A new giant theropod dinosaur track from the Middle Jurassic of the Cleveland Basin, Yorkshire, UK

Although dinosaur tracks from the Middle Jurassic of the Cleveland Basin are relatively common (Whyte and Romano 2001, 2002; Romano and Whyte 2003, 2012; Whyte et al. 2007; Romano et al. 2018), large, well-preserved theropod tracks are rare. Early records of the latter include those of Black et al. (1934, p. 299, fig. 20B) who documented a ‘25 inch’ (63.5 cm) long print from the Scalby Formation of Burniston Bay, 4.5 km north of Scarborough, and a slightly smaller print (61 cm), also from Burniston Bay, was recorded by Ivens and Watson (1994, p. 11) although no photographic record exists of this latter print. Earlier, in 1963, Mr C. R. Ivens had found a tridactyl print in South Bay, Scarborough, that was even larger (65 cm) (Ivens and Watson 1994, pp. 14–15). A concrete mould of this specimen (Accession No: SCARB:2004.2055) and corresponding plaster cast (Accession No: SCARB:2004.1833) are housed in the collections of Scarborough Museum. Since these early reports, the only record of large theropod prints were by Whyte et al. (2006) who described and figured a 61 cm long print also from Burniston Bay, and Whyte et al. (2007, fig. 8) who figured a slightly smaller and incomplete specimen. The specimen described by Whyte et al. (2006), and later refigured by Lomax and Tamura (2014, fig. 145) remains the only detailed modern description of a large theropod track from the Middle Jurassic of Yorkshire. It was assigned to Morphotype Bxviii (see below), based on the nomenclature adopted in Romano and Whyte (2003, pp. 211–212). The specimen is currently on display (Accession No: SCARB:2007.54) in the Rotunda Museum, Scarborough. Recently, another larger track of this type was discovered by one of us (RT) and rediscovered several months later (by MW), also from Burniston Bay (Fig. 1). In view of the rarity of these tracks, this specimen (Accession No: SCARB:2023.001) is described here and compared to those of previous records, along with an interpretation of the possible behaviour of the animal responsible.

Ichnogenus Megalosauripus Lessertisseur, 1955.

Ichnogenotype stated as Megalosauripus uzbekistanicus (Gabuniya and Kurbatov 1982) from the Upper Jurassic of Turkmenistan by Lockley et al. (1998, p. 324). But type ichnospecies identified as Megalosauripus maximus (Kuhn 1958) by Thulborn (2001, p. 218) from the Lower Cretaceous Bückeburg Formation, Harrl-Steinbruche, Germany.

Diagnosis of Megalosauripus (amended) given in Lockley et al. (1998, p. 324), and a more extensive revised diagnosis in Thulborn (2001, p. 218). See comments below under Ichnotaxonomy.

Megalosauripus ichnosp.

SCARB:2023.001 is a single underprint preserved as a positive hypichnion (hyporelief) on the underside of a 24–35 cm thick slab of silty sandstone c. 130 × 82 cm before trimming. The sandstone shows sedimentary features including cross-bedding, faint parallel lamination and sandstone clasts.

The fallen loose block on which the print occurs was found c. 40 m north of the steps at Crook Ness, Burniston Bay, 5 km north of Scarborough (Fig. 1). It was traced to a bed in the cliff face c. 1 m below the Burniston Footprint Bed (Hargreaves 1914) of the Long Nab Member of the Scalby Formation (Figs 2, 3). The bed was exposed in 1997 for 30 m along the cliff face as a lens-shaped unit, up to c. 35 cm thick, underlying a dark carbonaceous band and with a flattish top and irregular dinoturbated (sensu Dodson et al. 1980, p. 229; Lockley 1991, p. 215) base. It is of Bathonian (Middle Jurassic) age.

The large tridactyl weakly mesaxonic print (Figs 4, 5) is 80.0 cm long (FL = Footprint Length), from the claw imprint on the middle digit to the back of the metapodium (see later), and 41.0 cm wide (FW = Footprint Width). The digit divarication is c. 45°, although this was taken on the proximal part of the curved axis of the left digit (when viewed from below). If the distal part of this digit imprint was used, the divarication would be less than 30°. The two other digit imprints are long and straight, gradually narrowing and with a sub-triangular claw imprint clearly marked on the central digit. At the termination of the central and right hand digit imprints are long thin claw drag marks directed to the right. The 3.5 cm long central drag imprint is the longest (c. 15% total length of the track), and appears to originate to the right and below (stratigraphically higher than) the sub-triangular termination of the digit. The right side drag imprint is slightly shorter but extends along the same axial direction as that of the central digit. The left digit does not appear to show any claw drag imprint, but is poorly preserved. The print does not show any clear indication of a hallux.

