3
Finding food is essential for survival and through learning and experience such tasks will usually be mastered very well. Many birds need to learn which food items they can eat, where to find them and how to get them into their stomachs. Some foods are also hidden from view. They may need to be peeled and their outside appearance may be entirely different from the enclosed edible part (protected by rind, peel or hard shell). Other types of food must even be extracted from below the ground. The question is how such mastery works and what it implies. Furthermore, one would think that a bird needs to be aware of food shortages and of possible competition for food and may need to either plan ahead to obtain as much as possible for lean times and hide it (hoarding or caching), or migrate to new areas to avoid acute food shortages. Such food shortage may not be the result of seasonal change, as is the case in high latitudes of the northern hemisphere (with its reliable changes in temperature), but based on a decision at a time when it becomes clear, through certain weather conditions, that a food source or several ones have dried up. In Australia, this could be any part of the year in so many of its regions.
Type of foods eaten
As a general rule, we subdivide landbirds into groups by the food they eat. We talk about frugivores, nectarivores, insectivores, granivores, carnivores and omnivores. In other words, we generally accept that some species feed largely on seeds or fruit, others on invertebrates or other animals and another large group on nectar, while some are identified as just about being able to feed on anything. There are physical adaptations, particularly of the beak, that may reflect preferential foods. Carnivores have strong beaks curved at the end of the upper mandible so they can pierce a hide and tear flesh, while honeyeaters have finely curved beaks with long tongues that can be inserted into the neck of nectar-producing flowers, and so forth. Roughly, the beak can show on which food type a species relies, although some species may have an all-purpose beak.
But how true is all this in Australian birds? Australia’s flora and climate are fickle, unreliable and irregular (also referred to as environmental stochasticity) – for vast tracts of land there may be droughts for many years running or floods, fire, sandstorms, devastation of grasses and crops by locusts. There may be sudden frosts in areas that do not often get frost, delayed onset of rainy seasons or plagues that may devastate whole stretches of land, storms that unearth ancient trees and hailstorms that denude entire stretches of forest. Australia is the ‘land of the seven plagues’ and demands a range of adaptations that can somehow meet these challenges.
There are some specialists that defy the rules and have become so specialised that the very nature of their specialisation now threatens their survival. Black-cockatoos, Calyptorhynchus ssp. are a case in point, especially the glossy black-cockatoo, which is a very choosy feeder. In South Australia, glossy black-cockatoos feed almost exclusively on the seeds of the drooping she-oak (Allocasuarina verticillata), as they do in inland New South Wales, on Allocasuarina torulosa in Queensland and A. littoralis in eastern New South Wales and eastern Victoria (Forshaw and Cooper 2002). The South Australian species that now occurs only on Kangaroo Island (10 years ago numbering fewer than 200 individuals) does not even feed on every tree available of the same species but shows strong feeding bias towards larger trees and trees in which others have fed before (Pepper et al. 2000). The Gouldian finch feeds almost entirely on grass seeds, mostly taken from grasses in the genus Sorghum (Dostine and Franklin 2002). Both species are threatened with extinction. Hence, highly specialised species do exist, but it seems such specialisation comes with a price and it is almost always human intervention that has been the cause of their demise.
It will perhaps not come as a surprise to hear that a vast number of Australian birds are not feeders on exclusive diets, as we are often led to believe via the pet trade. There are two components to this – most vertebrates need all the same food groups that humans do, it just depends in what quantities and during which season they might be needed. Second, it may be one of the most fundamentally important adaptations of Australian landbirds that most of them can eat foods from a large variety of groups and that an astonishing number are either completely or almost completely omnivorous (as are humans) or can be so if the situation demands it (Barker and Vestjens 1990). Ravens and crows are largely seen as carrion feeders, but they also kill mice and lizards and kill even relatively large birds, such as pigeons, noisy miners and starlings. For several months a year, depending on season, they may mainly feed on grasshoppers, cicadas and beetles, but then have months and times in which they feed on seeds and fruit (Rowley and Vestjens 1973). Channel-billed cuckoos, for instance, that are largely frugivorous will also act like birds of prey, taking live nestlings, often at advanced age, from a whole range of bird nests such as noisy miners, apostlebirds and magpie-larks, as well as a whole range of large insects. Equally, the figbird, while largely frugivorous, also feeds on seeds, insects and nectar (HANZAB 2002). Even more surprising was the discovery that bird species that are not associated with nectar feeding, will actually take nectar. In the Top End of the Northern Territory, it was discovered that 21 species classified as insectivores, granivores and even frugivores were seen regularly drinking nectar (Franklin 1999). A second adaptation is that Australian birds are often on the move, as will be discussed below. The point to be made here is that many of them have made adaptations at species level to deal with the problems of a fickle food supply and it is thus no longer appropriate to refer to specific foraging strategies as food innovations if they have already become the standard repertoire of a species.
