sign in sign up
<<
Home , Brontornis, Gastornis

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/285787334

Terror cranes or peaceful plant-eaters: Changing interpretations of the palaeobiology of gastornithid

Article in Revue de Paleobiologie · December 2013

CITATIONS READS 6 80

2 authors:

Eric Buffetaut Delphine Angst French National Centre for Scientific Research University of Cape Town

439 PUBLICATIONS 7,190 CITATIONS 24 PUBLICATIONS 97 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Mass extinctions - drivers for evolution? View project

Mesozoic shark faunas of Southeast Asia View project

All content following this page was uploaded by Delphine Angst on 07 December 2016.

The user has requested enhancement of the downloaded file. 1661-5468

VOL. 32, N° 2, 2013 Revue de Paléobiologie, Genève (décembre 2013) 32 (2): 413-422 ISSN 0253-6730

“Terror cranes” or peaceful plant-eaters: changing interpretations of the palaeobiology of gastornithid birds

Eric Buffetaut1 & Delphine Angst2

Abstract Since the original description of a gastornithid in 1855, widely divergent hypotheses about the palaeobiology of these Early Tertiary giant birds have been put forward. Early interpretations, based on scanty material, often suggested aquatic habits. A faulty skeletal reconstruction by Lemoine in 1881 hindered the recognition of the close similarities between the European and the North American Diatryma. After the description of a nearly complete skeleton from in 1917, most reconstructions were based on it. The hypothesis of carnivorous gastornithids, on the model of the South American phorusrhacids, gained wide acceptance in the mid-20th century, and these giant birds were often considered as having competed with mammals. The alternative view of plant- eating gastornithids has gained ground since the 1990s, but their image as predators remains popular, partly on the basis of erroneous conceptions about their osteology. Contrary to what many reconstructions show, gastornithids had a short tarsometatarsus, a beak that was not hooked, and ungual phalanges that were more hoof-like than claw-like.

Keywords , Gastornis, Diatryma, palaeobiology, reconstuctions.

Résumé «Grues terrifiantes» ou paisibles végétariens: interprétations contradictoires de la palébiologie des oiseaux gastornithidés.- Depuis la première description d’un gastornithidé en 1855, des hypothèses très divergentes ont été proposées au sujet de la paléobiologie de ces oiseaux géants du Paléogène. Les premières interprétations, fondées sur un matériel fragmentaire, suggéraient souvent des mœurs aquatiques. La reconstitution erronée publiée par Lemoine en 1881 empêcha de reconnaître les ressemblances étroites entre le Gastornis européen et le Diatryma nord-américain. Après la description d’un squelette presque complet venant du Wyoming en 1917, la plupart des reconstitutions le prirent pour base. L’hypothèse de gastornithidés carnivores, sur le modèle des phorusrhacidés sud-américains, se répandit au milieu du XXe siècle, et ces oiseaux géants furent souvent considérés comme ayant été des concurrents des mammifères. La conception adverse suivant laquelle les gastornithidés se nourrissaient de végétaux a gagné du terrain depuis les années 1990, mais leur image de prédateurs demeure populaire, en partie à cause de conceptions erronées au sujet de leur ostéologie. Contrairement à ce que montrent nombre de reconstitutions, les gastornithidés avaient un tarsométatarse court, un bec qui n’était pas crochu et des phalanges unguéales ressemblant plus à des sabots qu’à des griffes.

Mots-clés Gastornithidae, Gastornis, Diatryma, paléobiologie, reconstitutions.

I. INTRODUCTION shows that interpretations have much varied over time, sometimes on the basis of erroneous reconstructions and Reconstructions of the giant Gastornis (frequently doubtful comparisons. The purpose of the present paper called Diatryma, a junior synonym) frequently feature is to critically examine these changing interpretations, in books and films about “prehistoric ”, where from a historical point of view, as a basis for further it is usually depicted as a ferocious carnivore preying investigations into the palaeobiology of Gastornis. on the small mammals of the Early Tertiary. The fact As first suggested by Coues (1884), and much more that an alternative reconstruction of the palaeobiology recently by Buffetaut (1997a, 2000, 2008), and as accepted of Gastornis, namely as a plant-eater, has also been by various authors (e.g. Mlíkovský, 2002; Mayr, 2009), put forward is often omitted, in favour of the far more Diatryma is herein considered as a junior synonym of spectacular interpretation as a “terror ” (Kurtén, Gastornis; both names are employed according to the 1971). A review of the interpretations of the habits and way they have been used by previous authors. diet of Gastornis since the first discoveries in the 1850s

