Experimental Ichnology -experiments with track and undertrack formation using emu tracks in sediments of different consistencies, with comparisons to fossil dinosaur tracks Cand. Scient. Thesis Jesper Milàn Geological Institute University of Copenhagen 13/5-2003 Experimental Ichnology -experiments with track and undertrack formation using emu tracks in sediment of different consistencies, with comparisons to fossil dinosaur tracks. Cand. Scient. Thesis Jesper Milàn Geological Institute, University of Copenhagen Contents Abstract.………………………………………………………………………………………………1 Introduction…………………………………………………………………………………………..2 History and introduction to experimental Ichnology…………………………………………..2 Purpose…………………………………………………………………………………………4 Terminology and definitions of tracks and trackways…………………………………………5 Part 1. Why use emus? The phylogenetic relationship between theropods, emus and modern birds…………………..7 The emu foot………………………………………………………………….………………..8 Morphology of the emu foot…………………………………………………………..8 X-ray photography……………………………………………………………………10 The problem of distinguishing between bird and dinosaur tracks…………………………….11 Introduction…………………………………………………………………………...11 Discussion…………………………………………………………………………….12 Conclusion……………………………………………………………………………13 Part 2. Field experiments with emu tracks Field observations of emu tracks and trackways with comparisons to fossil tracks and trackways….….....…………………………………………………………………………….14 Introduction…………………………………………………………………………..14 I Sediment grain-size analyses…………………………………………………………15 Walking cycle of an emu……………………………………………………………..15 Trackways…………………………………………………………………………….16 Divarication angle of emu footprints…………………………………………………18 Footprints in different substrates……………………………………………………..19 Sediment transport by the foot………………………………………………………..21 Didactyli in emu footprints…………………………………………………………...21 A track including metatarsus impression…………………………………………….22 Striations caused by skin tubercles…………………………………………………...22 Discussion and comparisons with fossil tracks and trackways………………………22 Conclusions…………………………………………………………………………..29 Part 3. Laboratory experiments with track and undertrack morphology Experiments with track and undertrack formation in layered cement………………….……..31 Introduction…………………….……………………………………………………..31 Definitions of undertracks……………………………………………………………32 Experiment 1…………………………….……………………………………………35 Results…………………………….…………………………………………………..37 Experiment 2………………….………………………………………………………43 Results………………………………………………………………………………...43 Experiment 3………………………………………………………………………….46 Results………………………………………………………………………………...46 Discussion…………………………………………………………………………….47 Conclusion……………………………………………………………………………49 Horizontal sections of an emu track and a theropod track exposed in horizontal and vertical view……………………………………………….………………………………………49 Introduction…………………………………………………………………………..49 Methods………………………………………………………………………………50 Results……………………………………………………………….………………..51 A theropod track exposed in horizontal and vertical section…………………………52 Discussion…………………………………………………………………………….53 II Conclusion……………………………………………………………………………53 Part 4. Fossil tracks and trackways Theropod tracks from Greenland……………………………………………………………...55 Introduction…………………………………………………………………………...55 The tracks……………………………………………………………………………..55 Grallator footprints…………………………………………………………………...56 Vertical sections through track NB1………………………………………………….59 Methods………………………………………………………………………………59 Results………………………………………………………………………………...60 Discussion…………………………………………………………………………….62 Conclusion……………………………………………………………………………64 A new track assemblage from Porto das Barcas, Lourinhã, Portugal………………………...64 Introduction…………………………….……………………………………………..64 The Lourinhã Formation……………………………………………………………...65 The track assemblage…………………………………………………………………65 Discussion…………………………………………………………………………….67 Conclusion……………………………………………………………………………68 General discussion…………………………………………………………………………………..70 Overall conclusions…………………………………………………………………………………73 Summary……………………………………………………………………………………………74 Acknowledgements…………………………………………………………………………………76 References…………………………………………………………………………………………..77 Appendix 1. Published work in connection with this study………………………………………...89 III Experimental Ichnology -experiments with track and undertrack formation using emu tracks in sediment of different consistencies, with comparisons to fossil dinosaur tracks. Cand. Scient. Thesis Jesper Milàn Geological Institute, University of Copenhagen Abstract To demonstrate the influence the substrate consistency exercises on the formation of tracks and undertracks, a series of field and laboratory experiments with emus was performed. The emu (Dromaius novaehollandiae) is well suited for comparative ichnological work as its feet bear a strong resemblance to the feet of Mesozoic theropods and further, the emu is adapted to a fully cursorial lifestyle. By encouraging the emu to walk in sediments of different consistencies from firm to fluid, a vide variety of track morphologies was obtained, similar to sedimentologically caused variations found in fossil tracks. To explore the formation of undertracks, experimentally obtained emu tracks emplaced in packages of layered, coloured cement of different consistencies, was subsequently sectioned vertically. The experiment demonstrated that the consistency of the substrate not only influences the morphology of the true tracks, but also exercises a strong control on the formation of undertracks. As a track not only represents the cast of the trackmakers foot, but in most cases also reflects the dynamic movement of the trackmakers foot, causes the trackmorphology to change with depth. This was effectively demonstrated by sectioning a deeply imprinted (4.c cm deep) emu track horizontally, which was subsequently compared with a Jurassic theropod exposed in both horizontal and vertical section. Examination of the preservation of four Upper Triassic theropod tracks from East Greenland, one examined in vertical sections, reveal strong similarities to the experimentally obtained tracks and undertracks. A new track assemblage from Portugal comprises tracks from three ornithopods and indications of varying water content in the sediments resulting in differences in preservation of the tracks. This makes the usage of experimental work very important when fossil tracks are described, as changes in sediment properties dramatically alter the appearance of both tracks and undertracks 1 Introduction ________________________________________________________________________________ Introduction “So complete is the agreement in all essential points between the footprints in the Triassic conglomerate and those of the living Emu, that, leaving all other consideration out of the question, one would not feel much hesitation in declaring for the Avian and, indeed, Ratitous character of the animal which produced the former” - W. J. Sollas 1879. History and introduction to experimental ichnology The main problem faced when studying fossil tracks and traces of animals is to identify the trackmaker. The problem becomes bigger when the tracks studied originate from extinct animals having no extant counterparts. However, by conducting comparative studies with tracks from extant animals with a comparable foot anatomy and similar lifestyle, much information about the fossil trackmakers can be inferred. Fossil footprints do not merely represent the tracks of the animals that passed by, but should be regarded as the dynamic by-product of the dynamic contact between the animal and its environment (Baird 1957). This implies that the morphology of the track is dependent on both the behaviour of the trackmaker and on the consistence of the sediment in which it trod. This understanding of the relation between behaviour and sediment is brought to the extreme by neoichnologist Tom Brown (1999), who claims not only to be able to identify any animal based on its tracks, but also to be able to tell the mood of the trackmaker from minute differences in how the pressure is applied in different parts of the footprints. Through time, several researchers have used recent animals with a comparable anatomy and lifestyle to help understand and interpret fossil tracks and traces. The first documented case, where tracks of recent animals were used to compare with fossil tracks was Sollas (1879), who used casts of footprints from emus, rheas and cassowaries to compare with the slender-toed tridactyl theropod footprints in the Triassic conglomerates of South Wales. Based on the close similarities he suggested that the footprints could originate from ancestors of ratitous birds (dinosaurs were only known from sparse material at that time). He further noticed that the track-morphology of the emu changed with the progression mode, so that in tracks where the emu was accelerating the metatarsal 2 Introduction ________________________________________________________________________________ pad were less impressed in the sediment, than when the emu was walking and thus the morphology of the tracks from the same trackmaker changed with mode of progression. In order to interpret the rich amphibian ichnofauna of the Permian Coconino Sandstone in Northern Arizona, a substantial amount of comparative work with salamander and reptilian trackways has been done through time. McKee (1947) performed a series of experiments in which he made different kinds of reptiles, mostly lizards, walk up the slopes on simulated dune foresets, similar to those found in the Coconino Sandstone. By varying the angle of the slope and the water content of the sand, from dry to saturated, McKee (1947) made convincing analogies to the different track morphologies found in the Coconino Sandstone. Peabody (1959) made an exhaustive research on the trackways of living salamanders for comparison with Tertiary salamander tracks from California. Brand & Tang (1991) used subaquaus salamander trackways to argue for an underwater origin for the otherwise considered aeolian Coconino Sandstone, but their arguments were heavily disputed (Lockley 1992, Loope 1992). Similar experiments with salamanders in substrates ranging from muddy to fine sand, level or sloping and with moisture contents from dry to submerged, clearly showed that the condition of the substrate is an important factor for the trackway morphology (Brand 1996). McKeewer & Haubold (1996) reclassified several Permian vertebrate trackways by demonstrating that several of the different ichnogenera erected through time, in reality, were sedimentological variations of no more than four valid ichnogenera. The fossil trackway, Pteraichnus, described by Stokes (1957) as the trackway from a pterosaur, were re-interpreted by Padian & Olsen (1984), who demonstrated with a recent caiman walking on soft clay, that Pteraichnus could as well be of crocodilian origin. Later, however, new research has suggested that at least some Pteraichnus trackways are of pterosaur origin (Lockley et al. 1995). Following the work of Sollas (1879) ratitous birds have been used, especially for compareison with small bipedal dinosaurs. Padian & Olsen (1989) used the tracks and trackway pattern of a rhea (Rhea americana) to infer stance and gait of Mesozoic theropods, and Farlow (1989) examined the footprints and trackways of an ostrich (Struthio camelus) and compared them with theropod tracks and trackways. Recently Gatesy et al. (1999) compared peculiar partly collapsed theropod tracks emplaced in deep mud from Jameson Land, East Greenland, with the tracks of a turkey walking in similar deep substrate, and found close similarities in the track 3 Introduction ________________________________________________________________________________ morphologies. Gatesy et al. (1999) therefore concluded that the foot movement of theropods during walking exhibited close similarities to the foot movements of modern birds. Despite of the gross overall similarities, the footprints of the large ratites differ significantly from each other when examined in detail, a phenomenon investigated by Farlow & Chapman (1997) and Farlow et al. (1997) who used field observations and casts of tracks from emu, ostrich, cassowary, rhea and the extinct moa to demonstrate how tracks from even closely related forms, exhibit differences significant enough to separate them in different ichnotaxa, had they been found as fossil footprints. Latest in connection with this study, Milàn & Bromley (2002a, 2002b) (Anonymous 2003) (Milan & Bromley in prep (see appendix)) has demonstrated several aspects of track formation related to substrate variations, by conducting field and laboratory experiments with emus. Purpose The field of tetrapod, and especially dinosaur ichnology has been the scene for much speculations and assumptions through time, but only little experimental work has been conducted to support the widely used ideas about track formation. Vertebrate trace fossils comprise not only footprints, although they by far are the topic to which most attention has been paid. Coprolites, fossil excrement, are commonly found, and especially dinosaur coprolites have gained some attention recently, in particular the purported tyrannosaur coprolite described by Chin et al. (1998). For a full account on the topic of dinosaur coprolites see Thulborn (1991) and on coprology in general see Lewin (1999). Fossil nests are also regarded as trace fossils, but only the nest itself is regarded a trace fossil, not associated eggs or embryos. The delicate wear marks found on gastrolites can also be described as trace fossils or bioerosion, while the gastrolith itself is not (R. G. Bromley pers. comm. 2000). The aim of this study is to conduct field and laboratory experiments with emu trackand undertrack formation and compare them with fossil dinosaur tracks and undertracks. Furthermore the control the substrate consistency exercise on the track morphology will be explored. The topics comprised in this work will be explored in four different parts. Each part will contain separate discussions and conclusions on their individual topics. Part 1 focuses on the emus as the best possible extant bird to use for comparison with non-avian Mesozoic theropods. This part will include the phylogenetic relationship between emus, 4 Introduction ________________________________________________________________________________ modern birds and non-avian theropods, a description of an emu foot, and a discussion of the difficulties in discriminating between bird and theropod tracks. Part 2 contains observational and experimental work conducted with emus in the field; This part will include walking patterns, differences in footprint morphology in different substrates, miscellaneous observations of emu tracks, and will conclude with comparisons of emu tracks with footprints and trackways from non-avian theropods and other dinosaur groups. Part 3 comprises experiments with track and undertrack formation demonstrated through serial making vertical sections through emu tracks sat in packages of coloured, laminated cement of different consistencies. Furthermore an emu track was sectioned horizontally to reveal how the shape of the track changes downward, this will be compared with observed erosional cuts through a Jurassic theropod track. Part 4 contains two cases, one with a description of four theropod footprints of the ichnogenus Grallator, from Jameson Land, East Greenland, which all show different modes of preservation, including one with impressions of the skin. Further one track has been sectioned vertically to reveal the subsurface deformation around the digits and the formation of undertracks in connection with the true track. The other case is a description of a new hithero undescribed ornithopod track assemblage from the Atlantic coast of Portugal, which has a variety of interesting sedimentological phenomena connected to the tracks. Terminology and definitions of tracks and trackways. The animal responsible for the tracks studied is termed the Trackmaker. When dealing with fossil tracks the trackmaker is usually unknown. Track refers to a single footprint from the footfall of either the manus or pes from the trackmaker. A succession of several tracks reflecting the progression of the trackmaker constitutes a trackway. Several tracks and trackways from different trackmakers constitutes a track assemblage. In order to describe the individual tracks, both palaeoichnological and neoichnological terminology will be adapted as the experiments contain both recent and fossil tracks. In general the palaeoichnological terminology will follow that of Lockley (1991) and neoichnological terminology will follow that of Brown (1999). The surface on which the trackmaker walks is termed the tracking surface (Fornós et al. 2002). The foot (manus or pes) of the trackmaker form a track in the tracking surface (Fig. 1). 5 Introduction ________________________________________________________________________________ The bottom of the track contains the direct impression of the trackmakers foot, and is termed the true track. If the sediment is soft enough to allow the foot to sink to a certain depth during the footfall, vertical, or near-vertical walls are formed from the true track up to the tracking surface; these are termed track walls (Brown 1999). If the track wall is sloping due to the dynamic movement of the foot during the foot-fall, the track at the tracking surface can appear larger than the true track, and is termed the overall track (Brown 1999). Surrounding the track on the tracking surface, a raised rim consisting of sediment displaced by the pressure of the foot, can be present. In layered sediments the weight of the trackmaker can cause disturbance of the layers subjacent to the tracking surface, and in that case cause the formation of undertracks. Undertracks and their formation is a phenomenon that has been the subject of much terminological and interpretational confusion through times (see pp 32-34: Figs. 28-36), but the majority of ichnologists today follow the terminology and definitions of Lockley (1991). The individual parts of a track can differ significantly, according to both the properties of the substrate and the mode of progression adopted by the trackmaker. 6 Part 1. Why use emus? ________________________________________________________________________________ Part 1. Why use emus? The phylogenetic relationship between theropods, emus and modern birds The birds, together with the crocodiles constitute the living member of the Archosauria. Although the birds and the crocodiles are each other’s closest living relatives, great morphological and functional differences separate them. The dinosaurs are closer related to birds than the crocodiles, and among the dinosaurs the theropods are closer related to birds than to the other dinosaur groups (Fig. 2A). The ratites are the largest extant ground living birds, and are all fully adapted to a terrestrial lifestyle. The wings of kiwis, emu and cassowary are reduced and almost appear as vestiges, while the rhea and the ostrich have large wings they use for display purposes. The tinamous birds, which constitute the sistergroup to the ratites within the palaoegnathae, have fully functional wings and are capable fliers (Davies 2002). The presence of flight-derived characters in the ratites, like the fused carpometacarpus, the presence of an alula in the rhea, the distal part of the tail fused into a pygostyle plus the fact that the ratites first evolved in the early Palaeocene, while ancestral birds have flown since the Late Jurassic show that their flightless-ness is secondary (Davies 2002). Flightless-ness among birds has evolved independtly several times. Almost as soon as the birds evolved the ability to fly, some groups lost it again and became terrestrial dwellers. The dromaeosaurs, who most researchers regard as the sister taxon to the birds (Holtz, Jr. 1994, 1998; Sereno 1997; Paul 2002), were incapable of flight despite their elongated arms and feathers (Ji et al. 2001), although some researchers consider the Dromaeosaurs and other derived groups of theropods as being secondarily flightless (Paul 2002). Archaeopteryx is by tradition considered the earliest bird, and thus is the definition of Aves all birds closer related to Archaeopteryx than to the Dromaeosaurs (Fig. 2B). The Cretaceous bird Rahonavis might be a more primitive bird than Archaeopteryx, as it still possesses the long stiff tail and the sickle claw on digit II as in the Dromaeosauridae (Holtz 1998). Flightless-ness first 7 Part 1. Why use emus? ________________________________________________________________________________ occurred in the Alvarezsauridae and later in Patagopteryx which both became long-legged cursorial forms. The Hesperornithiforms greatly reduced their wings, leaving only a splint of the humerus as an adaptation to a foot-propelled swimming and diving lifestyle. The ratites together with the tinamous birds constitute the group of Palaeognathe birds, of which the main character is the palaeognathus palate. The palaeognathe birds form a sister taxon to all other extant birds who is united by the neognath palate in the neognathae (Fig. 2C). Within the Neognathae a clade consisting of the Anseriformes (ducks and the like) and Galliformes (chickens and the like) forms the sister group to all other neognathe birds (Cracraft & Clarke 2001). It is interesting to note that among the extant birds, it’s the two basal clades, the palaeognathe birds and the Anseriformes and Galliformes who apparently have the least developed flight abilities and have the most earth bound lifestyles. The Palaeognathae and the Neognathae are the only two groups of birds to survive beyond the K/T boundary and thus constitute the extant crown group of theropods (Fig. 2). This makes the ratites the closest extant relatives of the Mesozoic theropods, with a comparable limb anatomy and cursorial lifestyle. The emu foot Morphology of the emu foot Like all birds, the emu walks in a digitigrade fashion, with the elongated tarsometatarsus lifted clear of the ground. Contrary to most birds, the ratites, to which the emu belongs, has reduced the foot to having only three digits, II, III and IV. Digit I, which in most birds are posteriorly directed and allows the foot to grip around an object like a branch in a tree, is lost in the large ground-living ratites, like the emu and cassowary, leaving their feet tridactyl (Fig. 3). The emu retains digit I in its embryonic state, but loses it again before hatching (Davies 2002). In the ostrich, the largest of the big extant ratites, also digit II is lost as an adaptation to high speed running leaving a foot only consisting of digits III & IV. Thus, the emu and the cassowary, of which the latter has an elongated specialized claw on digit II, and is furthermore not kept in captivity in Denmark, the emu foot remains the most useful for comparison with theropods, whose feet in most cases were unspecialised. For the description of the emu foot, the terminology of Lucas & Stettenheim (1972) will be used in the following. 8 Part 1. Why use emus? ________________________________________________________________________________ The tarsometatarsus of the emu consists of the fused metatarsals II, III & IV, fused together with the tarsals. The ankle joint of the emu, resembles the knee joint in being only suited to perform like a hinge and thus to only allow movement in the parasaggital plane. The hinge-like limb joints are a plesiomorphic character shared by the whole Dinosauria (Fig. 2A) (Christiansen 1997). The emu’s toes II, III &IV are homologous with the digits II, III & IV of the pentadactyl appendage, and the phalangeal formula is retained from their reptilian ancestors, as 3, 4 & 5, for digits II, III & IV. In birds who still possess digit I, the phalangeal count is 2, 3, 4 & 5, for digits I – IV (Lucas & Stettenheim 1972:65). In the emu foot, digit III is the longest and digits II and IV are of almost equal length, with digit IV only slightly longer despite that digit II consists of 3 phalanges whereas digit IV has 5 (Fig. 4A). The terminal phalanx of each digit bears a blunt claw. The integument on the sole of the emufoot consists of fleshy digital pads covering the joints between the phalanges. Each digital pad is separated from the next by a small gap, the interpad space, situated approximately at the middle of the two phalanges, the joint of which the pad covers. The joint between the basal phalanges and the tarsometatarsus are covered by a round pad, the metatarsal pad, which in the case of the emu is clearly separated from the other digital pads by a broad, deep interpad space (Fig. 4B). Since the digital pads cover the phalangeal joints, the number of digital pads corresponds to the number of phalanges in the foot. Digit II that consists of three phalanges thus has two digital pads covering the joints. Due to the shortness of the digit, however the interpad space is weakly developed. The terminal phalanx bears a claw. Digit III that has four phalanges bears three prominent digital pads, clearly separated by interpad spaces. Due to the shortness of the phalanges, digit IV, which consists of five phalanges has only what seems to be one long digital pad, weakly divided in two by a shallow part in the middle. Where digital pads in digits II and IV clearly reflect the number of phalanges in the digits, the pads on digit IV do not reflect the number of phalanges in the digit. If the number of digital pads should correspond to the number of phalanges the digital pad formula for the emu foot should be 2-3-4, for digits II, III and IV, but in reality it is 2-3-2. Most fossil theropod tracks exhibit the expected digital pad formula of 2-3-4 but exceptions exist. The theropod ichnotaxon Carmelopodus untermannorum described by Lockley et al. (1998) is characterized by having a digital pad formula of 2-3-3, instead of the normal 2-3-4 digital pad formulas for theropods. Padian & Olsen (1989) used the foot of a rhea to compare the number of 9 Part 1. Why use emus? ________________________________________________________________________________ digital pads in a track with the pedal skeleton, and showed that the number of digital pads in digit IV of the rhea does not correlate with the number of phalanges. The sole surfaces of the digital pads are covered with small, closely situated, horny tubercles of 1 mm size (Fig. 5A, B). Preserved skin impressions in dinosaur footprints have shown their feet to be covered with similar tubercles (Currie et al. 1991, Gatesy 2001). The dorsal side of the digits is covered with transverse overlapping scutes of almost the width of the digits (Fig. 5C). These scutes continue up on the dorsal side of the tarsometatarsus. The ventral side of tarsometatarsus, which in birds is only in contact with the ground during resting, are covered with small pointed scales that increase in size anteriorly. X-ray photography In order to obtain an image of the foot skeleton and an image of the skeleton with the soft parts in-situ, two X-ray photos was taken with different intensities (Fig. 6A, B). The equipment used was a KeveX micro focus X-ray source, which enables imaging of internal structures of even very small samples, unveiling details down to a few microns. The technique is based on an X-ray source with a very small focal spot, which allows a high magnification to be applied (projection radioscopy). The image is captured in real-time by an X-ray image intensifier, or other electronic xray imagining detector. The images are processed by Imagepro + 4.5 with contrast enhancing filters. The machine is normally used for scanning the internal parts of industrial components but is also very effective on organic materials. To produce an image with both soft parts and bones visible an intensity of 28.3 kv was used. To get an image with only the bones visible in much more detail, an intensity of 43.1 kv was used. The maximum size of the objects to be photographed in this machine is 13 x 9 cm that made it necessary to photograph the emu foot in four parts, which are afterwards pieced together electronically, to obtain a complete image of the foot. The x-ray photo taken with the high intensity show how the phalangial bones are articulated, and how the proximal phalanx 1 are significantly longer than the following in all three digits, especially in digit II where the length of phalanx 2 only is about 