Methods of obtaining food
Feeding techniques fall into several categories. One group stays put and expects to find the food within its own vicinity either on the ground, in the air or on trees and shrubs or a mixture of all three. These are largely territorial birds and those that keep loose home ranges. Feeding areas can also be divided vertically and some of these divisions can be very rigid indeed (called resource partitioning), meaning that some species may exclusively forage on the ground, others feed in the lower canopy and yet other, even closely related, species may exclusively feed in the upper canopy. Such feeding techniques are also affected by time of day. Some feeders are strictly diurnal (feed during the day), others crepuscular (i.e. feeding at dawn and dusk) and yet other species are exclusively nocturnal – all factors that may need to be taken into consideration.
Failure to take such resource partitioning into account has led to several disasters in biological control measures of introduced pests, as is well known. The cane toad is a prime example in Australia (feeding on the ground while the pests lived up the sugarcane shafts) and another, in Fiji, was the introduction of Australian magpies to control stick insect pests that threatened to destroy the coconut palm trees. However, magpies feed on the ground whereas the stick insects stay on the palm leaves. Magpies are strictly ground feeders and do not feed when roosting in trees (had currawongs been chosen, success would have been more likely). Magpies can be found on the island of Taveuni to this day and form an isolated population there (Kaplan and Rogers 2013).
Another group comprises food-switchers. Food-switchers appear to be very common in Australia and therein lies probably one of the main adaptations for a continent that often produces seeds, nuts and fruit at an erratic and unpredictable rate. For instance, adult currawongs are frugivores and also semi-nomadic. The pellets they produce contain the seeds that remain unharmed by digestion and then get ejected within a pellet that provides its own soil capsule to give a seed the chance to grow. However, their offspring first grow up entirely on a diet of invertebrates and even meat and eggs and then switch to fruit once fledged. But, while they continue to be an effective pest control of stick insects throughout adult life, they largely remain fruit eaters and seed dispersers from thereon (Buchanan 1989) although additional seasonal adjustments according to availability are also made (Wood 1998).
Honeyeaters, such as Lewin’s honeyeaters and noisy miners, known for their consumption of nectar, survive by consuming large amounts of invertebrates and, for some time in the year, may switch exclusively to one or the other food source and even defend it vigorously (Barker and Vestjens 1990). The common koel, a large cuckoo of the eastern seaboard, has its young raised largely by insectivorous hosts. A favourite host is the Australian magpie but there are others such as wattlebirds and currawongs. The moment the parasitic youngsters leave the nest, they feed on a diet of fruit. The female common koel tends to wait until the offspring is ready to fledge and then she collects the youngster after its first flight, presumably to enable the fledgling to recognise the very different food sources.
A number of studies confirm that food switching is widespread, especially at times when a typical food source is scarce. Interestingly, Franklin (1999) found that even insectivores may use nectar as a food source when insects are in short supply (such as during dry seasons) or simply when nectar is plentiful. Of the 21 species he identified feeding at the nectar source, one was a cockatoo (the little corella), two were parrots (red-winged parrot and northern rosella), and also striated pardalote, grey-crowned babbler, rufous whistler, spangled drongo and white-bellied cuckoo-shrike. Also seen were white-winged triller, three members of the Oriolidae (yellow oriole, olive-backed oriole and figbird), five Artamidae, including three species of woodswallow (white-breasted, black-faced and little woodswallow) as well as two butcherbird species (grey and pied butcherbirds), one corvid (Torresian crow), great bowerbird, double-barred finch and mistletoebird. That means that nearly half of the species recorded feeding at the nectar site were generally classified as insectivores, only a third as omnivores, and in this group particularly large birds as crows, bowerbirds and butcherbirds, and just two species of frugivores (mistletoe and figbird) and three species classified as granivore species such as corella, parrot and finch (Franklin 1999).
This kind of evidence makes an important contribution because the extraction of nectar is usually considered a specialised way of foraging. At the very least, the ability of many Australian birds to recognise, use and switch to unusual food sources is a remarkable adaptation to a difficult environment in which whole classes of food may at once disappear or be delayed by drought or fire, flood, storms or devastation by insect and other plagues.