1 CNRS, UMR 8538, Laboratoire de Géologie de l’Ecole Normale Supérieure, 24 rue Lhomond, F-75231 Cedex 05, . E-mail: [email protected] 2 Laboratoire de Géologie de Lyon: Terre, Planètes, Environnement UMR CNRS/Univ. Lyon 1/ENS-Lyon 5276, Dépar- tement des Sciences de la Terre, Faculté des Sciences, Université Lyon 1, Campus de la Doua, Boulevard du 11 Novembre 1918, F-69622 Villeurbanne Cedex, France. E-mail: [email protected] 414 E. Buffetaut & D. Angst

II. THE EARLIEST INTERPRETATIONS reconstruction impeded the recognition of the very close similarities between European and North American Gastornis remains were first discovered by the physicist gastornithids (Buffetaut, 1997b). Gaston Planté at Meudon, a western suburb of Paris, in 1855 (Buffetaut, 1997b). Although they were far from complete, consisting of an isolated tibiotarsus, to which a femur was soon added, they elicited much attention among the palaeontological community, because they indicated a very large bird of great geological antiquity (coming as they did from the basal , they were among the oldest bird remains known at that time). Speculations about the possible habits of Gastornis parisiensis were put forward in several of the papers that appeared in the 1850s about that giant bird. Whereas Valenciennes (1855) compared it with the albatross, Hébert (1855) considered that it was too heavy to have been able to fly. Prévost (1855) thought that it was as heavy as a horse, and able both to swim like a swan and to stand on one leg like a stork. The idea of a semi- aquatic Gastornis did not appeal to Owen (1856), who gave the first detailed and illustrated description of the tibiotarsus from Meudon; according to him it was more terrestrial than aquatic. Not surprisingly since they were based on the scanty original material from Meudon, the first interpretations of the habits of Gastornis were thus either vague or fanciful.

III. Victor Lemoine’s reconstruction

The first to propose a full skeletal reconstruction of Gastornis was Lemoine (1881). This was based on a number of bones from the Palaeocene of Cernay, near Reims in eastern France. Lemoine’s reconstruction showed a tall but rather slenderly built bird with a long neck, short wings and long legs (Fig. 1). This reconstruction was accepted by most authors for more than a century, until Martin (1992) showed that Lemoine Fig. 1: Lemoine’s reconstruction of Gastornis edwardsii had unwittingly incorporated in it skeletal elements (Lemoine, 1881, pl. XI). As shown by Martin (1992), belonging to various non-avian forms, including fishes a large part of the skeletal elements used for this and reptiles. Lemoine thought that the supposed wing reconstruction (notably in the skull and forelimb) were not avian. The tarsometatarsus is much too long. bones (which were shown by Martin to be entirely non- avian) indicated a paddle-like limb suggesting aquatic habits. On the other hand, the tarsometatarsus was reconstructed by Lemoine (1878, 1881) as very long and IV. Diatryma, an American gastornithid slender, but, as noted by Martin (1992, p. 100): “The tarsometatarsus illustrated by Lemoine was enormously Cope (1876, 1877) described as Diatryma gigantea a elongated by plaster restoration (Lemoine, 1878: pl. 3, fragmentary tarsometatarsus of a very large bird from figs. 1-3), and it includes fragments of a gastornithid the Early Eocene of New Mexico, and noted that its femur and tibiotarsus”. Complete tarsometatarsi (Fig. 2) closest ally would probably turn out to be the European from (Fischer, 1978) and France (Martin, Gastornis. For several decades, however, the North 1992) have shown that European gastornithids, like American material contributed little to an understanding their North American counterparts (see below) had short of gastornithid palaeobiology, because, although more and broad tarsometatarsi. The erroneously elongated fossils were discovered (Shufeldt, 1913), the available tarsometatarsus in Lemoine’s reconstruction may have specimens were too fragmentary to provide much influenced later reconstructions of gastornithids. Be information about the osteology of the American giant that as it may, for a long time Lemoine’s erroneous bird. Shufeldt (1913) thought it was a huge close “Terror cranes” or peaceful plant-eaters 415

to rheas, but admitted that many uncertainties remained about its anatomy. In addition, although Coues (1884) suggested that Diatryma and Gastornis were synonymous, possible affinities with the European giant bird were much disputed, and often rejected (Shufeldt, 1913), all the more easily that Lemoine’s faulty reconstruction was not questioned and obviously did not provide a reliable basis for comparisons (Buffetaut, 1997b).

