Bulletin 51

New Mexico Museum of Natural History & Science

A Division of the

DEPARTMENT OF CULTURAL AFFAIRS

Crocodyle tracks and traces

edited by

Jesper Milàn, Spencer G. Lucas,Martin G. Lockley and Justin A. Spielmann

Albuquerque, 2010

109

Milàn, J., Lucas, S.G., Lockley, M.G. and Spielmann, J.A., eds., 2010, Crocodyle tracks and traces. New Mexico Museum of Natural History and Science, Bulletin 51.

TRACES OF A LARGE CROCODYLIAN FROMTHE LOWER CRETACEOUS SOUSA FORMATION, BRAZIL

HEBERT BRUNO NASCIMENTO CAMPOS1, RAFAEL COSTA DA SILVA2 AND JESPER MILÀN3,4

1 Universidade Estadual Vale do Acaraú, Brazil, email: hebert_campos@hotmail.com;2 CPRM – Companhia de Pesquisa de Recursos Minerais – Serviço Geológico do Brasil, Departamento de Geologia,

Divisão de Paleontologia, Rio de Janeiro, RJ, Brazil, email: rcsilva@rj.cprm.gov.br, paleoicno@yahoo.com.br;3 Geomuseum Faxe, Østsjællands Museum, Østervej 2, DK-4640 Faxe, Denmark;

4 Department of Geography and Geology – Geology Section, University of Copenhagen,Øster Voldgade 10, DK-1350 Copenhagen K, Denmark

Abstract—Body imprints and tracks attributed to large crocodylians from the Lower Cretaceous Sousa Formationof Brazil are described and interpreted as having been produced in a subaqueous environment. In addition to thecrocodylian tracks, the assemblage also comprises isolated tracks from medium-sized theropods. The studiedcrocodylian traces are interpreted as subaqueous traces possibly produced by Mesoeucrocodylia crocodyles,during half-swimming and resting next to the margin of a lake. This is the first record of crocodylian traces in Braziland confirms the potential for finds of new ichnosites in the Rio do Peixe basins of northeastern Brazil.

INTRODUCTION

The Lower Cretaceous Rio do Peixe basins of northeastern Brazilhave the most representative Brazilian dinosaur ichnological assemblagewith more than 30 dinosaur tracksites identified in the Antenor Navarro,Piranhas and Sousa formations of the Sousa sub-basin and Uiraúna-Brejodas Freiras sub-basin (Leonardi and Carvalho, 2007). The Sousa Forma-tion is very rich in vertebrate ichnofossils, with a predominance of theropodfootprints (Leonardi, 1980, 1994). However, sauropod and ornitischiandinosaur tracks also occur (Leonardi and Carvalho, 2007), including theichnospecies Sousaichnium pricei Leonardi, 1979 (Iguanodontidae),Moraesichnium barberenae Leonardi 1979 (Ornithopoda) andStaurichnium diogenis Leonardi, 1979 (Ornithopoda). Tracks from sau-ropods, theropods and ornithopods are also known from the AntenorNavarro Formation (e.g., Carvalho, 2000b; Calvo et al., 2002), which isolder than the Sousa Formation, although tracks are not as abundant as inthe Sousa Formation. The ichnofauna of the Piranhas Formation hasbeen attributed to theropod and ornithopod dinosaurs (Leonardi andCarvalho, 2007). Dinosaur footprints were also reported for the Uiraúna-Brejo das Freiras basin (Carvalho, 1996), a northern sub-basin of the Riodo Peixe basin complex.

GEOLOGICAL SETTING

The area studied is located in the Sousa Formation of the Rio doPeixe basin at the Municipality of Sousa, Paraíba State, northeasternBrazil. The Tapera locality occurs between the coordinates UTM - 24Mand 0598220, N 9251674 and 24M and 0598222, N 9251744 (Fig. 1).The outcrop is exposed for about 200 m in a north-south direction and iscomposed of horizontal layers of laminated siltstone and fine sandstone(Fig. 2). The ichnofossils studied (Fig. 3) are found in the laminated beds,interpreted as the top of the Lower Sousa Formation. Below this layeroccurs a deposit of brown to green siltstone and mudstone. Mud cracks,wave ripples and horizontal stratification are present in the laminatedbeds.

