Australia: The Land Where Time Began

A biography of the Australian continent 

Dromornithidae - Mihirungs, the giant flightless birds of Australia

In Aboriginal legend of the people from western Victoria there is mention of a giant emu called Mihirung. The legend has now been found. It is Genyornis newtoni, the last member of the family Dromornithidae or thunderbirds to survive in Australia, apparently until the arrival of the Aboriginal People. This bird looked very much like a heavy ostrich.

Dromornithids are known only from Australia, mostly from the eastern half of the continent, though there are some known from Western Australia and Tasmania. They were very common at some Northern Territory sites, at some places comprising 60-70 % of the known fossil material. A fragment of a dromornithid-sized foot bone from Antarctica has been found, but its connection to dromornithids is uncertain.

The earliest evidence of Dromornithids are footprints in southeast Queensland that could be from these birds, but so far it's not certain. They have been dated to the Early Eocene, about 50 million years ago. The oldest identified bones of this family are found at Riversleigh, Queensland, from the Late Oligocene, about 25 million years ago. The complete skeleton of Genyornis newtoni was found at Lake Callabonna. The most recent species to be found was Genyornis newtoni, from Cuddie Springs north-central New South Wales that have been dated to 31,000 years ago. At this age they almost certainly inspired the Aboriginal legend of the Mihirung, as they would have been present when the first Aboriginal People arrived. From the fossil evidence so far it seems they were probably at the peak of their diversity in the Late Miocene when 3 species lived at the same time. 

They ranged from about the size of a cassowary up to possibly the largest bird that ever lived. Their appearance was similar to an emu, with powerful legs, reduced wings, and no sternum. The skulls, however, were very different from those of emus. Another difference was that the end bones of the toes looked more like small hooves than the claws expected in flightless birds.

The skulls are often missing from the fossils so there has been some confusion over the type of skull and beak, which in turn made it difficult to speculate on what they ate. The skull of some species, such as Bullockornis and Dromornis, have now been studied sufficiently to show that they were much larger and heavier than an emu skull. The narrow, deep bill made up 2/3 of the skull, with the front part being specialised for cutting and the back part for crushing. The muscle attachments showed that the muscles that operated the beak were large. They had a powerful bite. 

Other evidence of Dromornithids found include trackways, cranial endocasts, almost complete eggs and eggshell fragments, and gizzard stones.

The belief that Dromornithids were herbivores is based on the following points:

  • the lack of a hook at the end of the bill
  • the lack of talons on the toes
  • the association of gizzard stones with some specimens
  • the large number of individuals occurring together, suggesting flocking behaviour
  • the placement of the eyes on the sides of the skull.

Amino acid analysis of Genyornis eggshells also indicates that this species was herbivorous. Some have disputed this interpretation of the analysis, claiming the same results would be obtained if their prey were herbivores.

The size and heavy construction of the skulls of Bullockornis and Dromornis argue for something other than the normal herbivorous diet. It is possible that these 2 species were either carnivores or scavengers. The multi-part make-up and heavy structure of their beaks, as well as the heavy musculature, make it at least possible that they weren't herbivores. A problem with classifying these 2 species as herbivores is that no source of vegetable food requiring such large and powerful beaks is known from their environment. 

A feature of the skull that suggests possible herbivory is the placement of the eyes on the sides of the head so that they had a wide field of view, but possibly had a blind spot straight ahead, which wouldn't be expected in a hunter. It has been suggested that they could have been hunter/scavengers like hyenas. Scavenging would make use of the powerful beak, while not making it essential to hunt, though they could have hunted opportunistically. It is possible they may have been omnivorous.

It is known that while the vast majority of kangaroos are exclusively herbivores, at least a few became carnivores, so why not a few of the Dromornithids. Tree kangaroos are arboreal but they are still clumsy climbers, not yet having adapted to the arboreal lifestyle as completely as other marsupials such as possums. Maybe the placement of the eyes of the possible hunting dromornithids are in the same transitional stage.

