Naturwissenschaften (2007) 94:681 685 DOI 10.1007/s00114-007-0241-3 SHORT COMMUNICATION Bizarre tubercles on the vertebrae of Eocene fossil birds indicate an avian disease without modern counterpart Gerald Mayr Received: 6 February 2007 /Revised: 8 March 2007 /Accepted: 9 March 2007 /Published online: 4 April 2007 # Springer-Verlag 2007 Abstract Remains of fossil birds with numerous bony tubercles on the cervical vertebrae are reported from the Middle Eocene of Messel in Germany and the Late Eocene of the Quercy fissure fillings in France. These structures, which are unknown from extant birds and other vertebrates, were previously described for an avian skeleton from Messel but considered a singular feature of this specimen. The new fossils are from a different species of uncertain phylogenetic affinities and show that tuberculated vertebrae have a wider taxonomic, temporal, and geographic distribution. In contrast to previous assumptions, they are no ontogenetic feature and arise from the vertebral surface. It is concluded that they are most likely of pathologic origin and the first record of a Paleogene avian disease. Their regular and symmetrical arrangement over most of the external vertebral surface indicates a systemic disorder caused by factors that do not affect extant birds, such as especially high-dosed phytohormones or extinct pathogens. Keywords Fossil birds. Eocene. Vertebral tubercles. Paleopathology Introduction The fossil bird Idiornis tuberculata Peters, 1995 is a representative of the Idiornithidae, extinct relatives of the G. Mayr (*) Sektion Ornithologie, Forschungsinstitut Senckenberg, Senckenberganlage 25, 60325 Frankfurt am Main, Germany e-mail: Gerald.Mayr@senckenberg.de South American seriemas (Cariamidae), which were reported from the Paleogene of Europe and the Neogene of Africa (Mourer-Chauviré 1983, 2003). The species is known from a single, almost complete albeit jumbled skeleton from the Middle Eocene (about 47 Ma) of Messel in Germany. The holotype of I. tuberculata is most remarkable for the presence of numerous tubercles covering the cervical vertebrae and few parts of other bones. Peters (1995) assumed that they represent internal bone structures that were counterdrawn on the vertebral surface, but could not find an explanation for these formations, which have no counterpart among extant birds and other vertebrates. In this paper, I describe two new avian specimens from Messel and an isolated vertebra from the Late Eocene of the Quercy fissure fillings in France, which exhibit the very same vertebral tubercles. The new specimens indicate that these enigmatic structures are not a singular feature of the holotype of I. tuberculata but had a wider distribution among Eocene birds. Materials and methods The two specimens from Messel are deposited in the Forschungsinstitut Senckenberg, Frankfurt am Main, Germany and have the collection numbers SMF-ME 3548 and SMF- ME 10846a+b (Figs. 1a and 2a). The cervical vertebra from the Late Eocene locality La Bouffie of the Quercy fissure fillings is an uncatalogued specimen in the collection of Université Claude Bernard Lyon 1, France (Fig. 2d). Osteological terminology follows Baumel and Witmer (1993).
