4
Using and making tools has long been considered a hallmark of human superiority over other species, but we now know that some animals do this too (Lefebvre et al. 2002). Using tools is defined as the ability to recognise an object in the environment that has the potential to aid performance of a process of some kind (i.e. to assist in overcoming physical limitations). The idea of what is tool use and what is not has been a matter of longstanding debate. Generally, it is agreed that using a tool as an extension of limbs is ‘true’ tool use. Use of other items in a way not clearly serving as an extension of limbs is often referred to as proto-tool use. Vertebrates have been included or excluded from being classified as tool users on the basis of this definition. Birds were not all that well served by this definition because the best known example in the literature were the woodpecker finches (Eibl-Eibesfeldt 1961; Tebbich and Bshary 2004) that used the beak rather than their feet. Moreover, their tool use was not always seen as evidence of cognitive adaptive specialisation (Teschke et al. 2011). However, the reports about the imaginative use of tools in rooks, and of the tool use and its manufacture by blue jays (Jones and Kamil 1973), and much later by New Caledonian crows, became overwhelming evidence that birds can use tools and do so in a way that is not routine but requires some planning and thinking (Hunt and Gray 2003; Weir and Kacelnik 2006; Bird and Emery 2009).
Tool using in animals came into focus especially when it was discovered in the 1960s that chimpanzees select twigs, break them to a length suitable for inserting into termite nests and then retrieve termites to eat (Goodall 1986; McGrew 1998; Box and Russon 2004). To find that chimpanzees had some rudiment of tool using made an evolutionary link to humans. Tool using in great apes seemed further evidence of having a complex brain with an expanded neocortex (discussed in Chapter 2). In many ways, this singular observation changed the direction of research on animal behaviour and cognition. It seems a small event now, but in the 1960s it was an overwhelming discovery because it suggested that the categorical distinction we had made between animals and humans was incorrect (see also Seed and Byrne 2010).
The scientific community was even more stunned when, more recently, it was discovered that New Caledonian crows not only use, but also make, tools. It became clear that the crows were even more advanced than the chimpanzees in the sense that they could manufacture tools and thus another hallmark of ‘intelligence’ supposedly unique to humans tumbled in view of the evidence. Thus, the neocortex, as known in mammalian species, especially the enlarged version in great apes and humans, was not a precondition after all to perform complex cognitive tasks.
New Caledonian crows use various probing tools to obtain larvae from holes in trees (the larvae respond to the probing by biting into the twig and thus can be pulled out by the bird), and they manufacture a range of tools appropriate for various tasks. They cut twigs into the right length, give them the right shape with their beaks and even cut jagged-edged implements from the leaves of pandanus palms (Hunt 1996; Hunt and Gray 2003; Weir et al. 2002). Moreover, the crows store the tools they make in notch holes in trees and retrieve them for later use (Hunt 2000). This suggests that they have some notion of the function of the tools and can plan to use them in the future: a form of foresight that is another mainstay of higher cognition. If the chimpanzees had disproved that humans were unique in tool use, the crow disproved that manufacture and use of tools was a uniquely human ability. Moreover, it has since been shown experimentally that this innovative behaviour in birds is underpinned by cognition more complex than simple learning mechanisms (Taylor et al. 2010). Hence, the New Caledonian crows were singled out, international interest was raised to fever pitch and the findings were reported in prestigious journals such as Nature and Science.
However, they are not the only birds known to use tools, and many such accounts had been reported well before the reports of New Caledonian crows were published, although this was knowledge not widely shared in the academic community. As recently as 1964, a Dictionary of Birds published in England boldly proclaimed that there was only a single case of confirmed tool use and this was by New Caledonian crows (Thomson 1964). In 1977, Boswall, writing from England, reviewed the literature as it was available to him and mentioned some Australian species that were reported to have some rudimentary tool use (which happened to be identical to the tool using in chimpanzees). He summarised these early reports as follows:
Eastern Shrike-Tit Falcunculus rontatus was seen systematically examining a dead black wattle tree. The bird broke off the end of a twig with its bill, held the 2.5-inch probe sideways and then inserted it twice into a crevice. The twig was then dropped and the bird extracted a food item. A rather similar account of a Grey Shrike-Thrush Collucincla harmonica by Mitchell (1972) stated that the bird ‘probed the twig into the hole of a house brick that was among a pile of’ bricks resting in my garden ... After two or three probes ... an insect crawled out of the hole and the thrush dropped the twig, picked up the insect and flew off’. On one day in November 1971 Green (1972) and a colleague saw at least three (and probably more) Orange-winged sittellas Neositia chrysoptera using little strips of wood as probes. How a twig was obtained was not known, but each bird wielded it with the bill and placed it under the foot while picking up the grub that had apparently been dislodged. Although the sittellas carried the twigs in their beaks along a branch while they were hopping, they always dropped them before flying (Boswall 1977).