One of the arguments put forward as evidence against the hypothesis that they were carnivores is that they were too heavy to run down prey, so must have been either herbivores or scavengers. The emu, a known fast runner, has long, slender legs, though still powerful enough to disembowel an attacking dingo. But the legs of Dromornithids were much more heavily built. Biomechanical studies have been carried out on the legs of known Dromornithids and the conclusion drawn, based on the size of the muscle attachments and the estimated size of the muscles, was that they could indeed run much faster than originally assumed, and that at least some of the middle-sized species may have been able to outrun an emu. The emu achieves its speed by making its legs lighter, but the Dromornithids achieved the same goal by making the muscles larger. 


As with other Australian animal groups, dromornithids differ from giant flightless birds from elsewhere, being hugely enlarged geese. For some time it was thought that the dromornithids were related to the ratites, the large flightless birds that include emus, cassowaries, and ostriches. It is now mostly accepted the similarities in appearance between dromornithids and ratites is a case of parallel evolution resulting from their adopting of a similar lifestyle. As more research has been done on the skulls, which were not always present in the earlier discovered fossils, it is now thought that they branched off from the line that eventually led to the Anseriformes, which includes ducks and geese, at an early stage of their evolution.

An early anseriform-like bird from the Cretaceous, about 65 million years ago, Vegavis iaai, has been found in the Cape Lamb deposits on Vega Island, Antarctica. It is not believed to be on the direct line leading to anseriforms, but it demonstrates that birds similar to the line leading to the Dromornithidae existed in Antarctica prior to the separation of Australia from Antarctica.

5 genera and 7 species are currently known, with at least 1 new genus being currently studied.

Barawertornis tedfordi is the smallest known species. It was about a size of the present-day cassowary with a weight of about 80-95 kg. Late Oligocene-Early Miocene. QLD: Riversleigh
Ilbandornis lawsoni (Late Miocene. NT: Alcoota Station) and I. woodburnei (Late Miocene. NT: Alcoota Station) were of similar build and size to ostriches, having more slender legs than other dromornithids.
Bullockornis planei (Middle Miocene. NT: Bullock Creek) and Genyornis newtoni (Pleistocene. SA: Lake Callabonna; Baldina Creek; Mt Gambier; Naracoorte; Salt Creek. NSW: Cuddie Springs) were more heavily built. They were about 2-2.5 m tall and probably weighed up to about 220-240 kg.
Dromornis Stirtoni (Late Miocene. NT: Alcoota Station) is the largest known dromornithid, about 2-2.5 m tall and weighing up to about 500 kg.
Undescribed new genus and species: Late Oligocene-Middle Miocene. QLD: Riversleigh
Dromornis australis, Pliocene. QLD: Peak Downs

Dromornithidae indeterminate
Remains that can only be identified as belonging to this family are known from a number of localities. Some of these of particular interest are listed below.

Early Eocene (Redbank Plains Formation). QLD: Redbank Plains [foot impression] Mid-Tertiary. TAS: Endurance Tin Mine pit, 5 miles north of Pioneer [trackway] Late Oligocene/Early Miocene. SA: Muloorina Station: Snake Dam Locality [eggshell] Pleistocene. WA: Mammoth Cave

Bullockornis ("Bullock Bird")

Bullockornis planei (Rich,1979), the type species, was named after the Australian palaeontologist Michael Plane, who co-reported and described the Bullock Creek fossil locality (Plane & Gatehouse, 1968). The original find, a right femur, indicated the specimen was about the height of a modern ostrich, but much more massively built. The Bullock Creek Local Fauna from the Middle Miocene Camfield Beds, in the Northern Territory, has produced many specimens, among which are crania and upper and lower mandibles. Fossils from Bullock Creek are among the best preserved fossils from Australia. Some have been stained black or bluish-grey by manganese oxide, other are white or cream.

Murray & Vickers-Rich include some detailed data on the known bones of Bullockornis in their book, Magnificent Mihirungs. There is a great deal of variability in the size, and some variability of the shape, of the femur and tibia bones of B. planei. Most femurs and tibias display a degree of lateral compression, though some lack the compression, in which case they more closely resemble individuals of Dromornis stirtoni from the Late Miocene. The authors suggest that these bones that lack the compression that were found in the Bullock Creek deposit could possibly be a species of Dromornis, though they think the variation appears to be more continuous than would be expected for a different species. A characteristic that is considered to consistently distinguish Bullockornis from Dromornis is a femoral internal condyl that is significantly shorter anteroposteriorly, and with a less rounded posterior margin.