V. The South American phorusrhacids - A MODEL FOR GASTORNITHID PALAEOBIOLOGY?

Beginning in the late 1880s, palaeontological expeditions to Patagonia revealed that the South American continent had been inhabited during the Tertiary by giant flightless birds, the , or Stereornithes, as they were often called (Moreno & Mercerat, 1891; Ameghino, 1895; Andrews, 1899). These birds attracted much attention among the international palaeontological community, notably because of their huge skulls (often described at the time as as big as that of a horse). Their large, sharply hooked beaks and talon-like ungual phalanges clearly indicated a carnivorous diet (Ameghino, 1895). On the basis of size and of superficial similarities, possible affinities with Gastornis were sometimes mentioned. Lucas (1903, p. 556) suggested that Diatryma might be allied to the “Stereornithes”, basing his supposition on a comparison between the type of Diatryma gigantea and “an indifferent cast of ”. Interestingly, the sturdy tarsometatarsus of Gastornis/Diatryma does not resemble that of Brontornis, which is much more heavily built than other phorusrhacids and may not be a phorusrhacid at all (Agnolin, 2007). The idea of close relationships between gastornithids and phorusrhacids was eventually rejected by most authors (interestingly, it was recently revived by Bourdon & Cracraft, 2011). Nevertheless, the discovery of the Phorusrhacidae showed that very large carnivorous flightless birds, without any real equivalent in the modern world, had once existed, and this probably influenced later reconstructions of gastornithids, all the more so that the large beak of the latter appeared to be superficially similar to that of phorusrhacids. In many textbooks and popular publications, gastornithids and phorusrhacids are discussed together, thus suggesting analogies in biology and behaviour. However, they in fact differ in many points of their osteology, the phorusrhacids exhibiting various cursorial and predatory adaptations not seen in gastornithids (Fig. 3). Some commonly encountered inaccuracies in reconstructions of gastornithids (see below) may have been “borrowed” from features that are really present in phorusrhacids. Fig. 2: A Gastornis tarsometatarsus from the Palaeocene of Berru (eastern France) showing the real proportions of this relatively short and broad bone. Scale bar: 50 mm. 416 E. Buffetaut & D. Angst

Diatryma (although Matthew and Granger suspected that Lemoine’s reconstruction possibly was at least partly incorrect). The reconstruction by Matthew and Granger showed a rather stockily built bird, with a huge beak, quite different from the more slender bird, with a smaller skull, envisioned by Lemoine. The tarsometatarsi of the skeleton from Wyoming were incomplete, only the proximal and distal ends being preserved, and interestingly Matthew and Granger reconstructed them with an exaggeratedly elongate shaft, just as Lemoine had done for Gastornis in 1881. No reason was given by Matthew and Granger for that reconstruction, and it can only be surmised that they were influenced by the elongate tarsometatarsals of various cursorial large birds. Although Matthew and Granger (1917) rejected any close affinities between phorusrhacids and Diatryma (which they thought had to be placed in an of its own), they admitted that they shared parallel adaptations and that in particular the skulls were superficially alike (although they rightly noted that the beak of Diatryma lacks a decurved tip). Although they did not dwell much on the palaeobiology of Diatryma, they remarked (Matthew & Granger, 1917, p. 319) that it resembled Phorushacos “in size and proportions and was probably of similar habits”. What those habits were was not specified, but phorusrhacids were (and still are) generally considered as carnivorous. Matthew & Granger (1917) also published a life reconstruction (by Erwin Christman) of Diatryma steini in an Eocene landscape (Fig. 5B), using the as a guide for the external characters (e.g. the plumage). Fig. 3: Skeletal reconstruction of a phorusrhacid (“Pho­ This drawing served as a basis for many subsequent rorhacos inflatus”, now often called Patagornis reconstructions. The giant bird was the only marshi) from the of Patagonia, after Andrews depicted and no clues were offered about the possible (1899). Morphological differences with gastornithids diet of Diatryma. In accordance with the skeletal are indicated by numbers: 1: hooked preamaxilla; 2: long and slender tarsometatarsus; 3: claw-like ungual reconstruction, the tarsometatarsus was unduly elongate. phalanges. Compare with Fig. 4.

VII. Gastornithids as predators and as VI. The Diatryma skeleton from competitors of mammals Wyoming Matthew and Granger’s suggestion that Diatryma shared In 1916, a field party from the American Museum of similar habits with phorusrhacids does suggest that they Natural History discovered a nearly complete gastornithid considered this giant bird as a carnivore, although they skeleton in the Lower Eocene rocks of the did not mention a specific diet. Heilmann (1926, p. 173) of Wyoming. The specimen was described a later as was among the first to explicitly interpret Diatryma as a Diatryma steini by Matthew & Granger (1917). This find predator, writing that “in all probability it lived on small enormously increased our knowledge of the osteology of reptiles and mammals, which it could easily overtake gastornithids and made possible the first well supported and kill at a blow with its heavy beak”. The concept of skeletal reconstruction of one of these birds. However, carnivorous gastornithids was subsequently accepted by the reconstruction published by Matthew and Granger a number of authors (e.g. Simpson, 1950; Storer, 1960; (Fig. 4) not unexpectedly was thoroughly different from Kurtén, 1971; Gould, 1986). the fanciful reconstruction of Gastornis published by The reconstruction of gastornithids as large predators Lemoine (1881). As noted by Buffetaut (1997b), this then grew into the speculation that they had competed considerably delayed the recognition of the very close with mammals in the Early Tertiary. This was clearly similarities (amounting to identity at the generic level) expressed by Romer, beginning in the 1930s, in several between the European Gastornis and the North American of his influential books on vertebrate palaeontology “Terror cranes” or peaceful plant-eaters 417

Fig. 4: A: Reconstruction of the skeleton of “Diatryma” based on skeleton AMNH 6169 from the Lower Eocene of Wyoming, after Matthew & Granger (1917). The reconstructed shafts of the tarsometatarsi (arrows) are too long. B: Corrected reconstruction with shorter metatarsi.

