A NE\Ry PTEROSAUR LAGERSTATTE in N. E. BRAZIL (CRATO FORMATION; APTIAN, LOWER CRETACEOUS) : PRELIMINARY OB SERVATIONS

ORYCTOS,Vol. | :79 - 85,Octobre 1998

A NE\ryPTEROSAUR LAGERSTATTE IN N. E. BRAZIL (CRATO FORMATION; APTIAN, LOWER CRETACEOUS): PRELIMINARY OBSERVATIONS

David M. MARTILL1 and Eberhard FREY'z 'Department of Geology, University of Portsmouth, Portsmouth,PO1 3QL UK. 'zStaatlichesMuseum ftir Naturkunde Karlsruhe, GeowissenschaftlicheAbteilung, ErbprinzenstraBe13, D-7 6133 Karlsruhe, Germany.

Abstract : Although pterosaurs are rare in the Nova Olinda Member of the Crato Formation (Aptian, Lower Cretaceous,Araripe Basin, N.E. Brazil), their remains have been recoveredin increasing numbers over the last five years from the numerous stone pits between Nova Olinda and Santanado Cariri. Severaldistinct forms are known including tapejarids, a possible ornithocheirid and a possible azdharchid. Some specimensexhibit soft tissue preservation and structures previously unknown in related pterosaurs from other sites have been revealed. Preservationalstyles are variable in that some specimensare articulated, but examples also occur as partially articulated remains; associated,but disarticulated elementsand as isolated bones.

Key words: Pterosaur,Lower Cretaceous,Brazil, Lagerstdtte

Un nouveauLagerstâtte à ptérosauresau nord est du Brésil (Formation Crato, Aptien, Crétacéinférieur) : observationspréliminaires

Résumé : Les ptérosauressont rares dans le membre Nova Olinda de la Formation Crato (Aptien, Crétacéinférieur, Bassin de I'Araripe, Brésil). Depuis cinq ans, leurs restesont néanmoinsété retrouvésde plus en plus régulièrement dans les nombreusescarrières situées entre Nova Olinda et Santanado Cariri. Plusieurs formes différentes sont connues.incluant desTapejaridae, un possibleOrnithocheiridae et un possibleAzdharchidae. Certains spécimens pré- sententdes tissusmous préservéset des structuresjusque-làinconnues chez les ptérosaures.(traduit par la rédaction).

Mots clés : Ptérosaure, Crétacé inférieu4 Brésil, Ingerstritte.

INTRODUCTION Olinda and Tâtajuba (fig. 1), but rhis is probably arti- ficial since this region contains hundreds of small The Nova Olinda Member of the Crato Forma- stone quarries where the thinly bedded limestones are tion, Araripe Basin, North East Brazll (fig. 1), is dug manually for ornamental stone. This labour famous for the abundance and exceptional quality of intensive industry and the commercial value of the preservation of its fossil insects and plants (Martill, fossils assistin makitrg large numbers of fossils avai- 1993; Maisey, I99I). Remains of the gonorhynchi- lable almost daily. Other exposuresof the Nova Olin- form fish Dastilbe elongatus Silva Santos are very da Member elsewhere in the Chapada appear not to conrmor, but other fish species are rare (Brito et al. be so fossiliferous, and furthermore they are not so this volume). Although tetrapod remains are rare, accessible. they too ate well preserved despite some flattening. They include bird feathers (Martill &. Filgui era, Specimen numbers prefixed SMNK are housed 1994; Martill & Frey, 1995), crocodilians (Frey et al., in the StaatlischesMuseum ftir Naturkunde, Karlsru- in prep), probable rhynchocephalians (fig. 2), frogs he, ErbprinzenstraBe13, D-76I33 Karlsruhe,Germa- (fig. 3) and turtles (pers. obs.). The main fossil bea- ny; Specimens prefixed DNPM are held in the col- ring part of the sequenceappears to be restricted lar- lection of the Departmento Nacional Produçao Mine- gely to the region between Santana do Cariri, Nova raI, Rio de Janeiro, BrazTI.

79 ORYCTOS.Vol. 1. 1998

Fig. 1. Locality map showing the Chapada do Araripe in north east BrazII. The Crato Formation crops out around the edge of the Chapada,but is overlapped in the north west. Santanaon the map = the small town of Santanado Cariri, which along with Nova Olinda (a) is one of the most important sourcesof Crato Formation and SantanaFormation pterosaurs. The area indicated (b) yields pterosaursfrom the SantanaFormation.