1/3 of phalanx 1. In digit IV the combined length of phalanges 2-4 almost equal to the length of phalanx 1. Digit III has reduced the length of the phalanges gradually distally. On the x-ray photo taken with lower intensity the outline and some of the internal structure of the soft parts covering the bones are visible. With the fleshy parts in place, the digits approximately double their width. The pads covering the phalangeal 10 Part 1. Why use emus? ________________________________________________________________________________ joints are clearly seen on digit III as the swollen areas around the joints. On digits II and IV the interpad spaces are not visible from the angle shown by the photo. This corresponds with observations made by Farlow & Chapman (1997) who used xray photos of a rhea foot, to compare the soft parts of the foot, which would be reflected in a footprint, with the pedal skeleton. The problem of distinguishing between bird and dinosaur tracks Introduction When the first fossil dinosaur tracks were described in the early 1800´s, they were interpreted as the tracks of giant birds because of their close similarities to the tracks of extant ratites, especially the emu. This was a reasonable interpretation since dinosaurs as a group were not defined at that time. Hitchcock (1836) coined the name Ornithichnites for the tridactyl, slender-toed footprints from Connecticut Valley, and Sollas (1879) pointed out a number of similarities between Triassic footprints from South Wales and tracks of ratites, especially the emu, and suggested that the tracks could have been made by ancestors of the ratitous birds. These tracks have now all been shown to be the tracks of small theropod dinosaurs. Since all modern phylogenetic analyses have established beyond reasonable doubt that birds are the extant crown group of Theropoda (Holtz 1994, 1998; Sereno 1997; Makovicky & Sues 1998), birds (Aves) will herein be defined as the all animals closer related to Archaeopteryx than Dromaeosauridae (Fig. 2B). The track record of possible bird footprints dates back to the Late Triassic (Melchor et al. 2002), while the earliest bird, Archaeopteryx, first occurs in the Late Jurassic. Up through the Cretaceous, tracks with bird-like affinities (Fig. 7) become more common and several avian ichnogenera occur (Lockley et al. 1992). Small theropods and birds have very similar feet and footprints, thus making it difficult to determine the identity of the trackmaker. The following criteria erected by Lockley et al. (1992) and Lockley & Rainforth (2002) are normally used to distinguish bird tracks from dinosaur tracks. (1) A general resemblance to the tracks of Neornithines; (2) Small size, rarely in excess of 10 cm; (3) Slender digit impressions 11 Part 1. Why use emus? ________________________________________________________________________________ with indistinct differentiation of pad impressions; (4) Wide divarication angles (approximately 110120 degrees) between the outer digits, whereas it never exceeds 100 degrees in theropod dinosaurs (Currie 1981; Thulborn 1990); (5) Caudally directed hallux (digit I), oriented up to 180 degrees from middle forward digit, normally digit III; (6) Slender claws and (7) claws on the outer digits curving distally away from the middle digit. Another character, which has been suggested as distintion between bird and dinosaur tracks, is the shape and position of the metatarsal pad – the pad covering the metatarsal/phalangeal joint – that often misleadingly is termed the heel of the footprint. In birds this pad has a distinct rounded shape, possibly due to the fusion of the metatarsals that allowed less lateral movement of the bones in the metatarsal area of the foot (Farlow et al. 2000). While the metatarsal pad in extant birds consists of a single pad covering the metatarsal/phalangeal joint (Figs. 7 & 8), the metatarsal area in theropod tracks seems to consist of the phalangeal pads of digit IV which are impressed along its entire length, while a small gap separates digits III and II from the “heel” area (Fig. 9). The same asymmetry can be seen in tracks attributed to early tracks with ornithischian affinities like Anomoepus, and have been suggested to represent a primitive character for dinosaur tracks, while the condition with the single rounded metatarsal pad is suggested to be linked with the avian fusion of the metatarsus (Farlow et al. 2000). The pes skeleton of Archaeopteryx, the earliest known bird, shows beginning fusion of the proximal ends of the metatarsals, but since no footprints have been attributed to Archaeopteryx, it is not known what effect this incipient fusion of the metatarsals had on its footprint morphology. Discussion When examined separately, the characters used to distinguish bird tracks from theropod tracks seem to fail. Recent findings have rejected the character of the single rounded metatarsal pad as a character unique to bird footprints, as a small track approximately 5 cm in length, from the Lower Jurassic, well before any accepted skeletal bird remains occur, exhibits a single rounded metatarsal pad (M. A. Whyte & M. Romano pers. comm. 2001). The diminutive size of the track as a character linking a track to a bird can not be used as a single character to assign a track to a bird, as several theropods never exceeded the size of chickens, and thus, must have produced tracks of diminutive size. The divarication angle between digits II and IV, which have been regarded as a strong character to distinguish between bird and dinosaur tracks, has shown itself to be an unstable character. Among the extant flightless ratites the divarication angle ranges from an average of 46 degrees in the cassowaries, 57 in the greater rhea to 12 Part 1. Why use emus? ________________________________________________________________________________ 80 degrees in emus (Farlow et al. 2000) which all are divarication values well within the common range for theropod tracks. Even more bird-like tracks that pre-date birds occur in the Late Triassic redbeds of Argentina, and these show in all aspects a striking similarity to tracks of modern shore birds (Fig. 10). All tracks are of small size, ranging from 18.7mm to 27.5 mm in length, have long and slender toes, a single rounded metatarsal pad, divarication angles ranging from 87 to 137 degrees and some of the tracks even display the impression of a fully reversed hallux (Melchor et al. 2002). These new Upper Triassic tracks challenge every character established by Lockley et al. (1992) to distinguish bird from theropod tracks as they predated the earliest known widely accepted birds by some 60 million years. The small fossil Protoavis from the Upper Triassic is a possible candidate to early birdlike footprints as it possesses among several avian features a functionally tetradactyl foot with a reversed hallux (Chatterjee 1999). The avian status of Protoavis as proposed by Chatterjee (1999) is however not widely accepted (Witmer 2001), but avian or not, Protoavis shows, that already in the Upper Triassic there existed theropods having avian features in the feet, which would be able to produce birdlike footprints. To further complicate the picture, recent findings from France showed the existence of large flightless birds, at least the size of an ostrich, in the Upper Cretaceous (Buffetaut & Loeuff 1998). The tracks produced by a flightless bird of that size must be of comparable size to tracks produced by its contemporary medium-sized theropods. Conclusion The pronounced similarities between the footprints of birds and the small to mediumsized non-avian theropods, show that there is no way in praxis to effectively distinguish bird and theropod tracks from each other. This is evidenced by the presence of birdlike footprints having affinities to modern shorebirds in the Late Triassic, as well as small theropods having avian affinities in their feet and large flightless birds the size of medium-sized theropods in the Upper Cretaceous. The only way to be sure a bird-like fossil footprint really originates from a bird and not a non-avian theropod is if the track post-dates the Cretaceous – Tertiary boundary. 13 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Part 2. Field observations and experiments with emus Field observations of emu tracks and trackways with comparisons to fossil tracks and trackways Introduction The footprints of large flightless ratites have long been used for comparison with dinosaur tracks, especially the tracks of theropods and small bipedal ornithopods, because of the close morphological similarities in their tracks. The first record of comparisons between ratitous footprints and theropod tracks was Sollas (1879) who compared casts of emu and cassowary tracks with what he then believed to be tracks of giant extinct birds in the Triassic conglomerates of South Wales. Recently Padian & Olsen (1989) demonstrated that the stance and gait of theropods and small bipedal ornithopods were almost identical to that of the extant rhea, by comparing trackways from the rhea with those of the theropods. Farlow (1989) made similar observations of tracks and trackways of an ostrich, and noticed that an ostrich might not be the best of the large flightless birds to use for comparison with theropods, because of its didactyl foot. That a footprint representing the same action can differ significantly according to the type of sediment it is impressed in, was clearly demonstrated by Bromley (1996:157) who compared his own footprints from walking on photographic paper, dry sand, damp sand and on an intertidal mudflat. The emus used for the experiment belong to emu breeder Karin Holst, Mønge, who kindly allowed the use of her emus and helped with all the practical things involved in getting an emu to cooperate. Luckily the emus were domesticated and used to contact with humans, otherwise they can be very dangerous animals that actively will attempt to fight off intruders in their area. The emus are able to deliver a kick strong enough to seriously injure a human who gets in the way and male brooding emus are reported to attack from a distance of up to 100 m (Davies 2002:224). An earlier experimental session (Milàn Nielsen 2000) was attempted at Mogens Madsen, Kalundborg, a specialist in reptiles and exotic animals, whose emus was significantly less accustomed to human contact and did at first not tolerate other persons than the owner within the fencing, with the result that the owner had to stand between us and the emu most of the time, to prevent it from attacking. 14 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Sediment grain-size analyses The sediments used for the experiments were (1) the local soil in the emu enclosure, (2) beach sand and (3) a red haematite-coated sand from a local quarry. Each sediment was analysed using the methods advised by Tucker (2001) and the mean grain-size (Md), median grainsize (M) and degree of sorting (σΦ) were calculated for each sediment using the formulas by Folk & Ward (1957) (table 1). Median grain size, Md = Φ50 Mean grain size, M = (Φ16 + Φ50 + Φ84)/3 Sorting , σΦ = (Φ84 - Φ16)/4 + (Φ95 - Φ5)/6.6 Median grain size Mean grain size 2.2 2.18 1.4 1.56 2.4 2.38 Local soil Beach sand Red sand Sorting 1.57 = Poorly sorted 0.75 = Moderately sorted 1.42 = Poorly sorted Table 1. The median grain size, mean grain size and degree of sorting for the three soils used for track making. Degree of sorting is based on values determined by Folk & Ward (1957). The results (Fig. 11) show that the local soil is poorly sorted. The beach sand used is moderately sorted; the grains are subangular with a high sphericity. The red sand is poorly sorted, the grains are with and the grains are angular with high sphericity and a thin cover of haematite. The local soil admixed with various amounts of water was used to produce mud of different depths and consistencies (Figs. 18D-H) and casts of emu tracks impressed in deep mud (Figs. 23, 24, 44). The beach sand was used to record trackways and individual tracks (Figs. 14, 18C, 19, 22). Individual tracks were recorded in the red sand (Figs. 18B, 20). Beach sand from a second locality was used during earlier experiments (Figs. 15, 16, 17, 18A), but no grain-size data exists for that. Walking cycle of an emu In order to record the movement of the emu foot during a stride, a rapid series of photos of the legs of a bypassing emu was taken. The emu on the pictures was progressing at normal walking speed on firm soil along one of their preferred paths in the compound (Fig. 12). 15 Part 2. Field observations and experiments with emus ________________________________________________________________________________ In the following, the terminology for the description of foot movement is taken from Thulborn & Wade (1989) and Avanzini (1998); this terminology is applicable to all walking tetrapods. According to this terminology the movement of the foot during walking, can be divided into three phases. The touch-down phase (T), is the phase where the foot is extended forward and planted on the ground. This phase is followed by the weight-bearing phase (W), where the animal’s centre of gravity passes over the animal’s foot, which becomes impressed into the substrate. The last phase is the kick-off phase (K). In this phase the proximal parts of the foot are raised and the weight is transferred to the distal parts of the digits as the body moves forward and the foot subsequently is lifted towards a new T-phase. When the foot is lifted the digits converge and bend backward to a nearly vertical position while the foot is moved forward. Trackways In order to record continuing series of tracks or trackways, a lane of sand was laid out on one of the paths in the fencing regularly used by the emus. After the emus had been encouraged to walk and run through the sand, three trackways, two where the emus were walking and one where the emu was running were measured (Fig. 13). It was not possible with our available equipment to record the progression speed of the emus, but the normal walking speed of an emu is 7 km/hr, and they can reach running speeds of 45 km/hr and are able to maintain that speed for several kilometres if necessary (Davies 2002:22). The measurements used to describe trackways (Fig. 14) are stride length, i.e., the length between two subsequent footprints from the same foot, right - right or left – left; and the pace, i.e., the distance between the imprint from the left and right foot. Pace and stride length are measured between identical points in the two tracks, normally the metatarsal pad. Pace angulation: the angle between three consecutive footfalls, right-left-right or left-right-left, are used as a parameter for measuring the width of the trackway, how close to the midline of the trackway the trackmaker puts its feet. A pace angulation of 180° would imply that the animal put its feet directly in front of each other while walking, and then the trackway width would be equal to the width of the trackmaker´s foot. After measuring the trackways the average values for stride length, pace length and pace angulation were calculated (table 2), to allow comparison of the trackways. 16 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Trackway Max. Min. Mean Max. Min. Mean Max. Min. Mean B Stride length cm 112 95 105 Pace length cm 57 42 52 Pace angulation 167° 157° 170° C A 103 95 99 214 212 213 53 48 51 114 100 106 178° 152° 162° 146° 135° 142° Table 2. Selected measurements from the three recorded emu trackways. A trackway recorded during an earlier experiment session at Mogens Madsen, Kalundborg, captured an emu making a sharp turn and accelerating to run (Milàn Nielsen 2000). The working photograph of the trackway capturing the turn has captured seven footprints, two from the left foot and five from the right. When the emu made the turn it stepped off the prepared sand strip and out in the grass, which prevented further recording of tracks (Fig. 15). Tracks of the left foot are termed L1 and L2 and tracks from the right are termed R0 to R4. R0 and R1 are almost in contact with each other and originate from the emu´s reluctance to walk onto the sand, and thus first took a short and anxious step before it progressed. To clarify the foot movements during the turn, connecting lines between the consecutive (right-right and left-left) footprints were drawn. Approaching the turn the stride L1-L2 and R1-R2 are of equal length. The turn starts with the shorter stride R2-R3. The stride with the left leg crosses the right and the foot is planted outside of the prepared sand strip, preventing precise recording of its position. Right foot pushes backward (R4) and accelerates. The acceleration track differs significantly in shape from the track produced during normal walk. In the track representing normal walk, all three digits and the metatarsal pad are represented, the claw of digit III being slightly more deeply impressed than the digital pads (Fig. 16A). The acceleration track R4 (Fig. 16B), on the contrary, shows a pronounced unevenness in the depression of the digits. Digit III is the deepest impressed, and shallows proximally. The claw of the digit has left a clear cut through the sand. Digit IV is also most deeply impressed in the distal parts and shallows proximally. Only the claw and the distal digital pad of digit II are present in the track. The metatarsal pad is only represented by a very shallow depression 17 Part 2. Field observations and experiments with emus ________________________________________________________________________________ in the sand, and is almost indistinguishable from the surrounding sand. The track walls off the digits are sloping to each side; this shows that the divarication angle increased from the tracking surface to the bottom of the track by 15˚ during the acceleration (Fig. 17). The angle between digits III and IV increases by 27˚ from surface to the bottom while the angle between digits II and III decreases with 11˚ leaving a total increase in divarication angle of 15˚ from the surface of the track to the bottom (table 3). Interdigital angle II-III III-IV II-IV Surface 35˚ 27˚ 62˚ Bottom 24˚ 53˚ 77˚ Difference -11˚ 27˚ 15˚ Table 3. The interdigital angles of the acceleration track. Angles at the surface are significantly smaller than the angles at the bottom of the track. The divarication angle increases from 62˚ to 77˚ from the surface to the bottom of the track. Divarication angle of emu footprints In order to obtain an estimate of the variations in divarication angle of the emu foot during progression, the divarication angles were measured on each footprint in the recorded trackways (Figs. 12A-C, 14), and in the tracks emplaced in different sediments (Figs. 17A – G), a total of 30 footprints. The divarication angles ranged from 61° - 102°, with an average of 76.8° (table 4). Tracks Fig. 13A Fig. 13B Fig. 13C Fig. 15 (H1-5, V1-2) Fig. 18 (A-G) 61˚ 70˚ 78˚ 102˚ 77˚ Max. div. angle 102° Min. div. angle 61° Ave. div. angle 76.8° 69˚ 75˚ 82˚ 76˚ 63˚ Divarication angle 74˚ 78˚ 75˚ 77˚ 63˚ 95˚ 76˚ 64˚ 78˚ 79˚ 62˚* 63˚ 73˚ 90˚ Ave. div. angle fig. 13A Ave. div. angle fig. 13B Ave. div. angle fig. 13C 83˚ 84˚ 88˚ 91˚ 80˚ 80˚ 70.5˚ 73.8˚ 79.8˚ Table 4. The divarication angle measured in the trackways recorded in figs. 12, 14 and the tracks from fig. 18. The asteriks marks the acceleration track for which the surface value has been used. Lower table shows the average divarication values for all tracks measured and for the three trackways recorded. 18 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Footprints in different substrates Since a footprint can be described as the by-product of dynamic contact between an organism and its environment (Baird 1957), both the movement of the animal that produces the track, and the nature of the environment in which the animal trod, have considerable effect on the morphology of the track produced. To investigate this effect, eight different substrates were prepared in which the emus were encouraged to walk. The eight different substrates were; dry loose sand, damp sand, wet sand, firm mud, soft mud, deep soft mud, deep semi-fluid mud and deep fluid mud. The mud was produced by admixing the black organic rich soil from the emu-paddock with various contents of water, and the sand used was beach sand and red sand from a local quarry. At first it proved to be a problem to persuade the emus to walk on the prepared sediments. Emus act very suspiciously to changes in their environment, causing them at first to avoid the prepared areas where they were supposed to walk. But by placing the sediments on their preferred paths along the fencing (and encouraging them by holding a bucket of seed at the other end of the sand path) it proved much easier to make them walk through. After a period of getting used to walking in the sediments, the opposite problem arose, when the emus started constantly to walk back and forth through the sediments, so at times it was necessary to prevent them from overstepping and destroying the tracks once made. The emus were especially fond of the soft mud, in which they continuously walked back and forth. Dry loose sand The dry sand has no cohesive properties and the footprint collapsed immediately after the foot had been lifted. The tracks reveal little details except from the overall shape of the metatarsal pad and the digits. The digits appear broader and more rounded because of the collapsed track walls causes the overall track to appear significantly wider than the true track (Fig. 18A). Damp sand The track is clear and well defined in all respects. The shapes of the digits and the individual digital pads and claw impressions are clearly recognizable, even faint skin impressions are preserved in the prints (Fig. 18B). 19 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Wet sand The wetness of the sediment causes the sand to flow into the track, slowly obliterating the shape. The horizontal pressure applied to the sediment by the foot causes the sediment to crack in a radial fashion outwards from the track (Fig. 18C). Thin layer of soft mud The features of the track are very well preserved. The individual digital pads are clearly recognizable in the digits, and the small tubercles are preserved in most of the imprint. An amount of mud has been pressed up between the proximal parts of digits III and IV. After removal of the foot, the upward pressed mud collapsed down into the impression of digit IV, partly covering it (Fig. 18D). Deep firm mud The consistency of the mud is firm enough to allow the track walls to remain standing after the foot was lifted. The bottom of the footprint shows a perfect moulding of the sole of the foot, with every digital pad and impressions of the tubercles covering them (Fig. 17E). Deep semi-firm mud After the foot was lifted the softness of the sediment caused the track walls to slowly collapse over the digits. In the metatarsal area of the track, striations created by the tubercles during the impression of the foot are preserved (Fig. 18F). Deep semi-fluid mud The consistency of the mud caused the track walls to collapse over the digits while they were impressed in the sediment, causing in turn the material covering the digits to be lifted up by the foot and transported and dropped in front of the track. In the “heel” area of the tracks, striations in the track walls can be seen as a result of forward movement of the foot during the footfall. The consistency of the sediment causes it to slowly collapse, and thus destroy the shape, while an amount of ejected sediment is present in front of the track (Fig. 18G). 20 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Deep fluid mud In this sediment which had a semi-fluid consistency, the sediment collapsed over the digits during the T-phase, burying the digits but was too liquid to be lifted up by the foot. After the foot was removed, the surrounding sediment began to flow into the track, further obliterating the shape. A drag trace from the claw of digit III, produced when the foot is folded and lifted forward to a new T-phase, is present on the sediment surface in front of the track, together with drops of sediment ejected by the foot (Fig. 18H). Potters Clay In order to get a footprint with as many details preserved as possible, a fresh, severed emu foot was impressed in potters clay, the consistency of which make it the ideal material to preserve fine anatomical details. The individual digits are preserved in exquisite detail, with the tubercles on the digital pads and the interpad spaces clearly visible. Radial cracks formed in the area behind the metatarsal pad, in connection with the claw impressions and at the proximal parts of digits II and IV (Fig. 18I). Sediment transport by the foot While the emus walked through the sand and mudlayers it was noticed that a thin layer of sand or mud stuck to the sole of the emus foot (Fig. 19A). This was in some cases released in the following footprints. For an example where the emu walked out of the sand and onto firm soilground, its tracks became traceable for a distance as sand “shadows” of the footprint, originating from sand stuck to the foot and gradually released in the consecutive steps (Fig. 19B). This process is quite different, off course, from the transport process seen in semi-fluid mud. Didactyli in emu footprints On several occasions when the emus were walking on firm substrates like damp sand, it was noticed that digit II became less impressed in the sediment than digits III and IV, which always were impressed to about equal depth. In all tracks examined in a trackway produced by an emu walking in damp sand, digit II was less impressed than the other digits (Fig. 20A). The interpad space between the digital pads and the metatarsal pad left no impression in most of the tracks (Fig. 20B), indicating that the proximal part of digit II is held higher than in digits III and IV. In most tracks the impression of digit II is faint but recognizable. In one track however, the only 21 Part 2. Field observations and experiments with emus ________________________________________________________________________________ hint of digit II is a shallow pinch mark produced by the tip of the claw, which makes the track appear perfectly didactyl (Fig. 20C). That the emu carries the least weight on digit II is a feature it shares with the other ratites. The cassowary carries an elongated straight claw on digit II and in the ostrich the digit is totally absent as an adaptation to high-speed running. A track including metatarsus impression The emu, like all extant birds, has a fully digitigrade stance, with the elongated tarsometatarsus held at a steep angle to the ground. Only when the emu is resting, does it sit with the anatomical heel in contact with the ground (Fig. 21). While the emus were feeding on seed strewn on the ground it was noticed that they sometimes walked around on the metatarsus in a plantigrade fashion, producing very long tracks with full metatarsus impressions (Fig. 22). The metatarsus impression is deepest proximally, at the anatomical heel, and shallows distally towards the metatarsal pad. Striations caused by skin tubercles In the tracks formed in deep, firm to semi-firm mud, striations from the skin tubercles on the sides of the foot were preserved in some cases. The striations found originated from all parts of the foot, both the digits and the metatarsal area (Fig. 23). In many cases the claws left a clear striation mark as they sliced through the sediment (Fig. 24A). In some of the tracks made in deep substrates the striations from the skin tubercles change direction with depth, and thereby show the actual movement of the foot during walk. On digit II on one of the plaster casts, the striations, together with an elongated claw mark, show how the digit first was pressed down in the T- and Wphase and then dragged backwards and up during the K-phase (Fig. 24B). Discussion and comparisons with fossil tracks and trackways That the substrate has a great effect on the morphology of the footprint, and on the likelihood for the track to be preserved as a fossil track is discussed in detail by Allen (1997), who studied human and cattle tracks placed in sediments of different consistencies in the Severn Estuary, England. Allen (1997) observed that a human track printed in deep fluid to semi-fluid mud would