682 Naturwissenschaften (2007) 94:681 685 Fig. 1 a Specimen SMF-ME 10846a from the Middle Eocene of Messel, skull in lateral view; b detail of vertebrae with tubercles (framed area in a). hyo Hyoid apparatus, lrm left ramus mandibulae, rrm right ramus mandibulae. The cervical vertebrae are numbered. Coated with ammonium chloride. Scale bar=5 mm Results Specimen SMF-ME 10846a+b consists of a skull and the cranialmost six cervical vertebrae (Fig. 1a); in SMF-ME 3548, the skull is preserved, as well as the atlas, the axis, and fragments of the third cervical vertebra (Fig. 2a). The skull of SMF-ME 3548 measures 66.5 mm; in SMF-ME 10846, it is 56.5 mm long. Both specimens are, thus, considerably smaller than the skull of the holotype of I. tuberculata in which the cranium alone, i.e., the skull without beak, has a length of 55 mm (Peters 1995). They are of adult individuals because no sutures are visible between the cranial bones and the bone surfaces do not have a porous texture as in juvenile birds. The beak is shorter than the cranium, fairly straight, with a slightly hooked tip and sigmoidally curved cristae tomiales (SMF-ME 10846a). Although the exact extent of the narial openings cannot be discerned, these seem to have been fairly long. As in other Messel specimens, the palatal area is difficult to interpret, and only the ossa maxillaria, which are separated by a gap over most of their length, are clearly visible (SMF-ME 3548). In contrast to extant Cariamidae, the processus postorbitalis is very short. The processus zygomaticus, however, is wide and of similar shape to that of Cariama cristata (SMF-ME 3548). Welldeveloped processus basipterygoidei are clearly discernible in SMF-ME 3548 (Fig. 2a). Also in this specimen, the processus paroccipitales, i.e., the caudal walls of the cava tympanica, form a cone-like projection, just caudal to the alae parasphenoidales (Fig. 2a, b). The surface of the cranium and the lamina parasphenoidalis bears marked pores, which are best visible in SMF- ME 3548 (Fig. 2c). Among extant birds, these pores are very marked in the Andean Condor, Vultur gryphus, in which they are, however, still less distinct than in the fossils. The processus oticus of the quadratum (SMF-ME 3548) is mediolaterally wide and clearly divided into two capitula, although the incisura intercapitularis is very shallow. The rami mandibulae are slender in their distal half, and, most unusually, both specimens lack the distal portion of the mandible, which appears to have been broken and must have been very short. There are no fenestrae mandibulae, whereas these are present in extant Cariamidae and were also reported for I. tuberculata by Peters (1995). In both Messel specimens and in the Quercy vertebra, numerous spiny or knob-like tubercles are evenly distributed over most of the external surface of the axis (SMF-ME 3548) and the caudally adjacent four cervical vertebrae (SMF-ME 10846; Figs. 1 and 2). They are of varying size and shape and do not occur on the articular surfaces and within the canalis vertebralis. Some tubercles merge and form short ridges (SMF-ME 10846); others have a pointed, often slightly curved tip (SMF-ME 3548). All arise from the external vertebral surface, and in cases where they are broken, it can be discerned that they are solid structures. Only few tubercles occur on the corpus of the atlas (SMF- ME 3548), but several low ones are found on the basicranium around the foramen magnum, on the ossa exoccipitale et supraoccipitale (SMF-ME 3548).
Naturwissenschaften (2007) 94:681 685 683 Fig. 2 a Specimen SMF-ME 3548 from the Middle Eocene of Messel, skull in ventral view; b detail of cervical vertebrae with tubercles (large framed area in a); c detail of cranium showing marked pores (small framed area in a); d cervical vertebra from the Late Eocene locality La Bouffie of the Quercy fissure fillings in France in lateral and ventral view. arc Arcus atlantis, at atlas, ax axis, crp corpus atlantis, den (ax) dens of axis, lju left os jugale, lpo left processus oticus of quadratum, lrm pars Discussion The two Messel skulls exhibit a similar morphology and are probably from the same species; characteristic shared features are the short beak, the slender rami mandibulae, and the presence of marked pores on the dorsal surface of the cranial bones (Fig. 2c). As far as comparisons are possible, the specimens cannot be