Another observation to note here is the link that Ford and colleagues made between feeding regime and cooperative behaviour. They argued that there are no known cooperative species among seed eaters (such as Australian pigeons, and finches) or generally among frugivorous birds (such as parrots). They showed that cooperative breeders are found mainly among the insectivores generally and the species in these groups represent most passerine families in Australia (Ford et al. 1988). To quote:
It is among the insectivores generally that co-operative breeders are most frequent (20 out of 49 species). They include representatives of almost all the main passerine families found in Australia (Climacteridae, Acanthizidae, Maluridae, Meliphagidae, Eopsaltriidae and Corvidae sensu Sibley and Ahlquist 1985) and both large (e.g. Australian magpie, white-winged chough) and small species (e.g. thornbills, fairy-wrens) (p. 241).
Moreover, they found that cooperative breeders are not equally distributed across the continent but appear to have their highest density in Eastern Australia (Fig. 3.1).
One hastens to add that seed eaters and frugivores do not necessarily feed on their own. On the contrary, many species of pigeon, most finches, most parrots and cockatoos feed in family groups or flocks, changing sites together, advertising food sources and looking out for each other. Some, such as galahs, have structured in sentinel duties and even crèches to leave their youngsters behind in the care of a few adults. Hence, there are affiliative and social connections in feeding for the majority of all landbirds in Australia, with the exception of ratites, megapodes and most birds of prey.
Cognitive adaptations
One would expect that different cognitive mechanisms apply even within foraging behaviour: for instance, the extent of spatial memory needed for localised foods versus memory for food items distributed over vast areas may be different. It may be reflected in searching techniques and in switching from local memory to movement-based strategies (Sulikowski and Burke 2011a). On the other hand, exploration of new food sources may be required if the known food sources are scarce. Nicolakakis and Lefebvre (2000) surveyed the literature on western European birds for feeding and nesting innovations covering 30 years of journal publications. In a relatively small number of species in that region, they found as many as 507 described cases of feeding innovations and they further claimed that forebrain size was a second significant predictor for feeding innovations, while this was apparently not the case for nesting innovations. When Lefebvre et al. (2001) looked for feeding innovations in Australian birds, they found over 120 feeding innovations, but seemed to express some frustration with the material saying that such behaviour ‘should be featured in more notes’ but they did suggest that a larger neural substrate for processing and integrating information could be associated with innovative feeding. They further noted that, if all other things were equal, birds that have a broader diet (generalists) would respond more quickly to new feeding possibilities (opportunists) and show innovative solutions to feeding problems. There is a different question to consider: if Australian birds so readily switch to other food classes what does that mean about learning and memory? I would suggest that it has a great deal to do with it. Understanding what’s edible across food classes may require learning and memorising hundreds of items from different food classes and may involve exploration and discovery.
Fig. 3.1. Distribution of cooperative species across the Australian continent. (1) lowest concentration (fewer than 11 cooperative species per square km); (2) 11–14 species; (3) 15–18 species; (4) 19–22 species; (5) 23+ cooperative species. Ford and colleagues point out that area 5, the highest concentration of cooperative species, roughly coincided with least seasonal growth of plants. (Adapted from Dow 1980a and Ford et al. 1988.)
Manipulation to procure food
To be able to manipulate food in ways that aid the successful transfer of that food into the digestive tract of an animal seems very basic indeed. However, most Australian birds are altricial, meaning, as said earlier, that their young are often naked, helpless and entirely dependent on adults on hatching. Except for two orders, almost all landbirds belong into this category and need to be fed by adults, and will only gradually learn what food to eat and how. Skills are required to move a seed from the tip of the beak to the back of it so that it can be swallowed. Juveniles are usually not very accurate and often not quick enough to secure food that may be easily within reach, seeds may fall out of the beak or live food simply escapes from the puzzled bird. These are maturational aspects of development and learning and will be mastered once coordination has improved, but acquiring these techniques does not need to involve any complex cognitive aspects. Food identification, however, generally does involve learning and does so for most species.
Not all food comes in immediately edible form and some needs to be processed first. Even the simplest morsel of food may require some manipulation, such as removal of bark, husks or skins. For instance, parrots may dehusk seeds, peel shells from nuts or strip items off hard shells and skins, and they need to learn the techniques (and each step of this technique in the right order) or they would not be able to access the food.
Food extraction
Food extraction from under the ground or from a protective hard casing has been regarded as a mark of higher cognition in the primate literature. The discovery that some chimpanzee groups use hammer and anvil to crack open nuts has occupied the literature on cognition and tool use in primates for some time. Birds use tools as well, as will be elaborated in the next chapter, but even without tools they have shown remarkable ability and innovation in processing materials, such as shells and prey with tough skins or painful armour (Gibson 1986), activities for which both feet and beak are quite often needed to complete the task.
