of mind”.

Equally remarkable is the ability of some birds to adapt their behaviour to new circumstances. Crows are now known to be sophisticated manufacturers and users of tools. Crows also display ‘handedness’ – some hold the tools against their left cheek, and others hold them against the right cheek. The use of tools was previously thought to be restricted to humans and other primates.

In the film clip below, a crow first tries unsuccessfully to retrieve food from a container using a straight piece of wire; it then bends the wire against the side of the container, to make a hooked tool, which it uses to retrieve the food:

Crows can use a variety of tools, and can also select the tool best suited to the task at hand. In one study, captured crows were presented with pre-made tools of different diameters and tubes containing food that was accessible only through a small hole. The birds were seen to select the most appropriate tool, which they then used to retrieve the food. Furthermore, if the tools were tied together in a bundle, the crows were still able to select the right tool after they had untied the bundle.

Crows can also design new tools when they need to. In the wild, Gavin Hunt and Russel Gray, of the University of Auckland in New Zealand, have observed crows crafting hooked implements from leafy twigs. This was the first observed case of the crafting of tools by a non-human species.

The crows manufacture the tools in three stages, each of which involves a complex manipulation of the materials being used. First, they select the appropriate raw materials; normally, this is a fork formed from two twigs. The side twig is then broken off just above the junction of the fork, and discarded. The other twig is then broken off just below the junction. Finally, the bird uses its bill to remove the leaves, and to sculpt a fine hook from the end of the twig:

In the U. K., a group of rooks at a service station use co-operation to gain access to food that has been thrown into a rubbish bin. A pair of them first pulls the bin liner upwards, bringing the discarded food within their reach. Then, they either begin eating the food themselves, or throw it onto the ground, making it available to other crows in the group who have been waiting nearby.

Members of a crow population in Sendai City, Japan have devised an ingenious way to break open nuts: they perch over pedestrian crossings and drop the nuts in the path of oncoming cars. When the lights turn red, they swoop down and collect the contents:

In the past, researchers tended to think of the brains of birds and mammals in terms of divergent evolution. But it is becoming clear that this is probably a result of the nomenclature used in neuroanatomy (which was recently revised), and that there are comparable structures between the avian and mammalian taxa.

In humans, “intelligence” is associated with the prefrontal cortex. It is clear, though, that a laminated cortex is not, as previously thought, essential for compelx cognititive tasks. In the avian nervous system, the equivalent of the prefrontal cortex is a structure called the nidopallidum caudolaterale. And although there are significant differences between the gross structures of the bird and the mammalian brain, there are similarities in the connectivity of brain regions, in the neurotransmitters used and, of course, in the functioning of the cells.

Some researchers believe that primates – including humans – and corvids share a common cognitive “tool kit”, consisting of imagination, causal reasoning and predicting future events. In humans and birds, the neural substrates for these skills are found in the prefrontal cortex and the nidopallidum caudolaterale, respectively.

The cognitive capabilities of corvids seem to be closely tied to social complexity. The larger the group within which a bird species lives, the more adept are individuals of that species at using transitive inference to predict their position within a dominance hierarchy.

But how do birds achieve such complex cognition with relatively small brains? This is, as yet, unclear, although it has been suggested that larger numbers and greater densities of neurons are to be found in the brains of the more “intelligent” birds. This was the case in a preliminary comparison between corvids and pigeons, but it needs to be confirmed. One thing does seem quite clear: the corvids have a special place in the emerging field of cognitive ornithology.


7 thoughts on “Avian intelligence

  1. What a great post! I’ve been lurking around the science blogs for a few weeks now, just getting a feel for this corner of the blogosphere. I’m not quite ready to start talking yet, but I just had to let you know how much I’ve been enjoying your blog. You have a real knack for explaining things to a lay person like me.

  2. I’m so fascinated with this subject that I’ve been following your links through several layers. The most interesting articles seem to be the work of Emery and Clayton of Cambridge but everything I can find is behind a pretty pricey subscription wall. Do you have any freebie links to more of their work?

  3. A very interesting post, thanks! I would like to contribute one of my experiences, regarding the use of intelligent methods by small animals, mice in my case. Once, I observed two of them crack a hen’s egg in this way: One of them grabbed the egg with all its legs. The other took the first one’s tail in its mouth and started climbing up the wall’s rough surface, so they reached a height of about 30 cm over the floor. The first mouse then let go of the egg, which fell to the ground and cracked open, thus enabling them to eat it. This kind of synergy surely requires planning ahead to accomplish, more than that it requires a sense of teamwork and an exchange of information between the mice. I, for one, was very impressed with their clever plan!

  4. one little comment:
    There actually is a slight correlation between brain size (measured cranial capacity in MRI) and intelligence in humans – it is about 0,33. This data come year 2000 from metaanalysis of 36 researches by Vernon et al.
    Still, I doesn’t explain much of the IQ variability in population. But it’s one of the puzzle’s pieces. 🙂
    to Tasos Thomaidis:
    I heard such stories before :). Rats are belived to be capable of such coordinated teamwork. Amazing.

  5. The first demonstration of transitive inference in animals under controlled conditions was NOT in pinyon jays. It was in squirrel monkeys (McGonigle & Chalmers, 1977), and then in chimpanzees (Gillan, 1981) and pigeons (Tanaka & Sato, 1981). In 1990s, there were many studies on this topic in chimpanzees (Boysen et al., 1993), rhesus monkeys (Treichler & Van Tilburg, 1996), rats (Davis, 1992; Roberts & Phelps, 1994), and pigeons (von Fersen et al., 1991; Higa & Staddon, 1993; Sieman et al., 1996; Steirn et al., 1996).

  6. wonderful stuff! what interests me too is that there are some people for whom innovation at need is habitual, and enjoyed as a game; whereas some other people seem to have little or no inclination to develop novel solutions even when real-life need provides an incentive for them to do so, and even an unease or distaste at novel solutions devised and demonstrated to be effective by others. Is this ‘innovation-reluctance’ an exaggerated herd instinct.. such people needing to know that there is precedent before they will adopt a course of action?

Comments are closed.