The world’s oldest cerapodan ornithischian dinosaur from the Middle Jurassic of Morocco

1. Introduction

Cerapoda is a diverse clade of ornithischian dinosaurs with a global distribution. A major component of Cretaceous terrestrial ecosystems, early diverging cerapodans were bipedal, with forelimbs modified for grasping, but by the Late Cretaceous, hadrosaurids and ceratopsids had evolved obligate quadrupedality and sophisticated chewing mechanisms, and they became the dominant herbivores of the Northern Hemisphere [1–5].

Cerapoda is composed of two major clades: Ornithopoda, which includes the non-hadrosaurid iguanodontians and the duck-billed hadrosaurids, and Marginocephalia, which includes the horned, frilled ceratopsians and the dome-headed pachycephalosaurs [6–10]. Although the sister-taxon relationship between these two clades has not been questioned since the advent of cladistic analysis, a great deal of debate remains about cerapodan phylogenetic relationships. A series of small, bipedal ornithischians traditionally known as ‘hypsilophodontids’ are generally recovered as early diverging ornithopods [7,9–11], but Boyd [8] recovered several of them in a clade that formed the sister taxon of Cerapoda. The difference between these two topologies centres on the scoring of just a few characters related to the femur and dentition [12], and we consider these ‘hypsilophodontid’ taxa to be within Cerapoda, following most recent analyses (e.g. [11]). The enigmatic heterodontosaurids, whose members include some of the earliest-known ornithischians, were found as the basalmost clade of Ornithischia by most recent phylogenetic analyses [7–9] but they have also been recovered as a paraphyletic assemblage of early diverging pachycephalosaurs [10], and the idea that heterodontosaurids are related to marginocephalians [13] or lie elsewhere within Cerapoda [6] is not new. The presence of a Gondwanan clade of cerapodans, Elasmaria, that lies either within Ornithopoda [14] or outside of it [8] has also been posited, although the exact membership of that clade varies between analyses.

Cerapodans are well known from the Cretaceous Period, but their Jurassic record is much poorer. Several track sites (e.g. [15–17]) from the Middle Jurassic suggest that large-bodied ornithopods (probably iguanodontians) had already evolved by this time, but their body fossils remain elusive. The earliest definitive cerapodan body fossil is an iguanodontian femur, Callovosaurus leedsi [18], from the Callovian (Middle Jurassic) Oxford Clay Formation of the UK. The neornithischian Kulindadromeus, found in Siberia and preserving bristle-like integumentary structures [19], is of Bathonian age [20], and was recovered within Cerapoda by Dieudonné et al. [10], but has generally been considered as a non-cerapodan neornithischian [19,20]. More detailed descriptions and new specimens of this taxon are needed to adequately resolve its phylogenetic position. Ferganacephale, from the Bathonian Balabansai Formation of Kyrgyzstan, was described as a pachycephalosaur [21], but it is based on a few water-worn teeth, and its pachycephalosaurian status has been questioned [22,23]. Consequently, the oldest marginocephalian is currently Yinlong, from the Oxfordian Shishugou Formation of China [9,13]. To elucidate the early stages of the evolution of Cerapoda and to help resolve the numerous phylogenetic inconsistencies among different analyses, new specimens are needed, especially from historically undersampled localities.

The El Mers III Formation of the Middle Atlas Mountains, Morocco, was deposited on floodplains during the Bathonian [24,25]. The variegated green and red mudstones of the formation are extremely fossiliferous and have so far yielded the remains of the world’s oldest, and first African, ankylosaur, Spicomellus afer [26], and one of the oldest stegosaurs, Adratiklit boulahfa [27]. Here, we describe a fragmentary but definitive cerapodan from the El Mers III Formation. The specimen, the proximal end of a left femur, is the world’s oldest cerapodan and only the second recorded from the Middle Jurassic globally. The fragmentary nature of the material precludes naming it, but our ongoing work in the El Mers III Formation will hopefully reveal more complete specimens of this taxon in the future.

Institutional abbreviations—CMN, Canadian Museum of Nature, Ottawa, Canada; IVPP, Institute of Vertebrate Palaeontology and Palaeoanthropology, Beijing, PR China; MNA, Museum of Northern Arizona, Flagstaff, USA; NHMUK, Natural History Museum, London, UK; OUMNH, Oxford University Museum of Natural History, Oxford, UK; USMBA, Université Sidi Mohamed Ben Abdellah, Palaeontology Collection, Fes, Morocco.