The Lower Cretaceous Sousa Formation is interpreted asNeocomian age, Berriasian to lower Barremian (Carvalho, 2000a). TheUpper Sousa Formation is characterized by successions of siltstone andsandstone with horizontal stratification (Rocha and Amaral, 2006). Onthe outcrop, the siltstone layer is approximately 2.5 m thick and the finesandstone about 0.5 m. The lithology of the Sousa Formation suggestsdeposition by a low energy lacustrine environment with occasional flu-vial influence (Srivastava and Carvalho, 2002), and the preservation ofthe trace fossils is interpreted as a result of subsidence of the basin anddeposition of the layered siltstone.

DESCRIPTION OF THE TRACKS

The tracksite studied is named TASO-1/9 (TA = Tapera locality;SO = Sousa Municipality; 1/9 = imprints 1 to 9) and is composed ofseven footprints and two body imprints (Figs. 3-4). The footprint TASO-3 (Fig. 4) is the best preserved and is tetradactyl, ectaxonic and planti-grade to semiplantigrade, with preserved impressions of three phalangealpads in digit II, and four in digit III, which is the largest; the hypices areacute, and the digits exhibit pointed tips; the imprint of the fourth digit issmaller than the third and is separated by a larger interdigital angle; thesole is longer than wider and shows an irregular posterior margin. Thefootprint TASO-8 (Fig. 4) also has a large sole in relation to the digitalportion of the track and comprises three digit imprints with roundedextremities and hypicies, possibly due to poor preservation. FootprintTASO-6 (Fig. 4) is pentadactyl, ectaxonic, semiplantigrade, and the digitimprints progressively increase in length from digit I through IV, withdigit V being smaller and separated by a higher interdigital angle. Thehypices are rounded, and the extremities of the digits are pointed, exceptfor the fifth digit, which shows a rounded tip; the digits show a widerinterdigital angle than the other footprints, with total divarication angle

FIGURE 1. Location of the studied outcrop at Paraíba State, Brazil.

110

for digits I-V close to 180°.This imprint occurs close to another, the footprint TASO-5, which

comprises four digit impressions without a sole; the digit impressionshave pointed tips and internal marks that resemble phalangeal pads. Theimprint TASO-7 (Fig. 4) constitutes a solitary partial digit impressionwith a pointed tip and two phalangeal pads. The imprints TASO-2 andTASO-4 (Fig. 4) comprise irregular and amorphous marks without anymorphological characters attributable to tracks. The imprint TASO-1(Fig. 4) is a large, elliptical, elongated and slightly curved impression,which is deeper in the central portion. The imprint TASO-9 (Fig. 4) is anelongate, club-shaped mark. Deformation marks in the substrate, such aselevated rims and load deformation marks, can be observed around andinside the imprints. The tracks in the assemblage are set in an irregularfashion, and no apparent trackway pattern can be recognized, nor is itpossible to determine stride lengths or pace angulations.

Other similar marks can be seen on outcrop. Most of them showfeatures that resemble imprint TASO-1, with variations in shape andsize. These marks are usually elliptical or elongate, and none occursassociated with footprints or tracks. Beyond them, a few dinosaur foot-prints are also present on the outcrop. They constitute isolated andpoorly preserved tridactyl imprints, with acute hypices and digit tips.

DISCUSSION

It’s not possible to determine the travel direction of tracks TASO-1/9 or even to determine if they represent a sequence of imprints fromone or more animals. Indeed, the confusing pattern of the tracks is char-acteristic of subaqueous produced tracks. McAllister (1989) andMcAllister and Kirby (1998) established characters to recognize sub-aqueous footprints and many of them are found in the studied material,

such as reflecture of digits, elongation of traces, posterior overhang,impression of distal digits and unexpected configurations. The occur-rence of these characters allows us to interpret the tracks at TASO-1/9 tohave been produced under water. The deformation marks behind theimprints can be attributed to kick-off scours caused by the action ofswirls, created by the fast passage of the autopodia near the substrate,which suspend and redeposit the sediment just behind the footprint(McAllister and Kirby, 1998; Swanson and Carlson 2002).