Fig. 2. Photographfrom a slide of a probable rhynchocepha- Fig. 3. Frog from the Nova Olinda Member of the Crato lian from the Nova Olinda Member of the Crato Formation. Formation, Cearâ,north east Brazll. This specimenis housed Cearâ,north east Brazll. This specimenwas shown to one of in the palaeontologicalmuseum in the small town of Santana us (DMM) in 1996.Its presentwhereabouts is unknown. do Cariri. Scalebar =10 mm. Scalebar app. 45 mm.

80 MARTILL & FREY- NEV/PTEROSAUR LAGERSTATTE

STRATIGRAPHY A photograph of a fourth specimen with an appa- rent dentate beak was shown to us by a fossil collec- The Crato Formation is a series of laminated tor in Nova Olinda (Frey & Martill, 1994: fig.19) and limestones, silts and silty clays lying in places on the is now in the D.N.P.M. collection in Rio de Janeiro red clays or grey green silts of the Batatieras Forma- (Kellner, pers. com.). Frey &. Martill (1994) descri- tion (details of the stratigraphy and the nomenclature bed an almost complete pterosaur, lacking only the of underlying beds remain to be established) and on skull and some cervical vertebtae, with a wingspan late Proterozoic basement (Martill, 1993). It is over- calculated at 4.6 metres which was referred to the lain by a thick evaporite sequence;the Ipubi Forma- new taxon ArthurdacQlus conandoylei (SMNK IL32 tion, which in turn is overlain by the SantanaForma- PAL) and tentatively referred to the Ornithocheiridae. tion, well known for its fossiliferous carbonate More recently Campos and Kellner (1997) described concretions. The thickness of the Crato Formation is a new species of tapejarid pterosaur with extensive variable, but in the Nova Olinda region the formation soft tissue preservation.The specimenis discussedin reachesaround 20 metres.In the Nova Olinda region, more detail below. the base of the Crato Formation is marked by a minor A partial pterosaurian wing (fig. 4) comprising a disconformity where laminated limestones rest on a fourth metacarpal and first, second and third pha- sequence of greenish marls with rhizoconcretions langes of the wing digit (SMNK 2342 PAL) is parti- and red clays. Elsewhere, âs at Crato and Barbalha, cularly noteworthy, as the last two digits display the ((T)' there is a gradual transition from clays with thin shaped cross section typical of Azdharchidae limestones to laminated limestones. A sequence of (fig. 5) (Note, I-Inwin and Lu, 1997, place the Tapeja- laminated limestones, up to eight metres thick and ridae and the Azhdarchidae together in the Azdar- known as the Nova Olinda Member, is the most choidea). Unfortunately, the post cranial skeleton of important lithology palaeontologically (Martill & members of the Tapejaridae has not been described, Wilby, 1993). The Nova Olinda Member limestone is but an example of a small Tapejara wellhoferi from a millimetrically laminated, organic-rich carbon ate the SantanaFormation (SMNK II37 PAL) does not c(T)) which outcrops extensively on the north eastern have the shapedcross-section. flanks of the Chapada do Araripe. A similar limestone also occurs on the south east side of the Chapada between Porteiras and Caririmi- rim in Pernambuco. The age of the Nova Olinda Member remains to be established accurately, but palynomorph data suggestsan Upper Aptian age for the Crato Formation as a whole (Pons et al., 1990).

THE PTEROSAURS

Recently the remains of several pterosaurs have been collected from the numerous small stone quarries between the villages of Santana do Cariri, Tata- juba and Nova Olinda. Some of these specimens(ten are known to us) are of articulated remains and some Fig. 4. Partial wing of a pterosaur from the Nova Olinda show extensive areas of soft tissue preservation, Member of the Crato Formation (Aptian, Lower Cretaceous), including the preservation of soft tissue head crests Nova Olinda region, Cearâ, Brazll. Wing elements of the flight digit inlude the metacarpal,first, secondand third pha- never before seen in pterosaurs. langes. There are some small, disarticulatedelements asso- ((T)' Four specimens with cranial remains are known ciated with the wing finger. PhalangesII and III have the to us and of these, three can be referred to the eden- shapedcross-section typical of the Azhdarchidae. tulous Tâpejaridae. Scalebar = 50 mm.