(Romer, 1954, 1966). The following quotation (Romer, combined with the hypothesis of a carnivorous diet and 1954, p. 107) summarizes this viewpoint: soon became popularized by reconstructions showing “Diatryma, a fossil wingless bird from Wyoming, was a those giant birds chasing small mammals (usually early contemporary of the earliest horses. The horse, at that horses). The whole concept was summarized by Kurtén time, was the size of a fox terrier; this bird was seven (1971, p. 48-49), who described Diatryma and related feet tall. forms as “terror cranes” and noted: “Towering over the These giant early birds arouse one to speculation; their mammals of its day, this giant bird was fast enough to presence suggests some interesting possibilities – which outrun them all and powerful enough to kill any one never materialized. At the end of the Mesozoic, as we have of them. It is almost as if the ancient reptile lines was seen, the great reptiles died off. The surface of the earth making a final bid for the supremacy on land; for birds, was open for conquest. As possible successors, there of course, are descended from reptiles, and are cousins were two groups, the mammals – our own relatives – and to the and crocodiles”. According to Kay & the birds. The former group succeeded, but the presence Colbert (1965, p. 615), Diatryma was a large predatory of such forms as Diatryma shows that the birds were, at bird, which “was able to prosper only so long as the the beginning, their rivals. What would the earth be like meat-eating mammals remained relatively small”; once to-day had the birds won and the mammals vanished? ”. they became larger, they out-competed the giant birds. Although no predation is mentioned, the emphasis Interestingly, it was not until 1991 that an attempt clearly is on the size discrepancy between the giant birds was made to support the hypothesis of carnivorous and the early horses. However, although this seems to gastornithids on the basis of sound biomechanical have escaped Romer and many of those who followed evidence. Witmer & Rose (1991, p. 95) argued that the him, small early horses were not necessarily typical of safety factors in the construction of the skull of Diatryma the whole mammalian environment of gastornithids. For were too high for a herbivore and that “the traditional instance, in many Early Eocene formations in Europe, hypothesis of Diatryma as a carnivorous bird accords as and , Gastornis occurs together with well or better with the data at hand”. They considered that the pantodont Coryphodon, which was nearly the size of the very large biting forces exerted by the jaws of that a hippopotamus. Nevertheless, the idea of gastornithids giant bird enabled it to crush bones, and reconstructed as competitors of Early Tertiary mammals was taken up it as both an active predator and a scavenger (the by various authors (e.g. Colbert, 1945; Piveteau, 1951, scavenging hypothesis had already been put forward by 1955; Swinton, 1965; Beaumont, 1973; Desmond, Lanyon, 1963, and Fischer, 1978). 1975; Gould, 1986; Chatterjee, 1997). It was generally 418 E. Buffetaut & D. Angst

VIII. Gastornithids as plant-eaters been based on inaccurate representations of various elements of their skeleton. Some of these inaccuracies Witmer and Rose’s detailed biomechanical study are discussed below. was partly a response to Andors’s thesis on Diatryma (Andors, 1988; see also Andors, 1992), in which he reconstructed the giant bird as a graviportal herbivore IX. Some frequent inaccuracies in (or more accurately a folivore), on the basis of various gastornithid reconstructions osteological features and comparisons with living birds with a similar beak shape such as the takahe from New As noted above, many reconstructions of the Zealand. Andors’s hypothesis followed a suggestion by skeleton of gastornithids have suffered from various Watson (1976), and was to some extent supported by inaccuracies which in turn have sometimes led to flawed his classification of the Diatrymiformes as the sister- palaeobiological inferences. The worst mistakes were group of the , a group of plant-eating birds made by Lemoine (1881) and not only hampered the – thus somehow reviving Milne-Edwards’s idea that recognition of the very close similarities between the Gastornis was close to ducks (Milne-Edwards, 1867- European and North American gastornithids, but also 1868). Curiously enough, the hypothesis of affinities led to unwarranted palaeobiological conclusions, notably between gastornithids and ducks was shared by some about the forelimb which was considered by Lemoine authors who considered these giant birds as carnivores, to be somewhat paddle-like and suggestive of aquatic including Swinton (1965) and Desmond (1975, p. 156), habits. Later reconstructions, being mainly based on the the latter going so far as to mention “ostrich-sized geese” fairly complete skeleton AMNH 6069, from Wyoming, roaming the countryside in England and France during originally described by Matthew and Granger (1917), the Eocene. were more reliable but nevertheless contained various The discussion about the habits, and more specifically errors (Fig. 6), which are discussed below. the diet, of gastornithids is going on between tenants of carnivory and supporters of herbivory/folivory, and no consensus has yet been reached. However that may Exaggerated length of the tarsometatarsus be, any serious discussion of the palaeobiology of these giant birds must be based on an accurate knowledge of Lemoine’s reconstructions of the tarsometatarsus of their osteology. Previous palaeobiological interpretations Gastornis showed a long, slender bone (Lemoine, 1878, and life reconstructions of gastornithids have too often 1881). Martin (1992) showed that the length of the