The review also mentioned reports about some other Australian species (galahs, magpies, white-winged choughs, rainbow bee-eaters and black kites), but some of these findings will be discussed in the next chapter under play behaviour. As one may note by the tone of the report, at that time in avian research, the researchers themselves had not made the link between tool use and advanced cognitive ability; hence it remained somewhat a curiosity not worthy of international headlines. There are substantially more Australian species that in one way or another use tools or engage with objects in ways that are not part of their innate repertoire, as the previous chapter on foraging and food procurement has already made clear in the case of white-winged choughs and black kites.
Now we know that tool using in birds is an ability that is widespread in Australia, but little was known about this internationally. More recently, several researchers have set out to systematically document tool use in birds. In 2002, Lefebvre and colleagues searched for tool use by birds worldwide and they discovered that there are many passerines and raptors using tools or proto-tools. Indeed, they found that there were published accounts of over 100 avian species worldwide known to use tools and at least 20%, but possibly 25%, of all avian species so far studied and known to engage with objects are birds native to Australia. Lefebvre et al. (2002) concluded that tool using is more common in corvids and passerines than in other avian orders. Their study also surmised that using tools appears to correlate with having a larger brain.
Australian tool users
Later Bentley-Condit and Smith (2010) undertook the task of updating what was known about tool use in birds, also subdividing tool users into ‘true’ and prototype tool using and identifying its purpose in each case. On the basis of their summary report, the Australian species have been extracted and are listed below. Importantly, their table also indicates the purpose of the tool use.
The list is by no means complete (note that not all bowerbirds are listed, even though all bowerbird species build bowers, not just one or two species), nor is every use incorporated, but it is certainly the most comprehensive list so far published. The early publications by Chisholm are not listed, but they are perhaps a little more difficult to access internationally (Chisholm 1954, 1971a, 1971b, 1972). Among the tool users are some unexpected and little known Australian passerines such as the orange-winged or varied sittella (Green 1972), the crested shrike-tit (Noske 1985), the grey shrike-thrush (Reilly 1965) and better known ones such as the kookaburra (Roberts 1961) and butcherbirds (Sedgwick 1947), or birds of prey such as the black-breasted buzzard (Debus 1991; Pepper-Edwards and Notley 1991) and, as already mentioned, the black kite (Roberts 1982). There has been one published account of Australian magpies possibly also using a stick for probing for insects in a tree but such behaviour has not been widely confirmed (McCormick 2007) and is also not part of the table of documented cases.
A number of the birds listed in Table 4.1 have already been discussed in Chapter 3 because the tool use is related to procuring food. Use of sticks is of course very familiar to birds because so many species build nests out of sticks, and such use of sticks for anchoring, tie-down and wall support of a nest is usually discounted. Yet the familiarity with handling sticks may make it also simpler to use such sticks in different contexts (see Fig. 4.1). There are a few examples of tool using that are related to mate attraction, personal grooming (‘other’) and even to defence. Tool use for defence is known to occur in primates, such as throwing sticks at intruders. Brush turkeys use their feet to scratch up dirt and sticks, flinging them backwards and aiming them surprisingly accurately and with substantial force at an intruder, such as goannas on egg-stealing missions (Dow 1980b; Ehrlich et al. 1988). A more sophisticated use of a tool as a weapon was observed in a conflict between a Steller’s jay and an American crow in which the jay removed itself from the scene of the conflict in order to break off a stick that was then held up against the crow under attack (Balda 2007). Orangutans and chimpanzees may use sticks in a similar fashion (Kaplan and Rogers 2000) but, among birds, this is a rare observation indeed and, to my knowledge, has remained the only one so far.