81 ORYCTOS,Vol. 1, 1998

PRESERVATION

In most casestetrapod remains in the Crato For- mation are crushed flat, but aîe otherwise well-preserved. The skeletons are usually articulated, or partially articulated, but are rarely complete. The incompletenessis, in some cases,a result of careless collecting, in that in most of the examples known to us, the bones have continued beyond the boundaries of the slabs (e.g. figs. 6, l). Some disarticulation is evident but distances between bones are usually small suggestingthat some, if not all, of the disarticulation can be accounted for by gravitational col- lapse andlor microbial degradation.Some specimens may have been incomplete when they arrived on the lagoon floor due to decay or scavengingduring post- mortem drifting (Frey 8. Martill, 1994). Individual elements,including isolated skulls missing the lower jaws, also occur, but as these specimenshave often been obtained by museums from commercial fossil dealers through a chain of collectors and 'middle- men', their relationship to any other bones cannot be established. The bones are usually preserved as shiny, dark brown elementson slabs of buff coloured limestone. Unweathered slabs of Crato Formation limestone are bluish/grey, and bones here arepreser- ved as blackish elements, often with some carbona- ceous material (N.8. we have only observedthis lat- Fig. 5. Detail of secondphalanx of the flight finger of SMNK ter style of preservation in the fish Dastilbe). ((T)) 2342 PAL showing the shapedcross-sectional shape run- ning the lengh of the phalanx. Scalebar = 10 mm.

Fig. 7 . An indeterminate partial pterosaur skeleton from the Fig. 6. Holotype of Arthurdactylus conandoylei Frey and Crato Formation. This example is in the palaeontological Martill , 1994, SMNK II32 PAL from the Nova Olinda museum of Santanado Cariri, Cearâ,BrazII. One of the pes Member of the Crato Formation. Cearâ. Brazil. Scale bar = is articulated, but the wings show considerabledisarticula- 200 mm. tion. The body and skull is missitrg.Scale bar = 100 mm.

82 MARTILL & FREY -_ NE,WPTE.ROSAUR LAGERSTÂTTE f

Fig.8. Tentative reconstructionof the new tapejarid "Tapejara" imperator with soft tissue head- crest from the CratoFormation, basedon figures and photographs in Campos & Kellner 1997.

Frey & Martill (1994) described the post cranial anteriorly by an elongate bone (fig. 8). This crest skeletonof Arthurdactvlus conandoyli (SMNK II32 extends posteriorly some distance behind the skull PAL) in which only the neck and skull was missing where it is attachedto a posterior spina of the parie- (fig. 6). In this specimenmost of the post cranialske- tals. Two examples of another tapejarid species also letal elements show original bone to bone relation- display soft tissue head crests,but in these specimens ships. In an undescribed pterosaur specimen in the the crest is more erect and is at least four times the palaeontological museum of Santana do Cariri, the height of the skull and has a vertical (rather than post cranial skeleton of an indeterminate pterosaur arcuate) posterior margin (fig. 9) (Frey & Martill, in shows considerable disarticulation, but all bones in prep). As tn T. imperator these specimens display an close association.Some portions of this skeleton do anterior bony spine. In both examples there is a soft show true bone to bone relationships (fîg . 7), espe- tissue sheet anterior to the beak, and in one specimen cially in the region of the foot. Once again the skull, (SMNK 2343 PAL) there is a keratinous, hooked, as well as most of the vertebral column is missing. beak-like extension to the anterior tip of the skull which resemblesthe beak of accipital birds (fig. 10). The soft tissue portion of the crest contains verti- THE SOFT TISSUE PTEROSAURS cally orientated structural fibres which anchored onto the bony supporting crest below. At least three examples of pterosaurs with soft Such alarge structure on a very lightly construc- 'We tissues are known to us from the Crato Formation. All ted animal must have had an aerodynamic effect. of these are skulls of tapejarids and show evidence of speculatethat such a structure could have easily been soft tissue extensionsto a bony sagittal crest.At least used as a steering device (anteriorly placed rudder) two taxa are known. Campos &. Kellner described a provided that the animal flew at extremely low speed, new species of Tapejara (7. imperator Campos &. but it may also have passively generatedthrust, beha- Kellner, 1997; DNPM specimenrro. MCT I622-R) in ving like a sail (Frey & Martill, in prep.), which which an arcuate sheet of preserved soft tissue is would be a unique mode of aerial or possibly aquatic situated medially above the skull, and supported locomotion.