Fig. 5: A comparison of two reconstruction of “Diatryma”: A- after Fischer (1978), on the basis of specimens from the Middle Eocene of the Geiseltal; B- after Matthew and Granger (1917), on the basis of skeleton AMNH 6169 from the Early Eocene of Wyoming. “Terror cranes” or peaceful plant-eaters 419

running predators”. Curiously enough, although Andors (1988, 1992) correctly recognized that the tarsometatarsus as reconstructed by Matthew & Granger (1917) was too long, and published correct drawings of the pelvic limb, he reproduced the skeletal reconstruction by Matthew and Granger unchanged. While admitting that the tarsometatarsus of Diatryma was proportionately short, Witmer & Rose (1991, p. 116) considered that this feature was “simply plesiomorphic” and did not imply that this bird was graviportal. Their life reconstruction of Diatryma feeding on small mammals (Witmer & Rose, 1991, fig. 16) showed it with an unrealistically long tarsometatarsus. Graviportal habits were suggested for the gastornithid from the Geiseltal (“Diatryma geiselensis”) by Fischer (1978), who rightly insisted on the relative shortness of the tarsometatarsus and produced a life reconstruction of the bird with relatively short, robust pelvic limbs (Fig. 5). This restoration is certainly closer to the truth than those showing gastornithids with long tarsometatarsi, but not enough attention has been paid to it, and reconstructions Fig. 6: Reconstruction of the “terror crane Diatryma” attacking a small condylarth (Meniscotherium), after Kurtén of long-limbed, cursorial-looking gastornithids are still (1971, fig. 7). Erroneously reconstructed anatomical common (e.g. Paul, 2002). features include the excessively elongate tarsometatarsi Whatever its phylogenetic significance, the shape and (1), the hooked beak (2) and the talon-like ungual relative length of the tarsometatarsus are of obvious phalanges (3). importance for a reconstruction of the habits of Gastornis and certainly deserve more research from a biomechanical point of view. Despite Witmer and Rose’s comments, metatarsus in Lemoine’s reconstructions was artificially the proportions of the pelvic limb of Gastornis place it exaggerated by the incorporation of various elements that among the graviportal rather than the cursorial forms, did not really belong to that bone. Presumably, Lemoine, as defined by Storer (1960), as already noted by Andors who only had incomplete specimens at his disposal, was (1992). From that point of view, contrary to Simpson’s influenced by the long and slender tarsometatarsi of (1950) assertion, gastornithids differ considerably from present-day . On the basis of complete specimens the Phorusrhacidae, which have long, slender legs with a from France (Fig. 2), Martin (1992) was able to show long tarsometatarsus (with the exception of the robustly conclusively that the tarsometatarsus of Gastornis was built Brontornis, but see Agnolin, 2007, about the possible short and broad – a conclusion already reached by Fischer position of that outside the Phorusrhacidae). (1978), based on material from the Geiseltal identified by Fischer (1978) compared Gastornis with short-limbed him as Diatryma (Fig. 5A). such as Pachyornis (a conclusion also reached Interestingly, Matthew & Granger (1917), working on by Andors, 1992) and this comparison may be more the skeleton from Wyoming in which the tarsometatarsi relevant than supposed but weakly supported analogies were incompletely preserved, chose to reconstruct these with phorusrhacids. However, the implications of limb bones as relatively long and slender (Fig. 4). Writing a proportions for the possible diet of Gastornis are not few earlier, Shufeldt (1913) had been more prudent obvious: whereas Andors (1992) used the presumably and noted that it could not be determined whether the graviportal locomotion of gastornithids as evidence for tibiotarsus was long and slender, or short and thick as in a herbivorous diet, Fischer (1978) considered that these some moas. Despite the fact that more complete tibiotarsi heavy, non-cursorial birds were scavengers. were described by Sinclair (1928 – as Omorhamphus, considered as a juvenile Diatryma by Andors, 1992) and by Troxell (1931), showing that that bone was indeed Hook-shaped bill short and broad, the inaccurate skeletal reconstruction by Matthew & Granger (1917) continued to be reproduced Life reconstructions of gastornithids (e.g. Swinton, 1965, in countless publications, with implications about the pl. 7; Kurtén, 1971, fig. 7) frequently show them with locomotion of gastornithids. For instance, Simpson a beak ending in a decurved hook, as in birds of prey. (1950, p. 342) noted that “the phororhacoids in South Whereas phorusrhacids clearly had an upper jaw with a America, Diatryma in North America, and Gastornis in strongly downward curving hook-shaped tip (Ameghino, Europe all had cursorial proportions” and were “probably 1895; Andrews, 1899 – see Fig. 3), highly suggestive 420 E. Buffetaut & D. Angst