Table 4.1. Tool using Australian landbirds (extract from Bentley-Condit and Smith 2010)
Fig. 4.1. Striated pardalotes. They are said to use horizontal tunnels in the side of embankments and cliffs just as bee-eaters do, but I have seen them repair nests and in the last chapter one pair was described as using the understorey of a magpie nest. They are as competent with handling sticks as the tool users. Hence, it may not be correct to assume that the methods of nest building may be innate and that they cannot create different contexts and use sticks also as tool. (Image: Wikipedia commons)
Gannon (1930) was probably the first who noticed that male satin bowerbirds held a ‘bark-wad’ in their bills while working at their bower and ‘painting’ it. Marshall (1954) observed that males collected fragments of fibrous bark, manipulated them with the beak and so manufactured a small oval pellet with which, beak slightly open, the bird jabbed at, and painted, the individual twigs with the sides of its beak (Boswall 1977). Of course, bowerbirds actually construct bowers: structures that are exceptional in the animal kingdom and these will be discussed separately in the following chapter. The use of soft material is rarely mentioned as tool use, but here, using the bark as a paint brush, it is certainly true tool use. One might even think that this describes the use of leaves in shrike-tits. They have been seen collecting a bunch of leaves, holding them down with their left foot in order to wipe their beak clean (Noske 1985). Birds are meticulous about beak hygiene but usually do the wiping on a tree branch. Collecting leaves for this specific purpose when there are specific cleaning requirements is certainly an innovative way of dealing with a problem.
A tool use that has been observed in species worldwide is the use of stones to break open eggs. The only well-documented case of such use of stones by an Australian bird of prey is that of the black-breasted buzzards (Debus 1991). First they need to roll the egg into a position in which it cannot roll away. They take a stone in their beak, sit in front of the egg, sway their head and bash the egg, sometimes keeping the stone in the beak, sometimes releasing it as a throw against the egg (Fig. 4.2). By this means they can eventually succeed in cracking the shell of an emu egg. This is true tool use. Noisy pittas and other pittas are known to use logs or rocks as anvils to break open hard-shelled gastropods, such as snails, and arthropods (Schodde and Mason 1999) – again a behaviour that made instant international headlines when reported in chimpanzees and capuchin monkeys as stunning and cognitively very advanced behaviour, but had gone unnoticed in pittas and black-breasted buzzards and is not even mentioned among the list of tool-using birds.
Fig. 4.2. The true tool use of the black-breasted buzzard, using a rock and either hammering it against the strong shell of the emu egg or letting go of the rock with a powerful swipe of the beak while releasing it. In relation to the size of the bird, the egg is of substantial size and hardness, yet these buzzards have learned how to solve the problem and derive a substantial meal often in a harsh environment where food is difficult to find. (Adapted from Wikipedia commons.)
Stones are perhaps particularly remarkable as a tool because they bear no resemblance to sticks and nest building (i.e. to familiar use) and it would have needed considerable insight by a bird to arrive at the conclusion that the hardness of the material plus speed would create an impact strong enough to overcome the barrier of a shell.
Examples of methods of extraction were already discussed in the previous chapter. In some cases, then, tools are used not so much as cutlery but as weapons – spearing and throwing an object to procure a food item and hunting with weapons was seen as a major advance in evolution. As in primates, the use of tools by birds is passed on by learning, which is often referred to as cultural transmission.
The palm cockatoo is more unusual among the tool users in that its tool use is not related to food retrieval but to mate attraction and, territorial announcements, as well as to brooding safety. The male holds a stick in his foot and deliberately strikes it against a branch or tree trunk, accompanied by surprisingly high-pitched and somewhat puny vocalisations (Wood 1984; Bertagnolio 1994). Perhaps, the male compensates for lack of voice by the drumming as well as by fluffing up feathers, raising his crest and extending his wings when inviting a female to come to his territory. The wings are usually extended only when the female is in visual range (Fig. 4.3).
Fig. 4.3. The male palm cockatoo holds a stick in his left foot, clasping it between two digits and then raises it and, with some force, bashes the stick against a branch. It is not known whether he chooses a substrate for acoustic properties (i.e. near a hollow the ‘drumming’ would have more resonance) but such preference has been observed. (Adapted and partly redrawn from Eastman and Hunt 1966.)