All three digits show swellings, phalangeal pads/nodes, along their lengths. The central digit shows possibly three digital nodes. The left digit (seen from below) shows indistinct digital nodes, and the right digit two distinct phalangeal nodes. The specimen shows a very prominent posterior elongation with a rounded termination, the metapodium (Thulborn 1990, p. 79), that occupies nearly 50% of the total length of the print. There are striations on the surface of the print, running from the metatarso-phalangeal node to the rear of the metapodium. Oblique striations are present at the rear of the metapodium, and imprints of skin tubercles occur on the distal part of the metapodium and on the right digit. The deepest part of the print is around the metatarsal/proximal phalangeal nodes area, and the shallowest parts are occupied by the metapodium and digit imprints.

In the absence of a trackway it is difficult to be sure whether the track was made by a left or right foot. The direction of curvature of the claws is not always a reliable feature in determining this, while a more useful parameter is the posterior extension of the imprint of digit IV, produced by the metatarso-phalangeal node (Thulborn 1990, fig. 5.4). In the present specimen the right digit (viewed from below) appears to exhibit the more posterior node imprint. If this is so, then the print was probably made by a left pes. However, we regard this as rather tenuous evidence.

The track would be classified as a ‘giant’ track of Marty et al. (2018) when the length of the track is >50 cm. While the track is broadly similar to that described by Whyte et al. (2006) (Fig. 5b) in terms of size, it contrasts in a number of important features. The present specimen shows a considerably smaller divarication of the digits; c. 45° as against c. 70°, although this may be due to the strongly curved proximal part of the right digit in the latter. The claw marks are better defined in the present specimen, as is the distinct metapodium imprint. Whether these differences are significant enough to exclude the present specimen from being included within Morphotype Bxviii for specimens described and discussed by Whyte et al. (2006, p. 79), is perhaps more uncertain, and there remains the options to either modify the description of this morphotype, or propose and describe a new morphotype. Since at this time we have only a single specimen of this type from the Middle Jurassic of the Cleveland Basin, we prefer provisionally to include it within Bxviii, and to slightly modify the description of this morphotype (Whyte et al. 2006). In doing so, this approach avoids the ever increasing need to describe new morphotypes for any specimen that shows slight morphological differences from those specimens previously assigned to those morphotypes. The reasons for initially adopting the term morphotypes for the majority of the Cleveland Basin print types was explained in Romano and Whyte 2003 (p. 211). Since then, a number of the morphotypes have been assigned to formal ichnotaxa (Romano and Whyte 2010). The specimens figured by Black et al. (1934) and Whyte et al. (2007) (Fig. 5c, d), though lacking in detail, shows a divarication fairly similar to that of the present specimen, while the specimen described by Ivens and Watson 1994 (Fig. 5e) is too poorly preserved to allow a meaningful comparison.

The proposed revised description of Morphotype Bxviii is now:

Large mesaxonic print up to at least 55 cm long, and 40 cm wide with sub-triangular heel, with or without a prominent metapodium, consisting of distinctly clawed digit imprints with phalangeal pads. Divarication between 45–70°.

This is not a formal diagnosis, but serves to identify a number of prints from the Middle Jurassic of the Cleveland Basin that are distinct enough to warrant a separate grouping (Romano and Whyte 2003, p. 213).

The morphological differences in the present specimen to those seen in other specimens assigned to Morphotype Bxviii may be explained by: (1) different modes of print formation and/or substrate, (2) different behaviour at the time of formation, or (3) different foot morphology. In the first of these (1), the print described by Whyte et al. (2006) resulted from the foot being impressed through at least 2 cm of sandstone that produced digit imprints being bounded laterally by bulbous lobes of sediment that were transmitted during formation. No such lobes are present in the present specimen although the foot was also probably made on a sand substrate which then impressed through into the underlying finer grained sediment (see below). The possible different behaviour of the animal making the print (2) is discussed below (Behaviour). Finally, (3) may simply reflect different foot morphologies, hence a difference in species of trackmaker, a fact which at present we are unable to confirm.