Another species of dromornithid has been found in the Bullock Creek deposit that closely resembles Ilbandornis woodburnei from the Alcoota Local Fauna in the Camfield Beds from the Late Miocene. It has been tentatively called ?Bullockornis sp. The remains of this species are not as common in the Bullock Creek Despots as those of Bullockornis planei, the only fossil of this species to be found is a cranial wall fragment. The main feature of ?Bullockornis sp. distinguishing it from  Ilbandornis is slightly smaller size. Other distinguishing features of the ?Bullockornis sp. are a straight-sided femur shaft, a low, broad trochanter, and moderate width condyles and condylar crests that are slightly divergent. (Murray & Vickers-Rich, 2004).

Because of the combination of its large skull and possible carnivorous habits and its probable waterfowl relationships, Bullockornis was nicknamed 'The Demon Duck of Doom'.

It was assumed that the earlier and the smaller Bullockornis and Ilbandornis were also herbivorous. Then in 1997 a 38.5 cm long skull of Bullockornis planei was discovered. Birds have heads that are big enough to match the size of the food they normally eat, so if Bullockornis ate fruit they must have been enormous. Some think they ate tough plant material. It is thought by some that they may have been carnivorous. 


Anatomical evidence indicates some details of shape and posture, but the appearance of dromornithids is a problem when assembling skeletons for display because of the wide range of possible postures that can be characteristic of different bird species. An example is the differences in appearance of  between living species, the ostrich and the emu. The ostrich carries its head erect and emu carries its head below the level of its body much of the time. Because of this, the height quoted for the various species and specimens can vary according to the posture assumed for restoration. The 2 type of body plan adopted by the large ground birds are a narrow pelvis in the ostriches, emus and rheas, that reduces the mass behind the hip joint, and a broad pelvis in dromornithids, moas, elephant birds and diatrymatids. In the species with a relatively broad pelvis, the tarsometatarsus has been shortened, bringing the feet under the centre of mass, that has been repositioned (Andors, 1992). This applies to the bird when it is not moving, when it moves, because of their long legs with the centre of gravity being relatively high and must constantly be maintained, with the help of its long neck in the case of the large ground birds, the trunk rotates fore and aft around the hip joint and the legs allow a large range of postural arrangements due to their Z-configuration.

Proportional differences between the ratites and dromornithids are the most obvious differences. Some authors, such as Stephen Wroe, emphasise the large size of the dromornithid head. Murray & Vickers-Rich suggest the 50 cm Dromornis head doesn't appear unusually large when it is on a large bird that is about 2.7 m tall weighing between 400 and 500 kg. They describe the relatively tiny skull of the Giant Moa as a minuscule pinhead, giving it a very odd appearance.

A Dromornis is much larger than an ostrich, but its toes are shorter than those of the ostrich, which is considered to be a short-toed bird. Dromornithids also differed in having hoof-like unguals, with broad convex nails, on their toes, instead of the claws.

Murray & Vickers-Rich suggest some of the characteristics of dromornithids are anseriform-like, differing from the same characteristics in ratites. These include a relatively thicker neck with an open S-curved posture, a crop bulge that is well developed, more erect tail, and the body is more muscular with contours that are less angular. The neck muscles attach further up on the crown of the head. They also suggest the craniofacial joint, where the upper beak attaches to the forehead, would have been conspicuous in life, if it wasn't covered with feathers. Another anseriform feature found in dromornithids is extremely small eyes. They depict the small, muscular wings in the Z-fold pattern, rather than the hanging form as found in the casuariiform birds. According to Murray and Vickers-Rich, dromornithids could flap their small wings strongly, as indicated by the strong development of the bony crests and tuberosities where the muscles attached.