collapse and the sediment would gradually flow back and fill the track, obscuring it from the bottom. Such a track would, if fossilized, reveal little about the nature of the trackmaker, and may 22 Part 2. Field observations and experiments with emus ________________________________________________________________________________ be recognized only as a mass of disrupted sediments below a slight depression in the sediment surface (Allen 1997). Tracks in soft mud have a much less pronounced tendency to flow and collapse, but are generally poor in detail because of the tendency of the mud to adhere to the trackmaker´s foot and create adhesion spikes as the foot is withdrawn, leaving the track blurred. In some cases the movement of the foot left striations on the track walls. Tracks preserved in such sediments are likely to show only gross anatomical features (Allen 1997). In an ox track in stiff mud Allen (1997) observed a large amount of anatomical details preserved in the track and further observed that, with the initial drying of the sediment, dissication cracks formed in the sediment, localized and radiating outward from the track. Firm mud is, according to Allen (1997), likely to produce very well defined but shallow tracks, and if fossilized should preserve even fine details as skin impressions. Based on field observations of tracks in recent and Plio/Pleistocene sediments in Kenya, Laporte & Behrensmeyer (1980) described the connection between grainsize/watercontent of the sediment and the potential for track preservation. According to their observations, tracks will not be preserved in dry sediments, since dry sand is too loose to preserve tracks, and dry clay simply is too hard to allow formation of a track. If the sediment is water saturated, sand will be too loose to preserve tracks and clay will be too gooey. The optimal parameters for track preservation according to Laporte & Behrensmeyer (1980), is when the sediment is moist, and with a grain size between sand and clay. Diedrich (2002) demonstrated several different preservational variants of rhynchosaurid footprints caused by differences in water content of the sediments. Footprints made in dry subaerial sediments consist of little more than faint claw imprints. With increasing water content of the sediment, shallow tracks with skin texture preserved are found. In more water-rich, and thus softer sediments, the tracks become deeper and more blurred in shape and finally subaquatic tracks produced by a swimming trackmaker are found as elongated parallel scratch traces. These observations on the influence of substrate consistency on track morphology by Laporte & Behrnsmeyer (1980), Allen (1997) and Diedrich (2002), correspond largely with the results obtained from the experiments with emus walking in different substrates. The emu track in dry sand (Fig. 18A) shows little detail, but the gross overall shape of the track and has some small potential for being preserved. The damp sand (Fig. 18B) preserved anatomical details of the foot such as digital pads and even faint skin impressions. The chance of getting tracks preserved in damp sand is present since numerous tracks hawe been found in fossil aeolian deposits (McKeewer 1991; 23 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Fornós et al. 2002), and furthermore the consistency of damp sand proved rigid enough to allow plastercasts to be made of the footprints. The wet sand (Fig. 18C) produced deep, well defined tracks, but the wetness of the substrate caused the tracks to slowly flow together obliterating the shape of the track. The radial cracks formed around the track are, in this case solely caused by the horizontal pressure applied to the sediment, as the foot was touched down. A thin layer of soft mud (Fig. 18D) produced very detailed footprints with a high degree of anatomical details, such as digital pads and skin impressions. Tracks impressed in mud have a high preservational potential if allowed to dry a little before burial. That a thin layer of soft mud preserved the highest amount of anatomical details is confirmed by the observations of Laporte & Behrnsmeyer (1980), Allen (1997) and Diedrich (2002). The amount of details preserved in deep firm mud (Fig. 18E) is similar to that of a thin mud layer, and the fossilization potential is high. The deep track walls have preserved striations from the skin tubercles as the foot is moved through the sediment (Figs. 23 & 24), a phenomenon that has been found in fossil footprints (Figs. 25, 52 & 59). In deep mud of slightly softer consistencies (Fig. 18F) the track walls collapsed over the digit impressions after uplifting of the foot, leaving a track consisting of a triangular “heel” area and sometimes a round hole where digit III exited the sediment. This kind of track morphology is found in Triassic theropod tracks from Jameson Land, East Greenland, as described by Gatesy et al. (1999). By admixing the sediment further with water, giving the mud a semi-fluid consistency (Fig. 18G) the softness of the mud caused the track walls to collapse over the digits, embedding the foot in the sediment, which caused an amount of mud to be thrown up in front of the track during the K-phase. Tracks imprinted in this kind of sediment are not likely to reveal many details if fossilized, since the track slowly flow together after the foot has been lifted. In fluid mud (Fig. 18H) the sediment flow around the digits of the foot both during the T-phase and when the foot is lifted in the K-phase, causing little mud to be ejected. The track walls flow together quickly after the foot is removed making the track unrecognisable after a short period of time. A succession of sauropod tracks with evidence of progressively increasing water content of the sediment is described from the Upper Cretaceous of Bolivia. In some of the described trackways, the tracks gradually change morphology from well preserved to small, collapsed, rounded holes (Lockley et al. 2002). The track produced in potters clay (Fig. 18I) represent the optimal sediment for track preservation, and the fine grained texture of the clay preserves even minute details of the skin. Clay has a high potential for fossilization as it hardens as it dries. All dinosaur tracks with good skin 24 Part 2. Field observations and experiments with emus ________________________________________________________________________________ impressions are found in clay or very fine-grained sediments (Woodhams & Hines 1989; Currie et al. 1991; Gatesy 2001; McCrea et al. 2001). That the track from the same trackmaker can appear morphologically different as the substrate properties changes can cause confusion when fossil tracks are described. A revision of Permian vertebrate ichnogenera from Scotland and Germany, showed that many genera were simply substrate or behavioural variants of tracks from the same trackmaker (McKeewer & Haubold 1996). Same effect is clearly demonstrated by Bromley (1996:157), who made tracks with his own foot in different sediments, and in the recent trackway of an ox walking into a field of fluid earth (Bromley 2001). The formation of radiating fractures in the sediment around the emu tracks in sediments of rigid consistencies (Fig. 18C, I) and the tracks of Allen (1997) where the track acted as a localizing agent for desiccation cracks formed during initial drying of the sediment, is interesting since, in some occasions, similar radiating fractures have been described from fossil dinosaur tracks. On a track-bearing slab containing both ornithopod (Amblydactylus) and bird tracks (Ignotornis), from the Dakota Formation, Colorado, some of the large ornithopod tracks had radiating fractures around them (Lockley et al. 1989), and radiating fractures around depressions in sediments in the St. Mary Formation, Montana were interpreted as caused by footprints (Nadon 1993). In their description of dinosaur tracks on the Istrian peninsula, Croatia, Vecchia et al. (2001:Fig. 18) note fractures radiating out from the “heel” area of some weathered sauropod tracks, and discuss the possibility that it could be claw marks, but conclude that because of the number and the fact that the fractures points backward, it is weathered cracks in the sediment. It is highly probable that the cracks they describe in fact are radiating cracks, either originating from the footfall, or from initial formation of desiccation cracks, which after exposure have been altered by water erosion to the present appearance. Striations and skin impressions are seldom preserved in fossil footprints since these features only can be preserved in sediments of the right consistency. Furthermore, the skin impressions can only be preserved in the true track, or the natural cast hereof, and are easily destroyed by erosion. Skin impressions preserved in dinosaur tracks were depicted for the first time, but not described, by Hitchcock (1858:plate X), in a theropod track, Brontozoum (Eubrontes) giganteus, but have only recently been described in theropod footprints (Gatesy 2001). Skin impressions have been found on several occasions in well-preserved ornithopod tracks, 25 Part 2. Field observations and experiments with emus ________________________________________________________________________________ Caririchnium leonardii, from the South Platte Formation of Colorado (Currie et al. 1991) and in the Lover Cretaceous of East Sussex, England (Woodhams & Hines 1989). In both cases the impression was that of closely packed skin tubercles averaging 3-5 mm in size on the digits and up to 10 mm in the metatarsal region. Similar tubercles are found in Triassic theropod tracks in Jameson Land, Eastern Greenland. In some of the tracks the tubercles had left striations similar to those observed in the emu tracks, indicating the movement of the foot during the stride (Gatesy 2001). Probable ankylosaur footprints from the Dunvegan Formation, Alberta, show both manus and pes imprints with impressions of skin tubercles, and in one case the tubercles of the manus have left clear striations in the track walls (McCrea et al. 2001: Fig. 20.24a,b). A natural cast of an ichnite, Deltapodus brodericki, an ichnotaxon originally proposed by Whyte & Romano (1994) to be of sauropod origin, later reinterpreted by M. A. Whyte & M. Romano (pers. com. 2002) to be of stegosaur origin, from the Scarbourgh coast in England, show clear striations on the sides, and on the digits of the track. The tubercles responsible for the striations must have been about 1 cm wide (Fig. 25). The orientations of the striations show that they originated from the K-phase when the foot was lifted. Preserved skin impressions and striations found in Late Triassic theropod footprints from Jameson Land, East Greenland, show not only that the sole of the Late Triassic theropods feet were covered with skin tubercles, but have also been used by Gatesy (2001) to recreate the foot movement during the stride. The multidirectional striations preserved on the sides of the emu footprints (Figs. 23 & 24), show that it is possible to recreate the subsediment movement of the foot from the orientation of the striations. That the foot acts as a sediment-carrying device during walking has been rarely discussed in the literature; only Nadon (2001) mentions the effect as an explanation for the lack of skin details in most fossil footprints. His argument goes that the thin layer containing the actual impression of the skin tubercles is stuck to the underside of the foot, and only very rarely is released to produce a track with skin impressions. The observations from the emus (Fig. 19), confirms that the sediment covering the sole of the foot can be so thick that it would blurr the skin textures. Another interesting implication of the foot acting as a agent for sediment transport, is that bioturbation caused by walking vertebrates, in these cases not only serves to mix the sediment vertically, but that the transport of sediment stuck to the sole of the trackmaker feet acts as a horizontal component in the mixing of the sediment. 26 Part 2. Field observations and experiments with emus ________________________________________________________________________________ The measurements from the two trackways where the emu was walking differ little from each other, whereas the trackway from the running emu shows a pronounced change in the trackway parameters. The pace angulation in the two walking trackways is high, which gives a narrow trackway, but in the case with the running emu, the pace angulation is much lower resulting in a wider trackway. That the pace angulation is significantly smaller when the emu is running, directly contradicts the observation made in a theropod trackway from Arden Quarry, Oxfordshire, England by Day et al. (2002), where the pace becomes narrower, and thus the pace angulation higher, as the theropod increases its progression speed. The trackway originated from a large-sized theropod with an estimated hip height of 193 cm (Day et al. 2002). Not all trackways from running theropods are emplaced in a narrow pattern. Measurements made on a running trackway assigned to the ichnogenus Skartopus, which originate from a diminutive dinosaurian trackmaker with a foot length of about 5 cm, depicted by Wade (1989: Fig. 8.5) show that it places its feet in a broad trackway, with a pace angulation on average 147°. This is close to the pace angulation of 142° from the running emu, and might demonstrate that smaller animals do not need to place their feet beneath their centre of gravity while running, as larger animals do. If this were the case, the running pattern of the emu, with its foot-length of around 20 cm would then be comparable with small bipedal dinosaurs and not the larger ones responsible for the tracks as described by Day et al. (2002). James Farlow (pers. comm. 2002) has recorded trackways from running emus, and his experience is that pace angulations as low as 140° would be uncommon for running or trotting birds but not impossible. One of his recorded trackways had a stride length of just above 3 m and a pace angulation of 150°, but that was an unusually wide trackway. The possibility that the wide gauge trackway from the running emu is a result of the fact that we had to scare the emu a bit to get it to run, and thus represents an unnatural running pattern, must also be considered. A trackway from the Minnes Formation, Western Canada, shows a large theropod with a foot-length in excess of 50 cm, pace down to a halt and make a 90° degree turn to the right before it continued (Currie 1989). The track pattern during the turn is very similar to that of the emu making its turn (Fig. 26). As the animals approach the turn, the left foot is offset to the left (L2) and the last stride with the right foot (R2-R3) is significantly shorter than the previous R1-R2. The emu makes a sharper turn than the theropod causing R4 to be placed behind R3 and the stride L2 to the unrecorded L3 to cross behind R3. A similar foot pattern is found in Grayssac, France where a small theropod with a foot-length of only 11 cm makes a 180° turn (Griffiths et al. 2002). 27 Part 2. Field observations and experiments with emus ________________________________________________________________________________ The divarication angles measured from 30 tracks ranged from 61° to 102° degrees, with the average value of 76.8°. The changes in substrate did not seem to be the cause of the high variation in the divarication angles since the divarication angle could vary from 62° to 102° within the same trackway on the same sediment (Fig. 15). The minimum value of 61° was obtained from the running trackway and the maximum value of 102° was from a slow walk. The progression speed of the emu seems to have a small effect on the divarication angle, as the values for the running trackway (table 4) lie within a range of 17° between the minimum of 61° to the maximum of 78° degrees and an average divarication value of 70.5°, whereas the ranges for the two walking tracks were 20° an 31° for trackways B and C (table 4) and the average values for trackways B and C were 73.8° and 79.8°. This shows that the emu keeps a lesser angle between its digits and a more uniform divarication angle when it is running than when it walks. However, further recording of additional trackways is needed before any conclusions can be drawn on that. Measurements of 53 emu tracks by Farlow et al. (2000) gave an average divarication angle of 80.8°, with 55° as the minimum value and 119° as the maximum value, which gives a total range from minimum to maximum divarication angle of 69°. The average divarication angle of 76.8° and the minimum and maximum divarication angles of 61° and 102°, obtained in this experiment, correspond well with the values obtained by Farlow et al. (2000). The lower average divarication angle, compared with that of Farlow et al. (2000), is not that significant, since the individual angles differ considerably from each other, even in trackways on the same substrate during the same progression speed, that another sample probably would produce another average value. That the emu, when walking on firm substrates, carries the least weight on digit II and then produces tracks having little or no impression of the digit is interesting, since some dinosaur tracks have been reported that show various degrees of digit impressions. Thulborn (1990: 122) discussed the same substrate-related didactyli in emu footprints, and suggested that didactyl variants of both ornithopod and theropod footprints could be related to the consistency of the substrate rather than didactyl dinosaurs. A truly didactyl variant of a Eubrontes track is pictured in Lockley (1991), but it is due to a dinosaur that had a missing toe, as it is only on the right foot that the toe is missing. Some theropods, however, show clear adaptations in digit II that made them unsuitable for use in walking. The dromaeosaurs, troodonts and the primitive, dromaeosaur-like bird Rahonavis all carry an elongated sickle claw on a hyper-extensible digit II, used for killing prey. The length and shape 28 Part 2. Field observations and experiments with emus ________________________________________________________________________________ of the claw implies that the digit had to be lifted clear of the ground during walking, but no convincing tracks, except perhaps for one Deinonychosauripus (Zhen et al. 1986. ref. in Zhen et al. 1989), have been attributed to the sickle-clawed theropods. A yet undescribed trackway from Crayssac, France, shows perfectly didactyl theropod tracks, but in this case the tracks apparently consists of digits II and III and show no trace of digit IV (Peter Griffiths 2002, pers. comm.). This contradicts what would be expected for didactyl theropod footprints, as all specialization in theropod digits occurs in digit II. That the emu, in spite of its elongated metatarsus, is able to walk plantigrade, although ineffectively, is interesting since plantigrade theropod trackways found along the Paluxy River, Texas show that, at least the theropods responsible for these tracks, were able to perform plantigrade walk at normal speed (Kuban 1989). The Early Jurassic ichnogenus, Anomoepus, which is referred to small ornithopod dinosaurs, includes a variant where the trackmaker is resting with the metatarsus fully impressed in the sediment, as well as ischiadic, and belly impressions and five-fingered manus imprints (Hitchcock 1858: plate VIII, IX). Interestingly the metatarsus impressions in these resting traces (Fig. 27) is impressed in the sediment in a way similar to that of the emu resting track (Fig. 22), with the proximal part of the metatarsus most deeply impressed, shallowing distally. In this case the metatarsal joint is not represented by any pad impressions. Similar resting traces are not known from theropods. Unfortunately the other parts of the emu body failed to leave recognizable traces when the emu was resting, as a comparison of the other parts of the emu resting track with the associated body imprints in the Anomoepus resting track would be interesting. Conclusions By conducting practical field experiments with living animals, valuable insight into the processes of trackmaking and track preservation can be obtained, which is hard to obtain from laboratory experiments. This is due to the fact that the formation of a track is the result of the dynamic interaction between the trackmaker and the substrate. The close similarities between the tracks of emus and especially tracks of theropods and small ornithopods make emus ideal for comparative track studies. Other features of the emu foot bear close resemblance to features in the feet of other dinosaur groups like the thyreophorans and the larger ornithopods, in that the skin texture in their footprints in all cases are that of closely packed tubercles. 29 Part 2. Field observations and experiments with emus ________________________________________________________________________________ The divarication angle is not a reliable value to discriminate between tracks, as the angle varies significantly in tracks within the same trackway (Table 4), and even at different depths in the same track as the case with the acceleration track (Fig. 17). The consistency of the substrate exercises a strong control on the track morphology and the amount of anatomical details preserved in the tracks. That means that extreme attention should be paid to the sedimentary context in which the tracks appear. 30 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Part 3. Laboratory experiments with track and undertrack morphology Experiments with track and undertrack formation in layered cement Introduction During recent years, several authors have documented that the physical properties of sediments exercise a strong control over the morphology of vertebrate tracks impacted within them (McKeewer & Haubold 1996; Bromley 1996, 2001; Allen 1997; Gatesy et al. 1999; Gatesy 2001; Nadon 2001; Fornós et al. 2002; Melchor et al. 2002) . In marine invertebrate trace fossils, differences in substrate consistency can cause trace fossils produced by the same tracemaker to appear morphologically sufficiently different that they are placed within different ichnogenera (Bromley 1996). Permian vertebrate trackways of varying morphologies from Scotland have been reclassified, as it turned out that, what had been described as trackways produced by different taxa of animals were the result of differences in sediment consistency, rather than the prints of different trackmakers (McKeewer & Haubold 1996). Triassic theropod tracks from Jameson Land, East Greenland, which display a variety of morphological variation resulting from differences in sediment consistency and depth (Gatesy et al. 1999), range from deeply collapsed tracks that preserve hints of the subsurface foot movements, to shallow tracks in thin layers of mud that preserve exquisite impressions of the skin (Gatesy 2001), similar substrate-related changes in track morphology are described from a sauropod trackway where the tracks changes from well-defined to collapsed within the same trackway (Lockley et al. 2002). Small Late Triassic bird-like footprints from La Rioja, Argentina, are interpreted as having formed in and around small ponds, and display lateral morphological and preservational differences according to local changes in the wetness of the sediments around the ponds (Melchor et al. 2002) and Bromley (1996) demonstrated with his own foot, how different his tracks appeared when emplaced on photographic paper, wet sand, dry sand and in the deep sticky mud of an intertidal mud flat. A dramatic example of changes in substrate consistency that not only altered the trackway morphology significantly, but ultimately caused the trackmaker´s death, is described by Bromley (2001) in the trackway of a recent musk ox (Ovibos moscatus) in East Greenland. The 31 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ animal walked into soil liquified by solifluction, where it presumably got stuck and in the attempt to get free, broke or damaged a leg and continued for a while in a broad zig-zagging “trackway” at the end of which the carcass was found. Vertical sections through vertebrate tracks have previously been used by Avanzini (1998) to clarify the walking dynamics and animal-substrate interactions of small theropod dinosaurs, and to investigate how track morphology varies with emplacement depth throughout the track (Jackson 2001, 2002). In further examples, fossil goat tracks in Pleistocene aeolianites from Mallorca are exposed, eroded to different depths and in a variety of preservational states, all displaying different morphologies (Fornós et al. 2002). Using the feet of emus, three experiments with track morphology and undertrack formation are performed, by making tracks in packages of six layers of cement each of different colour, with each layer of 1 cm. in thickness. The aim of these experiments is to investigate how changes in substrate consistency influence the track morphology, using the method developed by Milàn & Bromley (2002b), both on the sediment surface (true track), and the subsurface sediment deformations and undertrack formation. To reveal the subsurface deformations and the undertracks the cement blocks containing the tracks were sliced vertically. In order further to be able to study the formation of undertracks at subjacent horizons, another experiment was carried out using a package of alternating sand and cement layers to allow horizontal splitting of the package. Definitions of undertracks Several theoretical models for undertrack formation have been proposed during the years (Hitchcock 1858; Leonardi 1987; Thulborn 1990; Lockley 1991), while actual experimental work on the factors effecting the formation of undertracks has only recently been conducted (Allen 1997; Brown 1999). The concept of undertrack formation was first to be discussed by Edward Hitchcock (1858) who, during his work with his presumed bird footprints from the Connecticut Valley, New England, described how the same track could be found in stacked successions at several different levels below each other. His figure (Fig. 28) shows his interpretation of how an undertrack is formed below a tridactyl foot. Hitchcock does not use the term undertrack, neither does he coin any other names for the phenomenon in the book. It is however unclear in the illustration which is the 32 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ level that represents the true track and if some of the layers represent the sediment covering the true track. He notes in the text (Hitchcock 1858: 33) that the tracks seem to enlarge downward and that details can appear at lower levels that are not present at higher levels. Another example (Fig. 29) shows how a track is supposed to be transmitted down through a sediment package at an oblique angle if the sediment surface the animal walks upon is sloping. However, although he seems to have got the concept of undertracks basically right, some of his interpretations of undertracks are based on tracks from different animals found on different horizons. Some of the illustrated examples show quite different footprint morphologies at different depths (Hitchcock 1858:plate XIX). That details of the foot can appear and disappear again at different levels suggests that in reality these are tracks from different animals at different times, which happened to be preserved above each other. Especially one exceptionally well preserved stack of slabs which Hitchcock named “Fossil Volume No. 27/4” (Hitchcock 1858: p. 33, plate LII Fig. 6) shows two small tracks repeated on five layers below each other. Hitchcock notes that there seems to be a slight variation in the size and placement of the tracks on the successive horizons, and the tracks on some horizons seem to advance relative to the others, and on another horizon seem to be displaced sideways. These observations suggest that the “Fossil Volume” most likely does not represent a series of undertracks from the same animal, but instead represents a stacked succession of footprints from several different animals at different times. Goldring & Seilacher (1971) investigated the different morphologies of fossil limulid trackways and concluded that the morphological differences were caused by the different sized walking appendages that penetrated the sediment to different depths. When the trackways were exposed at different levels, the morphology would then reflect how deeply the different parts of the walking appendages had been impressed into the sediment (Fig. 30). Goldring & Seilacher (1971) used the term “undertracks” for the part of the tracks preserved below the sediment surface. Leonardi (1987: plate VI) presented very realistic model for undertrack formation, but commented on it only very briefly using the term “subtrace” or “ghost print” (Fig. 31). However, schematic as the model is, it incorporates a lot of accurate features found in tracks and undertracks. Surrounding the track is a raised rim of upward displaced material, formed by the vertical pressure the foot-fall exercised on the sediment. Pittman & Gillette (1989) depicted a cross section of a sauropod track that clearly shows how the weight of the animal has deformed the underlying layers down to a depth of 20 cm below the true track. They coined the term “ghost prints” for these subsurface deformations which, 33 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ as they explain would have resulted in vague but recognizable tracks if they had been exposed at the surface. Thulborn (1990) further developed the models for undertrack formation and presented two scenarios that he termed “underprints” and “transmitted prints” or “ghost prints”. “Underprints” sensu Thulborn (Fig. 32) are produced when the foot penetrates down through the layers in laminated sediments and are filled in with sediment of another consistency. In this model the sediment is supposed to be split at successively deeper levels to reveal less complete sections of the footprint. The other model employing by the two terms “transmitted tracks” and “ghost prints” occurs when the weight of the foot does not penetrate the layered sediments but deforms them under the foot (Fig. 33), thereby making it possible to split the sediment package at successively deeper levels and reveal correspondingly shallower and less detailed versions of the whole footprint. Lockley (1991) further discussed the phenomenon of undertrack formation, which according to him corresponds to Thulborn´s (1990) terms “Transmitted prints” or “Ghost prints”. In Lockley´s (1991) model the undertrack is formed when the weight of an animal deforms the layers below the surface that the animal trod upon, and thus makes it likely to expose the same track at different horizons. This model takes into account the problem with the tracks from animals of different size or animals that bear different weight on their fore and hind limbs. The tracks from a mixed ichnofauna with both large and small animals will only be preserved on the original surface, while only the tracks from the larger animals will be preserved on lower horizons, since the smaller animals fail to deform the lower lying layers, because of their light weight (Fig. 34). The same can happen with a quadroped animal that bears different weight on fore- and hindlimbs like most of the derived ornithopods. Their trackways will seem quadropedal on the original surface but bipedal on the lower surfaces. The first systematic experiments with undertrack formation were carried out by Allen (1997), who conducted exhaustive laboratory experiments with trackmaking for comparison with subfossil ox tracks. The experiments were conducted by impressing a cylinder, roughly the shape of the foot of an ox, into layered packages of plasticine (Fig. 35), which were subsequently sliced vertically. These experiments helped to clarify the origin of the various deformation structures found in association with fossil tracks, although Allen (1997) did not use the term undertracks or related terms to name the deformation structures occurring in the layers subjacent to the track. Neoichnologist Tom Brown (1999), performs what he calls “layer cake” experiments to describe how the “Concentric rings of distortion” around a track progress in three dimensions. 