of carnivorous habits, the skull of Gastornis does not describes them as “rather short, pointed, moderately show such a bony structure (Fig. 4). The premaxilla has decurved and destitute of prominent flexor tubercles; a triangular tip which is not markedly deflected ventrally the unguals of digits II-IV (especially III) are broad and (Matthew & Granger, 1917; Fischer, 1978), and in this decidedly hoof-like, whereas the ungual of digit I is regard resembles some of the Australian dromornithids, relatively slender. The morphology of the unguals of the notably Bullockornis (Murray & Vickers-Rich, 2004) anterior toes indicates that the foot of Diatryma was better much more than the phorusrhacids (compare Fig. 3 suited for cursorial ground activity than for perching or and 4). The juvenile gastornithid described by Sinclair for grasping prey”. Recently described footprints referred (1928) as Omorhamphus does show a premaxillary tip to gastornithids from the Early Eocene of Washington that is slightly curved, but not really hook-shaped. As State, USA, do not indicate raptor-like claws (Mustoe et noted by Witmer & Rose (1991), in some birds the horny al., 2012). If gastornithids were indeed predators, they rhamphotheca can increase the curvature of the beak, but must have used their beak much more than their toes for in fact it mostly accentuates the shape of the premaxillary the capture of prey, whereas in phorusrhacids both were bone, and it seems unlikely that Gastornis possessed a certainly involved. strongly hook-shaped horny beak. However, as pointed out by Witmer & Rose (1991), some carnivorous birds, such as storks, do not have a hooked beak, and Fischer X. Conclusions (1978) compared the shape of the beak of gastornithids with that of the carrion-eating crow (Corvus crassirostris), From this brief review, it appears that interpretations a scavenger in which the beak is not hooked. Therefore, of gastornithid palaeobiology have frequently been reconstructions showing Gastornis with a phorusrhacid- influenced by inaccurate reconstructions, and that some like strongly hooked beak are probably not correct, but of these interpretations have become entrenched in both this in itself does not imply that it was herbivorous. scientific and popular literature, without being much questioned. In particular, once the idea of gastornithids as large predators feeding on small mammals, and Claw-like ungual phalanges more generally as competitors of mammals during the Early Tertiary, had been proposed during the first half Some life reconstructions (e.g. fig. 7 in Kurtén, 1971) of the 20th century (notably by Romer), it soon became show gastornithids with strongly curved, claw-like toes considered as well established scientific truth and was (Fig. 6). Kurtén’s reconstruction clearly implies that widely disseminated in scientific papers, textbooks and these talons were used by the bird to attack its prey popular publications (and more recently in films). The (in his figure a small condylarth). In fact, unlike the later hypothesis of herbivorous gastornithids, although phorusrhacids, which had hooked ungual phalanges very it was supported by evidence that could not easily be similar to those of raptors, with strong flexor tubercles dismissed (notably in Andors’s works) and has been (Ameghino, 1895), gastornithids had unguals that were favoured by various authors (e.g. Mayr, 2009), is not not much curved (Fig. 7). Andors (1988, p. 253-254) gaining acceptance easily, perhaps because it runs against conventional wisdom and is less spectacular than the image of giant birds tearing apart hapless mammals. At least, some authors now present both hypotheses (Feduccia, 1999; Paul, 2002; Wallace, 2004). This is not to say that the matter has been definitively settled in favour of the herbivory (or folivory) of gastornithids and that tenants of a carnivorous diet are just clinging to an outdated concept. The question remains open, although it is clear that if gastornithids were predators, their adaptations to that mode of life were significantly different from those of other carnivorous giant birds such as phorusrhacids. Other options have been put forward, such as scavenging (Fischer, 1978), although it should be admitted that it is difficult to find instances of exclusive scavengers among living terrestrial tetrapods. Whether gastornithids fed on animals (alive or dead) or on plants, or possibly on both, remains undecided, and it is hoped that renewed investigations, based at least in part on new Fig. 7: Left foot bones of Gastornis, showing ungual phalanges which are not hook-shaped (cast of Gastornis skeleton, techniques (such as isotopic geochemistry) will shed AMNH 6169, from the Early Eocene of Wyoming, in light on that question. Whether gastornithids competed Museo de La Plata). with mammals for “supremacy” on the continents in the “Terror cranes” or peaceful plant-eaters 421