assigned to any of the caudalis of left ramus mandibulae, max os maxillare, nas os nasale, pac (ax) processus articularis caudalis of axis, par processus paroccipitalis, pbp processus basipterygoideus, pv (ax) processus ventralis of axis, rpo right processus oticus of quadratum, rps rostrum parasphenoidale, rrm right ramus mandibulae, trd fragment of third vertebra, zyg processus zygomaticus. Specimens in a c were coated with ammonium chloride. Scale bars=5 mm species already described from Messel. They are too small to belong to I. tuberculata, but there are undescribed idiornithid species of corresponding size in Messel (Peters 1991 and author s own observations). Unfortunately, the single known skull of Idiornis is the cranium of I. tuberculata, which allows limited comparisons with the new specimens. Peters (1995) noted that processus basipterygoidei are not present in I. tuberculata,
684 Naturwissenschaften (2007) 94:681 685 but the corresponding area is too poorly preserved for a definitive assessment. These processes are absent in the putative idiornithid Elaphrocnemus (Mayr and Mourer- Chauviré 2006) and extant Cariamidae, but occur in the Phorusrhacidae (e.g., Sinclair and Farr 1932), which are other extinct members of the Cariamae, the clade to which Idiornithidae and Cariamidae belong. The mandible of I. tuberculata exhibits a large fenestra mandibulae (Peters 1995), which is also present in extant Cariamidae but absent in the Messel specimens described in the present study. Although a reliable classification of the Messel specimens is not possible, their morphology does not preclude an assignment to Idiornis, whose cranium is also unusually sculptured (Peters 1995). Certainly, however, they do not belong to I. tuberculata, so that tuberculated vertebrae are now known to occur in more than one Eocene bird species. Idiornithidae are among the more abundant birds in the Quercy locality la Bouffie (Mourer-Chauviré 2006), which makes it likely that the tuberculated vertebra from this site (Fig. 2d) is from an idiornithid species. Again, however, the only known idiornithid vertebrae are those of I. tuberculata, which, owing to their poor preservation, do not allow detailed comparisons with the la Bouffie specimen. Earlier assumptions that the tubercles are internal bone structures (Peters 1995) are disproved by the new fossils, which clearly show that they are situated on the bone surface. Complete fusion of the tarsalia indicates that the holotype of I. tuberculata is the skeleton of an adult individual, as are the specimens described in the present study, and the tubercles are, thus, no unusual ontogenetic feature (contra Peters 1995). The Quercy fissure fillings and Messel represent fossil deposits of completely different origin, i.e., fossiliferous phosphorites of terrestrial origin in the former case and a lacustrine, bituminous oilshale in the latter, and an origin due to an unusual taphonomic or diagenetic process can be ruled out. Even if all specimens with tuberculated vertebra can be shown to belong to the Idiornithidae, which is arguably not a safe assumption, it is not very likely that presence of tubercles is a distinct anatomical character of this avian taxon. In the living animals, the vertebral tubercles were embedded in the cervical musculature, and there is no obvious function they could have fulfilled. Tubercles further occur on few other bones of the holotype of I. tuberculata, which are not functionally correlated with the vertebrae (quadratum, pelvis, and tibiotarsus; Peters 1995). Limb bones of the Idiornithidae are fairly abundant in the Quercy deposits (Mourer-Chauviré 1983) but do not exhibit similar tubercles, which further indicates that these structures are of pathologic origin, documenting an avian disease in the early stages of neornithine diversification (Ericson et al. 2006). In addition to the tubercles, both Messel skulls are unusual for the presence of numerous well-marked pores on the dorsal surface of the cranium (Fig. 2c), which represent the openings of Volkmann s canals of the Haversian system and transmit blood vessels from the periosteum into the bones. Because similarly marked pores are unknown from extant birds, a correlation with the tubercle formation seems possible, and the Haversian system, which occurs in many lamellar and compact bones including the vertebrae, may have been involved