Magpies are very keen on bone marrow and, very occasionally, they may have access to a bone that happens to have been fractured. I have witnessed magpies going to great length to manoeuvre the open-ended side of a mammalian bone into a position from which they are able to insert their beak into it, either by holding one foot firmly on the bone or even by positioning it in a vertical position, such as on a ledge or a rock and then grasping it with one foot, while feeding. The impaling of prey on thorns by many of the shrikes (Laniidae) is well known (Smith 1973). Other birds, such as butcherbirds, store food in the forks of branches in trees and shrubs (Walters 1980); and several other birds may use spines or forked sticks to anchor a carcass while they flay it with the bill (Sedgwick 1947).
A good deal of manipulation is also involved in the mussel-eating behaviour of white-winged choughs that was recorded by Hobbs in great detail at Marthaguy Creek north of Warren (New South Wales). The 19 choughs he observed were not only all very successful in opening the mussels but the mussels also had to be dug from the mud at the banks of the river. They were not visible to the foraging birds. The choughs thus needed to know where to find them, how to extract them and then how to access the food. Yet all 19 were successful in pulling out mussels hidden several centimetres deep in the mud. They then went to great trouble to clean off the mud and then hammered the shells with their beak. In some cases, the procedure of opening the shells also involved tool use if a bird found another mussel or empty shell. The description is worth quoting here in full:
During this activity the birds eventually came across empty shells. Immediately, if the bird was carrying a mussel, it dropped it and picked up the empty valve. Holding it so that it pointed downwards, with the convex side away from its breast, the bird repeatedly struck the unopened mussel with it. Usually the blows were made with a slight side-to-side movement of the head so that the mussel was struck a double blow; when broken valves were used, direct downward blows were made. The blows did not appear to be aimed at any particular part of the mussel. Occasionally, an unopened mussel was used as a hammer, also with an attempt at the double stroking blow; sometimes the target was missed; at others the bird simply dropped the ‘tool’ on the mussel, after holding it high with outstretched neck (Hobbs 1971, p. 84).
Today we would describe this as complex tool use and argue that such a problem-solving activity requires innovation and, in order to do the task, at least the very first bird of the group that performed it was doing it by its own wits (complex cognition). The others may have learned this by association, of course, but it is clear that the innovation opened up an entirely new food source and one that was nutritious.
Using tools will be described in the next chapter in more detail but here is perhaps also the moment to mention prey dropping as a strategy when all manipulation does not suffice to extract the edible part. The best known example is that of a raptor called a Lammergeier, that takes very large bones to the sky, drops them over rocky outcrops and thus breaks them to access bone marrow (Huxley and Nicholson 1963), or gulls that fly high to break shells (Maron 1982; Switzer and Cristol 1999) or common ravens that drop walnuts (Cristol and Switzer 1999). Another variant: on one occasion I saw a black kite take to the air with a live lizard in its talons. Once high in the sky, the prey was dropped and the kite immediately flew after it and consumed it on the ground. Whether this was an accidental loss of the prey item or a preference of the kite to feed on a meal that was dead, cannot be established. However, in all cases of food droppings, the birds managed to consume the items using gravity itself to solve the problem of extracting the food or to make it easier to consume the food.
Another unusual case of preparing food for consumption was described by Hindwood in 1946. In his garden in Sydney, New South Wales, he noticed that shrike-tits were not just occupying a tree but systematically cleaning up caterpillars of a moth, Mnesampela privata. The problem with this eucalypt species (Eucalyptus cinerea) is that it produces a highly concentrated oil containing cinerol (or eucalyptol). Such a potent and poisonous concoction that does not harm the caterpillars but, accumulatively, could harm the birds, would normally be ample protection for the caterpillars. However, shrike-tits have found a way around this problem by removing the alimentary canal (the entire digestive tube) that contains the cinerol concentrates. Hindwood writes:
(Prizing open their webbed shelters …) a bird will extract one of the larger caterpillars with its powerful beak; it will then fly to a more stable perch, hold the grub against the branch with one foot, straighten it out, break the skin behind the head and extract the alimentary canal, which usually comes away entirely … (Hindwood 1947).
The removal of the alimentary canal makes the food safe for the birds. How they acquired the skill of extracting the distasteful, and possibly fatal, part of the caterpillar’s body is impressive and not known. This is problem solving at its best.
To feed on something poisonous requires some substantial feeding innovation. The cane toad is Australia’s curse because nothing has so far stopped its catastrophic advance from Queensland across the entire state and much of the tropical end of Australia, having now reached New South Wales and being already ensconced in Lismore and the Northern Rivers area. The toad kills everything in its path that tries to eat it, because it is poisonous at every life stage and has sufficiently toxic external poison glands to kill snakes, lace monitors and even mammals such as eastern quolls. Even eggs and tadpoles that are eaten kill fish and birds, leaving a trail of dead native animals. The same is true for reptiles, including crocodiles, and small carnivorous mammals (Beckmann and Shine 2011). The northern quoll is near extinction because it had fed on cane toads and was killed at an alarming rate. A new conservation program in Kakadu National Park has taken the effective route of hand-raising young quolls and teaching them to avoid cane toads. It seems that this program has so far had the greatest success towards saving this native species. It is known that bee-eaters are seriously compromised by cane toads. The problem is that the poison glands are located on both sides of the toad’s head, thus are most effective when an animal tries to capture the toad from behind.