34 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ These experiments are conducted by applying thin alternating layers of soil and flour, in which he afterward steps and carefully cuts vertical slices through (Fig. 36). As the above-mentioned examples show, there have been many contradictional definitions of undertracks through time. The definition of undertracks used by Goldring & Seilacher (1971) is comparable to Thulborn´s (1990) “Underprints”, as they represent the true track exposed at different depths as a result of erosion. The “Transmitted prints” or “Ghost prints” sensu Thulborn (1990), the “Ghost print” sensu Pittman & Gillette (1989) the “Subtrace” or “Ghost print” sensu Leonardi (1987) and the “Concentric rings of distortion” sensu Brown (1999) equals Lockley´s (1991) definition of undertracks as they represent the deformation exercised by the weight of the foot on the layers subjacent to the foot and does not incorporate the true track. Goldring & Seilacher´s (1971) use of the term undertrack does not correspond to Lockley´s (1991) definition of an undertrack. Despite of all this confusion and contradictions among earlier authors, most authors today working with vertebrate ichnology, follow the undertrack definition sensu Lockley (1991). Whenever the term undertrack is used in the following, it will refer to the definition of Lockley (1991). Experiment 1 Four substrate packages were prepared, each containing six layers of coloured cement. The cement was prepared in four mixtures, each having different water content, in order to obtain substrates of consistencies ranging from firm to semi-fluid (Table 5). Package and track number. 1 2 3 4 Millilitre water/kg cement 129 143 157 171 Table 5. The water:cement ratio for the four sediment packages expressed in millilitre of water per kilo cement. Tracknumbers 1-4 corresponds to figure numbers 37-40. The cement was applied in successive layers of about 1 cm thickness to form a package of six layers. To be able to distinguish the individual layers of cement from each other, the layers were coloured with either red or black iron oxide or were left naturally grey. The succession of layers is from the bottom up: grey- black- red- grey- black- red. Next a fresh, severed foot of an emu 35 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ (Dromaius novaehollandiae) was impressed in the package of cement with the approximate weight of an adult emu, 40 kg. A real emu foot was used in the experiments, since it has several advantages over a model. The digits have all the anatomical details found in tridactyl theropod dinosaur footprints, including the small tubercles in the skin, and the configuration of the digital pads around the phalangeal joints. The emu foot is tridactyl, consisting of digits II, III and IV. Digit I, the hallux, which in modern birds is posteriorly directed and used for grasping branches, and occurs uncommonly as posterolateral traces in dinosaur footprints (Irby 1995). It is absent in all the ratites, except for the kiwi (Davies 2002). The terminology of Thulborn & Wade (1989) and Avanzini (1998) is adapted here to describe the foot movement. According to this terminology the movements of the foot during the walk can be divided into three phases (Fig. 12). In the touch-down phase (T), the foot is extended forward, and planted on the ground while the digits diverge. Next is the weight-bearing phase (W), where the metatarsus is moved forward and the animal’s centre of gravity passes over the foot, which becomes impressed into the substrate. Last is the kick-off phase (K). In this phase the proximal parts of the foot are raised and the weight is transferred to the distal parts of the digits as the body moves forward and the foot subsequently is lifted in preparation for a new T-phase. When the foot is lifted, the digits converge and bend backwards to a nearly vertical position while the foot is moved forward. In order to make the tracks as authentic as possible these movements were carefully mimicked with the severed emu foot during the experiments. Because the emu foot is fresh, containing freely moving sinuses and flexible joints, its behaviour during the experiment mimicked a living foot. When weight is applied to the middle digit III in the T-phase, the outer and inner digits II and IV spread out, and when the metatarsus is lifted and moved forward in the K-phase, the digits converge and fold backwards as in the living emu, thus making the footprints very authentic. After hardening of the cement, the blocks containing the tracks were cut in slices 11 mm thick, perpendicular to the length axis of digit III. It was not possible to cut thinner sections without risking that the slices would break when handled afterwards. Eleven representative slices from each block were selected, representing sections from the tip of the claw of digit III to the proximal part of the metatarsal pad. The slices from each footprint were scanned directly on a flatbed scanner at 300 dpi, the glass plate protected with an overhead projector film, using alcohol as contact medium between 36 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ the overhead film and the cement. The occasional unevenness of the thickness of the cement layer is due to difficulties in applying the soft cement in thin layers. The observations of the live emus in the compound treading in various substrates showed that the emus did not change gait or foot posture when treading in soft contra firm substrate. I.e. they remained “flat-footed” and did not, for example tread on tip-toes in soft, sticky mud. Therefore the experiment with a dead foot will provide reliable results. Results Track 1 (Fig. 37) This track was emplaced in the firmest mixture of cement and water. The track is well defined with a high quality of anatomical details preserved. The individual claw imprints and digital pads are clearly defined. There is no skin texture preserved, owing to the stickiness of the cement, which caused the cement to adhere to the sole of the emu foot and thus create numerous small adhesion spikes at the bottom of the print. Skin impressions in fossil footprints are rarely preserved, although they are not uncommon in the right sedimentological contexts (Currie et al. 1991; Gatesy 2001). Slices 1 & 2 Sections through the claw of digit III. In slice 1 only the tip of the claw has left a slight cut in the surface layers. Slice two is through the middle of the claw, which has cut straight through the upper layers, the displaced material having been pressed down into the lower layers, forming a V-shaped undertrack in the middle grey layer. Slices 3, 4 & 5 Sections through digit III. The track walls are vertical and well-defined. The bottom of the digit imprint is flat due to the flattening of the digital pad when impressed in the substrate. The undertracks formed below the digit are rounded in shape and become progressively shallower and wider with depth. In slice 5 the claw of digit II has cut through the upper layer and formed a shallow undertrack in the middle grey layer. 37 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Slices 6, 7 & 8 Sections through the parts of the track with all three digits represented. The digits have formed recognizable undertracks down in the lower red layer. The undertracks becomes progressively shallower and wider with depth. The spike in the imprint of digit IV in slice 7 is material ejected by the claw as the foot were lifted. The depth and width of the track decreases proximally. Slice 9 Section through the shallow interpad space separating the digits from the metatarsal pad. Individual digits are only represented by shallow depressions, the central being the deepest. The shape of the surface track is recognizable in the undertrack at the middle grey horizon. Slices 10 & 11 Sections through the rounded metatarsal pad. The metatarsal pad, despite its deep impression in the true track, has only caused the formation of a shallow undertrack. Track 2 (Fig. 38) The slightly wetter consistency of this mixture caused the track walls to converge after withdrawal of the foot. This partial collapse of the track walls has left a track that at the surface appears to have very slender toes. The number and arrangement of the digital pads and claws are still evident in the track despite the partly collapsed state. Slices 1, 2 & 3 The claw has cut straight through the upper layers. In slice 2 a deep V-shaped undertrack is formed in the lower layers. Minimal upward transport of sediment has made ridges higher than original surface. Slices 4 & 5 Slices through digit III. The track walls gently converge upward in slices 4 and 5, causing the track to appear narrower at the surface than at the bottom. The rounded undertrack in the layers below the digit is deep and rounded in shape and becomes progressively shallower and wider with depth. 38 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Slice 6 The main feature in this slice is digit III which has left a deep, broad impression with slightly converging track walls. The claw of digits II has cut through the upper layers and produced a shallow undertrack in the middle grey layer. Digit IV is only represented as a shallow pinch in the surface and a very shallow undertrack and vide undertrack in the middle grey layer. The undertracks formed by digit III are well-defined. The material displaced by this digit has formed raised rims on both sides of the digit. Slices 7 & 8 Section through all three digits. The track walls in the digit impressions are converged and overhanging. The displaced sediment between the digits has been pressed up between the digits. The rim of raised material between the digits is cracked at the surface parallel to the pressure exercised by the digits. The undertracks are well defined and recognizable down in the lower black layer. Slice 9 Section through the shallow interpad space separating the digits from the metatarsal pad. The undertracks in this section are recognizable down in the lower black layer. Slice 10 & 11 The rounded metatarsal pad has left a bowl shaped depression. The undertracks in section 11 are especially well-developed and consistent in shape and dimensions with the true track, except for becoming shallower and wider downward. Track 3 (Fig. 39) The softness of the sediment caused the track walls to collapse above the digits while the foot was temporarily impressed within the cement. During the lifting of the foot, the material covering the foot was dragged up and forward, forming ridges at the surface in front of digits II and IV and a raised rim around the exit hole for digit III. Anatomical details of the track are hard to recognize in the true track. The existence of claws on digits II & IV is only hinted at by narrow cuts in the surface and the claw impression in digit III is partly covered by the rim of displaced material 39 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ deposited around the digit. The shallower area of the broad interpad space separating the digits from the metatarsal pad is vaguely defined in the true track. During slicing the block containing the track cracked along the length axis of digit III. The vertical white line in the middle of each section is reflections from the glue. Slice 1 The claw has cut straight through the layers, down to the lower black layer. The upper black layer has been dragged upward during the lifting of the digit and forms a raised rim together with the displaced red surface layer. The undertracks along the lower horizons are V-shaped, and become successively wider and shallower with depth. Slices 2 & 3 Sections through the distal part of digit III. The digit has pressed an amount of the red surface layer down in the bottom of the track where it forms a drop like inclusion lined with thin layers of the subjacent layers. The raised rim on both sides of the track consists of both the displaced red surface layer and material from the upper black layer that has been dragged up when the foot was lifted. Undertracks are formed in the lower red and black layers. Slices 4 & 5 The subsurface structures created by the proximal end of digit III are almost similar to the structures produced by the distal part, except that the drop like inclusion in the bottom of the digit imprint is almost detached from the surface layer. The ejected material on the surface originated from material lifted and dragged forward by digits II and IV. A clear undertrack of the digit is formed in the lower red and black layers. Slice 6 The claw of digit IV has cut down through the upper red and black layers and formed an undertrack down in the lower red layer. The deformation produced by digit III is similar to that of the previous slices. Digit II is indirectly represented with a depression of the layers created in front of the digit. The raised rim on the surface is ejected material from the forward movement of the foot. 40 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Slices 7 & 8 All three digits is present in this section, digit II by a deep cut from the claw and digit IV as a broad rounded depression. The undertracks follow the contours of the foot nicely and become successively broader and shallower downward. The surface red layer between the digits has been scraped off during the forward movement of the foot and deposited in front of the digits in slice 5. The true track in slice 8 is relatively better defined than the true track in the other sections, as this part is not covered by cement dragged forward by the digits. Slice 9 The shallow interpad space reveals at the surface faint hints of the division in three digits, while the undertracks especially in the upper black and grey horizons hint rather better at the existence of three digits better. Slices 10 & 11 The rounded imprint of the metatarsal pad is well-defined and not collapsed as the rest of the track. Successively shallower undertracks are formed down to the lower black layer. Track 4 (Fig. 40) The high water content produced a slurry-like, semi-fluid mixture, which caused the cement to flow over and cover the digits while the foot was impressed. During lifting of the foot, the collapsed material was dragged forward and deployed in front of the digits. Subsequent flow of the cement left a true track with very low topographical relief. Slice 1 There is no direct deformation by the digits, but the upper black layer and the red surface layer have been dragged upward by the nearby movement of the digit and are deposited as a raised rim on the surface. Slice 2 The claw of digit III has cut the through the layers down in the lower black layer. The cut is infilled with material from the upper black layer. 41 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Slices 3 & 4 Sections through the distal part of digit III. The digit has down-pressed an amount of the red surface layer, which appears as a drop like inclusion at the bottom of the imprint. The subjacent layers form a thin lining around the depressed red layer. The lower layers appear with a raised rim on both sides of the digit, making the shape of the undertracks appear in raised relief, especially along the lower black horizon. Slices 5, 6 & 7 The digit impression is filled with the upper red surface layer, lined by thin layers of the subjacent black layer. A well-defined undertrack is formed in the lower grey and black layers. A raised rim of the red surface layer, displaced by the forward movement of the foot, is deposited on the surface. Slices 8 & 9 Sections showing all three digits. The deformation caused by digit III is similar to that of slices 5 to 7. In section 9 the impression of digit III becomes shallower which marks the start of the interpad space in the proximal part of the digit. Digits II and IV has formed undertracks down in the lower black layer. The displaced sediment has risen up between the digits above the level of its original surface. The surface material between the digits has been scraped off by the forward movement of the foot, in the case of slice 9 down to the lower black layer. Slice 10 The shallow interpad space between the digits and the metatarsal pad are well defined at the surface, as the proximal part of the track is less collapsed and not as effected by the collapse as the distal end of the track. The undertracks become shallower and wider downwards and is recognizable to the lower black layer. Slice 11 The impression of the metatarsal pad is the deepest part of the track, as is less effected by the collapse and deformation of the track as than the other parts. The undertracks are formed down in the lower grey layer. 42 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Experiment 2 Two tracks in packages of layered cement, produced during the initial phases of the experiments, turned out to be indeed very interesting as failures in the mixtures resulted in unexpected applications. The two tracks are hereafter referred to as track 5 and 6. The mixing and appliance of the cement is identical to that of experiment 1. In this case however the agents used to colour the cement were blue powder-based colour, and red waterbased paint. Results Track 5 The water: cement ratio for the mixture of the sediment in which track 5 produced was 143 mL/kg, similar to that of track 2. The true track on the surface appears well defined with distinct impressions of the individual digital pads and claws (Fig. 41A). The track walls have converged somewhat after removal of the foot due to the softness of the sediment, as was the case in track 2 (Fig. 38). The block containing the track was cut in six slices perpendicular to the length axis of digit III (Fig. 41B). Slice 1 Slice 1 is cut just in front of the claw of digit III, and shows very little disturbance of the layers. Only a slight depression on the surface is present and a small amount of the grey surface layer has been dragged down in the blue layer by the claw, although the claw imprint is not present in the slice. Slice 2 The claw has cut right through the upper layers and pressed the displaced material down in the lower layers creating a V-shaped undertrack in the lower layers. The width of the undertrack is almost double the width of that of the true track on the surface. 43 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Slice 3 Track walls are vertical in this section through the middle of digit III. The layers bend nicely around the digit and are uncut but thinner. The undertrack of the digit in the lower layers becomes progressively wider and shallower downwards. Slice 4 This section shows undertrack formation from all three digits. The sediment between the digits was pressed up due to the vertical pressure exercised on the sediment by the foot. Again the undertrack becomes wider and shallower with depth. Slice 5 Section through the shallow area in the track through the interpad space separating the digits from the metatarsal pad. Individual digits in the undertrack are only recognizable in the upper layer. The undertrack in the lower layers consists of a bowl shaped depression. Slice 6 The rounded metatarsal pad has left a V-shaped depression at the surface. A less depressed undertrack is formed in the upper blue layer. The lower layers are deformed in the same bowl shaped way as in slice 5. Undertracks The water-based paint used to colour the red cement layers had an unforeseen effect on the cement. The red layers never really hardened and would easily crumble when touched by hand. This enabled the red layer to be dissolved in water and removed, revealing a complete subsurface horizon with the undertrack exposed. Both the cast of the undertrack in the former red horizon (Fig. 41C), and the undertrack in the lower grey layer (Fig. 41D) is revealed in this way. The cast of the undertrack represents a surface approximately 2 cm below the tracking surface. The digits are clearly recognizable in the cast as they are the deepest impressed in the cement, compare with slices 1-6 (Fig. 41B). Individual features of the digits like digital pads and claw imprints are not recognizable in the undertrack. The metatarsal area is only represented in slight relief but still recognizable. At the surface at approximately 4 cm depth, the undertrack is much less well defined, but digit III is still represented by a prominent depression, although 44 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ significantly wider than in the true track. Digits II and IV are harder to recognize, but shallow depressions in the cement hint at their existence. The overall track shape is hard to recognize at this depth. Track 6 The water: cement ratio for this sediment is 171 mL/kg, identical to the mixture used in track 4. The true track as it appears at the surface is collapsed due to the softness of the cement (Fig. 42A). The soft cement that collapsed over the digits during the depression of the foot, was dragged forward when the foot was lifted and deployed in front of the digits, forming mounds in front of digits II and IV and a rounded mound in front of and around digit III. Four slices cut through the block containing the track reveal how the subjacent layers are deformed and the undertracks are formed (Fig. 42B). Slice 1 Section showing the part of the track where the claw of digit III has cut through the layers. The claw has cut through the three upper layers and formed undertracks down in the lower layers. During lifting of the foot an amount of the soft cement has been dragged upwards causing the formation of a raised rim in the upper blue layer on both sides of the digit. Slice 2 The up- and downward movement of the digit has produced a zone of vertical mixing of the layers. An amount of the upper layers appears as a raised rim on both sides of the digit and in the undertracks. A “drop” of the red surface layer is enclosed in the bottom of the digit imprint. A wide undertrack is formed in the bottom blue layer. Slice 3 The claws of digits II and IV have cut the upper layers, forming narrow imprints in the sediments. The structures formed by digit III are similar to those in slice 2. A clear undertrack showing all three digits has been formed in the lower red and blue layers. 45 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Slice 4 A rounded depression in the surface represents the rounded metatarsal pad. The shallow interpad spaces between the digits have formed a shallow undertrack with vague impressions of the three digits. Undertracks As in the case of track 5, the red water-based paint used to colour the cement, prevented the cement from hardening properly and made it possible to dissolve it in water afterwards. The hereby exposed surface at approximately 4 cm depth (Fig. 42C), displays a clear undertrack comprising three well defined digits. Digits II and IV are the least well-defined but in digit III the claw impression is recognizable and even to a small degree the division of digital pads. Experiment 3 In order to create a package of sediment that would be split along successive horizontal planes, the experiment using the mixture used for track 1 (129 mL water/kg cement) was repeated, this time in a package of five alternating layers of wet sand and cement, each of 1 cm thickness. The sand was admixed with the same quantity of water than the cement. After hardening of the cement, the sand layers were washed out and the missing layers recreated with “Wacker Silicone” moulding rubber. This enabled the package to be split along four horizontal planes, each exposing an undertrack at a different horizon subjacent to the true track and thereby allowing an investigation of the loss of information occuring with increasing depth below the substrate surface. Results Tracking surface. The true track (Fig. 43A) is well-defined, with a high level of anatomical details preserved. The claw imprints and the imprints of the individual digital pads are well preserved. In the proximal part of the track, in and around the metatarsal pad, small patches with impressisons of the tuberculous skin are preserved. 46 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ 1cm depth At 1 cm below the surface (Fig. 43B) a clear undertrack has been formed. Most features of the true track, such as claw imprints and digital pads, are visible in the undertrack. The radial distribution of the pressure from the foot causes the undertrack to appear wider and more rounded in shape. 