Even more remarkably, the palm cockatoo manufactures its drumming tools and fashions the tools precisely to needs. These tools are clearly seen as tools by the cockatoo because the sticks are either refashioned into nesting material (described in the next chapter) or they are tools taken from one nesting hole to another (HANZAB 1999). Fashioning a tool to precise specifications and reuse of a tool is regarded as very advanced and complex tool use. The palm cockatoo’s tool use is the most complex tool use known in birds in the wild and is matched only by the New Caledonian crow. Again, the latter was reported in Nature, while the incredible tool use in palm cockatoos known since the beginning of the 20th century had gone largely unnoticed and is still not fully explored.
Indeed, the many studies of discovery of tool using in Australian birds published throughout the 20th century have barely rated a mention in the international literature and were certainly not making news in leading journals as has the tool use of New Caledonian crows. A recent paper still continues to speak of the ‘phylogenetic rarity of animal tool use’ (Rutz and St Clair 2012) but, on closer inspection, as the last two chapters have tried to show, this is not quite true. Australian birds are avid tool users and show behaviour that presupposes learning, insight or innovation and it occurs in a range of avian families. One suspects that such information is still not shared knowledge internationally.
To be sure, the idea of cognition in birds is a relatively recent one and it was only then that tool using suddenly became important as a potential indicator of complex cognition. It is now repeatedly questioned whether, in fact, this is an index of complex cognition and not simply adaptive, i.e. not reflecting special levels of cognition (Biro et al. 2013). Rutz and St Clair (2012) argued that captive-bred, hand-raised New Caledonian crows demonstrated a strong genetic predisposition for basic tool use and manufacture, thus proposing that this behaviour is an evolved adaptation. They reconstructed a scenario suggesting that a common ancestor of New Caledonian crows, originating from a (probably) non-tool-using South-East Asian or Australasian crow population, colonised New Caledonia after its last emersion several million years ago and they found different conditions there that may have fostered the evolution of tool-using traits. For instance, the presence of desirable, but out-of-reach, food, lack of direct competition for such resources, low levels of predation risk and the evolution of a prolonged juvenile development, enabling learning of complex behaviour, all combined to lead to the complex tool-using behaviour now seen in this species (Rutz and St Clair 2012). So, somehow, they argue both explanations at the same time (genetic predisposition and learning). It is not clear where this genetic predisposition might come from and what kind of changes in the forebrain would have been necessary. The argument may work for a single species but it is a little more difficult to pursue the same argument for a large number of very different species. And even if what they argue is correct in this case (although only speculative at the moment), much of what we call higher order cognition in humans is evolved behaviour and now part of our general spontaneous behaviour. As Patrick Bateson rightly said, we can do many clever things without thinking about them (Bateson and Marneli 2007). Other researchers now claim that there is no insight involved in the behaviour of New Caledonian crows because they take so easily to the task (Taylor et al. 2012). Another study (Wimpenny et al. 2009) put New Caledonian crows to further tests by asking them to use three different tools sequentially. The crows solved the problem but the conclusion of their paper is worth quoting in full:
While the ability of subjects to use three tools in sequence reveals a competence beyond that observed in any other species, our study also emphasises the importance of parsimony in comparative cognitive science: seemingly intelligent behaviour can be achieved without the involvement of high-level mental faculties, and detailed analyses are necessary before accepting claims for complex cognitive abilities (Wimpenny et al. 2009).
While one detects a general reluctance to accept that birds could be ‘bright’, one wonders how much more scientific scrutiny needs to be employed – already much more intense than it had ever been in great apes – to consider high-level mental faculties in birds as more than just a possibility. After all, in human evolution it is always stressed that the innovation of tool use was one of the most crucial milestones in the evolution of a larger brain. For instance, Iriki and Taoka (2012) proposed a triadic niche construction to explain human brain evolution, consisting of ecological, neural and cognitive factors. They write:
Hominid evolution has involved a continuous process of addition of new kinds of cognitive capacity, including those relating to manufacture and use of tools and to the establishment of linguistic faculties. The dramatic expansion of the brain that accompanied additions of new functional areas would have supported such continuous evolution. Extended brain functions would have driven rapid and drastic changes in the hominin ecological niche, which in turn demanded further brain resources to adapt to it (Iriki and Taoka 2012).