There appear to be some invertebrates that are not affected by the cane toad, such as wolf spiders, tarantulas, meat ants and water beetles. Another natural solution seems to have offered itself via two species of birds. Apparently, both Torresian crows (Donato and Potts 2004) and the currawong have learned to kill the toad by flipping it over on its back (see Fig. 3.2), and ripping open its soft underbelly with their beaks, exposing the internal organs, which are then a safe meal (Debus 2012). Two species of raptor, black kites and whistling kites, have also been identified as having found a way of bypassing the poison by consuming only the tongues of cane toads. However, they tend to take only carcasses. Beckmann and Shine (2011) concluded that the invasion of cane toads thus constitutes a novel prey type for scavenging raptors, rather than a threat to population viability. Both the techniques are remarkable feeding innovations and there is general hope among conservationists and the wider public that these innovations may spread and might become more widely established in native bird populations.
Fig. 3.2. A cane toad has been successfully dissected by the crow. The innovative technique is to throw the cane toad on its back where the poison glands are and thus unable to do harm. The bird does not swallow the carcass but feeds only on the inner parts. (Adapted from image 4786, taken in Mackay by Ian Sutton on 23 June 2008. http://www.flickr.com/photos/22616984@N07/2603282907/in/photostream.)
It is unusual for landbirds to attempt to catch fish, yet that is precisely what Australian ravens have been seen doing in Western Australia and successfully so. They stand on a rock in a flowing creek and catch fish in their beak while close to the surface and taking them to land for consumption (HANZAB 2006a, p. 700). They also used an anvil to steady snails so that the shells can be broken (Rowley and Vestjens 1973).
A food innovation that has occurred in Australia, and is generally relatively rare, requires collaboration or mutual agreement (mutualism) between birds and mammals. Bradshaw and White (2006) discovered that Torresian crows had commenced cleaning ticks off Indonesian cattle, called banteng (Bos javanicus). Banteng were introduced to Northern Australia about 160 years ago and today the original 20 individuals have increased to about 5000 in the Arnhem Land area. Bradshaw and White (2006) described the elaborate, rather than incidental, set of circumstances in detail. In short, the crows approached the female banteng, somehow indicating their intention. The banteng female then rolled onto her back and held her legs up, straining to hold her position, so that the crows could get to the belly and the area between belly and leg. The crows then proceeded to quickly peck at the exposed areas, the authors assuming that the crows extracted ticks and the cow then rolled back onto her belly. Obviously, this is a feeding innovation that is quite dangerous because the crows are not flying onto the back of their chosen species but under the belly. It therefore requires a very clear understanding between the two species: (a) that one means no harm and (b) that it is of benefit to both parties. The entire interaction is thus quite remarkable.
It has been said that food extraction (i.e. finding food that is not visible, such as digging out roots, as great apes do) requires not just skill but cognitive abilities – knowing where to look and how to proceed and even the right timing for extraction. It requires the ability to learn and to remember the crucial aspects of the extraction process. It cannot be innate because the scenario is a complex interplay of timing, positioning, and use of visual or auditory cues that have to be learnt. Magpies regularly extract food from under the ground. They cannot look for it but they can listen for it. These are the larvae of the scarab beetles that make minute gnawing sounds and vibratory movements under the surface. Indeed, a study in the early 1980s (Floyd and Woodland 1981) showed conclusively that magpies find scarab larvae by sound alone (and in some cases some vibratory cues) and not by olfactory or visual cues, as starlings and American robins do when they find earthworms, by spotting minute mounds of dirt heaped above the ground and left behind by the worms while they moved beneath the surface (Heppner 1965; East and Pottinger 1975). This listening method of extractive foraging is not innate. It is a skill that fledgling magpies have to learn and, while they do so, they walk close to the adult and eventually link the sound to the food by watching the female deliver a powerful jab with her beak into the soil and then bring up a juicy grub in the beak (Kaplan 2008b).
Procuring food by unusual means
Even at the most basic level, one would expect that foraging involves information processing and often at least a limited amount of decision making. First, one would predict systematic differences in the manner in which different resources are exploited (techniques for gathering). Second, one would expect birds to be risk sensitive (Bateson and Kacelnik 1998). This may demand vigilance and awareness of possible danger, perceiving the direction from which such danger could come and planning an escape route. Third, one would expect an awareness of risk arising from competitors for a food source, be they conspecifics or heterospecific (Dow 1977).