2 cm depth The undertrack at 2 cm depth (Fig. 43C) is less well defined although still recognizable as a track. The division of the digital pads is still faintly seen in digit III. Claw impressions are faintly recognizable in all three digits and the round metatarsal pad is still a prominent feature of the track. 3 cm depth At this horizon (Fig. 43D) the undertrack is very shallow. While digit III is almost impossible to distinguish, digits II and IV and the metatarsal pad are still faintly recognizable. 4 cm depth At 4 cm depth (Fig. 43E) only a very shallow and faint impression of the metatarsal pad and digits II and IV are distinguishable at this horizon. Discussion The thickness of the cement package prevented the foot from penetrating the cement to deeper levels than were observed, which means that the experiments must be compared with a sedimentary environment consisting of a six cm layer of soft mud on a firm base. This situation is not uncommon in track bearing sedimentary environments, which often consists of floodplain or similar deposits in connection with fluvial systems (Nadon 1993, 2001), and can thus be considered as models of realistic tracking environments. In fact the bottom layer, supported on the uncompressable base, has resisted deformation and its upper surface has acted to some extent like a tectonic décollement surface. This is most clear in figs. 38-40. Here, the overlying layers are well deformed and the undeformed contact with the bottom layer must have acted as a basement thrust surface on a very small scale. 47 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ By comparing the results from experiments 1 and 2, section by section in the different substrate consistencies, it becomes clear how great an influence the substrate consistency has on the morphology of both the true track and the formation of undertracks. Where the true track is welldefined in track 1 (Fig. 37) and in the track from experiment 3 (Fig. 43A), a high degree of anatomical details are preserved, such as number and arrangement of digital pads. With a slightly higher water content the track walls converge after the foot is lifted, leaving a true track which displays apparently much thinner and more slender digits (Figs. 38 & 41) than the tracks from firmer substrates. The tracks obtained with higher water content in the mixtures, collapse during the stride that causes the material covering the digits to be ejected and deployed in front of the digits, blurring the shape of the true track. In firm sediments (Figs. 37, 38, 41 & 43) the force applied to the substrate by the foot is transferred radially outwards, giving the undertracks a wider and more rounded appearance than the true track. Similar observations of track and undertrack morphology was obtained by Allen (1997) by impressing a cylinder into coloured plasticine. In softer sediments, the track walls collapsed during the formation of the tracks, leaving true tracks with very few details preserved. However, the sections through the tracks emplaced in soft and semi-fluid cement (Figs. 39, 40 & 42), show that the undertracks formed below the level of the collapsed track walls. These undertracks actually retain the shape of the foot better than the true track, which is collapsed and disturbed. This is best seen in track 6 (Fig. 42C), which represents an undertrack exposed at a surface of approximately 4 cm depth. The succession of undertracks obtained by experiment 3 (Fig. 43A-E) clearly demonstrate undertrack formation in firm sediments, where the undertrack becomes successively wider, shallower and less detailed with depth. That undertracks can differ significantly from the true tracks and in some cases directly cause misidentification of the trackmaker is not a phenomenon that is restricted to vertebrate tracks. Arthropod trace fossils like limulid tracks, at different states of erosion, and when exposed as undertracks have been misidentified on several occasions as tracks of vertebrates (Goldring & Seilacher 1971). The morphological variety of the experimentally produced cement tracks bears a strong resemblance to the Triassic theropod tracks form Jameson Land, East Greenland. These were interpreted as having been emplaced in mud of different thicknesses and water content, and accordingly displayed a range of preservational states from clearly recognizable with impressions of the skin preserved (Gatesy 2001), to deeply collapsed tracks formed in thicker semi-fluid mud 48 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ (Gatesy et al 1999). Similar collapsed and semi collapsed theropod tracks are known from an Early Cretaceous littoral mudflat in Greyssac, France (Griffiths et al. 2002). In both cases the resemblance to the herein obtained results is striking and this indicates the validity of this experimental procedure. Conclusion Changes in water content, and thus the firmness of the sediment, dramatically alter the appearance of both true tracks and undertracks. Although not a naturally occurring sediment, footprints formed in packages of coloured cement can provide reliable examples of changes in the morphology in tracks and undertracks as the sediment properties change. The similarity of the herein described experimentally obtained tracks and theropod tracks from sediments of various properties, proves the value of this method to experimentally investigate the connection between sediment consistencies and the resulting differences in track and undertrack morphology. This connection between sediment properties and track morphology is important to bear in mind when studying fossil footprints. This is because local changes in sediment properties do occur, and tracks often are found exposed to various degrees, in different eroded states, or as undertracks at various subjacent horizons. By applying experimental methods such as these it is not only possible to predict the shape of tracks in different taphonomical states, but also to determine the consistency of the sediments in which tracks are found. Horizontal sections of an emu track and a theropod track exposed in horizontal and vertical view. Introduction Fossil footprints should not be regarded as only two-dimensional structures. In many cases the tracks are formed in sediments soft enough to allow the trackmaker’s foot to sink to a considerable depth, capturing not only the shape of the foot, but also parts of the foot movements as described by Gatesy et al. (1999). If the sediment is firm enough to maintain the shape after the foot has been lifted, infillings of the tracks can produce complex 3D structures with dramatic changes in overall track morphology according to the depth at which they are exposed. 49 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ In field exposures of fossil tracks, the bedding plane in which they were imprinted has commonly been eroded to various degrees, and may thus not reflect the actual surface the animals walked on. Thulborn (1990) used the term “underprints” to describe the case where a track with its natural cast still in place is exposed eroded to different levels. The appearance changes with depth since different parts of the foot are impressed to different levels in the sediment. The aim of this study is to investigate the depth-related changes in footprint morphology, by cutting horizontal sections through the cast of an emu (Dromaius novaehollandiae) track formed in deep sticky mud, and by studying a Middle Jurassic theropod track from the Scarborough Coast, England. The theropod track is exposed both as an eroded horizontal part and in vertical section. Methods In order to investigate the effect of depth on footprint shape, plaster casts of deep emu tracks were cut in horizontal sections. Sediment of a consistency rigid enough to allow the formation and preservation of tacks of usable depth were prepared by admixing the thick organic rich soil in the emu compound with water, to produce a thick sticky mud. Afterwards the emus were encouraged to walk through the prepared mud. Casts of selected tracks, representing normal progression were made using Plaster of Paris. The track selected for sectioning is a right pes and has digit III most deeply impressed to a depth of 4.5 cm, followed by the metatarsal pad. The impression of Digit II is in this case slightly deeper that digit IV. To simulate the condition where a track is preserved in a sedimentary package with the sedimentary infilling (natural cast) still inside, the cast of the track was imbedded in plaster of a different colour to allow discrimination of the track and the surrounding matrix. This made it possible to cut horizontal sections through the tracks, to simulate the condition where the track with its sedimentary infilling is exposed eroded to different depths, and thus to describe the morphological differences occurring in the track with depth. The width of the blade of the saw used was 3 mm, leaving the distance between every other section 3 mm, while the other sections were maintained at approximately 10 mm to prevent breakage of the slices. The pictures of the two halves of the slices were flipped mirror wise; in order to present the surfaces as cuts seen from the same angle. The following measurements were taken for each section: Track length, measured from proximal end of the metatarsal pad to tip of digit III. Track width, measured between the tips 50 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ of digits II & IV. Width of digits, measured in the middle of each digit. Results are shown in Table 6. Results The cast of the track (Fig. 44A) is well defined, and shows evidence of the dynamic movement of the foot during the stride, evidenced by sloping of the track walls in the metatarsal area, resulting from the forward movement of the foot in the touch-down phase (T-phase sensu Thulborn & Wade 1989). A long drag mark from the claw of digit III is projecting distally away from the tip of the digit, originating from the forward movement of the foot while it is lifted towards a new T-phase. The first section (Fig. 44B) is cut approximately 3 mm below the original sediment surface. The shape of the track is partly obscured because of unevenness in the original sediment surfaces, making it difficult to obtain measurements. Because of the drag mark from the claw of digit III, the track extends beyond the edge of the slab. In section 2 (Fig. 44C) the length of the drag mark still prevents measurements of the track length itself. Sections 3 & 4 (Fig. 44D, E) show the track with a complete outline. The track width does not change in the two sections, but the track appears 7 mm shorter and digit III becomes 1 mm thinner from section 3 to 4. In sections 5 & 6 (Fig. 44F, G) the track becomes divided as the sections go through the deep interpad spaces dividing the digital pads from the metatarsal pad. In Section 5, digit IV is disconnected from the metatarsal pad while digits III and II still are connected to the metatarsal pad by a thinner area. Section 6 shows complete separation of digits II and IV from the metatarsal pad and digit III is still connected by a narrow area. Digit III becomes separated from the metatarsal pad in sections 7 & 8 (Fig. 44H, I) and the impression of digit IV disappears in section 7, as it was less impressed in the substrate than digits II & III. Digit II, however, is only present with a small impression in section 7, and disappears completely in section 8, which represents the very bottom of the track, where only the distal part of digit III and the metatarsal pad are represented, being the parts of the foot carrying the most weight during the stride. 51 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ Section 1 (Fig. 44B) 2 (Fig. 44C) 3 (Fig. 44D) 4 (Fig. 44E) 5 (Fig. 44F) 6 (Fig. 44G) 7 (Fig. 44H) 8 (Fig. 44I) Depth below Footprint surface, mm length, mm 3 6 14 17 25 28 38 41 230+ 230+ 211 204 199 196 189 185 Footprint width, mm 181 181 174 174 149 140 69 33 Width digit II middle, mm 34 26 26 24 24 9 - Width digit III middle, mm 45 44 44 43 42 42 36 33 Width digit IV middle, mm 34 34 32 30 29 23 - Table 6. Measurements of sections 1-8. Footprint length is measured from the proximal end of the metatarsal pad to the distal end of the claw of digit III. The footprint length in sections 1 & 2 are in excess of the length of the slab due to the elongated drag mark from the claw of digit III. The digit width is measured from the middle of the digits. The missing measurement in section 1 is due to unevenness in the sediment surface which obliterated the shape of digit II. Missing measurements in sections 7 & 8 are due to the different penetration depths of the digits in the substrate, compare with Fig. 44A-I. Theropod track exposed in horizontal and vertical section A Middle Jurassic theropod track found approximately 2.6 km north of Scarborourgh, just south of Crook Ness Steps, at the northeast coast of England (Fig. 45), is preserved with the natural cast still filling the print. The specimen was found in-situ exposed near the waterline. The track occurs in finegrained sandstone and is infilled with similar sand but of a slightly lighter colour, allowing discrimination of the track from the surrounding rocks (Fig. 46). The track is that of a medium sized theropod as evidenced by the long slender toes with clear claw impressions, the total length of the track including the parts covered by overlying sediments is estimated to 35 cm. Only two toes are visible when seen on the horizontal surface (Fig. 46A, B). The length of the left digit exposed and the position and size of the right digit suggests that the digits visible are digits II & III or III & IV. The claw of digit III in the track is turned right compared to the direction of the digit, which suggests that the track originates from a left foot, as the claw in digit III in most theropod tracks is turned inward toward the midline of the trackway (Lockley 1991), and in this case identifies the exposed digits as digit II & III. The rock containing the track is cut by a crack, along which the rock on one side is eroded to a horizontal plane exposing the distal parts of the track. The rock on the other side of the crack is exposed as a higher horizon, above the track. This difference is erosion depth has produced 52 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ a natural vertical section of the track (Fig. 46C). The vertical section of the track reveals that the shape of the track changes with depth, as the track walls are partly collapsed above digit III, which would give it a much narrower appearance if exposed at a higher horizon. Digit IV is only present in the vertical section of the track as the horizontal surface does not represent an actual bedding plane, but is a random erosional plane running diagonally through the track. Discussion The emu track sections show that the area of the tracks is significantly larger in the sections cut near the surface, than in sections from the bottom of the print. In this case the firmness of the mud prevented the track walls from collapsing after withdrawal of the foot. The larger track outline in the upper track sections is a combination of two factors. The first is that the pressure transmitted to the sediment by the foot is directed radially away from the digits, giving the deformation a horizontal as well as a vertical component (Allen 1997). The second factor is the forward movement of the foot during walk, which in deep sediments causes the track walls in the distal and proximal parts of the track to appear sloping. The area around the metatarsal pad and the proximal end of the digits tend especially to increase in the surface layers of the track. This is because of the narrow gap between the proximal end of the digits, which, when the foot is moved forward and up, will push the surrounding sediments causing the track to appear larger in that area. The theropod track (Fig. 46A-C) shows a different change in track outline with depth, as the sediments were partly collapsed over the digits before infilling occurred. Had the horizontal section of the track been exposed at a higher level, all three digits of the track would have been present, but the digits would have appeared significantly thinner due to the partial collapse of the track walls. Conclusion There is a significant difference in the shape of both the investigated emu- and theropod track, according to the level at which the track is exposed. In both cases this is due to the properties of the sediments in which the tracks were emplaced. The emu track, made in deep sticky mud, becomes narrower with depth and the individual digit impressions become detached from each other nearer the bottom of the print and the length and width of the track decreases significantly with depth. The theropod track, if exposed at the original tracking surface, would have displayed much narrower digit impressions than those here exposed by the eroded part of the track, and would 53 Part 3. Laboratory experiments with track and undertrack morphology ________________________________________________________________________________ also have occurred tridactyl as the impression of digit IV would have been present, as evidenced by the vertical section. This study has shown that depth-related changes in the shape of tracks occur and can cause the same track to significantly alter its appearance with depth. For this reason, it is important to bear in mind when describing fossil footprints, in which the sedimentary infill still is in place, that the track might appear different if exposed at different levels. This effect is particularly troublesome in relation to ichnotaxonomic treatment of tracks. 54 Part 4. Fossil tracks and trackways ________________________________________________________________________________ Part 4. Fossil tracks and trackways Theropod tracks from Greenland Introduction Four Upper Triassic theropod footprints of the ichnogenus Grallator, from the upper part of the Fleming Fjord Formation, Jameson Land, East Greenland have been examined, with special attention paid to the preservation of anatomical details and to the properties of the sediments in which they are impressed. When an animal walks, the weight of the animal not only affects the surface it is walking upon, but subjacent horizons also are subject of deformation caused by the pressure from the foot transferred radially outwards during the footfall (Allen 1997). To investigate this effect a theropod track from a laminated mud/silt stone is sectioned vertically, in order to reveal possible undertracks and other subsurface deformations caused by the trackmakers foot. The tracks The tracks were collected by Lars Clemmensen and Niels Bonde from the track bearing beds of the upper part of the Fleming Fjord Formation, where several tracks and trackways have been found (Jenkins et al. 1994, Clemmensen et al. 1998, Gatesy et al. 1999, Gatesy 2001). In the following the two tracks collected by Niels Bonde will for be termed NB1 and NB2, and the two tracks collected by Lars Clemmensen will be termed LC1 and LC2. The Upper Triassic Fleming Fjord formation consists of a well-exposed 200-300 m thick succession of lake deposits. The upper part of the formation, the Carlsberg Fjord beds of the Ørsted Dal Member is 80-115 m thick and is composed of structureless red mudstones, rhythmically broken by thin greyish siltstones (Clemmensen et al. 1994). The thin siltstone beds represent episodes where the mudflats were flooded by lake water of a depth sufficient enough to allow formation of small wave ripples. The flooding was followed by periods of subaerial exposure and desiccation. The tracks of the formation are all found at these horizons, on which numerous small theropod footprints are preserved as well as rarer footprints of larger prosauropods (Jenkins et al. 1994; Clemmensen et al. 1998; Gatesy et al. 1999 & Gatesy 2001). 55 Part 4. Fossil tracks and trackways ________________________________________________________________________________ Grallator footprints Grallator (Fig. 47) is one of the oldest scientifically named tetrapod ichnogenera, first described by Hitchcock (1858) as the footprints of large flightless birds under the name Ornithichnites. The ichnogenera spans from the Late Triassic throughout the Jurassic, and originates presumably from several different small to medium sized theropods. Grallator tracks are mostly tridactyl, consisting of digits II, III & IV In rare occasions traces of digit I, the posteromedially orientated hallux, is preserved. Since digit I is situated on an elevated position on the metatarsus (Christiansen 1997), impressions of the digit mostly occur in tracks sunken deeply into sediments as those described by Gatesy et al. (1999). Grallator, together with the larger ichnogenera Anchisauripus and Eubrontes, is now believed to constitute an ontogentic series rather than the tracks of different trackmakers, as the size differences in the tracks is according to the changes in foot dimensions through allometric growth (Olsen et al. 1998). The individual digits in Grallator footprints have well defined digital pads and impressions of long, slender claws, nicely reflecting the phalangeal skeleton inside (if the preservation allows recognition of them). Digit II, which consists of three phalanges, has two prominent pads covering the joints. The three phalangeal pads cover the four phalanges in Digit III. The claw of digit III is offset towards the midline of the trackway. Digit IV consists of four to five small phalangeal pads, in most cases fused to a lesser number due to the shortness of the phalanges. Contrary to the emu footprints and the other theropod footprints with more birdlike affinities, Grallator footprints do not possess the pronounced pad covering the phalangeal – metatarsal joint; instead, the proximal pads of digit IV extend backward of digits III and II, making the proximal end of the track asymmetrical. The claw of digit III is always directed inward towards the midline of the trackway (Lockley 1991; Farlow et al. 2000). The asymmetry of the proximal end of the foot and the orientation of the claw of digit III are important indicators of right or left feet if only single footprints are examined, as these characters often can be recognized in even weathered and eroded tracks. The makers of Triassic Grallator tracks are supposed to be small to medium sized Coelyphysis - like theropods. NB1 At the surface the track appears shallow (14 mm at the deepest) and at first glance hardly recognizable (Fig. 48). This is due to erosion and to the fact that parts of the track still retains some of the infilling material. The track is approximately 234 mm long and 140 mm wide, with an 56 Part 4. Fossil tracks and trackways ________________________________________________________________________________ divarication angle of 56 degrees, the shallowness and the eroded state making accurate measurements difficult to obtain. When used in the formula developed by Alexander (1976), the length of the footprint suggests that the trackmaker had a hip height of 936 mm, from which an estimated length of the animal, supposed it fits the Coelyphysis bodyplan, would be around 4 m (P. Christiansen pers. comm. 2002). No division of phalangeal pads is visible and the shape of the track appears rounded with short, blunt digits. The outline, however, has retained the slight turn of the claw of digit III and the asymmetry in the proximal end of the track, identifies the track as a right pes. A raised rim of displaced material is present at the proximal part of digit II, suggesting the trackmaker made a slight sideways movement of the foot during progression. Lateral movements in the proximal parts of theropod feet during walk, have previously been described by Avanzini (1998), and in that case the movement occurred inward contrary to track NB1 in which the movement was outward. The slab containing the track is a reddish mudrock 45 mm thick on average. The surface of the slab is cut by weathered polygonal dissection structures and is covered by a thin haematite coating. The underside is covered with numerous small Diplichnites trackways preserved in hyporelief, and the arthropod resting trace Rusophycus, indicating a fluviatile and possibly lacustrine environment before the track-bearing layer was deposited (Bromley & Asgaard 1979). NB2 Track emplaced in deep soft mud, which has been pressed up between the digits and collapsed after withdrawal of the foot, causing identification of individual characters of the track to be difficult (Fig. 49). The rear margin of the track is not preserved as the part containing it was not collected. The outline of the digits appears blurred on the surface. Interestingly, a backward-projecting impression is present behind the supposed digit II and indicates the presence of digit I. Due to its elevated position on the metatarsus, the hallux only touches the ground if the foot sinks to a certain depth in the sediments, as illustrated by Gatesy et al. (1999). If the projection is in fact digit I, then the track is from a right foot. Due to the missing rear margin of the track, the track length is estimated at 160 mm, giving a trackmaker with a hip height of 640 mm. The divarication angle between digits II & IV is 75 degrees. 