A subaqueous track can be produced in two ways. The first isduring half-swimming in shallow water, with the digits touching thebottom while the body and tail float. In this kind of track, the occurrenceof traction footprints associated with drag marks is typically due tovariation in the size of the trackmakers relative to the depth of the watersheet (e.g., Foster and Lockley, 1997; Silva et al., 2007, 2008). Trackswith a similar origin were described by McAllister and Kirby (1998),Sedor and Silva (2004), Avanzini et al. (2005) and Silva et al. (2008,2010). The second way to produce subaqueous tracks is during swim-ming close to the bottom, with the whole body submerged. The presenceof ripple marks associated with the imprints suggests that the trackswere produced in shallow water during half-swimming and resting, whichis reinforced by the occurrence of the body impressions. The imprintsare cut by mud cracks, which demonstrates that the surface was subaeri-ally exposed and dried after the tracks were produced. Indeed, the tracksstudied can be assigned to the semi-terrestrial tracks described by Silva etal. (2008), which can be defined by the prevalence of traction tracks, andalso associated with swimming traces and drag marks, while mud cracksmay be present.

According to Srivastava and Carvalho (2002), the Sousa Forma-tion was deposited by a low energy lacustrine system under fluvial

FIGURE 2. General view of the Tapera outcrop, Lower Cretaceous Sousa Formation.

111

influence. Thus, the tracks studied attributed to crocodylians probablyoriginated from the margins of a drying-out lake, and the better-pre-served dinosaur tracks would have been produced in the later stages ofdrying out. Carvalho (2000a) has interpreted the essentially microclasticsequence of the Sousa Formation as sediments from a lacustrine, swampyand meandering-braided fluvial paleoenvironment. The lakes along whosemargins dinosaur footprints were produced were small, temporary, hotand shallow, with alkaline waters (pH between 7 and 9), as suggested bythe study of conchostracans.

The tracks studied occur close to dinosaur footprints that havebeen assigned to theropods (Campos, 2004). However, tracks TASO-1/9 are clearly different from theropod dinosaur tracks, which are typicallytridactyl and digitigrade, in characters such as having a larger sole area,shorter digits in relation to the sole, higher number of digits and thepresence of associated body impressions. The elongated, curved shapeof track TASO-1 is consistent with belly imprints from extant crocodiles(Milàn and Hedegaard, this volume). Similarly, the tracks TASO-1/9share no similarities with the ornithopod and sauropod tracks recordedelsewhere in the Sousa or Antenor Navarro formations (Leonardi andSantos, 2006). The subaqueous character of tracks TASO-1/9 is anotherelement that permits us to deduce the trackmaker identity. Theropod andornithopod dinosaurs can produce subaqueous footprints during half-swimming, as evidenced by several records (e.g., Coombs, 1980; Gierlinskiand Potemska, 1987; McAllister, 1989). However, these tracks usuallyconsist of elongated drag marks from the digits. As the trackmaker ap-proaches the margin of the water body, and shifts to normal walkingposture, the tracks begin to display the typical characters of those dino-saur groups, such as being digitigrade and tridactyl. However, it is stillmore plausible to assign the tracks to an aquatic trackmaker than to an

FIGURE 3. General view of the track assemblage TASO-1/9.

unusual occurrence of normally terrestrial animals (Lockley and Rice,1990).

Among the Lower Cretaceous aquatic animals capable of produc-ing traces in continental water bodies are turtles and crocodylians. How-ever, turtle footprints are symmetrical, with short digits of similar di-mensions, and usually only the anterior portion is preserved, and thetracks are set in a broadly paced, alternating pattern (e.g., Avanzini et al.,2005). In addition, footprints of the manus and pes of turtles are virtu-ally indistinguishable, whereas crocodylians have strong heteropody.The body impressions recorded within the studied material are also verydifferent from what would be expected from a turtle. The chaotic patternof the tracks TASO-1/9 is interpreted as evidence that the tracks weremade during swimming and/or resting on the bottom, and that the manusand pes were not the main propellants, but were only used as auxiliarypropellants and in changes of direction.