There remains the question as to whether the print may be assigned to any named ichnotype. The specimen described by Whyte et al. (2006) was compared to prints of Eubrontes and Megalosauropus (Thulborn 1990; Lockley et al. 1996). Later Lockley et al. (1998) Lockley and Meyer (2000) and Thulborn (2001) made the case that Megalosauripus is the correct spelling for Jurassic material while Megalosauropus is reserved for Cretaceous material. While not wishing to comment here on this particular nomenclatorial discussion, it is clear that differences of opinion exist in the diagnosis at ichnogeneric level. Thus, for example, Lockley et al. (1998, p. 324) stated that Megalosauripus is a tridactyl track, while Thulborn (2001, p. 218) considered it to be tetradactyl. The specimen described here shows features in common with some of the forms assigned to this ichnogenus.Megalosauripus is the most commonly quoted ichnogenus for Late Jurassic large theropod tracks (Rauhut et al. 2018, and references therein) and includes the ichnospecies M. uzbekistanicus and M. teutonicus (Lockley and Meyer 2000; Cobos et al. 2014; Marty et al. 2018).

Lockley et al. (1998, figs 4, 10), Lockley and Meyer (2000, figs 7.2–7.6) and Thulborn (2001, figs 1–4) figured examples of Megalosauripus that have features in common with the Burniston print; in particular, the size and pronounced metapodium. Yet the divarication of digit imprints II to IV in the revised diagnosis for Megalosauripus (and specifically M. maximus) given by Thulborn (2001, p. 218) is 65°–70°, which although considerably greater than in the present specimen, is similar to that recorded by Whyte et al. (2006) for their specimen. No divarication angle is given in the diagnosis by Lockley et al. (1998). However, ‘pronounced irregularity has been noted in trackways attributed to Megalosauripus’ (Lockley and Meyer 2000, fig. 7.7), and Thulborn (1990, fig. 5.5) illustrated how variation in footprint width of Neotrisauropus deambulator, those of a theropod from the Lower Jurassic of Lesotho, may occur along a single trackway. Also, the metapodium in the present specimen of Megalosauripus ichnosp. constitutes nearly 50% of total footprint length. Until more material is known from the Scarborough region, together with a trackway, such as that known for M. transjuranicus (Razzolini et al. 2017), and elsewhere (Razzolini et al. 2016), a final assignment must remain tentative.

Rauhut et al. (2018) figured and described two morphotypes of giant theropod footprints from the Kimmeridgian of Asturias, north Spain. These authors recognize tracks that are weakly mesaxonic (their Morphotype A) in which specimens reach up to 80 cm in length and have a moderately long metapodium (Rauhut et al. 2018, fig. 5), and those (their Morphotype B) that are strongly mesaxonic, much longer than wide and with a relatively longer metapodium (Rauhut et al. 2018, fig. 6B, E). Sub-horizontal striations are present in the walls of the digit imprints in these Asturian examples (Dra Laura Piñuela; pers. com. to MR) that may indicate forward movement of the foot that resulted in the elongated metapodium. These Spanish specimens show claw impressions which vary from narrow and short to wide and long. The present specimen from Yorkshire shows more in common with those Asturian specimens of Morphotype B, although contrasts in being more weakly mesaxonic.

There remains the discussion as to how the present specimen should be assigned. We have chosen Megalosauripus ichnosp. based on previous records of prints assigned to this ichnogenus and large theropods in general have pronounced (sub)triangular claws (Marty et al. 2018) as in the present specimen and are unlike those figured for the North American ichnogenus Dilophosauripus (Lockley and Hunt 1995, fig. 4.7B; Milner et al. 2009, fig. 6E). As Rauhut et al. (2018, p. 14) point out, until a further detailed revision of the ichnogenus Megalosauripus is undertaken, the assignment of their Asturian material to that ichnotaxon must remain tentative. In the case of whether tridactyl or tetradactyl prints are characteristic of this ichnogenus, preservation must also be taken into account. A shallow print onto a firm substrate will not necessarily leave an imprint of the fourth digit, or hallux, while a more deeply impressed print into softer substrate may allow the hallux to make contact with the sediment surface. As discussed below, the present print was made on unconsolidated sediment and penetrated the substrate to a depth whereby the presence of a fourth digit/hallux would have been expected to have been made. However, it should be noted that a minor protuberance posterior to the right digit (viewed from below) could indicate a possible hallux, but it is impossible to confirm.