57 Part 4. Fossil tracks and trackways ________________________________________________________________________________ The slab containing the track is composed of reddish-brown laminated siltstone and has small current ripples on the upper surface. The track has fractured along the length axis of all three digits making it possible to investigate the sediment deformation around the digits. The sediments are directly disturbed and mixed by the foot to 45 mm depth and undertracks are formed to 56 mm depth. The most interesting feature found in connection with the track is the presence of striations in the track walls of digit III (Fig. 52A, B). The striations originate from the small skin tubercles that covered the sole and sides of the digital pads of small theropods (Gatesy 2001). The width of the striations suggests a size for the skin tubercles of 1.5-2 mm, very similar to those found in the extant emu (Fig. 52C) The striations demonstrate a forward and upward movement of the foot and were created as the foot was dragged up through the soft mud. Skin impressions and striations in dinosaur tracks are of rare occurrence, as they are delicate structures that are preserved only in sediments of the right consistencies, and easily are destroyed by erosion. Impressions of tuberculous skin and striations produced by the tubercles as they slide through the sediment are previously known from ornitopod tracks (Currie et el. 1991), in tracks observed in vertical cliff sections (Difley & Eckdale 2002) and from thyreophoran tracks of probable ankylosaur affinity (McCrea et al. 2001; M. A. Whyte & M. Romano pers. Comm. 2002). Theropod tracks with skin impressions are previously only described from tracks collected in the same area as the tracks described herein (Gatesy 2001), but Hitchcock (1858, plate X) depicted what appear to be skin impressions in the metatarsus area of a Brontozoum (Eubrontes) giganteus track, without mentioning them in the text. LC1 This track originated from a left foot, identified by the asymmetry in the proximal end and the orientation of the claw of digit III (Fig. 50). The length is 213 mm and the width is 149 mm, suggesting a trackmaker with a hip height of around 850 mm. The divarication angle for digits II & IV is 50 degrees. The track was emplaced in relatively shallow mud, which has caused the digits to be only shallowly connected. The individual digital pads are recognizable in digits II and III, while digit IV still contains some of the covering sediments, hindering identification of individual digital pads. Impressions of long slender claws are present in digits II & III while the proximal part of digit IV is partly infilled with clay. A prominent rim of material displaced by the digits is present 58 Part 4. Fossil tracks and trackways ________________________________________________________________________________ between digit IV and III, caused by the upward and forward movement when the foot is lifted. Low rims of displaced material are present along the outline of the track. LC2 The slab contains the distal parts of digits II, III and IV of a small theropod track (Fig. 51). The sediment is finely laminated red and yellow-brown mudstone, 130 mm thick in the part containing the digit impressions. On the tracking surface two systems of dissication cracks are present, one system of small polygons 10-20 mm in size and a larger system of polygons 100-150 mm in size. A raised rim of displaced material surrounds the distal parts of the digits. The claw of digit III is set off to the right of the length axis of digit III indicating that the track is a left pes. Only the distal digital pad and a small faint claw mark of digit II are present. Digit III has three well-defined digital pads, the distal one bearing a clear claw impression. Only the medial part of digit II is present, there being a possible division in digital pads present. No parts of the metatarsal pad or the proximal end of the digits are present in the slab, but by extrapolating the length of the digits a total footprint length of 160 mm is estimated, which implies a trackmaker with a hip height of 640 mm. By extrapolating the length axis of digits II & IV backwards a divarication angle of 43 degrees between the digits is obtained. A shallow but well-defined undertrack containing all three digits is present on the underside of the slab that represents a horizon 130 mm below the tracking surface. No claw traces are present in the undertrack, but what appears to be the middle part of each digit has left the deepest impression as undertrack; it is also evident that the middle digital pad in digit III is the deepest impressed in the true track. This suggests that the trackmaker carried the most weight on the middle of the digits during the K-phase. Vertical sections through track NB1 Methods In order to investigate the formation of undertracks and other subsurface deformation structures resulting from the footfall, the slab containing track NB1 was cut in vertical slices perpendicular to the length axis of digit III. The saw cuts were 2 cm apart, after subtracting the 3 mm thickness of the saw blade, each slice is 17 mm thick. The slices are numbered 1 to 11 counting 59 Part 4. Fossil tracks and trackways ________________________________________________________________________________ from the proximal end of the metatarsus area, to the tip of digit III (Fig. 53). Subsequently, each slice were polished and submerged for 20 minutes in acetic acid in order to enhance the surface contrasts of the slices and allow examination of the fine structures. Each section was afterwards digitally photographed. Digital colour and contrast enhancing was subsequently done using Corel Photo-paint 10. The slab consists of three distinct layers. The bottom of the slab consists of 8-10 mm of plane-laminated reddish to brown mud, above that is a 22 – 24 mm layer of small-scale (4 –5 mm) current ripples. The 15-18 mm upper layer is almost structure-less with faint laminations, and occasional quarts grains in mm size. Results Section 1 The very proximal part of the metatarsus area has caused a slight depression in the surface of the slab. Undertrack formation is vaguely present in the upper weakly laminated layer. Section 2 The true track is formed down in the upper weakly structured layer, and a very shallow undertrack is present in the middle current-rippled layer. A raised rim of displaced surface material is present at the right side of the footprint, originating from a sideways and outwards movement of the foot during the stride. Sections 3-5 The bottom of the track is flat, revealing no division of the digits. The prominent, raised rim of displaced material is revealed in section to be a deep structure, bending and partly mixing the layers down in the middle current-rippled layer. Section 6 This marks the transition where the digits diverge from each other. No signs of the individual digits are present at the bottom of the track, but the undertrack along the surface of the current-rippled layer shows the initial division between digits II & III. The sediment layers at the 60 Part 4. Fossil tracks and trackways ________________________________________________________________________________ right side of the track are disturbed and partly mixed by the sideways movement of the foot, making identification of individual layers difficult. Section 7 The bottom of the track now shows the initial division between digits II & III. The section further reveals that digit II is partly filled with sediment and thereby hardly recognizable at the surface. Digits III & IV are not divided at the bottom of the track, but the undertrack reveals the division as a raised area pressed up between the digits. A prominent fault, cutting the layers, is present projecting down- and outward from the bottom of the right side of the footprint. Section 8 The impression of digit II is hardly recognizable at the surface, as it is filled with sediment and only appears as a broad, shallow area. Digit IV divides from digit III at the surface. The undertrack formed along the top of the current-rippled layer show clear impressions of all three digits. Sections 9 -11 These sections only show impressions of digit III. The bottom of the digit impression is cowered by a thin layer of the sediment fill. A prominent rim of displaced material is present on either side of the digit. Section 9 A shallow undertrack of the digit is present in the current-rippled layer of the slab. The rim of displaced sediment on either side of the digit is also recognizable in the undertrack. Section 10 The layers below the digit seem to be disturbed on the right side, suggesting a sideways movement of the digit during the kick-off. The disturbance is present down to the currentrippled layer. 61 Part 4. Fossil tracks and trackways ________________________________________________________________________________ Section 11 This section passes through the tip of the claw impression of digit III where the slender, sharp claw has cut down through the sediment and left a narrow hole, traceable down in the current-rippled layer. The cut from the claw is filled with the same, slightly lighter coloured, sediment as the bottom of the other sections of digit III. The upper part of the imprint is broader and more rounded at the bottom, suggesting that the foot was emplaced flat on the ground, and the deep cut from the claw originated from the kick-off. Discussion The tracks described above reflect sediments having four different properties. These properties have caused the tracks to display different appearances depending on the differences in substrate consistency. Track LC1 is shallow but very well defined, containing many recognizable features such as claws, digital pads and the division of the digits. LC2 is shallow and has very well defined impressions of the individual digital pads and claws. The weathered, and partly infilled state of track NB1 hinders any inferences about the originally preserved anatomical details in the track. However examination of the vertical slices of the track shows that the true track is relatively shallow. Except for the claw imprint, which protrudes down to 22 mm, the track does not exceed 16 mm in depth, as measured from the present sediment surface. The deep soft sediment of track NB2 has caused the track walls to collapse, blurring the shape of the track and causing significant sediment disturbances between the digits. However, the deep mud, which is responsible on one side for obliterating the shape of the track, has on the other hand preserved some of the three dimensional movement of the foot. The preserved striations from the skin tubercles on both sides of digit III show how the foot was moved forward and upward in the stride. The divarication angles between digits II & IV in the four tracks lies well within the normal range for theropods as documented by Farlow et al. (2000), with the value of 75 degrees for track NB2 in the high end of the range reported for theropods. At first glance NB1 could be suspected to be an undertrack because of its shallow appearance and the vaguely defined rounded shape of the digits. However, the presence of infilling of a slightly different colour and especially the deep cut from the claw of digit III revealed in 62 Part 4. Fossil tracks and trackways ________________________________________________________________________________ section, proves that the track is in fact a true track. The rounded appearance of the track at the surface must be the result of secondary erosion after the track had been exposed at the surface. The asymmetry of the footprint with the raised rim of displaced material on the right side of the proximal part of the track, in fact displays many similarities with the tracks described by Avanzini (1998). In that case the asymmetry occurred in the opposite side of the track. Avanzini (1998) used this deformation to infer about the walking dynamics of theropod dinosaurs and incorporated a slight inwards turn of the foot in his reconstruction of theropod foot movements during a stride. The sections herein described, directly contradict this model as the deformation here occurs in the opposite side of the footprint. An origin of the deeply disturbed fault-like structure in the sediments below the raised rim of displaced material could be ascribes as digit I (the hallux) being dragged through the mud. However, this can be ruled out as the deformation occurs in the side of the track opposite to the expected position of the hallux, which in theropods is posteromedially orientated. Furthermore the structure goes beyond the level of deformation caused by the foot, which contradicts the elevated position of the hallux in theropods (Christiansen 1997). A possible origin of the structure is that it is a fault formed during the formation of the raised pad of displaced material caused by a sideways and downwards movement of the foot during the stride. That microfaulting occur in sediments in connection with track formation has been established experimentally by Allen (1997) and Jackson (2002). However, the uniform bending of the layers down into the structure (slice 7) does not correspond with what would be expected of a fault. Instead the architecture suggests that the structure is in fact similar to the structures termed Cave and Cave-in by Brown (1999:60-61). Cave is the situation where a slight turn of the trackmakers foot during the stride causes the foot to slightly undermine the track wall, leaving parts of the track wall as an overhang, whereas Cave-in is the situation where the overhang afterwards collapses over the Cave. The differences in foot length of the four tracks described herein, suggests that they represents tracks from theropods of three different sizes. NB2 and LC2 are from a small animal with a an estimated hip height of around 640 mm, and tracks NB1 and LC1 are produced by larger animals with hip heights of 850 and 930 mm. Only one partial theropod skeleton of uncertain affinities has been found in the formation (Jenkins et al. 1994), but its femur length of 330 mm suggests a loosely estimated hip height of 650 mm, corresponding well with the estimated size of the trackmaker responsible for tracks NB2 and LC2. Tracks NB1 and LC1 were produced by larger 63 Part 4. Fossil tracks and trackways ________________________________________________________________________________ theropods, and can hereby supply the skeletal record with data about contemporary theropods of either other species or ontogenetic stages. Conclusion The originally different sediment properties exercised a strong preservational bias of anatomical details in the four footprints. The tracks made in shallow mud, LC1 and LC2, are welldefined and show preservation of a high degree of anatomical details. In contrast track NB2 show few details in terms of number and arrangement of digital pads and claw impressions preserved, but instead the sediment was of a consistency sticky enough to preserve delicate details such as the movement of the skin tubercles on the track walls of digit III. The method of examining vertical sections of slabs containing tracks can provide much information unobtainable from study of the true tracks. In the case described herein the sectioning of the track could confirm that the track was a true track and also provide additional data about the length and width of the digits. Successful sectioning of tracks is strongly depending on the lithology of the rock. Undertracks and other deformations are best observed in well-laminated rocks, whereas a homogenous rock is unlikely to provide much information in section. A new track assemblage from Porto das Barcas, Lourinhã, Portugal Introduction In the autumn of 2001, a new small exposure of a track-bearing bed was discovered in the coastal cliffs of the Upper Jurassic Lourinhã Formation at Porto das Barcas near Lourinhã, Portugal, approximately 70 km north of Lisboa (Fig. 54). The part of the Lourinhã Formation exposed at Porto das Barcas is Kimmeridgian to Tithonian. The Lourinhã Formation contains a very rich vertebrate fauna consisting of fish, turtles, mammals (Antunes 1998), crocodiles (Mateus 2002), sauropods (Bonaparte & Mateus 1999; Mateus & Antunes 2000a,), theropods (Mateus 1998; Mateus & Antunes 2000a, 2000b, 2000c), ornitopods (Mateus & Antunes 2000a, 2001) and thyreophoras (Mateus & Antunes 2000a) as well as dinosaur nests, eggs and embryos (Antunes et al. 1998, Mateus et al. 1998). The formation 64 Part 4. Fossil tracks and trackways ________________________________________________________________________________ further contains numerous plant fragments (Pais 1998). Tracks of ornithopods and theropods have previously been collected from loose beach blocks and are on display at Museu da Lourinhã. Due to the exposed location of the tracks and the immediate danger that the tracks would be destroyed by erosion from the Atlantic Ocean, the tracks were collected and are now part of the Museu da Lourinhã collection, collectively numbered ML 557. The Lourinhã Formation The Lourinhã Formation is part of the Lusitanian Basin and consists of approximately 140 m of terrestrial sediments, deposited during the initial rifting of the Atlantic in the Kimmeridgian and Tithonian. The sediments consist mainly of thick red and green clay layers, interbedded of massive fluvial sandstone bodies and heterolithic horizons. The sandstone bodies appear as horizontally extensive and lenticular beds; some is traceable for several kilometres along the sections exposed along the coast. The sandstone lenses have been interpreted as distal alluvial fan facies originating from periods of extensive faulting (Hill 1989). Two thin carbonate intercalations (0.5 m) mostly consisting of the bivalve Isognomon, are recognizable throughout the whole formation. The upper one marks the Kimmeridgian – Tithonian boundary (Leinfelder 1987). The trackway is situated on the surface of a clay layer in the thick channel complex approximately 30 m below the Kimmeridgian – Tithonoian boundary (Fig. 55). The track assemblage The exposed surface in the track-bearing layer contains the natural cast of five tracks (Fig. 56), all pes prints. No associated manus prints were present. The tracks were originally emplaced on the surface of a 20 cm thick layer of clay, overlying a massive fluvial sandstone bed. Another fluvial sandstone bed overlies the clay layer with the tracks, and it is sand from this period of flooding that has in-filled the tracks and produced the casts. The tracks were numbered from 1 to 5. Tracks 1, 2 and 3 originates from the same animal, and constitute a small trackway consisting of two right (tracks no. 1 & 3) and one left (track no. 2) pes. Track 4 is located between tracks 1 and 2 and is orientated approximately perpendicular to the travel direction of tracks 1 to 3. Track 5 is from a significantly smaller animal, with a length of foot only 25 cm. The depth of the tracks increases northward, with track 5 being the shallowest and track 1 the deepest (Fig. 57) (Table 7). 65 Part 4. Fossil tracks and trackways ________________________________________________________________________________ Length, cm. Width, cm. Depth, cm. Track 1 33 30 12 Track 2 32 32 12 Track 3 33 37 9 Track 4 33 32 10 Track 5 25 23 7 Table 7. Measurements of length, witdth and depth for tracks 1 to 5. Track 1 is still partly embedded in the clay layer in which it originally was emplaced. The clay layers are bent around the track. A curious feature in connection with tracks 1 and 2 is the presence of the casts of fractures in the sediment radiating outward from the tracks (Fig. 58). Track 5 is less eroded than the others and has a well-defined outline. The track walls are sloping as if the trackmaker was sliding in the stride. Striations from the skin are preserved in the track walls (Fig. 59). A sixth track from the same horizon was observed, exposed at approximately 20 m further to the south, but was impossible to reach because of the steepness of the cliff section there. Half a kilometre to the north, several tracks are exposed as casts protruding beneath the same horizon, but all are heavily eroded and only recognizable as rounded casts. Several trampled intervals and cross sections of tracks can be recognized for several kilometres along the same horizon. The tracks are too poorly preserved to be assigned to a specific ichnogenus but their overall shape, having approximately equal length and width, with broad blunt toes without distinct claw impressions, clearly assign them to ornithopods (Moratella et al. 1988). The speed of progression of a dinosaur (or any other walking animal) can be calculated by the formula proposed by Alexander (1976): Velocity m/s = 0.25g0.5 x SL1.67 x h-1.17, where SL is the stride length between two consecutive footprints from the same foot, in this case tracks 1 and 3, which has a stride length measured from “heel” to “heel” of 1.71 m. The hip height, h is measured from the acetabulum (hip socket) to the ground when the leg is held in normal walking posture. In cases where the hip height is unknown Thulborn (1990: 251) advises to use an approximated hip height of 5.9 x foot length for ornithopods with a foot length in excess of 25 cm. In this case the estimated hip height of the trackmaker would be 195 cm (Fig. 60). By applying these values in the velocity formula, a progression speed of 0.89 m/s = 3.2 Km/h is obtained, which is a slow walking speed for an animal of this size. The progression speed of the other two individuals is impossible to calculate as they are represented by single tracks only. 66 Part 4. Fossil tracks and trackways ________________________________________________________________________________ Discussion The radiating fractures in the sediment around tracks 1 and 4 could suggest that the consistency of the sediment was rigid enough to crack in reaction to the outward force exercised on the sediment by the pressure of the foot. Similar cracks in the sediment, radiating outward from dinosaur tracks are described in ornithopod footprints from Colorado by Lockley et al. (1989) and in sauropod tracks from the Istria Peninsula, Croatia by Vecchia et al. (2001). Recent experiments with emus walking in damp sand produced the same type of radiating fractures that were formed during the weight-bearing phase of the stride, where the full weight of the animal is applied to one foot. Another explanation for the radiating fractures in connection with the tracks is that of Allen (1997), who demonstrates how the tracks of an animal can act as an agent for the initial formation of desiccation cracks in the sediment surface. If that is the case, the tracking surface must have been exposed for some time before the flooding and burial occurred, but not long enough to the development of a system of polygonal desiccation cracks. The infilling of the tracks is horizontally layered, indicating a passive infilling of a pre-existing depression. If the track was emplaced during the deposition of the sand the infilling would have been convoluted or disturbed (Nadon 1993, 2001). That the tracks in the assemblage become deeper northward indicates an increasing wetness of the sediment causing the trackmaker to sink deeper into the substrate. This effect of substrate consistency affecting the track depth is known from other tracksites. At the Purgatory River tracksite in the Morrison Formation, Lockley (1986) uses the depth of sauropod tracks to interpret the palaeo-topography and palaeo-water table. A titanosaurid trackway from the Upper Cretaceous of Bolivia displays an example where the tracks pass from uncollapsed to completely collapsed within a short distance (Lockley et al. 2002), also evidence of a rapid increase in sediment water content. Track 5 is the least impressed track in the assemblage, but as it originates from a smaller animal, the lesser depth of this track should be explained by the lighter trackmaker, which did not sink as deep at the others heavier animals. However the more well-defined shape, the preserved sloping track walls and the presence of striations in the track walls of track 5 suggests that either the sediment had a locally firmer consistency or more plausibly that the animal responsible for track 5 was crossing the area at a later time where the sediment had dried to a firmer consistency. 67 Part 4. Fossil tracks and trackways ________________________________________________________________________________ Another way to explain the sloping track walls in track 5 could be they were a result of a secondary tectonic disturbance of the clay layer from sliding between the two massive overand underlying sandstone bodies. This explanation can be ruled out, as the other four tracks display no signs of such a tectonic disturbance. Therefore the sloping track walls most likely must be the result of a sideways movement of the trackmaker’s foot during the stride in a sediment firm enough to preserve the shape and details of the track. Skin impressions in dinosaur footprints are rare as they are delicate structures easily destroyed by erosion, or not preserved at all due to the properties of the sediment. They are, however, found in footprints from several groups of dinosaurs, ornithopods (Currie et al. 1991), theropods (Gatesy 2001) and thyreophorans (McCrea et al. 2001) and occur either as impression of the sole of the trackmaker’s foot, in all cases shown to be the impression of skin covered with small tubercles, or as striations in the track walls produced by the tubercles as they are dragged through the mud (Difley & Ekdale 2002, McCrea et al. 2002). The striations in the track walls of track 5 were probably produced when the foot was lifted out of the mud. The absence of associated manus imprints in this case suggests that the trackmakers in this case were progressing in a bipedal stance. Ornithopods are believed to have been occasional bipeds, as evidenced by several trackways (Moratella et al. 1992, 1994). Ornithopods bore the least weight on their forelimbs which were therefore less impressed in the sediment than the hind limbs, which resulted in the manus tracks commonly being small or absent in ornithopod trackways. In this case, however, the depth of the tracks and softness of the sediment should have preserved associated manus imprints if the trackmakers had had a quadropedal stance. Another explanation for the missing manus impressions could be that the smaller manus imprints were overstepped and destroyed by the larger pes of the ornithopods. Conclusion The above-described sedimentological features in the track assemblage make it possible to infer about the properties of the substrate both during and after the trackmakers passed by. The sediment was softer toward the north as evidenced both by the deepening of the tracks from the same animal, i.e, tracks 1 to 3, and track 4 which crosses the trackway constituted by tracks 1 to 3, lying between tracks 1 and 2. Track 5 is inconclusive in this connection as it clearly originates from a smaller animal. But the well-defined shape and the preserved striations from the skin show that it was impressed in a firmer substrate than the other tracks. 68 Part 4. Fossil tracks and trackways ________________________________________________________________________________ By combining the sedimentological phenomena associated with the tracks, the track assemblage can provide information of the timing of the animals that passed through and the subsequent flooding and burial of the surface. If assuming the radiation fractures around tracks 1 and 4 in fact represent initial desiccation cracks, and the relatively better preservation of the smaller track 5 is due to a later emplacement in a dryer sediment, then the track assemblage demonstrates an example of an exposed clay surface with tracks from three different animals, crossing the surface to different times. The surface was exposed long enough to start the initial forming of desiccation cracks in the surface. 