83 ORYCT'OS.Vol. 1. 1998

DISCUSSION

The Nova Olinda Member of the Crato Forma- tion is already well known as a fossil conservation Lagerstâtte, having yielded one of the most diverse Cretaceous non-amber entomofaunas (Grimaldi, 1990) and a wide variety of other organisms, some with soft tissue preservation and colour patterning (Martill & Frey, 1995). Although pterosaur remains are rate in the Nova Olinda Member, recent efforts by quaffy workers to obtain specimens for profit has demonstrated that pterosaurs at least occur regularly, if not abundantly. Although the specimens are rare, they appear to be more abundant than pterosaurs from the Solnhofen Limestone and like examples from this deposit, may show evidenceof soft tissue preservation.At present, pterosaur diversity appearsnot be as greatas in other pterosaur bearing Lagerstâtte such as the Solnhofen Limestone of B avaria or the SantanaFormation, also Fig. 9. Skull and soft tissuehead crest of a new tapejaridpte- from the Araripe Basin. However, this low diversity rosaur from the Nova Olinda Member, Crato Formation, of is more likely a reflection of the relatively small Nova Olinda region, Cearâ,BraziI. The head crest has been cut off at number of specimens so far reported. Continued col- the edge of the slab, but a conservativeestimate of its shapesuggests it would be four to five times the height of lecting from this spectacular fossil deposit will the skull. Specimennumber SMNK 2344 PAL. almost certainly yield more pterosaur remains, and Lens cap = 50 mm dia. with the preservation of soft tissues, we expect the Crato Formation to gain in importance for pterosaur studies in the future.

Fig. 10. Distal portion of the premaxilla of a new tapejarid from the Nova Olinda Member, Crato For- mation, of Nova Olinda region, Cearâ, BrazlI. This specimen (SMNK 2343 PAL) shows an anteriorly placed skin "flap" and a "horny" beak - like covering which projects both anteriorly and ventrally beyond the premaxilla and would have overlappedthe denta- ry when the beak was closed. Scalebar = 50 mm.

84 MARTILL & FREY- NEW PTEROSAURLAGERSTÂTTE

ACKNOWLEDGEMENTS REFBRENCES BRITO, P. M ; MARTILL, D. M. & WENZ, S. 1998.A semionotid fish from (Aptian, Thanks to Dr Paulo Brito (UERJ) and Dr Beri- the Crato Formation Lower Cretaceous) of Brazll: palaeoecologicalimplications. Oryctos, | : 37-42. mar Filgueira (D.N.P.M. Fortaleza) for their assistan- CAMPOS, D. A. & KELLNER, A. W. A. L991. Short note on the ce in the field. Thanks to Mr Robert Loveridge for first occuffence of Tapejaridae in the Crato Member (Aptian), photography; Dr David Unwin for helpful discus- Santana Formation. Araripe Basin, northeast Brazll. Annais sions and Dr Mike Barker for critical reading of the Acadamiebrasileira, Ciencia,69 (1) : 83-87. manuscript. We especially would like to thank the FREY, E. & MARTILL, D. M. 1994.A new pterosaurfrom the Crato Formation (Lower staff of the Museo do Paleontologia, Santanado Cari- Cretaceous, Aptian) of BrazlL. Neues Jahrbuch Geologie und Pcilaontologie,Abhandlungen, 194 : ri fiir for kindly allowing us to photograph specimens in 319-4r2. their care and M. Julian Hume for allowing us to GRIMALDI, D. A. (Ed.). Igg}.Insects from the SantanaFormation, reproduceFig. 8. Lower Cretaceous,of Brazll. Bulletin of the American Museum of Natural History, 195 : 1-191 . MAISEY, J. G. (ed.) L99I. Santanafossils, An illustrated Atlas. Tropical Fish Hobbyist Publications,Neptune City, N. J. : I-459. MARTILL, D. M. 1993. Fossils of the Santana and Crato Formations,Brazil. PalaeontologicalAssociation, Field Guides to FossilsSeries. London. 5 : I-92. MARTILL, D. M. & FILGUIERA, J. B. M. 1994. Fossil feather from the Crato Formation of Brazll. Palaeontology,3T : 483-481. MARTILL, D. M. & FREY, E. 1995. Colour patterning preservedin Lower Cretaceousbirds and insects: The Crato Formation of N.E. Brazil. Neues Jahrbuch fu, Geologie und Pcilaontologie, Monatshefte; 118-128. MARTILL D. M. & WILBY, P. R. 1993. Stratigraphy. Chapter 3 In MARTILL, D. M. Fossils of the Santana and Crato Formations, Brazil. PalaeontologicalAssociation, Field Guides to Fossils Series,London, 5 : l-I92. PONS, D., BERTHOU, P.-Y. & CAMPOS, D. A. 1990. Quelques observations sur la palynologie de l'Aptien supérieur et de l'Albien du Bassin d'Araripe (N.E. du Brésrl).Atas do Simpôsio sobre a Bacia do Araripe e Bacias Interiores do Nordeste, Crato, 14 a 16 de Junhode 1990 : 241-252. UNWIN, D. M. 8. LU, J. 1997. On Zhejiangopterus and the relationship of Cretaceouspterosaurs. Flistorical Biology, t2 : r99-2r0.

Note reçuele 3l-07-1997 acceptéeaprès révision le l5-03-1998