A large footprint with clawed, padded digits and low divarication (FL>FW) suggest the maker was a very large, probably megalosaurid theropod. By their very nature at the top of a food chain (Cobos et al. 2014), their numbers would be expected to be considerably less than those lower down the food chain and consequently would produce fewer footprints that potentially may be fossilized (Romano et al. 2018, fig. 6). This may explain the general paucity of these prints in the Middle Jurassic of the Cleveland Basin. Based on the full footprint length (80.0 cm), using the morphometric ratio approach, where hip height (h) = 4.9 footprint length (FL) for large theropods (Thulborn 1990, p. 251), the hip height of the maker of the present specimen would be an estimated c. 4.0 m. If allometric equations are used (Thulborn 1990, p. 254) the hip height would be slightly less. A more recent paper by Henderson (2003) suggested that a factor of 4FL as originally suggested by Alexander (1976) should be used to determine hip height. If this ratio is followed the present specimen would have had a hip height of around 3.2 m. However, all these estimates include full extent of the metapodium in the calculation of FL (see Behaviour section below). Therefore, if the metapodium imprint at the back of the print is not taken into account, and that the true length of the foot (print) is c. 60 cm, then the hip height of the maker would probably be between 2.5–3.0 m, which we feel would be more accurate.

Irrespective of whether the assignment to Megalosauripus ichnosp. is correct, it is proposed that the possible maker of the print was something akin to Megalosaurus. The type species, Megalosaurus bucklandii, was thoroughly revised by Benson (2010) who provided a new diagnosis and described additional material all known from Middle Jurassic (Bathonian) sites in England (Molnar and Farlow 1990, p. 191; Whyte et al. 2006, p. 80; Benton and Spencer 1995). Although speculative, it is possible that the described print could have been made by a Megalosaurus-like theropod based not only on the size and morphology of the print, but also the age which is similar to the stratigraphic range known for Megalosaurus (Benson 2010).

It is generally recognized that an animal showing various behaviour patterns may produce traces that are different, and that these may be assigned to different ichnotaxa. A well-documented example of this in the invertebrate fossil record are the ichnogenera Rusophycus, Cruziana and Diplichnites produced by trilobites and possibly by the same individual (Crimes 1970; and references therein). Examples from dinosaur traces are known where habitual bipeds have rested/crouched during more normal locomotion and left a variety of additional impressions that have been interpreted as having been made by the metatarsals, ischial callosity, manus and tail (Thulborn 1990, figs 4.13, 4.14; Lockley and Meyer 2000, fig. 5.9; Lockley et al. 2003; Milàn et al. 2008, fig. 3; Milner et al. 2009, fig. 4). These impressions may have been given different ichnofossil names from those tracks made by the same animal during walking or running.

The pronounced metapodium of the present example was made by the sole of the foot. But does this print represent the pes imprint of this bipedal dinosaur during ‘normal’ locomotion, or that of a resting/crouching animal? The absence of a well-marked metapodium in the previous specimen(s) assigned to Morphotype Bxviii may simply be due to these animals adopting a more digitigrade as against plantigrade posture of the present specimen, which may reflect different speeds of locomotion. In general, three-toed dinosaurs were digitigrade and prints left by such animals would normally just record the medial to distal ends of the digits. However, with a different posture or locomotion speed the rear parts of the digit may be imprinted, and with a plantigrade posture a heel-like impression may result (Thulborn 1990, fig. 4.6). Without the benefit of a trackway, this cannot be confirmed. Another possibility of recording a metapodium would result from a deeply impressed print made by a digitigrade animal, where the front of the foot sunk more deeply into the substrate while the rear of the foot made a shallower impression in the substrate. Clearly the state of the substrate is of importance and whether the animal was moving over firm or soft ground. Since the deepest part of the present print is the central region this does not appear to be the case with this specimen.