69 General discussion ________________________________________________________________________________ General discussion The individual cases presented in the previous four parts, have all dealt with different aspects of experimental and comparative Ichnology. Part 1 explored why birds, and especially the ratites to which the emus belong, are the best candidates to simulate the tracks of Mesozoic theropods. The ratites represent the most primitive linage of extant birds; although secondarily flightless, they are now adapted to a fully cursorial lifestyle. Among the ratites the emu has proven to be the best candidate to compare with theropods, as their feet are tridactyl, contrary to the reduced didactyl foot of the ostrich. The stubborn and often uncooperative mind of the emus and related birds, however, makes successful field experiments extremely difficult, unless one has the good fortune to have access to animals customized to human contact. To distinguish between bird and theropod tracks has proven to be difficult if not impossible in the Mesozoic, as all the track characters supposed to indicate an avian origin of the tracks have proven to be inconclusive, as (1) Mesozoic theropods is known from diminutive size, (2) some late Mesozoic birds reached approximately the size of an ostrich, and (3) the avian track characters is found in various combinations plus (4) that extremely birdlike tracks are found which predates the earliest birds with about 60 million year. This further validated the use of birds and especially the ratites to simulate Mesozoic theropods. The method of conducting field experiments with emus has proven very useful as it thereby is possible directly to observe how changes in tracking substrate effects the track morphology. The usage of a living animal to make the tracks, gives the advantage that all the dynamic aspects affecting the track formation, which are impossible to fully recreate using laboratory models, is captured in the tracks. Even the use of a fresh emu foot in the cement experiments, cannot fully recreate the complex dynamic interaction between the substrate and a living animal, despite that the foot movements were carefully mimicked during the experiments. The field experiments with the emu and the laboratory experiments with emu tracks in layered cement with different water content and thus different consistencies, clearly demonstrates that the properties of the sediments exercise a strong control on the morphology of the tracks emplaced within them. All cases show that the higher the water content of the sediment, the more the track seems to collapse and destroy anatomical information of the trackmaker´s foot (Figs. 37-43). The four examined Grallator tracks from Greenland further supports this, as they represents tracks sat in 70 General discussion ________________________________________________________________________________ sediments of four different properties, preserving different amounts of anatomical details of the foot (Figs. 48-51). These observations are in full agreement with Gatesy et al´s (1999) observations on semi-collapsed theropod tracks from Greenland and the different preservation of rhynchosaurid footprints described by Diedrich (2002). The track assemblage from Portugal demonstrates that variations in sediment properties do occur on a very local scale, causing differences in the preservation of consecutive footprints from the same track assemblage (Figs. 57-59). That the sediment properties can cause the tracks from the same trackmaker to appear significantly different has long been recognized among invertebrate ichnologists, as many invertebrate ichnogenera are merely sedimentological variations of the actions of the same trackmakers (Bromley 1996). It is first recently that vertebrate ichnologists have begun to consider that sedimentological variations can cause significantly different variations among tracks from the same trackmaker (Brand 1996; Gatesy et al. 1999; Diedrich 2002). The method of studying fossil footprints in vertical sections, proved very useful as it helped reveal undertracks and subsurface deformation structures associated with the footprint, that otherwise is unobtainable, if the track is studied only at the surface. The experimentally produced tracks in layered cement, demonstrated effectively how the consistency of the substrate affected the morphology of the undertracks (Figs. 37-43). The tracks from Greenland were all formed in relative thin layers of mud overlying a firm base, as was the case in the experimentally obtained tracks. On that account a direct comparison of the vertical slices through the experimentally obtained tracks and the vertical slices of the Grallator track from Greenland is validated. The undertracks present below the Grallator track (Fig. 53) bear the strongest similarity with the experimentally obtained undertracks from the driest mixtures of cement and water (Figs. 37, 41), which also corresponds with the fact that the driest mixtures produce the shallowest tracks. The track NB2 (Fig. 49) represents deeper, softer sediment that collapsed after withdrawal of the foot, as in the case of the wetter experimental mixtures (Figs. 39, 40, 42) and the emu tracks from deeper softer sediments (Figs. 18F-H). The emu tracks in sediments of different consistencies (Fig. 18) and the experimentally obtained tracks (Figs 37 - 42) all show similar trends in collapsing and changes in track morphology with increasing sediment wetness. The four Grallator tracks from Greenland (Figs. 48 - 51) display the same kinds of changes in morphology as do the tracks described by Gatesy et al. (1999) from the same area. This shows that the morphology of the overall track on the 71 General discussion ________________________________________________________________________________ surface can be used to estimate the original consistency and wetness of the sediment, at the time the trackmakers made their tracks. Recently, Nadon (2001) argued that most tracks found, indeed are true tracks and not undertracks. The argument goes that the lack of details in tracks are due to subsequent erosion of the track, or that the thin layer containing the skin impression and the fine details had stuck to the trackmakers foot and was carried with it. The experimental work with the emus supports these ideas, as in some instances a layer of the surface sediment did stick to the sole of the emu foot and were transported away from the tracks (Fig. 19). However in many instances skin impressions were preserved in exquisite detail, in the tracks made in mud of different consistencies (Fig. 18D, E). The experimental work with layered cement however, clearly shows that undertracks are prone to form in substrates of ideal consistencies for trackmaking (Figs. 37-43), and well-developed undertracks are also revealed in the sediments beneath the sectioned Grallator track (NB 1) form Greenland (Fig. 53) and on the underside of the slab containing track LC2 (Fig. 51). At the surface, track NB 1 had the appearance of an undertrack, as could be expected by comparison with the cement experiments (Fig. 43). When examined in vertical sections, the surface appearance of the track turned out to the result of erosion and the partial presence of the original sedimentary infilling (Fig. 53). The sections however, also show the presence of well-developed undertracks below the footprint. The herein obtained results neither fully refute nor support the idea of Nadon (2001). Both the experimental tracks and the real tracks examined, showed well-developed undertracks, and that erosion can alter the appearance of a true track to that of an undertrack. The field experiments further demonstrated that the sediments in some cases did stick to the sole of the trackmakers foot and was transported, and thereby removing the upper layer of the track which contained the skin impressions and fine details of the foot. 72 Overall conclusions ________________________________________________________________________________ Overall conclusions 1. The adaptations, although secondary, to a cursorial lifestyle in the ratitous birds, together with the close similarities in footprint morphology, make emus in particular very useful as substitutes for non-avian theropods in experimental ichnology. 2. There is no reliable way in praxis to discriminate between bird and theropod tracks in the Mesozoic, as no definite avian characters can be confirmed. 3. The morphology of emu tracks displays a great morphological variation; even within tracks from the same animal. Especially the divarication angle can vary significantly in consecutive footprints. 4. The consistency of the sediments exercises a strong control over the morphology of the track as a whole, and especially the amount of anatomical details preserved within it. 5. Striations from skin tubercles can provide information of the sub-sediment movements of the trackmakers foot during the stride, both in modern and fossil footprints. 6. Undertracks tend always to be more rounded, shallower and less detailed than the true track, unless the nature of the surface sediment is so fluid that the true track collapses completely, in which case the undertrack actually can retain the shape of the track better than the true track. 7. By examining the sedimentological context of a track, the taphonomy of the track can provide information about the properties of the sediments the track were created in and the timing of the subsequent burial of the track. 8. When studying fossil tracks, special attention should be paid to the original properties of the sediments in which the tracks were emplaced, as the morphological variations caused by the different sediment properties causes tracks from the same trackmaker to appear significantly morphologically different, and this could give rise to the erection of different ichnogenera. 73 Summary ________________________________________________________________________________ Summary The traditional problem in ichnology is to interprete fossil tracks and traces without knowing the nature of the actual trackmaker. In vertebrate Ichnology, however, the case is normally easier as it often is possible to find animals with similar foot morphology and expected mode of life to those responsible for the fossil tracks. Tracks from the ratitous birds, and especially the emu, have proven the best candidate for comparison with tracks from non-avian theropods, as both foot morphology and lifestyle of the animal are comparable. Recent finds of small, indeed very birdlike tracks from the Upper Triassic has proven that it is very difficult, if not impossible to discriminate between supposed tracks from Mesozoic birds and tracks from non-avian theropods, as track characters normally thought to be of definite avian origin, now appear already in the Upper Triassic, well before the origin of the earliest birds. Field experiments conducted by encouraging emus to walk through sediments of different water contents, and thus different consistencies ranging from dry sand to deep fluid mud demonstrate the control the sediment properties exercise on the footprint morphology. The tracks emplaced in firm to semi-firm mud preserved the highest amount of anatomical detail, while tracks in deep fluid mud collapsed immediately after withdrawal of the foot, obliterating the shape of the footprint. It was observed that emus in some instances during resting or feeding walked in a plantigrade stance, producing highly elongated tracks with impression of the full metatarsus, tracks that bear a strong resemblance the Jurassic ichnogenus Anoemoepus. When walking on firm ground the emu carry the least weight on digit II, which in some instances fail to leave an impression in the substrate, resulting in a didactyl track. The divarication angle between digits II and IV in the emu footprints varies greatly, even from track to track within the same trackway. It was noticed, however, that the divarication angle was more uniform in trackways where the emu was running than in trackways resulting from walking. To investigate the formation of undertracks a series of laboratory controlled experimental emu tracks were made in packages of layered cement with different water contents, and thus different consistencies. Subsequent vertical slicing of the packages containing the tracks revealed that the formation and morphology of the undertracks differ according to the sediment consistency, where the best defined undertracks form in relative firm sediments. However, in tracks 74 Summary ________________________________________________________________________________ made in soft sediments, which at the surface was totally collapsed and hardly recognizable, the undertracks retained the shape of the foot better than the true track. To simulate the condition where a track with its sedimentary infilling is eroded to different degrees, a plaster cast of an emu track was cut in successive horizontal sections. This experiments demonstrated how the morphology/outline changes downward through the sections of the track. The most pronounced changes are in the shape and width of the digit impressions, and that the digit impressions detach from each other at a certain depth. The results of this are compared to similar depth related changes in the track morphology of a Jurassic theropod track exposed, in both horizontal and vertical view due to erosion. This is important to bear in mind when describing fossil tracks, that a track can exhibit great morphological variation when exposed at different horizons. Examination of four Triassic theropod tracks from Jameson Land, East Greenland, show different degrees of preservation according to differences in both the original sediment consistencies and the degree of later erosion. The best preserved track originally emplaced in a thin layer of soft mud, exhibits detailed impressions of the arrangement and number of digital pads, while another track originally emplaced in deeper soft mud is collapsed and hardly recognizable as a theropod track. The collapsed track, however, has preserved striations from the fine skin tubercles, covering the theropods foot, on both sides of digit III. The striations bear a strong similarity to striations from the skintubercles covering the foot an emu. A theropod track, which at the surface had the undetailed and rounded appearance of an undertrack was cut in vertical sections. The vertical sections revealed that the track was indeed a true track and that the rounded and undetailed appearance was caused by erosion and partly infilling of the track. The sections shoved the narrow cut from the claw on digit III and the undertracks formed below the track hinted the division of the digits better than in the eroded true track. A newly discovered track assemblage from Porto das Barcas, Lourinhã, Portugal comprises five tracks from three ornithopods walking in different directions. The tracks in the assemblage become successively deeper northward indicating that the sediments were softer in that direction when the animals walked there. The two deepest tracks in the track assemblage are surrounded by fractures radiating outward from the track, and the shallowest track has preserved striations from skinturbercles. Comparisons between the experimental work and the examined fossil tracks, clearly demonstrate a strong linking between the properties of the sediments and the preservation potential 75 Summary ________________________________________________________________________________ of tracks emplaced within it. When describing fossil tracks, strong attention should be paid to the taphonomical state of the track, as erosion as well as differences in the original sediment properties can cause tracks from the same trackmaker to appear sufficiently different to give rise to different ichnogenera. Acknowledgements First a special thanks to my supervisers Eckart Håkansson and Richard G. Bromley, who allowed me total freedom to conduct the research for my work. Arne Nielsen provided immense help with the practical aspects of the cement experiments. Special thanks go to the emu owners Karin Holst and Mogens Madsen, who kindly allowed us to work with their emus and provided valuable expertise in handling them. The Danish Museum of Hunting and Forestry, Hørsholm, Denmark, kindly provided me with a spare emu foot. Ole Bang Bertelsen, Geological Institute, Copenhagen provided help with photographic work, and showed patience when I brought a still bleeding emu foot into his photolab. Octavio, Horacio and Isabel Mateus and everybody at GEAL Museum da Lourinhã are thanked for hospitality, scientific stimulus and good company. Niels Bonde and Lars B. Clemmensen, Geological Institute, Copenhagen provided me with access to their track material collected in Greenland. Per Christiansen, Zoological Museum, Copenhagen answered willingly a lot of questions about dinosaurs, and James O. Farlow shared his experiences with handling emus and ostriches with me. InnospeXion ApS, Kisserup, Denmark are thanked for giving me access to high-tech X-ray equipment. Martin A. Whyte, Mike Romano, Simon J. Jackson & Daniel J. Elvidge from the Sheffield Dinosaur Track Research Group provided excellent company and inspiring discussions and access to track material. Anne Mehlin Sørensen is thanked for her critical reading of the manuscript. Niels Just Rasmussen, Bettina M. Sørensen, University of Copenhagen & Gilles Cuny, Geological Museum Copenhagen, provided practical bird-minding help during the fieldwork. And finally a big thank to Polly, Oliver and three other unnamed emus who more or less willingly provided the footwork for the field experiments. 76 References ________________________________________________________________________________ References Alexander, R. McN. 1976. Estimates of speeds of dinosaurs. Nature 261: 129-130. Allen, J. R. L. 1997. Subfossil mammalian tracks (Flandrian) in the Severen Estuary, S.W. Britain: mechanics of formation, preservation and distribution. Philosophical Transactions of the Royal Society London, B 352: 481-518. Anonymous [Milàn, J & E. Håkansson,] 2003. Ny viden om dinosaurspor. National Geographic Danmark, 3(2003): 9. Antunes, M. T. 1998. A new Upper Jurassic Paulchoffatiid multituberculate (Mammalia) from Pai Mogo, Portugal. Memórias da Academia de Sciencias 37: 125-153. Antunes, M. T., P. Taquet & V. Ribeiro. 1998. Upper Jurassic dinosaur and crocodile eggs from Pai Mogo nesting site (Lourinhã- Portugal). Memórias da Academia de Sciencias 37: 83-100. Avanzini, M. 1998. Anatomy of a footprint: Bioturbation as a key to understanding dinosaur walk dynamics. Ichnos 6(3): 129-139. Baird, D. 1957. Triassic Reptile Footprint Faunas from Milford, New Jersey. Bulletin of the Museum of Comparative Zoology 117: 449-520. Bonaparte, J. & O. Mateus. 1999. A new diplodocid, Dinheirosaurus lourinhanensis gen. et sp. nov., from the Late Jurassic beds of Portugal. Revista del Museo Argentino de Ciencias Naturales 5(2): 13-29. Brand, L. R. 1996. Variations in salamander trackways resulting from substrate differences. Journal of Paleontology 70(6): 1004-1010. Brand, L. R. & T. Tang. 1991. Fossil vertebrate footprints in the Coconino sandstone (Permian) of Northern Arizona: Evidence for underwater origin. Geology 19: 1201-1204. 77 References ________________________________________________________________________________ Bromley, R. G. 1996. Trace Fossils – Biology, Taphonomy and Applications. 2nd edition. Chapman & Hall. 361 pp. Bromley, R. G. 2001. Tetrapod tracks deeply set in unsuitable substrates: Recent musk oxen in fluid earth (East Greenland) and Pleistocene caprines in Aeolian sand (Mallorca). Bulletin of the Geological Society of Denmark 48: 209-215. Bromley, R. G. & U. Asgaard. 1979. Triassic freshwater ichnocoenosis from Carlsberg Fjord, East Greenland. Palaeogeography, Palaeoclimatology, Palaeoecology 28: 39-80. Brown Jr., T. 1999. The science and art of tracking. Berkley. 219 pp. Buffetaut, E. & J. Le Loeuff. 1998. A new giant ground bird from the Upper Cretaceous of southern France. Journal of the Geological Society, London 155: 1-4. Chatterjee, S. 1999. Protoavis and the early evolution of birds. Palaeontographica Abteilung A 254: 1-100. Chin, K., T. T. Tokaryk., G. M. Erikson & L. C. Calk. 1998. A king-sized theropod coprolite. Nature 393: 680-682. Christiansen, P. 1997. Hindlimbs and feet. In Currie, P. J. & K. Padian (eds.) 1997. Encyclopedia of Dinosaurs. pp 320-328. Academic Press. 869 pp. Clemmensen, L. B., D. V. Kent & F. A. Jenkins Jr. 1998. A Late Triassic lake system in East Greenland: facies, depositional cycles and palaeoclimate. Palaeogeography, Palaeoclimatology, Palaeoecology. 140: 135-159. Cracraft, J. & J. Clarke. 2001. The basal clades of modern birds. In Gauthier, J. & L. F. Gall (eds.). New perspectives on the Origin and Early Evolution of Birds. pp. 143-156. Yale Peabody Museum. 613 pp. 78 References ________________________________________________________________________________ Currie, P. J. 1981. Bird footprints from the Gething Formation (Aptian, Lower Cretaceous) of Northeastern British Columbia. Journal of Vertebrate Paleontology 1(3-4): 257-264. Currie, P. J. 1989. Dinosaur footprints of Western Canada. In Gillette, D. G. & M. G. Lockley (eds.). 1989. Dinosaur tracks and traces. pp 293-300. Cambridge University Press. 454 pp. Currie, P. J., G. C. Nadon & M. G. Lockley. 1991. Dinosaur footprints with skin impressions from the Cretaceous of Alberta and Colorado. Canadian Journal of Earth Science 28: 102-115. Davies, S. J. J. F. 2002. Bird Families of the World: Ratites and Tinamous. Oxford University Press. 310 pp. Day, J. J., D. B. Normann., P. Upchurch & H. P. Powell. 2002. Dinosaur locomotion from a new trackway. Nature 414: 494-495. Diedrich, C. 2002. Vertebrate track bed stratigraphy at a new megatrack site in the Upper Wellenkalk Member and orbicularis Member (Muschelkalk, Middle Triassic) in carbonate tidal flat environments of the western Germanic Basin. Palaeogeography, Palaeoclimatology, Palaeoecology 183: 185-208. Difley, R. L. & A. A. Ekdale. 2002. Footprints of Utah´s last dinosaurs: track beds in the Upper Cretaceous (Maastrichtian) North Horn Formation of the Wasatch Plateau, central Utah. Palaios 17: 327-346. Farlow, J. O. 1989. Ostrich Footprints and trackways: Implications for Dinosaur Ichnology. In Gillette, D. D. & M. G. Lockley (eds.). Dinosaur Tracks and Traces. pp. 243-248. Cambridge University Press, Cambridge. 454 pp. Farlow, J. O. & Chapman, R. E. 1997. The scientific study of dinosaur footprints. In Farlow, J. O. & M. K. Brett-Surman (eds.). 1997. The Complete Dinosaur. pp 518-553 Indiana University press. 752 pp. 79 References ________________________________________________________________________________ Farlow, J. O., S. M. Gatesy., T. R. Holtz Jr., J. R. Hutchinson & J. M. Robinson. 2000. Theropod Locomotion. American Zoologist. 40: 640-663. Farlow, J. O., J. McClain & K. Shearer. 1997. Intraspecific and interspecific variability in foot and footprint shapes in ground birds: implications for the ichnology of bipedal dinosaurs. Journal of Vertebrate Paleontology 17(3): 45A Folk, R. L. & W. Ward. 1957. Brazos river bar: a study in the significance of grain size parameters. Journal of Sedimentary Petrology 27: 3-26. Fornós, J. J., R. G. Bromley., L. B. Clemmensen & A. Rodriguez-Perea. 2002. Tracks and trackways of Myotragus balearicus Bate (Atiodactyla, Caprinae) in Pleistocene aeolianites from Mallorca (Balearic Islands, Western Mediterranean). Palaeogeography, Palaeoclimatology, Palaeoecology 180: 277-313. Gatesy, S. M. 2001: Skin impressions of Triassic theropods as records of foot movement. Bulletin of the Museum of Comparative Zoolology 156(1): 137-149. Gatesy, S. M., K. M. Middleton., F. A. Jenkins jr & N. H. Shubin. 1999. Three-dimensionel preservation of foot movements in Triassic theropod dinosaurs. Nature 399: 141-144. Goldring, R. & A. Seilacher. 1971. Limulid undertracks and their sedimentological implication. Neues Jahrbuch Geologie und Paläeontologie Abhandlingen 137: 422-442. Griffiths, P.J., J-M. Mazin, J-M. & Billon-Bruyat. 2002. The agility of theropod dinosaurs when turning: evidence from trackways from Grayssac, France. In Norman, D. & P. Upchurch (eds.). Symposium of Vertebrate Palaeontology ad Comparative Anatomy, SVPCA 50, Abstract Volume. 11-13 September 2002, Cambridge. 1p. Hill, G. 1989. Distal alluvial fan sediments from the Upper Jurassic of Portugal: Controls on their cyclicity and channel formation. Journal of the Geological Society, London 146: 539-55. 80 References ________________________________________________________________________________ Hitchcock, E. 1836. Ornithichnology,-Description of the Foot marks of Birds, (Ornithichnites) on new Red Sandstone in Massachusetts. American Journal of Science 29: 307-340. Hitchcock, E. 1858. Ichnology of New England, A report on the Sandstone of the Connecticut Valley Especially its Fossil Footmarks. Boston: W. White (Reprinted by Arno Press in the Natural Sceince in America Series). 220 pp + plate I – LX. Holtz, T. R., Jr. 1994. The phylogenetic position of the Tyrannosauridae: implications for theropod systematics. Journal of Paleontology 68: 1100-1117. Holtz, T. R. Jr. 1998. A new phylogeny of the carnivorous dinosaurs. Gaia 15: 5-61. Irby, G. V. 1995. Posterolateral markings on dinosaur tracks, Cameron Dinosaur Tracksite, Lower Jurassic Moehave Formation, northeastern Arizona. Journal of Paleontology 69(4): 779-784. Jackson, S. 2001. Reconstructing and interpreting 3-dimensional structures associated with dinosaur footprint formation from tracks from the Middle Jurassic of Yorkshire. 45th Annual Meeting of the Palaeontological Association. Abstracts. Geological Museum, University of Copenhagen, 15th-19th December 2001. Jackson, S. 2002. How to make dinosaur tracks: interpreting dinosaur footprint formation and preservation using laboratory controlled simulations. The Palaeontological Association Newsletter, 51:81. Jenkins, F. A., N. H. Shubin., W. W. Amaral., S. M. Gatesy., C. R. Schaff., L. B. Clemmensen., W. R. Downs., A. R. Davidson., N. Bonde & F. Osbæck. 1994. Late Triassic continental vertebrates and depositional environments of the Fleming Fjord Formation, Jameson Land, East Greenland. Meddelelser om Grønland, Geoscience 32: 1-25. Ji, Q., M. Norell., K.-Q. Gao., S.-A. Ji & D. Ren. 2001. The distribution of integumentary structures in a feathered dinosaur. Nature 398: 1084-1088. 81 References ________________________________________________________________________________ Kuban, G. J. 1989. Elongate dinosaur tracks. In Gillette, D. & M. G. Lockley. (eds.), Dinosaur tracks and traces. pp 57-72. Cambridge University Press. 454 pp. Laporte, L. F. & A. K. Behrensmeyer. 1980. Tracks and substrate reworking by terrestrial vertebrates in Quaternary sediments of Kenya. Journal of Sedimentary Petrology 50(4): 1337-1346. Leinfelder, R. R. 1987. Multifactorial control of sedimentation patterns in an ocean marginal basin: the Lusitanian Basin (Portugal) during the Kimmeridgian and Tithonian. Geologische Rundschau 76(2): 599-631. Leonardi, G. (ed.) 1987. Glossary and manual of tetrapod footprint palaeoichnology. Dept. Nacional de Producao Mineral, Bracil. 75 pp. Lewin, R. A. 1999. Merde, Excurtions into Scientific, Cultural and Socio-Historical Coprology. Aurum Press, London. 164 pp. Lockley, M. G. 1986. The palaeobiological and palaeoenvironmental importance of dinosaur footprints. Palaios 1: 37-47. Lockley, M. 1991. Tracking Dinosaurs. Cambridge University Press. 238 pp. Lockley, M. G. 1992. Comment and reply on “Fossil vertebrate footprints in the Coconino sandstone (Permian) of Northern Arizona: Evidence for underwater origin”. Geology 20: 666-667. Lockley, M. G., A. P. Hunt., M. Paquett., S-A. Bilbey & A. Hamblin. 1998. Dinosaur tracks from the Carmel Formation, Northeastern Utah: implications for Middle Jurassic palaeontology. Ichnos 5: 255-267. Lockley, M. G., T. J. Logue., J. J. Moratella., A. P. Hunt., R. J. Schultz & J. W. Robinson. 1995. The fossil trackway Pteraichnus is pterosaurian, not crocodillian: implications for the global distribution of pterosaur tracks. Ichnos 4: 7-20. 82 References ________________________________________________________________________________ Lockley, M. G., M. Matsukawa & I. Obata. 1989. Dinosaur tracks and radial cracks: unusual footprint features. Bulletin National Science Museum Tokyo, Serie C 15(4): 151-160. Lockley, M. G. & E. C. Rainforth. 2002. The track record Mesozoic birds and pterosaurs: An ichnological and paleoecological perspective. In Chiappe, L. M. & L. M. Withmer (eds.). 2002. Mesozoic birds above the heads of dinosaurs. pp 405-420. University of California Press. 520 pp. Lockley, M. G., A. S. Schulp., C. A. Meyer., G. Leonardi & D. K. Mamani. 2002. Titanosaurid trackways from the Upper Cretaceous of Bolivia: evidence for large manus, wide-gauge locomotion and gregarious behaviour. Cretaceous Research 23: 383-400. Lockley, M. G., S. Y. Yang., M. Matsukawa., F. Fleming & S. K. Lim. 1992. The track record of Mesozoiz birds: evidence and implications. Philosophical Transactions of the Royal Soceity London B 336: 113-134. Loope, D. B. 1992. Comment on “Fossil vertebrate footprints in the Coconino sandstone (Permian) of Northern Arizona: Evidence for underwater origin”. Geology 20: 667-668. Lucas, A. M. & P. R. Stettenheim. 1972. Avian Anatomy, Integument. Part I. Agricultura Handbook 362. U.S. Dept. Agricultura, Washington D.C. 340 pp. Makovicky, P. J & H-D. Sues. (1998). Anatomy and phylogenetic relationships of the theropod dinosaur Microvenator celer from the Lower Cretaceous of Montana. Am. Museum Novit 3240: 116. Mateus, I., H. Mateus., M. T. Antunes., O. Mateus., P. Taquet., V. Ribeiro & G. Manuppella. 1998. Upper Jurassic theropod dinosaur embryos from Lourinhã (Portugal). Memórias da Academia de Sciencias 37: 101-110. Mateus, O. 1998. Lourinhanosaurus antunesi, a new Upper Jurassic allosaurid (DinosauriaTheropoda) from Lourinhã, Portugal. Memórias da Academia de Sciencias 37: 111-124. 83 References ________________________________________________________________________________ Mateus, O. 2002. On a large crocodile vertebrae of cf. Machimosaurus from the Late Jurassic of Portugal. Il Congresso Ibérico de Paleontologia, Libro de Resúmenes pp 73-75. Mateus, O. & M. T. Antunes. 2000a. Late Jurassic dinosaurs of Portugal. Abstracts of the 1st Symposium of European Dinosaurs, Dusseldorf, Germany. Mateus, O. & M. T. Antunes. 2000b. Torvosaurus sp. (Dinosauria: Theropoda) in the Late Jurassic of Portugal. Livro de Resumos do I Congresso Ibérico de Paleontologia pp 115-117. Mateus, O. & M. T. Antunes. 2000c. Ceratosaurus (Dinosauria: Theropoda) in the Late Jurassic of Portugal. Abstract volume of the 31st International Geological Congress. Rio de Janeiro, Brazil. Mateus, O. & M. T. Antunes. 2001. Draconyx loureiroi, a new Camptosauridae (Dinosauria: Ornithopoda) from the Late Jurassic of Lourinhã, Portugal. Annales Paleontol 87(1): 61-73. McCrea, R. T. M. G. Lockley & C. A Meyer. 2001. Global distribution of purpoted ankylosaur track occurrences. In Carpenter, K (ed.). The Armoured Dinosaurs. pp 413-454. Indiana University Press. 526 pp. McKee, E. D. 1947. Experiments on the development of tracks in fine cross-bedded sand. Journal of Sedimentary Petrology 17(1): 23-28. McKeever, P. J. 1991. Trackway preservation in aeolian sandstones from the Permian of Scotland. Geology 19: 726-729. McKeewer, P. J. & H. Haubold. 1996. Reclassification of vertebrate trackways from the Permian of Scotland and related forms from Arizona and Germany. Journal of Paleontology 70(6): 1011-1022. Melchor, R. N., S. D. Valais & J. F. Genise. 2002. Bird-like fossil footprints from the Late Triassic. Nature 417: 936-938. 