Murray and Vickers-Rich have suggested that as dromornithids were closely related to magpie geese, dromornithids may have had similar colours and patterns seen in present-day geese and ducks. Some similar features may be brightly coloured bills and reddish skin around the eyes and bill base. They speculate the, as in magpie geese, the bill may have been red. The highly social species, most ducks and geese, often use plumage stages and pattern signals (semantic colouration) for communication. If dromornithids shared this plumage characteristic they may have had contrasting plumage markings. As in many ducks and geese, Murray & Vickers-Rich suggest they may have had sexually dimorphic colours and patterns. As with other ground birds, especially those lacking the ability to fly sufficiently to escape predators, have cryptic colouration. The authors suggest this may not have helped the dromornithids very much because of their large size.


Wroe (1999) has suggested that Dromornithids were carnivores, or possibly scavengers. Murray & Vickers-Rich (2004) give reasons in Magnificent Mihirungs why they believe that if Dromornithids ate meat it was as opportunistic scavengers, as are many living birds, such as carrion crows, that often swallow bones that are crushed in their gizzards. They claim the swallowing of objects such as bones would have been a problem for Dromornithids because of the narrowness of their throats, confined dorsally by a shallow palate, ventrally by hyoid apparatus, large tongue and hypoglossal muscles, and laterally by deep maxillopalatines They also believed the oral cavity of Dromornithids was not well designed for very large objects such as bones with irregular shapes. They concede the large beak of Dromornithids would have been capable of cracking tough objects such as bones, using an example of a small bird, the hawfinch, that cracks olive stones and cherry pits with its beak, to demonstrate the potential was present in the Dromornithids. They believe that even if Dromornithids were opportunistic scavengers, they were not designed as specialised scavengers or carnivores. They believe there is no good reason for concluding that the undoubted power of the beaks of Dromornis or Bullockornis were specialised for bone reduction.

They further claim that if the beaks were not used for bone reduction, they were the wrong shape for scavengers. Avian scavengers don't crush bones, as mammalian scavengers do, they all have narrow heads, often with long beaks that can reach into confined spaces among ribs. The Dromornithid beaks were broad, blunt-tipped and deep, all features that would have made the picking of flesh from bones difficult. Murray & Vickers-Rich concede that they may have eaten small prey and occasionally scavenged, but were not obligate scavengers or carnivores.


Murray & Vickers-Rich (2004) have argued against any of the Dromornithids being carnivores. In their book, Magnificent Mihirungs, they give a detailed account of the functioning arrangement of the jaws of Dromornithids. They list a number of features that give some indication of the feeding habits of Dromornithids. These include the form of the beak, deep and narrow, blunt, with a wide overhanging nail with no rostral hook, and with 3 functional areas, the tip is designed for gripping, the middle portion for crushing, and the back portion for shearing. Other anatomical features they describe that are connected with feeding are structural features of the skull and jaws, and the muscles and nerves associated with these features, and a well developed nerve supply and muscles to the tongue. The mandibular tomial margin (cutting edge of a bird's beak) is well above the centre of jaw rotation, another feature found in birds with beaks having a high level of manipulatory ability with food. They suggest that these features indicate these birds were capable of complex manipulation of food items, as is found in parrots, but not in most other birds.

Other features they mention are the small orbits on the sides of the head, with optic lobes reduced relative to the cerebral hemispheres, and short, broad toes that have hoof-like toenails. They believe the eyeballs protruded enough to allow some forward vision, but the visual fields probably didn't overlap, and their optic lobes were relatively smaller than those of domestic ducks, though the authors believe they didn't have poor eyesight, but it probably lacked the acuity present in the eyes of predatory species.

These features would normally be considered to be more likely in a herbivore, the small eyes on the sides of the head, with reduced associated optic lobes, and the lack of the claws on the toes or a rostral hook on the beak. The say these features do not conclusively rule out the eating of meat, citing the Kea Parrot of New Zealand. These parrots appear no different from other parrots, but have adopted an omnivorous type of feeding that includes some meat, such as small prey and incapacitated larger prey.