The prize for the most unusual way of procuring food (and tool using), would have to go to the black kite, however, and possibly to other birds of prey of the Top End. Kites, widespread also elsewhere in the world, occupy a vast area of Queensland and the Northern Territory and can be found in scrub and dry inland areas, as well as in coastal areas of the Top End. A newspaper article in the Tennant Creek & District Times reported on 10 May 2013 that a Mr Bob Gosford from an Ethnoornithology Study and Research Group was going to give a paper at the Society of Ethnobiology in Texas the coming week. He was to report eye-witness accounts of black kites that picked up smouldering firesticks and carried them in the air to an unburnt section of the bush, then dropped the stick and waited at the edge of the ensuing fire for small prey to be flushed out of the grasses. He said that there was a growing body of evidence to suggest that some raptors are active promoters, preservers and users of fire in the Australian landscape. A fire fighter from Tennant Creek confirmed this too, saying ‘It’s a nightmare for fire fighters but it’s pretty smart of the birds’.
Another unusual behaviour of food procurement in black kites involves ‘fishing’. There was one case of bread dunking observed in the black kite on the banks of the Lennard River at the camping ground of Windjana Gorge National Park in the West Kimberley of north-western Western Australia (Roberts 1982). The bird used the dunked bread in order to lure prey to within the birds’ reach. It is really a remarkable tool use and one of the few examples of ‘baiting’ behaviour known worldwide, reviewed by Boswall (1977): he reported cases then that included a green heron in Florida using pellets of fish food and three others using bread scraps to lure fish within range of their bills. Another was a green heron, in Kenya, also baiting fish with bread. A captive sun bittern in Britain baited fish with mealworms. Other known cases of baiting events observed are in Kenya by a pied kingfisher and a Squacco heron. Hence, baiting appears to be rare and the reported cases are all species specialised in foraging in or near water. However, kites are not fish hunters. They are largely known as scavengers and, apart from feeding on carrion, might hunt for a range of small vertebrates. The kite was observed for over an hour while it repeatedly tried to secure a fish or crayfish. The kite dropped some bread in the shallow part of this river and then waited until fish or freshwater crayfish (cherabins, Macrobrachium sp. about 20 cm long) had assembled and then tried to catch a live prey in one of its talons. It eventually succeeded in catching a crayfish (Roberts 1982). Given its normal range of feeding, this behaviour seems a highly unusual and innovative method to procure food.
Black kites detect potential food sources from the air, gliding on thermals. However, in a separate incident, they displayed yet other skills as I observed myself. On our way to our field research station in far west New South Wales (half way between Bourke and Wilcannia), we stopped for a camping night near Brewarina on the Barwon River. Our galah, a formerly abused bird that had lost the ability to fly because of old wing fractures, always accompanied us on these long trips and I had put the bird into a nearby native shrub so it could feed and climb while we were getting the campsite ready for the night. I had noticed two black kites circling above and swiftly spiralling down towards us and I motioned them to go away with wide sweeps of my hand. The birds seemed to respond and were no longer seen and we concentrated on our tasks at hand. It was almost pure chance that I looked around and found that the two kites, far from having disappeared, had not just come close but were on foot less than a metre away from the galah. The galah had not raised the alarm either because of their cautious approach. When I looked around, they were even taking cover behind a rock and my walking towards them first make them peep at me and only when I had nearly reached them did they very leisurely fly off. The entire scene had the mark of a clever and unmistakable hunting strategy. While kites often walk across the ground to collect lizards and even insects, such a deliberate act of ambush on foot was a remarkable experience and in this case, it was a shared ambush by two kites. Not even the skilful ambush hunter, the goshawk, would try to do this. Galahs are a favourite food for many birds of prey. Even small raptors, such as black-shouldered kites can be seen pursuing them in flight. To approach within a metre of us, though, seemed very intentional and carefully planned. Their behaviour resembled more that of a wild feline planning an ambush than a bird of prey and, given all the other examples of tool use and food procurement observed in this species, it is not exaggerated to suggest that the black kite is a particularly versatile and innovative hunter, with a large repertoire of strategies at its disposal.