84 References ________________________________________________________________________________ Milàn Nielsen, J. 2000. Tetrapodspor og sporserier. Unpublished Bac. Scient. Thesis. Geological Institute, University of Copenhagen. 42 pp. Milàn, J. & R. G. Bromley. (Submitted). The impact of sediment consistency on track- and undertrack morphology: experiments with emu tracks in layered cement. In Rainforth, E.C. & McCrea, R.T. (eds). Lithichnozoa: Fossil vertebrate footprints. Indiana University Press. Milàn, J. & R. G. Bromley. 2002a. The influence of substrate consistency on footprint morphology: field experiments with an emu. The Palaeontological Association Newsletter 51: 30. Milàn, J. & R. G. Bromley. 2002b. One foot – many tracks; experiments with undertrack formation in layered cement. In Norman, D. & P. Upchurch (eds.). Symposium of Vertebrate Palaeontology and Comparative Anatomy, SVPCA 50, Abstract Volume. 11-13 September 2002, Cambridge. 1p. Moratella, J. J., J. L. Sanz & S. Jimenez. 1988. Multivariate analysis on Lower Cretaceous dinosaur footprints: Discrimination between ornithopods and theropods. Geobios 21(4): 395-408. Moratella, J. J., J. L. Sanz & S. Jimenez. 1994. Dinosaur tracks from the Lower Cretaceous of Regumiel de la Sierre (province of Burgos, Spain): inferences on a new quadrupedal ornithopod trackway. Ichnos 3: 89-97. Moratella, J. J., J. L. Sanz., S. Jimenez & M. G. Lockley. 1992. A Quadropedal ornithopod trackway from the Lower Cretaceous of La Rioja (Spain): Inference on gait and hand structure. Journal of Vertebrate Paleontology 12(2): 150-157. Nadon, G. C. 1993. The association of anastomosed fluvial deposits and dinosaur tracks, eggs and nests: Implications for the interpretation of floodplain environments and a possible survival strategy for ornithopods. Palaios 8: 31-44. Nadon, G. C. 2001. The impact of sedimentology on vertebrate track studies. In Tanke, D. H. & K. Carpenter (eds.). Mesozoic Vertebrate Life. pp 395-407. Indiana University Press. 577 pp. 85 References ________________________________________________________________________________ Olsen, P. E., J. B. Smith & N. G. McDonald. 1998. Type material of the type species of the classic theropod footprint genera Eubrontes, Anchisauripus and Grallator (Early Jurassic, Hartford and Deer Basins, Connecticut and Massachutts, USA). Journal of vertebrate Paleontology 18(3): 586601. Padian, K. & P. E. Olsen. 1984. The fossil trackway Pteraichnus: Not pterosaurian, but crocodillian. Journal of Paleontology 58(1): 178-184. Padian, K. & P. E. Olsen. 1989. Ratite Footprints and the Stance and Gait of Mesozoic theropods. In Gillette, D. D. & M. G. Lockley (eds.). Dinosaur Tracks and Traces. pp 231-242. Cambridge University Press. 454 pp. Pais, J. 1998. Jurassic plant macroremains from Portugal. Memórias da Academia de Sciencias 37: 25-48. Paul, G. S. 2002. Dinosaurs of the Air, the evolution and loss of flight in dinosaurs and birds. The John Hopkins University Press. 460 pp. Peabody, F. E. 1959. Trackways of living and fossil salamanders. Publications in Zoology. University of California Press 63(1): 1-72. Pittman, J. G. & D. D. Gillette. 1989. The Briar Site: A new sauropod dinosaur tracksite in Lower Cretaceous beds of Arkansas, USA. In Gillette, D. D. & M. G. Lockley (eds.). Dinosaur Tracks and Traces. pp 313-332. Cambridge. 454 pp. Sereno, P. C. 1997. The origin and evolution of dinosaurs. Annual Review of Earth and Planetary Science 25: 435-489. Sollas, W. J. 1879. On some Three-toed Footprints from the Triassic Conglomerate of Southern Wales. Quarterly Journal of the Geological Society of London 35: 511-517. 86 References ________________________________________________________________________________ Stokes, W. L. 1957. Pterodactyl tracks from the Morrison Formation. Journal of Paleontology 31: 952-954. Thulborn, R. A. 1991. Morphology, preservation and palaeobiological significance of dinosaur coprolites. Palaeogeography, Palaeoclimatology, Palaeoecology 83: 341-366. Thulborn, T. 1990. Dinosaur tracks. Chapman and Hall. 410 pp. Thulborn, R. A. & M. Wade. 1989. A footprint as history of movement. In Gillette, D. & M. G. Lockley (eds.) Dinosaur tracks and traces. pp 51-56. Cambridge University Press. 454 pp. Tucker, M. E. 2001. Sedimentary Petrology 3rd ed. Blackwell Science. 262 pp. Vecchia, F. M. D., G. Tuni., S. Venturini & A. Tarlao. 2001. Dinosaur tracks in the upper Cenomanian (Late Cretaceous) of Istrian Peninsula (Croatia). Bollettino della Societa Paleontologica Italiana 40(1): 25-53. Wade, M. 1989. The stance of dinosaurs and the Cossack Dancer syndrome. In Gillette, D. G. & M. G. Lockley (eds.). 1989. Dinosaur tracks and traces. pp 73-82. Cambridge University Press. 454 pp. Witmer, L. M. 2001. The role of Protoavis in the debate on avian origins. In Gauthier, J and L. F. Gall (eds). New Perspectives on the origin and early evolution of birds. pp 537-548. Yale Peabody Museum 613 pp. Whyte, M. A. & M. Romano. 1994. Probable sauropod footprints from the Middle Jurassic of Yorkshire, England. Gaia 10:15-26. Woodhams, K. E. & Hines, J. S. 1989. Dinosaur footprints from the Lower Cretaceous of East Sussex, England. In Gillette, D. G. & M. G. Lockley (eds.). 1989. Dinosaur tracks and traces. pp 301-307. Cambridge University Press. 454 pp. 87 References ________________________________________________________________________________ Zhen, S., L. Jianjun., R. Chenggang., N. J. Mateer & M. G. Lockley. 1989. A review of dinosaur footprints in China. In Gillette, D. G. & M. G. Lockley. (eds.) 1989. Dinosaur tracks and traces. pp 187-197. Cambridge University Press. 454 pp. 88 Appendix 1. Published work ________________________________________________________________________________ Appendix 1. Published work connected to this study The following is a list of published articles and talks comprising various topics of tetrapod Ichnology related to my research during the thesis. Papers Milàn, J. & R. G. Bromley. (Submitted). The impact of sediment consistency on track- and undertrack morphology: experiments with emu tracks in layered cement. In Rainforth, E.C. & McCrea, R.T. (eds). Lithichnozoa: Fossil vertebrate footprints. Indiana University Press. Milàn, J. (in press). Dolkhale, flyveøgle eller hoppende dinosaur? VARV (in Danish) Anonymous [Milàn, J & E. Håkansson,] 2003. Ny viden om dinosaurspor. National Geographic Danmark, 3(2003): 9. (in Danish) Milàn, J. & R. G. Bromley. 2002. The influence of substrate consistency on footprint morphology: field experiments with an emu. The Palaeontological Association Newsletter 51: 104 Milàn, J. 2002. Emu- og rovdinosaurspor. VARV 2: 18-20. (in Danish) Milàn, J. 2002. De første spor på land –hvem satte dem egentligt?. VARV 2: 11-17. (in Danish) Milàn, J. 2002. På sporet af dinosaurerne. Dinosaurliv, tillæg til Dagbladet Politiken.16/3-2002. pp10-11. (in Danish) Milàn, J. 2001. Dinosaurspor. Naturens Verden 84(11,12): 22-31. (in Danish) Milàn Nielsen, J. 2001: På sporet af dinosaurerne. VARV 2: 3-7. (in Danish) 89 Appendix 1. Published work ________________________________________________________________________________ Milàn Nielsen, J. 2001: ”Dino Tracking”-på sporet af fortidens kæmper. Hovedområdet. Det Naturvidenskabelige Fakultet ved Københavns Universitet. 2: 19-20. (in Danish) Abstracts and talks Milan, J. (Submitted). Vertical sections through a theropod track – revealing the hidden ichnological information. SVPCA, 51st Symposium of Vertebrate Palaeontology and Comparative Anatomy, 17- 21 September 2003, Oxford. Milàn, J. 2002. Emu og rovdinosaur fodspor: eksperimentel og sammenlignende ichnologi. Palaeontologisk Klub, Geological Institute, University of Copenhagen 26/11-2002. Abstract 1 p. (in Danish) Milàn, J. & R. G. Bromley. 2002. One foot, many tracks: experiments with undertrack formation in layered cement. In Norman, D & Upchurch, P. (eds.). Abstract volume, SVPCA, 50th Symposium of Vertebrate Palaeontology and Comparative Anatomy, 11-13 September 2002, Cambridge. Milàn, J. 2002. Dinosaur and bird tracks, comparisons and confusions. In Bird Evolution, Abstract Volume. Palaeontologisk Klub, Geological Institute, University of Copenhagen. Abstract 2 p. Milàn, J. 2002. Emu tracking – applications for dinosaur ichnology. Speech at the Joly Society, Trinity College, Dublin, Ireland 11/4-2002. Abstract 1 p Milàn, J. 2002. I dinosaurernes fodspor. Speech at Folkeuniversitetet Ålborg. 27/2-2002(in Danish) Milàn, J. 2002. I dinosaurernes fodspor. Speech at Folkeuniversitetet Odense. 26/2-2002(in Danish) Milàn Nielsen, J. 2001: I dinosaurernes fodspor. Speech at Dansk Naturhistorisk Forening 11/102001, Zoological Institute, University of Copenhagen. Abstract 1 p. (in Danish) 90 Appendix 1. Published work ________________________________________________________________________________ Milàn Nielsen, J. 2001: En emu træder i det! –eksperimenter med spormorfologi udført med emu i sedimenter med forskellig konsistens. In Sporfossiler, Kalksedimenter, Rhodos, Abstract Volume, Palaeontologisk Klub, Geological Institute, University of Copenhagen. Abstract 2 p. (in Danish) 91 Tracking surface Overall track Raised rim Track wall True track Undertrack Figure 1. The terminology used to describe a track. Figure 2. The phylogenetic relationships of the ratites. (A) Simplified cladogram of Archosauria. Birds and crocodiles are the only groups living today. (B) Cladogram of Aves. The palaeognate and neognathe birds constitutes the extant crown group of Aves. The dromaeosaurs are the sister group of Aves and are thus closer related to birds than to other theropods. (C) Cladogram of the major groups of modern birds. The ratites together with the tinamous birds comprises the Palaeognathae. The Neognathae comprises a group formed by the Galliformes and Anseriformes and the group to which all other extant birds belong, Neoaves. Based on Dingus & Rowe (1998), Holtz, Jr. (2000), Cracraft & Clarke (2001) & Davies (2002). A B C Figure 3. Feet of the three biggest extant ratiteous birds. (A) Right foot of an emu. The emufoot is tridactyl with long sharp claws. (B) Left foot of an ostrich. Digit III is the main weight-bearing digit in the didactyl foot. Digit II is lost in the ostrich, and digit IV is reduced and bears no claw. (C) Left foot of a stuffed cassowary from the Sheffield City Museum. Digit II- in front- bears a highly elongated sharp claw. Cassowary foot equals the emu foot in size. Figure 4. (A) The skeleton of the emu foot, with the terminology used in the text. Skeleton redrawn from x-ray photo. (B) The skeleton of the emufoot superimposed on an emu footprint to show the correlation between the footprint features and the skeleton. The number of digital pads in digit II and III corresponds the number of phalangeal joints in the digits. The three short phalanges in the middle of digit IV fails to leave impressions of individual digital pads, and only two weakly divided digital pads are visible in the footprint. The rounded metatarsal pad are separated from the digital pads by a deep broad interpad space. Interpad spaces are marked with shaded signature. The phalangeal skeleton is redrawn from x-ray photo (Fig. 3A) and the footprint is redrawn from a footprint impressed in potters clay. A B C Figure 5. (A) The ventral side of the emu foot. The digital pads and the interpad spaces are covered with small horny tubercles. (B) Close up photo of the basal digital pad on digit III, showing the closely packed skin tubercles. The size of the tubercles is about 1 mm. (C) The dorsal side of the foot with the broad overlapping transverse scutes. Photos by Ole Bang Bertelsen. A B Figure 6. X-ray photos of an emu foot, taken with different intensities. (A) Image taken with 43,1 kv. Only the bones and the ligaments connecting them are visible at this intensity. (B) Image taken with 28,3 kv. At this intensity the soft parts of the foot are visible together with the bones, the latter only visible as a black shadow, revealing little details. 5 cm Figure 7 Figure 8 Figure 7. The Lower Cretaceous bird footprint Aquatilavipes curriei from Alberta. Dimensions L x W approx. 8 x 10 cm, divarication angle 120 degrees (From McCrea & Sarjeant 2001). Figure 8. Plaster cast of recent emu track. Dimensions L x W 19.1 x 15.2 cm, divarication angle 85 degrees. 5 cm Figure 10 Figure 9 Figure 9. Typical theropod (Grallator) track. Dimensions L x W approx. 15 x 7 cm, divarication angle 38 degrees (after Olsen et al. 1998). Figure 10. Track with birdlike affinities from the Late Triassic of Argentina, divarication angle 120 degrees (from Melchor et al. 2002). Local soil Phi unit Mesh size µm Weight g Weight percent Cum. percent 0 1000 16.6 8.3 8.3 1 500 27.7 13.9 22.2 2 250 51.3 25.7 47.9 3 125 41.4 20.6 68.5 4 63 38.1 19.0 87.5 <5 <63 24.1 12.5 100.0 % Grain size (phi units) Beach sand Phi unit Mesh size µm Weight g Weight percent Cum. percent 0 1000 12.1 6.1 6.1 1 500 28.1 14.1 20.2 2 250 87.4 43.7 63.9 3 125 62.9 31.5 95.4 4 63 5.2 2.6 98.0 <5 <63 4.8 2.4 100.4 % Grain size (phi units) Red sand Phi unit Mesh size µm Weight g Weight percent Cum. percent 0 1000 12.4 6.3 6.3 1 500 17.8 9.0 15.3 2 250 49.8 25.2 40.5 3 125 62.3 31.5 72.0 4 63 31.7 16.0 88.0 <5 <63 23.9 12.1 100.0 % Grain size (phi units) Figure 11. Grain size graphs and sorting schemes for the three different sediments used for trackmaking. A C E B D F . 12. Figure The walking cycle of the emu. (A) The left foot is put forward and down in the T-phase, while the right foot is in the end of the W-phase. (B) Left foot is in the W-phase with the weight of the animal centred directly above the foot. The right foot is in the end of the K-phase, only touching the ground with the tip of digit III. (C) The left foot is in the end of the W-phase, while the right foot is moved through the air towards a new K-phase. (D) The metatarsal area of the left foot is lifted in the K-phase while the right foot is in the start of the W-phase. (E) Left foot is lifted clear of the ground and forward towards the T-phase, the right foot is in the middle of the Wphase with the weight centred directly above the foot. (F) Left foot enters a new T-phase and right foot is lifted in a new K-phase completing the cycle. A B C Figur 13. Emu trackways. (A) Trackway of emu running at moderate speed. Notice the low pace angulation, average 142° and the wide trackway. (B & C) Trackways of walking emus. The pace angulation are higher than in the running emu, average 170° (B) and 162° (C). Trackway A are drawn to different scale than B and C. Scale bars equals 50 cm. Trackways redrawn from photos. Figure 14. The measurements used to describe fossil and recent trackways, exemplified with a theropod, Eubrontes, trackway. Figure modified after Lockley (1991). Figure 15 A Figure 16 Figure 15. Trackway where an emu makes sharp turn to the right and starts to run. Notice how the stride length decreases just before the turn is initiated. Track R4 is the acceleration track where the emu kicks off to run. Trackway redrawn from photo. Figure 16. Plaster casts of emu tracks. (A) Track produced during normal walk is equally impressed and well-defined. (B) The track where the emu accelerates is deeply impressed in the distal parts of the digits, while the metatarsal pad is almost non-present in the track. The digits have diverged during the formation of the track causing the trackwalls of the digits to slope. Figure 17. The divarication angle increases from the surface to the bottom of the track. Black lines represent length axis of the digits at the surface and red line at the bottom. B A D G B C E F H I Figure 18. The different appearances of emu footprints according to in which sediment they are impressed. See text for details. (A) Dry sand. (B) Damp sand. (C) Wet sand. (D) Thin soft layer of mud. (E) Deep firm mud. (F) Deep semi-firm mud. (G) Deep semi-fluid mud. (H) Deep fluid mud. (I) Potters clay. A B Figure 19. Sediment transport by the foot. (A) The sediment sticks to the sole of the emu´s foot, carrying an amount with it at each step, which is gradually replaced with new sediment. (B) A “shadow” of the track is produced when the emu walks out from the sand, and into firm soil where the sand gradually is released from the sole of the foot. Photo A by Richard G. Bromley. A B C Figure 20. Various degrees of impression of digit II in emu tracks made in damp sand during normal walking. (A) Digit II impressed to approximately equal depth of digit III & IV. (B) Digit II only vaguely impressed, leaving a faint trace of the digit and the claw. (C) Digit II not impressed in the sediment, only a small mark from the tip of the claw hints the existence of the digit. Photos by Jesper Milàn. Figure 21. Emu resting in plantigrad position, with the anatomical heel in contact with the ground. Photo by Richard G. Bromley. Figure 22. Cast of emutrack with full metatarsus impression. The track is obtained while the emu were feeding on seeds strewn on the ground. Scalebar 10 cm. Photo by Ole Bang Bertelsen. Figure 23. Plaster cast of emu footprint with striations on the sides of digit III and on the metatarsal pad. The striations on the proximal part of digit III and on the metatarsal pad are formed by the skin tubercles during the T-phase, while the striations on the distal pad is created during the K-phase. Photo by Ole Bang Bertelsen. A B Figure 24. Striations from movement of digit II in a track sat in deep mud. (A) Striations on the side of the digit and the broad mark from the claw that has sliced through the sediment as the digit was moved backwards and up in the K-phase. (B) Multidirectional striations on the side of the digit capturing the movement of the digit as it was first pressed down in the T- and Wphase, and afterwards dragged backwards and up in the K-phase. Figure 25. Striations produced by skin tubercles in Deltapodus Brodericki. Natural cast of a supposed stegosaur ichnite with well-preserved striations on the sides and the digits of the track. The orientation of the striations show they were formed while the foot was lifted and moved forward. L3 R4 R3 L2 R2 L1 R1 A B Figure 26. (A) Trackway of a theropod making a sharp turn to the right. Notice the similarity in footprint pattern in the trackway of the turning emu (B). Homologous tracks in the emu trackway and theropod trackway are numbered equally. Theropod trackway redrawn after Currie (1989). Figure 27. Left footprint with metatarsus impression from an Early Jurassic Anomoepus major resting trace. The ichnogenera belong to a small ornitopod dinosaur. The proximal part of the metatarsus is deepest impressed like in the resting trace from the emu. Dotted line indicates missing or weakly impressed parts. Compare with resting track of emu (Fig. 22). Figure redrawn after Hitchcock (1858:plate VIII). Figure 28 Figure 29 Figure 28. Hitchcock´s model of deformation of the sediment layers below a tridactyl foot. Figure from Hitchcock (1858). Figure 29. Model explaining how the weight of an animal can be transferred to lower layers if the animal walks on a sloping surface. Figure from Hitchcock (1858). Figure 30. Limulid “undertracks”. The walking appendages has penetrated the sediment to different depths creating morphologically different trackways when exposed at different levels below the original sediment surface. Figure from Goldring & Seilacher (1971). Figure 31. “Subtraces” or “ghost prints” as illustrated by Leonardi (1987). Figure 32. “Underprints” censu Thulborn (1990). After infilling, the track can be found exposed to different depths due to erosion. If different parts of the track are unevenly depressed, the track will appear morphologically different with depth. Figure from Thulborn (1990). Figure 33. “Transmitted tracks” or “Ghost tracks” The weight of the trackmakers foot deforms the subjacent sediment layers in which a stack of successively shallower and less detailed tracks form. Figure from Thulborn (1990). Figure 34. Lockley´s (1991) illustration of the undertrack concept. This model is identical to that of Thulborn (1990). Figure from Lockley (1991). Figure 35. Experiment with trackformation in layered plasticine, notice the prominent undertracks formed below the track. Figure from Allen (1997). Figure 36. Vertical section through a human footprint emplaced in a package of alternating soil and flour to demonstrate the “Concentric rings of distortion” around a track. This phenomena is identical to “undertracks” censu Lockley (1991). Figure from Brown (1999). III 1 2 3 4 II IV 5 5 6 7 8 9 6 10 11 7 1 4 8 2 9 3 10 11 Figure 37. Track emplaced in cement package with water: cement ratio of 129 mL/kg. Arrows and numbers correspond to the described slices. All sections in frontal view. See text for details. III 1 2 3 4 II 5 IV 5 6 7 8 9 6 10 11 7 1 8 2 9 3 10 4 11 Figure 38. Track emplaced in cement package with water: cement ratio of 143 mL/kg. Arrows and numbers correspond to the described slices. All sections in frontal view. See text for details. III 1 2 3 5 4 5 IV 6 II 7 8 6 9 10 11 7 1 2 8 9 3 10 4 11 Figure 39. Track emplaced in cement package with water: cement ratio of 157 mL/kg. Arrows and numbers correspond to the described slices. All sections in frontal view. The vertical white line in the middle of sections is glue as the block cracked during the slicing. See text for details. III 1 2 3 4 5 5 6 II 7 IV 8 9 10 6 11 7 1 8 2 9 3 10 4 11 Figure 40. Track emplaced in cement package with water: cement ratio of 171 mL/kg. Arrows and numbers correspond to the described slices. All sections in frontal view. See text for details. III 1 2 II 1 IV 3 4 5 6 2 A III IV II 3 4 C III II IV 5 6 D B Figure 41. Track produced in a water: cement mixture of 143 mL/kg. (A) The track walls in the true track have converged after the foot is lifted, making the digit impressions appear narrower than they actually are. (B) Sections through the cement package revealing the undertracks. The number on slices correspond to numbers on the true track (A), see text for details. After removal of the red layer two surfaces with undertracks were exposed. (C) One surface with the cast of the undertrack as seen on the underside of the upper blue layer, (D) and one with the undertrack in the lower grey layer. See text for details. 1 1 2 2 3 3 4 4 A B III II IV C Figure 42. Track produced in water: cement mixture of 171 mL/kg. (A) The true track on the surface is collapsed and distorted by the sediment displacement by the foot movement. (B) Sections through the track reveal the subsurface deformation and formation of undertracks. Number on slices correspond to numbers and arrows on true track, see text for details. (C) The undertrack exposed at the lower blue surface, notice that in this case the undertrack actually presents the shape of the foot better than the true track. A B C D E Figure 43. Track and undertrack formation at five subjacent horizons. (A) True track at the surface. (B) Undertrack at 1 cm depth. (C) Undertrack at 2 cm depth. (D) Undertrack at 3 cm depth. (E) Undertrack at 4 cm depth. III IV II A B C D E F G H I Figure 44. (A) Cast of the emu track used in the experiment. Notice the presence of the drag mark from the forward movement of the claw of digit III. (B, C) Horizontal sections through the track near the surface. The drag mark from digit III´s claw prevents measurements of the track length. (D, E) Well-defined tracks with complete outlines from sections cut at 14 & 17 mm depth. (F, G) The individual digits detach from each other at depths of 25 & 28 mm. (H, I) At the bottom of the footprint at depths of 38-41 mm only the deepest impressed parts of the foot has left impressions. Digit IV disappears first, followed by digit II, leaving a track consisting of only the metatarsal pad and digit III. * Scarborough Figure 45. Location map. III II A B Overlying sediment IV Track exposed in vertical section Track exposed in Horizontal section II III C Figure 46. (A) Theropod track with preserved sedimentary infilling partly exposed by erosion of the embedding rock. (B) Outline of the horizontal section of the track highlighted. (C) Headon view of the track revealing that different parts of the track are exposed in both horizontal and vertical sections. The sediment has collapsed above digit III, causing dramatic alternations of apparent digit width with depth. Outline of the track is highlighted by solid line and the dotted line represents the original tracking surface. Track width III II IV Track lenght Divarication angle A B Figure 47. (A) Ideal Grallator track superimposed by the pedal skeleton. Shaded area represents the typical features of the track. The track is a right foot. Notice the asymmetry of the proximal end of the track, and that the claw of digit III is offset towards the midline of the trackway. (B) The measurements used in describing the footprints. Figure after Olsen et al. (1998). III IV II A B Figure 48. Track NB1. (A) The weathered appearance of the track prevents recognition of any but gross morphological details. (B) Outline with track features highlighted, and the raised rim of displaced material. Scale bar 10 cm. III IV II I A B Figure 49. Track NB2. (A) The soft sediment in which the track is impressed has caused the track to partly collapse after withdrawal of the foot. The slab containing the track has fractured along the length axis of digits II, III & IV. (B) Interpretative drawing of the track outline, notice the presence of the small digit I. Arrows indicate areas with skin striations preserved. Scale bar 10 cm. III II IV A B Figure 50. Track LC1. (A) Shallow but well defined track from a medium-sized theropod. Individual digit impressions are only vaguely connected. (B) Interpretative drawing of outline and morphological features of the track. Scale bar 10 cm. III II IV b A B Figure 51. Track LC2. (A) Distal part of tridactyl track. Individual phalangeal pads are well defined on the digits. Clear claw impressions present on digits III & IV. (B) Interpretative drawing highlighting anatomical features of the track. Scale bar 10 cm. B A C Figure 52. Striations from the skin tubercles preserved in track NB2, the striations appear in the trackwalls on both sides of digit III. (A) Striations on left side of digit III. (B) Striations from the right side. The upward and forward orientation shows that the striations were formed while the foot was being lifted out of the sediment. (C) Striations of similar size and appearances in the trackwalls of the cast of an emu track sat in deep sticky mud, striations on the metatarsal pad and in the proximal end of digit III were formed during the touch-down of the foot, whereas striations in the distal end of the digit were produced during the lifting of the foot. The striations in the emu and the theropod track is of equal size, approximately 1.5 - 2 mm vide. 11 10 9 8 7 Figure 53. Sections of track NB1. The sections are numbered 1 to 11 from the proximal end to the tip of digit III. See text for details. Continues on next page. 6 5 4 3 2 1 Figure 53 continued. Porto * Lourinhã Lisboa Figure 54. Location map. m Figure 55. Simplified sedimentological log of the cliff section exposed at the track locality. Silhouette of track indicates the track-bearing horizon. Log by Jesper Milàn & Sverre Jensen. Figure 56. The trackway as it appeared in-situ, located on the underside of a sandstone bed. Photo Sverre Jensen. Figure 57. Map of the track assemblage. (A) Tracks 1 to 3 forms a small trackway. Track 4 is from an animal crossing the path of tracks 1 to 3. Track 5 is from a smaller animal that made a sliding movement of the foot during the stride. Notice the presence of radiating fractures in the sediment around tracks 1 and 2. (B) The track-bearing horizon in horizontal view. Notice the northward deepening of the tracks. Figure 58. Close up photo of track number 1 from the trackway. Part of the track is still covered by the clay in which it was impressed. The clay layers are bent around the track. Notice the layered infilling of the track and the prominent casts of the radiating fractures in the sediment around the track. Figure 59. The cast of track 5 after removal of the embedding sediment. Notice the asymmetry of the track showing that the trackmaker has made a sideways movement of the foot. Striations in the trackwalls indicate the presence of tuberculous skin. Hip height Figure 60. The measurements used for calculating the progression speed from a trackway.