Murray & Vickers-Rich refute the claim that the heads of Dromornithids are large for their body size, claiming that their heads were in the same relative proportion to their body size as is found other Anseriformes, but the negative allometric relationship between the neurocranium and the mandibles', the area for jaw muscle attachment in magpie geese is much greater in Dromornis or Bullockornis. The magpie goose has about 4 times the muscle insertion area relative to its body mass, even though in Bullockornis the deepening of the skull increases the muscle insertion area. Murray & Vickers-Rich say their head size to body size proportions should not be compared to those of ratites, such as the emu, as the head to body proportions of ratites differ from those of Anseriformes such as ducks and geese. In spite of the great size difference between magpie geese and Bullockornis, according to Murray and Vickers-Rich, if muscle strength is proportional to the square of its cross-sectional area, the jaw muscle power of the magpie goose is about equal to that of Bullockornis, considering their relative body masses. 

In Magnificent Mihirungs, Murray & Vickers-Rich have compared the beak of Dromornis Stirtoni with the snout of Alkwertatherium webbi, a zygomaturine diprotodontid. There are obvious similarities between the contours and sizes of the mandible tips of dromornithids and the zygomaturine marsupial. Rich (1980) has described the dromornithids as "the diprotodontids of the bird world!", because of their abundance, size and biodiversity. The design of their eating apparatus also makes them similar to diprotodontids. Both the beak of Dromornis stirtoni and snout of Alkwertatherium webbi have a broad, rounded tip, and transition to shearing surfaces in similar positions, the tomial overlap in the beak and the sectorial premolars in the snout. The mandible tips of the dromornithids appear to be close analogues of the incisor/premolar cropping complexes found in Diprotodontids. Murray & Vickers-Rich has suggested that this apparently considerable overlap of the diets, possibly primarily foliovorous, suggesting convergent trophic morphology, may have contributed to the extinction of Dromornis stirtoni, the largest known Mihirung, possibly being out-competed by the Diprotodontids.

Like other Aboriginal stories set in the Dreamtime, the stories of Mihirungs, what may have been Dromornithids, possibly Genyornis, have a level of fear associated with them. In the stories of the giant emu they say it killed and ate a many people. If the stories are based on actual encounters with a giant bird, such as a species of Genyornis, it would suggest that the Mihirungs could possibly be at least part-time carnivores or scavengers, and had the potential for aggression. Some evidence, such as seen at Lake Callabonna, where 3 Genyornis were trapped in the mud, suggest they were flocking birds. The aggressive behaviour of flocks of geese has been used as living security systems. If a single large living goose can be a treat to intruders, what sort of threat could a 2.8 m Genyornis be when defending its territory/chicks/eggs. Modern cassowaries, much smaller than a Genyornis, can be very aggressive when a human is perceived as a threat, and can inflict serious injuries with a kick of their powerful clawed feet. The 'Be cass-o-wary' section of this link gives some idea of how dangerous a smallish giant bird can be. Herbivore or carnivore, if the Aboriginal stories are based on actual encounters, it could indicate they were very dangerous birds to encounter. There is evidence of their eggs being cooked. Taking eggs from a Genyornis nest must have been a perilous undertaking.

In the Riversleigh deposits many dromornithid bones have been found associated with aquatic animals such as lungfish, turtles and crocodiles. This has led to the suggestion that they may have spent some time in water, as do the living cassowaries that are good swimmers.

Birds from Riversleigh Fossil Deposits

Sources & Further reading

  1. Dinosaurs of Australia and New Zealand, John A. Long, UNSW Press, 1998
  2. Peter F. Murray & Patricia Vickers-Rich, Magnificent Mihirungs: The Colossal Flightless Birds of the Australian Dreamtime, Indiana University Press, 2004
  3. Mary E White, After the Greening, The Browning of Australia, Kangaroo Press, 1994


  1. Dromornithidae images
  2. Anseriformes
  3. Definitive fossil evidence for the extant avian radiation in the Cretaceous
  4. Vegavis



Author: M. H. Monroe
Last Updated 14/04/2009


Giant extinct birds
Banded stilts - breeding on playas
Opportunistic feeding-Breeding Behaviour
Journey Back Through Time
Experience Australia
Aboriginal Australia
National Parks
Photo Galleries
Site Map
                                                                                           Author: M.H.Monroe  Email:     Sources & Further reading