in tubercle formation (tubercle formation may possibly be better understood by future histological examination of the specimens for which these are not available). No comparable disease patterns are, however, known from extant birds (e.g., Gylstorff and Grimm 1987) and other vertebrates. Exostoses caused by bacteria (e.g., mycobacteriosis) or retroviruses (e.g., osteopetrosis; Silverman 1976; Brothwell 2002) are irregular, locally concentrated, and clearly distinguished from the structures found in the fossils. The fairly regular shape and distribution of the tubercles, and the fact that their occurrence is symmetrical on both halves of the pelvis of I. tuberculata (Peters 1995: Fig. 6), is more suggestive of systemic disorders, e.g., metabolic or endocrinal dysfunctions. However, although an oversupply of vitamin D 3 or increased estrogen production owing to tumors may lead to hyperostosis in adult birds (Krampitz et al. 1983; Stauber et al. 1990), in extant Neornithes, systemic disorders do not result in the formation of even roughly similar tuberculated structures. As it can be assumed that the physiology of Paleogene representatives of crown group Neornithes did not substantially differ from that of all extant avian taxa, the tubercles must have been caused by factors that do not affect extant birds, such as especially high-dosed phytohormones or extinct pathogens. Acknowledgment I thank C. Mourer-Chauviré for the permission to study the Quercy vertebra and S. Schaal and E. Brahm for the loan of the Messel specimens. I am further indebted to C. Brause, E. F. Kaleta, N. Kummerfeld, M. Lierz, and M. Pees for the discussion on the eventuality of a pathologic origin of the tubercles and for suggesting some disease patterns. S. Tränkner took the photographs for Figs. 1 and 2a c. Comments by C. Mourer-Chauviré, U. Göhlich, and two anonymous referees improved the manuscript. References Baumel JJ, Witmer LM (1993) Osteologia. In: Baumel JJ, King AS, Breazile JE, Evans HE, Vanden Berge JC (eds) Handbook of avian anatomy: nomina anatomica avium. Pub Nuttall Ornithol Club 23:45 132 Brothwell D (2002) Ancient avian osteopetrosis: the current state of knowledge. Acta Zool Cracov 45(Spec Iss):315 318 Ericson PGP, Anderson CL, Britton T, Elzanowski A, Johansson US, Källersjö M, Ohlson JI, Parsons TJ, Zuccon D, Mayr G (2006)
Naturwissenschaften (2007) 94:681 685 685 Diversification of Neoaves: integration of molecular sequence data and fossils. Biol Lett 2:543 547 Gylstorff I, Grimm F (1987) Vogelkrankheiten. Ulmer, Stuttgart Krampitz G, Enbergs H, Petersen BK (1983) Knochen-und Knorpelaufbau, Kalkstoffwechsel. In: Mehner A, Hartfiel W (eds) Handbuch der Geflügelphysiologie. Karger, Basel, pp 163 217 Mayr G, Mourer-Chauviré C (2006) Three-dimensionally preserved cranial remains of Elaphrocnemus (Aves, Cariamae) from the Paleogene Quercy fissure fillings in France. N Jb Geol Paläontol Mh 2006:15 27 Mourer-Chauviré C (1983) Les Gruiformes (Aves) des Phosphorites du Quercy (France). 1. Sous-ordre Cariamae (Cariamidae et Phorusrhacidae). Systématique et biostratigraphie. Palaeovertebrata 13:83 143 Mourer-Chauviré C (2003) Birds (Aves) from the Middle Miocene of Arrisdrift (Namibia). Preliminary study with description of two new genera: Amanuensis (Accipitridae, Sagittariidae) and Namibiavis (Gruiformes, Idiornithidae). Memoir Geol Surv Namibia 19:103 113 Mourer-Chauviré C (2006) The avifauna of the Eocene and Oligocene phosphorites du quercy (France): an updated list. Strata sér 1 13:135 149 Peters DS (1991) Zoogeographical relationships of the Eocene avifauna from Messel (Germany). In: Bell BD, Cossee RO, Flux JEC, Heather BD, Hitchmough RA, Robertson CJR, Williams MJ (eds) Acta XX Congressus Internationalis Ornithologici. New Zealand Ornithological Congress Trust Board, Christchurch, pp 572 577 Peters DS (1995) Idiornis tuberculata n. spec., ein weiterer ungewöhnlicher Vogel aus der Grube Messel (Aves: Gruiformes: Cariamidae: Idiornithinae). Cour Forsch.-Inst Senckenberg 181:107 119 Silverman FN (1976) Virus diseases of bone. Do they exist? The Neuhauser Lecture. Am J Roentgenol 126:677 703 Sinclair WJ, Farr MS (1932) Aves of the Santa Cruz Beds. Rep Princeton Univ Exp Patagonia 1896 1899 7:157 191 Stauber E, Papageorges M, Sande R, Ward LJ (1990) Polyostotic hyperostosis associated with oviductal tumor in a cockatiel. J Am Vet Med Assoc 196:939 940