Dunking behaviour
Dunking of food may be regarded as an odd category to be placed under tool use but Bentley-Condit and Smith (2010) included it under tool using because it may be involved in food preparation or aid in capturing food or be undertaken for other reasons that also facilitate additional actions or plans. Dunking seems a widespread behaviour. It has been documented in just about all corvids both in Australian and in other corvid species worldwide (Goodwin 1986; McMillan 1992; Slee 1992; Reid and Reid 1996) and in sea- and shorebirds worldwide (Morand-Ferron et al. 2004). However, gulls have not been mentioned in any of these records but I noted potato-washing behaviour in native seagulls in Broome, a town in the Kimberley in the very northern part of Western Australia (Kaplan 2007b). Here, seagulls frequented a favourite holiday resort that served chips next to the pool. Many guests ended up not eating the entire serving of chips and left these on the table for a waiter to clear. Regularly, the birds would arrive and clear the chips off the table but not immediately ingest the salted potato strips at the table. Each chip was picked up separately, taken individually to the edge of the pool, dunked and then thrashed vigorously from left to right several times, then laid down carefully next to the pool, inspected, sometimes resulting in a repeat of the process. Only then did they proceed to swallow it.
While this is not a case of ‘true’ tool use, it is clear that using the water has a specific and desired outcome: probably to clear the food of salt. It is a behaviour that has been observed in other birds in Europe, such as in common sandpipers (Simmons 1950), in grackles (Wible 1975) and also in blackbirds (Watkin 1950). The same behaviour has been reported in monkeys, such as Japanese macaques, in capuchin monkeys, Cebus apella, and crab-eating macaques (Visalberghi and Fragaszy 1990) and now also in great apes. Interestingly, in primates food washing is viewed as representing local culture (Allritz et al. 2013). We now have so many examples worldwide of birds and primates washing food that perhaps this behaviour, and what it represents, needs to be revisited.
However, none of the Australian parakeets, parrots, lorikeets and cockatoos are listed in any tables, even though they dunk seeds, bread, vegetables and fruit as standard behaviour and this occurs without exception in any of the Australian species, admittedly not observed in the wild but in wild birds when transiently captive (i.e. when recovering from injury and then released). And it was not dependent on breeding status or having to raise offspring. It could be a playful behaviour or may show a standard behaviour for a group of birds that tend to occupy vast stretches of the inland in which water is often a scarce commodity. Most parrots feed regurgitated seeds to their offspring and any extra fluid would be advantageous. Princess parrots, for instance, may nest some 30 km away from the nearest waterhole (Eastman and Hunt 1966).
Dunking may just be a way of softening food and some field studies of Carib grackles (Morand-Ferron et al. 2004) seem to support this view because many individuals observed chose to eat the morsel once it had been dunked and soaked. I have observed dunking for later consumption in princess parrots and would not be surprised if more species have acquired this behaviour. It has been documented in Australian ravens that have been seen not only dunking bread in shallow water but standing on it so the bread would not float, then removing the bread from the water and consuming it (HANZAB 2006a, p. 701).
In other instances, dunking is not necessarily for the benefit of the morsel holder, but for offspring. Koenig reported in 1985 that dunked morsels were delivered directly to nestlings of the Brewer’s blackbirds (not found in Australia). In this case, it would be true tool use because the birds had learned that certain materials such as hard bread could soak up and hold water and thereby could be used like a sponge for water carrying and delivery (Koenig 1985). There is at least one observation reported that the Australian raven may have used bread for the same purpose and in the same manner (HANZAB 2006a, p. 701) as the Brewer’s blackbird. This behaviour certainly is a remarkable feeding innovation because it solves an important problem of supplying fluids when the food itself is dry and water a scarce commodity. The latter is certainly the case in Australia. Yet the extent and functions of dunking are not altogether clear in Australian species. To my knowledge, at least in psittacine species, the behaviour has not been documented in the wild.