This pattern of alternating body and tail impressions associatedwith more or less randomly associated manus and pes prints are consis-tent with what is observed in subaqueous traces from extant crocodiles,where the main propellant is the long, laterally compressed tail, and themanus and pes only occasionally touch the bottom (Milàn and Hedegaard,this volume). Thus, Mesoeucrocodylia crocodylians seem to be the mostplausible candidates for the makers of tracks TASO-1/9, when the pat-tern and morphology of the tracks and the paleoenvironmental contextare taken into consideration.

Fossil crocodylian tracks are relatively rare, and there are only afew reports in the literature. An important record is the ichnogenusHatcherichnus sanjuanensis from the Upper Jurassic of Utah, USA(Foster and Lockley, 1997). This ichnogenus is characterized by thepresence of imprints of the manus and pes associated with drag marks

112

FIGURE 4. Interpretive drawings and details of the track assemblage TASO-1/9, attributed to a crocodylian trackmaker.

from digits and possibly the body and tail, very similar to what is ob-served in terrestrial tracks from extant crocodylians (Farlow and Elsey,this volume; Milàn and Hedegaard, this volume). Subaqueous tracksattributed to theropods by McAllister (1989) were restudied and as-signed to crocodylians by Lockley (1991). This material consists mainlyof drag marks of digits and cannot be directly related to the track TASO-1/9.

Foster and Lockley (1997) compared Hatcherichnus sanjuanensisdirectly with modern crocodylians assuming, based on Steel (1973), thatthere’s no significant differences between the hand and foot amongMesoeucrocodylia taxa. The same assumption will be used here. The

crocodylian manus is pentadactyl, with the inner three digits completeand well formed, increasing in length from I to III, with digits IV and Vbeing shortened and more slender (Romer, 1956). Although quite large,TASO-6 displays affinities with a crocodylian manus imprint in beingpentadactyl with five widely diverging short digits, with a total divarica-tion angle of around 180° (Richardson et al., 2002; Erikson, 2005; Avanziniet al., 2007, Farlow and Elsey, this volume; Milàn and Hedegaard, thisvolume). The crocodylian pes is tetradactyl, with digit V absent. DigitsI-III increase in length, and digit IV is shortened and more slender. Asnoted by Brinkman (1980), crocodylians can use a digitigrade position atthe start of the sprawling step, but also can locomote under a more

113plantigrade position, producing sole impressions. Thus, tracks TASO-3and TASO-8 can be interpreted as pes impressions of crocodylians.

It is difficult to estimate the size of the trackmakers, especiallybecause there may be allometric growth, with the manus and pes becom-ing shorter in relation to the body during development on the animal(Farlow and Britton, 2000). Based on Foster and Lockley (1997) andFarlow and Britton (2000), we tentatively estimate the length of thetrackmakers as being five or six meters.

There are no skeletal records of crocodylians in the Sousa Forma-tion, so a direct comparison with osteological material cannot be made.This is the first ichnological record of crocodylians from Brazil, so thepresent work constitutes an important contribution to the knowledge ofCretaceous vertebrate faunas of South America.

CONCLUSION

The track assemblage consists of isolated tracks from medium-sized theropods and various traces from large crocodylian trackmakers.The studied crocodylian traces are interpreted as a subaqueous trackproduced possibly by Mesoeucrocodylia crocodylians, during half-swim-ming and resting next to the margin of a lake. This is the first record ofcrocodylian tracks in Brazil.

ACKNOWLEDGMENTS

Thanks to Mr. Tiago José de Castro and Mr. Carlos Magno BezerraCortez from 14th District of the Departamento Nacional da ProduçãoMineral (DNPM), Natal, Rio Grande do Norte, Brazil for providingsupport of the fieldwork. The research of Jesper Milàn was supportedby the Danish Natural Science Research Council, Denmark. We thankFelipe Vasconcellos and Peter Falkingham for their critical reviews andconstructive comments on our manuscript.