Early Tertiary also is a moot point. Beyond the slightly Mexico. Proceedings of the Academy of Natural Sciences anthropomorphic undertones of the question, it should of Philadelphia, 28(2): 10-12. be remembered that the giant birds may have competed Cope, E.D. (1877) - Report upon the extinct Vertebrata obtained with mammals just as well whether they were herbivores, in New Mexico by parties of the expedition of 1874. Reports upon United States geographical surveys west of , or carnivores. the one hundredth meridian, 4(2): 1-370. Coues, E. (1884) - Key to North American Birds (2nd ed.). Estes & Lauriat, Boston: 863 p. ACKNOWLEDGMENTS Desmond, A. (1975) - The hot-blooded dinosaurs. Blond & Briggs, London: 238 p. Federico Agnolin (Buenos Aires) and Cécile Mourer- Feduccia, A. (1999) - The origin and evolution of birds. 2nd Chauviré (Lyon) provided useful comments on the edition. Yale University Press, New Haven & London: original version of this paper. 466 p. Fischer, K. (1978) - Neue Reste des Riesenlaufvogels Diatryma aus dem Eozän des Geiseltales bei Halle (DDR). Mitteilun- gen aus dem Zoologischen Museum in Berlin, Annalen für References Ornithologie, 2: 133-144. Gould, S.J. (1986) - Play it again, Life. Natural History, 95(2): Agnolin, F. (2007) - Brontornis burmeisteri Moreno & Merce- 20-26. rat, un Anseriformes (Aves) gigante del Mioceno Medio de Hébert, E. (1885) - Note sur le tibia du Gastornis parisiensis. Patagonia, Argentina. Revista del Museo Argentino de Comptes Rendus de l’Académie des Sciences, Paris, 40: Ciencias Naturales, Buenos Aires, 9(1): 15-25. 579-582. Ameghino, F. (1895) - Sur les oiseaux fossiles de Patagonie. Heilmann, G. (1926) - The origin of birds. Witherby, London: Boletín del Instituto Geográfico Argentino, Buenos Aires, 208 p. 15: 501-602. Kay, M. & E.H. Colbert (1965) - Stratigraphy and life history. Andors, A.V. (1988) - Giant groundbirds of North America John Wiley & Sons, New York, London, Sydney: 736 p. (Aves, Diatrymidae). Unpublished Ph.D. thesis, Columbia Kurtén, B. (1971) - The age of mammals. Weidenfeld & Nicol- University, New York: 577 p. son, London: 250 p. Andors, A.V. (1992) - Reappraisal of the Eocene groundbird Lanyon, W.E. (1963) - Biology of Birds. The Natural History Diatryma (Aves: Anserimorphae). In: Campbell, K.E. Press, New York: 175 p. Papers in Avian honoring Pierce Brodkorb. Lemoine, V. (1878) - Recherches sur les oiseaux fossiles des Los Angeles County Natural History Museum Bulletin, 36: terrains tertiaires inférieurs des environs de Reims, 1. F. 109-125. Keller, Reims: 69 p. Andrews, C.W. (1899) - Extinct birds of Patagonia. - 1. The Lemoine, V. (1881) - Recherches sur les oiseaux fossiles des skull and skeleton of Phororhacos inflatus Ameghino. terrains tertiaires inférieurs des environs de Reims, 2. Transactions of the Zoological Society of London, 15(3): Matot-Braine, Reims: 75-170. 55-86. Lucas, F.A. (1903) - Notes on the osteology and relationships of Beaumont, G. de (1973) - Guide des vertébrés fossiles. Dela- the fossil birds Hesperornis, Hargeria, Baptornis and chaux & Niestlé, Neuchâtel: 479 p. Diatryma. Proceedings of the United States National Bourdon, E. & J. Cracraft (2011) - Gastornis is a terror bird: Museum, Washington, 26: 545-556. new insights into the evolution of the Cariamae (Aves, Martin, L.D. (1992) - The status of the late birds Neornithes). Society of Vertebrate Paleontology 71st Gastornis and Remiornis. In: Campbell, K.E. Papers in Annual Meeting Program and Abstracts, Las Vegas: 75. Avian Paleontology honoring Pierce Brodkor. Los Angeles Buffetaut, E. (1997a) - New remains of the giant bird Gastornis County Natural History Museum Bulletin, 36: 97-108. from the Upper Paleocene of the eastern Paris Basin and Matthew, W.D. & W. Granger (1917) - The skeleton of Dia- the relationships between Gastornis and Diatryma. Neues tryma, a gigantic bird from the Lower Eocene of Wyoming. Jahrbuch für Geologie und Paläeontologie, Monatshefte, Bulletin of the American Museum of Natural History, New Stuttgart, 3: 179-190. York, 37(11): 307-326. Buffetaut, E. (1997b) - L’oiseau géant Gastornis: interpréta- Mayr, G. (2009) - fossil birds. Springer, Berlin: tion, reconstitution et vulgarisation de fossiles inhabituels 262 p. dans la France du XIXe siècle. Bulletin de la société géolo- Milne-Edwards, A. (1867-1868) - Recherches anatomiques et gique de France, Paris, 168(6): 805-811. paléontologiques pour servir à l’histoire des oiseaux fos- Buffetaut, E. (2000) - Are Gastornis and Diatryma congene- siles de la France (tome l). Masson, Paris: 175 p. ric? Vertebrata Palasiatica, suppl. vol. 38: 3. Mlíkovský, J. (2002) - birds of the world. Part 1: Buffetaut, E. (2008) - First evidence of the giant bird Gastornis Europe. Ninox Press, Praha: 405 p. from southern Europe: a tibiotarsus from the Lower Eocene Moreno, F.P. & A. Mercerat (1891) - Catálogo de los pájaros of Saint-Papoul (Aude, southern France). Oryctos, 7: fósiles de la República Argentina conservados en el Museo 75-82. de La Plata. Anales del Museo de La Plata, Paleontología Chatterjee, S. (1997) - The rise of birds. The Johns Hopkins Argentina, La Plata, 1: 1-68. University Press, Baltimore & London: 312 p. Murray, P.F. & P. Vickers-Rich (2004) - Magnificent Mihirungs Colbert, E.H. (1945) - The book. American Museum - The Colossal Flightless Birds of the Australian Dream- of Natural History, New York: 156 p. time. Indiana University Press, Bloomington: 416 p. Cope, E.D. (1876) - On a gigantic bird from the Eocene of New Mustoe, G.E., D.S. Tucker & K.L. Templin (2012) - Giant 422 E. Buffetaut & D. Angst