Other agents or uses
A strange category which tends to be included under tool using is for personal hygiene and removal of ectoparasites. It is called ‘anting’ because the bird deliberately steps next to or into an ant’s nest, opens its wings and allows the ants to crawl all over its body or picks up ants and rubs them over specific parts of its body. The former is called ‘passive anting’ and the latter ‘active anting’. This has been observed in over 200 passerine species around the world (e.g. Craig 1999; Wiles and McAllister 2011). In Australia, detailed description exist for the drongo, the chestnut-breasted mannikin, serviced by green tree ants Oecophylla smaragdina (Valentine 2007), and Chisholm has described the behaviour in quite a few other Australian species, including the emu, satin bowerbird, superb fairy-wren, yellow-tailed thornbill, Lewin’s honeyeater, red-browed finch, pied currawong and turquoise parrot (Chisholm 1959).
More unusual is the observed behaviour of two Australian birds of prey: one, as already mentioned, is the unusual use of sticks to ignite bushland observed in black kites; the other is the use of sticks for medicinal purposes in wedge-tailed eagles. The wedge-tailed eagle sometimes uses fresh branches of shrubs which appear to have specific fly repellent qualities and take these to the nest where they are draped into a section of the nest and doing this may ease the burden of flies, attracted by the smell of rotting meat leftovers. Such medicinal use of plants has been documented to occur in some primates, especially in chimpanzees. In the same way, anting has been regarded as a form of self-medication (Revis and Waller 2004; Weldon 2004).
Another use of sticks is purely social and its description as a tool can at best be regarded as symbolic. This has been described in black-faced woodswallows. For night roosting, they huddle together as a group but to be part of the cluster each individual bird, on arrival, has to pass an entry test by carrying a stick. When accepted with a chorus of greetings, then the bird drops the twig. Each bird flying to the roost does the same. Immelmann (1960) noticed a bird that was flying towards the roosting perch without a twig but then, just before reaching it, veered away, flew to the ground and procured a twig, then flew up towards the tree branch, presented the twig, was greeted with a chorus and then dropped the twig, joining in a chorus for the next arrival. It seems that the twig serves the same symbolic purpose as a membership card of a club. It thus seems to have a purposeful social function. It is not too far removed from the gift giving rituals in which some birds engage which, quite often, consists either of something edible or a twig or grasses (as in grebes) suggestive of nest building. The symbolic use of a twig in woodswallows is an example of the use of twigs furthest removed from any discernible practical use, but it has a function that, in primates, is generally referred to as the rudiments of culture.
Means-end test
Tool use has been tested repeatedly on many vertebrates. One of the simplest is called the means-end test. Means-end problems are devised to test an animal’s ability to understand the connection between two objects either by insight, also called means-end knowledge, or by trial and error. It involves comprehension of spatial relationships. The reasons for using such a test may vary, but, whatever methodology is used, it is used to discover cognitive abilities in the animals tested. Some tools used in the laboratory are artefacts, such as presenting an animal with a rake (hardly a natural tool available in nature), with which it could obtain food if the animal was able to figure out that the rake could be used for that purpose.
The idea of using an implement to be able to access a food item has many variants. It does not have to be a rake, but could be done with strings. Such tests, for example asking the animal to attempt pulling a string to get to a favourite food tied to it at the end, or providing a tool that can rake in the food, have been used in comparative psychology for over 100 years as seemingly simple tests (Fig. 4.4). Strangely, they have mostly been applied to mammals and primates but, increasingly, bird species such as ravens have also been tested (Heinrich 1995; Heinrich and Bugnyar 2005). Bernd Heinrich tested his captive-raised ravens by providing a vertical string hanging from a branch and testing the ravens to find a way to retrieve the item. They could do so at once, and without training, by raising the string little by little, holding each bit that was raised under foot so that eventually the morsel of meat was at perch height and could be ripped off the string. The next test used two vertical strings crossing over with only one containing food and being located on a direct line of the other string’s tie at the perch level. In order to succeed in this task, the ravens thus first had to see or understand that the food at the end of one string was not in the same position on a vertical plane as the one without food and hence had to identify the food-carrying string. They performed very well in this task too.