REFERENCES

Avanzini M., García-Ramos J. C., Lires J., Menegon M., Piñuela L., andFernández L. A., 2005, Turtle tracks from the Late Jurassic of Asturias,Spain: Acta Palaeontologica Polonica, v. 50, p. 743–755.

Avanzini, M., García-Ramos, J. C., Lires, J., Piñuela, L. and Lockley, M. G.,2007, Crocodylomorph tracks from the Late Jurassic of Asturias (Spain):Ichnos, v. 14, p. 143–153.

Brinkman, D., 1980, The hindlimb step cycle of Caiman sclerops and themechanics of the crocodile tarsus and metatarsus: Canadian Journal ofZoology, v. 58, p. 2187–2200.

Calvo, J. O., Campos, H. B. N., Kellner, A. W. A., Saraiva, A. Á. F., 2002,Floresta dos Borbas, a new locality with dinosaur tracks in the SousaBasin, Paraíba, Brazil: Arquivos do Museu Nacional, v. 60, p. 277–284.

Campos, H. B. N., 2004, A new archosaur tracksite from the Early Creta-ceous of Sousa Basin, Northeasten Brazil: Paleo Nordeste, Resumos, p.3.

Carvalho, I. S., 1996, As pegadas de dinossauros da Bacia de Uiraúna-Brejodas Freiras (Cretáceo Inferior, Estado da Paraíba): Boletim do 4º Simpósiosobre o Cretáceo do Brasil, p. 115–121.

Carvalho, I. S., 2000a, Geological environments of dinosaur footprints inthe intracratonic basins of northeast Brazil during the Early Cretaceousopening of the South Atlantic: Cretaceous Research, v. 21, p. 255–267.

Carvalho, I. S., 2000b, Huellas de saurópodos de la Formación AntenorNavarro (Cretácico Temprano de la Cuenca de Sousa), Serrote do Letreiro,Paraíba, Brasil: Ameghiniana, v. 37, p. 353–362.

Coombs, W. P., 1980, Swimming ability of carnivorous dinosaurs: Science,v. 207, p. 1198–1200.

Erikson, B. R., 2005, Crocodile and arthropod tracks from the Late Pale-ocene Wannagan Creek fauna of North Dakota, USA: Ichnos, v. 12, p.303–308.

Farlow, J. O. and Elsey, R., 2010, Footprints and trackways of the Americanalligator, Rockefeller wildlife refuge, Louisiana: New Mexico Museum ofNatural History and Science Bulletin, this volume.

Farlow, J. O., and Britton, A., 2000, Size and body proportions in Alligatormississippiensis: Implications for archosaurian ichnology: Paleonto-logical Society of Korea, Special Publication, v. 4, p. 189–206.

Foster J. R., and Lockley M.G., 1997, Probable crocodilian tracks and tracesfrom the Morrison Formation (Upper Jurassic) of eastern Utah: Ichnos,v. 5, p. 121–129.

Gierlinski, G., and Potemska, A., 1987, Lower Jurassic dinosaur footprintsfrom Gliniany Las, northern slope of the Holy Cross Mountains, Po-land: Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen, v.175, p. 107–120.

Leonardi, G., 1979, Nota preliminar sobre seis pistas de dinossaurosOrnithischia da Bacia do Rio do Peixe, em Sousa, Paraíba, Brasil: Anaisda Academia Brasileira de Ciências, v. 51, p. 501–516.

Leonardi, G., 1980, Dez novas pistas de dinossauros (Theropoda Marsh,1881) na Bacia do Rio do Peixe, Paraíba, Brasil: Actas II CongresoArgentino de Paleontología y Bioestratigrafía y CongresoLatinoamericano de Paleontología, Buenos Aires, 1978, T.I., p. 243–248.