Thus, the presence of the metapodium may result from behavioural activity when the animal rested by squatting or laying on the ground. However, Lockley and Meyer (2000, fig. 7.7B) figure a trackway attributed to Megalosauripus in which the pes prints show elongation of the back of the print, and another of Eutynichnium lusitanicum from the Late Jurassic of Portugal (Lockley and Meyer 2000, fig. 7.1) with pronounced metapodium elongation of the pes prints. Both of these trackways appear to be the result of a normal gait, and of an animal walking with a more plantigrade posture. Razzolini et al. (2016, fig. 6) figured a variety of tracks assigned to Megalosauripus from the Middle Jurassic of Portugal which exhibited a range of shapes at the posterior end that included some with an extended metapodium. With just a single print described here, it is not possible to be certain which of the above possibilities applies here, but the authors provisionally prefer to propose a walking locomotion of the maker, though do not rule out resting/crouching activity. It is also significant that those striations observed on the print extending along the length of the print may indicate some forward slipping and so exaggerate the length of the metapodium.

Finally, another interesting interpretation suggested by Dr Neil Clark one of the reviewers, is that the print may indicate that the animal maker originally made the elongated metapodium in a crouched position, then slid forwards, and finally stood more upright resulting in a more typical theropod digitigrade overprint before moving off. This is similar to what was interpreted for the resting trace described by Milner et al. (2009).

Tridactyl tracks that have a significantly elongated imprint to the rear of the print, such as those described by Romano and Citton (2016, fig. 1), Milner et al. (2009, fig. 4) and Lockley and Meyer (2000, fig. 5.9) are quite distinct from the track described here. These tracks are generally interpreted as resulting from a crouching stance where the elongated imprint at the rear resulted from contact with the substrate with the ischial callosity (Milner et al. 2009) or tarsals/metatarsals. The example given in Milner et al. (2009, fig. 3) appears to show a resting/crouching behaviour within a trackway of more normal gait, but the latter prints still exhibit a relatively pronounced metapodium imprint and not the digitigrade print of other dinosaur bipeds.

The general palaeoenvironment of the Long Nab Member in the Cleveland Basin is now generally understood as a low-lying coastal plain and fluvial complex (Alexander 1989). The bed from which the footprint was found comes from a sequence within Burniston Bay that Slater et al. (2017) recovered a palynological assemblage of spores and pollen that indicated a diverse vegetation dominated by Araucariacites within conifer pollen assemblages, and no evidence of marine influence.

The print is preserved at the base of a thin sandstone bed that wedges out to the north and south, and is here interpreted as a crevasse splay deposit. Current work by one of the authors (MR) on a similar unit in the middle of Burniston Bay indicates that such units may sometimes have had a N or NW source. The internal features of the bed indicate that the unit was deposited as a single event. The bed shows cross-bedding on a 12 cm scale that indicates a current indicative of the more proximal part to the flow. Cross-sections through the bed clearly show disruption of the sedimentary features within the bed as the foot penetrated the bed causing loading features (dinoturbation) at the base and fragmentation of the unit resulting in the sandstone clasts. It is not clear however, whether (a) the animal making the prints walked on the surface of the crevasse flow deposit after it had been deposited and settled and before it was consolidated within the interdistributary bay or lagoon, or (b) whether the animal traversed the area during the emplacement of the flow, or even (c) after the crevasse flow had been covered by further fine-grained sediment. The last scenario may explain the incorporation of the muddy layers within the disturbed parts of the sandstone bed.

Special thanks go to Mark, Aaron and Shae Smith from Redcar for cutting and transporting the block on which the specimen occurs into more manageable sizes before removal from the beach. Thanks to John Oxley MBE FRS for helping with photography & survey. Also thanks to the Rotunda Geology Group for their help and support during the removal of the print. MR thanks Dr Laura Piñuela of MUJA, Asturias, Spain, for the helpful constructive comments on the print. Dr Neil Clark and an anonymous referee are thanked for their perceptive comments. Thanks also to Matthew Holmes for creating figure 2 and modifying figure 1.

JGH: formal analysis (equal), investigation (equal), writing – original draft (lead), writing – review & editing (equal); MR: writing – original draft (supporting); DRL: data curation (equal), investigation (equal), writing – review & editing (equal); RT: investigation (supporting); MW: resources (equal)

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.