Geographical mobility as a feeding strategy
A third, and probably by far the largest, group of Australian birds has to seek new food and water supplies and this can take many forms. Some may be locally mobile, semi-nomadic or fully nomadic or others may be seasonally migratory. Silvereyes (Fig. 3.3), which are omnivorous feeders, are interesting because some small populations are sedentary and stay at a location while other populations completely vacate the south in autumn and head north to Queensland. This species has caused people to wonder how a bird so small can traverse the vast body of sea between Tasmania and the mainland and that other populations travel the thousands of miles from Victoria to Queensland – and it has become a model species in Australia for migratory birds (Wiltschko et al. 2006, 2009; Deutschlander et al. 2012). Migratory behaviour in the high latitudes has been chiefly associated with seasons and temperature changes. Northern hemisphere birds by and large, have to move from their summer grounds and move south, many of which will go as far as Africa or, if flying from North America, as far south as Mexico or the northern parts of South America.
Fig. 3.3. Silvereyes – remarkable migrants. Silvereyes are very small and, as other small migratory birds, they impress us with their ability to fly thousands of kilometres. They are an important study species in searching for establishing what precisely guides them to find their way. (NB: not all silvereyes are migratory.)
Migratory behaviour has not quite generated the same degree of excitement in Australia, partly because the migratory sea- and shorebirds that arrive from as far away as Russia (even Siberia), China or Japan only tend to overwinter in Australia and then return to their main breeding and feeding grounds back north of the equator. Further, because landbirds remain on the continent, it does not necessarily conjure up the same sense of distance as do the migratory feats of small songbirds flying all the way from England to Africa, or from northern parts of North America to Mexico or even Columbia to overwinter. These journeys seem much more dramatic in terms of distance than the journeys across the Australian continent (although these distances are often no smaller) or the possibly short hops from the Top End of Australia to New Guinea. However, Australian nomads and even some birds not thought of as migratory at all, may cover distances that are substantially larger than most typical northern European, and even some American, migratory routes (although some of these are extraordinarily extensive). We know from banding records that wedge-tailed eagles, otherwise largely sedentary, may travel quite far – at least juveniles do. Two juveniles were clocked off in different years as having covered 856 km and 868 km, respectively, in their journeys established from banding records (HANZAB 1993, p. 168). Indeed, as was noted by Griffioen and Clarke (2002), there are large-scale movements across the continent evident from the latest Bird Atlas data (Barrett et al. 2003).
Spatial memory and hiding food
To understand how evolution has shaped animals’ behavioural and cognitive mechanisms, much research has focused on spatial learning and memory. The ecology of species has been used to predict differences in spatial cognition as a function of species (Olson 1991) and context (Sulikowski and Burke 2007; reviewed by Sherry 2006) and, indeed, there are differences.
There are several very different types of feeding strategies that do require spatial memory of some kind. One strategy relates to knowing when and where fruit might be ready to be harvested (relevant to the nomads, migrants and locally mobile species) and, secondly, micro spatial memory is needed for a local, even territorial, bird that has hidden food itself and needs to be able to retrieve it. One is an extra-group strategy and another an in-group strategy, meaning that location is either in a large scale or confined locally.
Another special subcategory is the question of memory of depletion and replenishment of a food source. Nectar production is quite unusual in that, once taken, it is only depleted for a short time and then the plant produces more the next day or several hours later on the same day. For efficient foraging, it is a question of remembering which flower had been visited and which still has nectar to harvest. The latter is discussed in the literature under win-stay and win-shift categories. Win-shift biases refer to the tendency to spontaneously avoid the food source rather than to return to locations where food has recently been found. Experiments have reported win-shifting behaviour in five families of Australian nectarivorous birds. In some of these families of the large group of Australian honeyeaters (Meliphagidae), it was assumed that a win-shift bias had become part of a general, even innate, response to feeding on nectar, but this is not always the case, as shown in other species that equally feed on nectar. In a captively reared group of rainbow lorikeets, for example, no bias was found and yet they foraged efficiently, had good spatial search ability and above chance performance for both shift and stay contingencies at long and short delays (Sulikowski and Burke 2011b). This suggests that it is a skill that requires decision making, memory and some active thought processes, at least in this species. Moreover, the question of motivation or intention may also determine performance. This was made evident in a study of spatial cognition and feeding in noisy miners (Sulikowski and Burke 2011a). The researchers tested this omnivorous honeyeater on both baited and unbaited nectar feeders and also on invertebrate provisions. When they were set up to search for invertebrates, the miners made a mistake by occasionally revisiting unbaited feeders but if the task was solely to remember baited and unbaited feeders, any errors disappeared.
Caching birds such as the Florida scrub-jay (Dally et al. 2006a, b), and sometimes currawongs, are said to have evolved a particularly good spatial memory needed when storing food and for finding that food again. In birds with caching behaviour, it has also been noted that ravens, for instance, are able to assess the food according to its perishability and will retrieve the items that spoil fastest first (Clayton and Krebs 1994). In birds that cache their food, the part of the brain used in the task (the hippocampus) increases in size when the bird is given the opportunity to cache.