Leonardi, G., 1994, Annotated Atlas of South America Tetrapod Footprints(Devonian to Holocene) with an appendix on Mexico and CentralAmerica. Brasília: CPRM. 248 p.

Leonardi, G., and dos Santos, M. F. C. F., 2006, New dinosaur tracksitesfrom the Lower Cretaceous basin (Paraíba, Brasil): Studi Tridentino diScienze Naturali, Acta Geologica, v. 81, p. 5–21.

Leonardi, G., and Carvalho, I. S., 2007, Dinosaur Ichnocoenosis from Sousaand Uiraúna-brejo das Freiras Basins, Northeast Brazil; in Carvalho etal., eds., Paleontologia: Cenários de Vida. Editora Interciência, p. 363–377.

Lockley, M. G., 1991, Tracking Dinosaurs: Cambridge University Press,Cambridge, 238 p.

Lockley, M. G., and Rice, A., 1990, Did “Brontosaurus” ever swim out tosea?: evidence from brontosaur and other dinosaur footprints: Ichnos, v.1, p. 81–90.

McAllister, J. A., 1989, Dakota Formation tracks from Kansas: implica-tions for the recognition of tetrapod subaqueous traces; in Gillette, D. D.and Lockley, M. G., eds., Dinosaur tracks and traces. Cambridge Univer-sity Press, New York, p. 343–348.

McAllister, J. A., and Kirby, J., 1998, An occurrence of reptile subaqueoustraces in the Moenkopi Formation (Triassic) of Capitol Reef NationalPark, south central Utah, USA: National Park Service, Technical Re-port 98/1, p. 45–49.

Milàn, J. and Hedegaard, R., 2010, Interspecific variations in tracks andtrackways of extant crocodylians: New Mexico Museum of NaturalHistory and Science Bulletin, this volume.

Richardson, K. C., Webb, G. J. W. and Manolis, S. C., 2002, Crocodiles:Inside out. A guide to the crocodilians and their functional morphology.Australia, Surrey Beatty and Sons, 172 p.

Rocha, D., and Amaral, C., 2006, Geologia da Bacia do Rio do Peixe: ServiçoGeológico do Brasil, p. 1–12.

Romer, A. S., 1956, Osteology of the Reptiles. University of Chicago Press,Chicago, 772 p.

Sedor F. A., and Silva R. C., 2004, Primeiro registro de pegadas deMesosauridae (Amniota, Sauropsida) na Formação Irati (Permiano Su-perior da Bacia do Paraná) do Estado de Goiás, Brasil: Revista Brasileirade Paleontologia, v. 7, p. 269–274.

Silva, R. C., Sedor, F. A., and Fernandes, A. C. S., 2010, Ichnotaxonomy,functional morphology and paleoenvironmental contexto ofMesosauridae tracks from Permian of Brazil: Revista Brasileira de

114Geociências, in press.

Silva R. C., Carvalho I. S., Fernandes A. C. S., and Ferigolo J., 2007,Preservação e contexto paleoambiental das pegadas de tetrápodes daFormação Santa Maria (Triássico Superior) do Sul do Brasil; in CongressoBrasileiro de Paleontologia, 20, Paleontologia: Cenários da Vida. EditoraInterciência, Rio de Janeiro, p. 525–532.

Silva R. C., Ferigolo J., Carvalho I. S., and Fernandes A. C. S., 2008, Lacertoidfootprints from the Upper Triassic (Santa Maria Formation) of South-ern Brazil: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 262,p. 140–156.

Srivastava, N. K. and Carvalho, I. S., 2002, Bacia do Rio do Peixe (Paraíba):XLI Congresso Brasileiro de Geologia, João Pessoa. Guia de Excursão.João Pessoa(PB): Sociedade Brasileira de Geologia/Núcleo Nordeste, v.1, p. 1–25.

Steel, R., 1973, Crocodylia. Handbuch der Paläoherpetologie, part 16, 116p.

Swanson, B. A., and Carlson, K. J., 2002, Walk, wade or swim? Vertebratetraces on an Early Permian lakeshore: Palaios, v. 17, p. 123–133.