Fig. 4.4. Means-end test design. The design for means-end tests can be very simple and has been used in laboratories for at least 100 years. One string with a favourite food item attached can be placed vertically or horizontally as above and the food placed nearest the observer but the end of the string for pulling at the opposite distant end. Unless the onlooking animal causally understands the relationship between the string and the food, it will pick the nearest string (b) to pull but not the one attached to the food (a).
In another study (Schuck-Paim et al. 2009), the means-end tests were applied to psittacine species and the tests had the added difficulty that the task was mediated via a demonstrator first. Their sample included two Australian species: the rainbow lorikeet and the sulphur-crested cockatoo. The sulphur-crested cockatoo was also tested in a similar experiment but for different purposes (Magat and Brown 2009). This will be discussed in more detail in Chapter 11.
Schuck-Paim and colleagues suggested an interesting hypothesis. They proposed that problem-solving ability (as the means-end test represents) is related to social complexity. In order to prove their idea, they selected species with varying degrees of social complexity (i.e. specific traits of social organisation). The first two they chose were highly social, living in large fluid groups, as is true of the South American spectacled parrotlet which, like the Australian galah, not only lives in large groups but also has a structured crèche system. As the second variant, they chose the Australian rainbow lorikeet, which also lives in highly social and fluid groups but without a crèche system. Finally, they selected two species characterised by small and stable group composition with one stable (dominant) breeding pair: the South American green-winged macaw and the Australian sulphur-crested cockatoo. The researchers then presented all with the same experimental design, asking of them not simply to do an individual means-ends test but to do so via observation, designating an observer and a demonstrator. Their hypothesis was confirmed: the parrotlets performed best, followed by the rainbow lorikeets, both of which followed exactly what the demonstrator had done, while the macaw and the cockatoo made more errors. In other words, the more complex the social system, the more likely it was that the bird could solve the problems and learn from demonstrators (Schuck-Paim et al. 2009).
Recently, Colbert-White et al. (2013) successfully conducted some vertical single string pulling experiments with Harris’s hawks, adding that birds of prey have rarely been tested, but this species was capable of performing this task. One ought to add that the Harris hawk is one of the most social birds of prey and the only one described in detail so far that hunts collaboratively (Dwyer and Bednarz 2011), informing, rather than contradicting, the view that social complexity may be related to problem-solving ability and learning from a demonstrator.
Our laboratory also tested captive zebra finches and they too succeeded in means-end tests. This finding is important because these tests get us out of the more typical parrot/corvid constellation of laboratory testing. Suffice it to say here that we were impressed by the quick way in which zebra finches were able to solve the single string problem and quickly learned from conspecifics to pull the correct string. Penny MacIntyre, who completed some of these experiments in our laboratory in 2009 at the University of New England, suggested quite rightly that the task would be rather natural for zebra finches (i.e. not ‘intelligent’) because it was similar to the habit of zebra finches landing on long reeds and needing to reach the seeds at the end of the reed by bending them down by the weight of the bird (pers. comm.). However, if intelligence is defined as the ability to apply past knowledge to new problems, then the quick adaptation is rather clever. Tests in our laboratory also included means-end experiments in the field using completely untested and untrained wild magpies. Magpies solved both the single string and double cross string tasks with ease. However, these comparative tests may tell us more about the ecological relevance of the tests than about intelligence.
The skills described in this chapter may have little to do with insight or cognitive complexity. Critics may call this adaptive behaviour or argue most of it could be learned by associative learning. Taylor and colleagues have therefore tested whether New Caledonian crows can learn and understand the functional properties when confronted with a novel tool type (Taylor et al. 2011). They gave them a choice to raise floating food in a tube to levels at which they could reach it. They learned the properties of objects very quickly and chose pebbles, not wood, and dropped these into the tube. New Caledonian crows have no specific experience with stones and pebbles so this was indeed a novel problem. This behaviour is clearly beyond associative learning because the choices the birds made required first an understanding of the causal relationship between weight and sinking of objects and changes in water level.
Tool use, then (and any variants of object manipulation) is far more widespread among birds than previously thought, and certainly prevalent in Australian birds, especially when all food innovations, many of which include tool use or manipulation of some kind, are included. Some other behaviour involving tools will be discussed in the next chapter under nesting and building, and further examples of object manipulation among native species will be explored a little further when play behaviour is discussed.