Eocene bird footprints from northwest Washington, USA. Sinclair, W.J. (1928) - Omorhamphus, a new from Palaeontology, London, 55: 1293-1305. the Lower Eocene of Wyoming. Proceedings of the Ameri- Owen, R. (1856) - On the affinities of a large extinct bird (Gas- can Philosophical Society, Philadelphia, 67(1): 51-65. tornis parisiensis Hébert) indicated by a fossil femur and Storer, R.W. (1960) - Adaptive radiation in birds. In: Marshall, tibia discovered in the lowest Eocene formation near Paris. A.J. (ed.). Biology and comparative physiology of birds, Quarterly Journal of the Geological Society, London, 12: vol. 1. Academic Press, New York & London: 15-55. 204-217. Swinton, W.E. (1965) - Fossil birds. British Museum (Natural Paul, G.S. (2002) - Dinosaurs of the air. The Johns Hopkins History), London: 65 p. University Press, Baltimore & London: 460 p. Troxell, E.L. (1931) - Diatryma, a colossal heron. American Piveteau, J. (1951) - Images des mondes disparus. Masson, Journal of Science, New Haven, 22(127): 18-34. Paris: 157 p. Valenciennes, A. (1855) - Remarque à 1’occasion d’une com- Piveteau, J. (1955) - Oiseaux. In: Piveteau, J. (ed.), Traité de munication de M. Lartet sur le tibia d’oiseau fossile trouvé Paléontologie. Tome V. Masson, Paris: 994-1091. à Meudon. Comptes Rendus de l’Académie des Sciences, Prévost, C. (1855) - Nouveaux documents sur le gisement du Paris, 40: 583-584. Gastornis parisiensis et considérations générales sur les Wallace, D.R. (2004) - Beasts of Eden. University of California vestiges laissés par les oiseaux dans les terrains des divers Press, Berkeley: 340 p. âges. Comptes Rendus de l’Académie des Sciences, Paris, Watson, G.E. (1976) - And birds took wing. In: Colbert, E. H. 40: 616-619. & S.L. Fishbein (eds). Our continent, a natural history of Romer, A.S. (1954) - Man and the Vertebrates. Vol. I. Penguin North America, National Geographic Society, Washington Books, Harmondsworth: 198 p. D.C.: 98-107. Romer, A.S. (1966) - Vertebrate paleontology, 3rd edition. The Witmer, L.M. & K.D. Rose (1991) - of the jaw University of Chicago Press, Chicago & London: 468 p. apparatus of the gigantic Eocene bird Diatryma: implica- Shufeldt, R.W. (1913) - Extinct ostrich birds of the United tions for diet and mode of life. Paleobiology, Boulder, States. Aquila, Budapest, 20: 411-422. 17(2): 95-120. Simpson, G.G. (1950) - Are nonflying wings functionless? Science, Washington, 112: 342. Accepté janvier 2013

View publication stats