The best-known example of caching abilities is the North American Clark’s nutcracker that has the extraordinary capacity to remember the location of tens of thousands of its seed hiding spots. It does so by using cognitive spatial maps: that is, using geometry rather than simply remembering the details or landmarks. In case of the nutcracker, it seems that evolution has adapted its hippocampal regions to do this feat and it is clear why – harvest time of pine seeds is confined to a very short period in autumn but winter is long and the snow-cover of half a year would make it impossible for the birds to survive without this source. By harvesting as many pine nuts as possible, the bird can use these caches as a food supply for the remainder of winter. Such caching skills have not developed in tropical birds for obvious reasons: food supply tends to be ongoing and foods may rot very quickly or be taken by insects, especially ants.
Experiments using western scrub-jays suggest though that there may be more at work than merely memory of geometry. To hide food in the presence of others is risky because somebody else may remember the site and subsequently steal it. The ability to remember exact and general location of the hidden items for days has been tested in a number of Northern American birds such as pinyon jays. Bednekoff and Balda (1996) found that the birds watching someone hide a piece of food even 2 days later could remember the location of caches made by the observed bird, and even 7 days later searched in the correct general locations. No studies have been conducted in Australian native species and we thus cannot tell whether there are any species that show similar ability both in caching and in retrieval and in the tactical game of rehiding a cache when the owner of the morsel had seen a conspecific watch it doing the hiding.
There have been increasing number of reports of Australian crows and ravens caching food in the wild (HANZAB 2006a, p. 701) using different techniques and approaches but often involving digging and certainly hiding the food for later consumption. This applies to all five Australian corvid species (Secomb 2005a,b). Otherwise, the information of caching in Australian species is rather scanty. There is a published Australian report on a pair of nesting peregrine falcons seen caching their food (Cameron and Olsen 1993). On many occasions I have observed Australian magpies making half-hearted attempts at hiding food but usually deciding to eat it rather than hide it, even when it was already hidden (i.e. they retrieved it themselves, ran off with it and then, seemingly running out of ideas of where to next hide it safely, consumed the morsel) but occasionally, it seems, that magpies also cache their food (Rollinson 2002). Bowerbirds, currawongs and magpies are known for object caching, but food caching has been reported in bowerbirds only for the Papuan McGregor bowerbird that is not found in Australia (Pruett-Jones and Pruett-Jones 1985).
One anecdotal event can be recounted here involving a particular currawong, a bird that was hand raised because it had been orphaned. When this bird was old enough to be released into the garden (and basically ready to leave), I showed it a place where I was hiding grapes in a small brown dish: a real hiding place on the ground behind one of the aviaries but close to the house and further hidden by a ledge. I placed grapes there every day for a week and then only once every 2 days, slowly diminishing the amount, as is practice in so-called soft releases of hand-raised birds. I observed it taking the grapes every day and then the bird eventually disappeared from the premises. Its right wing had been broken and had mended in a slightly distorted way so that wing feathers protruded on the right wing even when the wings were folded in the resting position, but did not affect its flight, however. The bird was therefore clearly distinguishable from other currawongs. After a 5-year absence, this bird turned up in our backyard again. While it was unlikely that any other birds would have the precise distorted position of the right wing as the one I had hand-raised, I wanted to put the bird to a test. Hence, I put grapes in the hiding place that I had used years before. Because I was hiding the food behind an aviary with a solid wall, the bird, still sitting on a tree branch some 20 m away, could not see the food and the exact location to which I was taking it. I walked away from the aviary and then, in full view of the bird, demonstratively went back inside the house and watched. Barely a minute later, the bird swooped down without hesitation, hopped across to the edge of the aviary and retrieved all the grapes from underneath the ledge. The action was so quick and executed without the slightest hesitation that there was absolutely no doubt in my mind that this was the self-same bird released 5 years earlier. While caching behaviour in currawongs has been described (Prawiradilaga 1994), retention of memory of location involving a time lapse of 5 years is an extremely impressive performance. After this interaction, the bird stayed and with her partner successfully raised three offspring on the property and then they all disappeared. Given the anecdotal evidence we already have of ravens and magpies, there are obviously ample opportunities to test Australian species systematically.
Food extraction, spatial memory and food innovations have been considered hallmarks of cognitive complexity and in many Australian species there is compelling evidence of incredible feats on many of these performances. Timmermans and colleagues (2000) have shown some time ago that feeding innovation rates in birds and the size of the avian forebrain are linked. Some of the foraging habits are even accompanied by tool use and this will be the focus of the next chapter. Tool use, in a way, was one of the most common categories of behaviour described in primates and birds as a clear sign of cognitive behaviour. It is evidence of forebrain activity in birds, as indeed has been shown in humans.