2. Material

3. Description

In anterior view, the femoral head (figure 1A,1, hd) projects dorsomedially from the shaft, and the surface connecting the medial shaft margin and ventral surface of the head is a gentle, continuous concave curve, offsetting the head from the shaft on a distinct neck. A distinct neck between the femoral head and shaft is not present in the earliest-diverging ornithischians and non-cerapodan neornithischians (e.g. Lesothosaurus, NHMUK PV RUB 17 [30,31] Laquintasaura, NHMUK PV R 36796 [32,33], Hexinlusaurus [34]; Sanxiasaurus [35] or in thyreophorans (e.g. Scutellosaurus, MNA V.175 [36], Scelidosaurus, NHMUK PV R 6704 [37]), but is present in all cerapodans [8] (e.g. Hypsilophodon, Parksosaurus, Haya and Minimicursor; [37–40]). The dorsal and medial margins of the head are rounded in anterior view, although the medial margin is slightly eroded and thus appears somewhat flattened. A genuinely flattened ventromedial margin of the femoral head, often marked ventrally by a distinct ridge, is present in early diverging ankylopollexian iguanodontians, including Cumnoria (OUMNH J3303 [17]), Camptosaurus, Uteodon [41,42] and Mantellisaurus (NHMUK PV R 5764; NHMUK PV R 11521 [43]) and is also present in Callovosaurus leedsi (NHMUK PV R 1993). As this feature is absent in dryosaurids (e.g. Valdosaurus, NHMUK PV R 184, 185 [44], Dysalotosaurus, NHMUK PV R 6861), its presence in Callovosaurus calls into question the dryosaurid identification of the specimen by Ruiz-Omeñaca et al. [18]. The head is separated from the greater trochanter by a trough-like depression (the trochanteric fossa or fossa trochanteris; figure 1A, tr), a feature which is absent in early diverging ornithischians and non-cerapodan neornithischians (e.g. Lesothosaurus, Hexinlusaurus, Laquintasaura, Sanxiasaurus [30,32,34,35]), but present in all cerapodans ([7,9,10]; e.g. Hypsilophodon, Parksosaurus, Haya and Minimicursor; [38–40,45]). Although the dorsal surface of the greater trochanter (figure 1A, gt) is eroded, it appears unlikely that the greater trochanter would have reached the same height as the dorsal surface of the head in anterior view. This, and the gentle curvature of the head’s ventral margin, gives the head the appearance of projecting strongly dorsomedially when viewed in anterior or posterior view (figure 1A,B). The combination of a dorsomedially projecting femoral head with the low greater trochanter is an unusual feature within Neornithischia, as, in general, the greater trochanter projects to the same level as the head. However, a similar combination of characters is observed in the neoceratopsians Montanaceratops [46], Auroraceratops (E Morschhauser 2023, personal communication), Archaeoceratops (IVPP V 11115) and the early diverging cerapodan Thescelosaurus (NHMUK PV R 3675 [47]). The anterior surface of the shaft is convex, and an irregular ridge (figure 1A, ar) is present extending from the anterior surface of the greater trochanter ventrally. The anterior trochanter is not evident, although it is likely that it was present as a distinct, finger-like process that arose ventral to the preserved portion of the femur and has not been preserved. This indicates that the lesser and greater trochanters were unfused and separated by a deep cleft, and also that the lesser trochanter must have arisen from a relatively distal position on the proximal femur. Fusion of the greater and lesser trochanters is variable in many ornithischians and may be related to ontogenetic stage in some taxa (S.M., R.J.B. and P.M.B., personal observation), but the distally located position of the base of the lesser trochanter is unusual.

• Download figure
• Open in new tab
• Download PowerPoint

In proximal view (figure 1C), the anterior surfaces of the femoral head and shaft are more or less flat: the head does not extend far anteriorly, although it is slightly abraded. However, there is still a clear constriction between the head and greater trochanter (figure 1C), a feature that is absent in non-cerapodan neornithischians (e.g. Hexinlusaurus; [7,9]). The posterior surface of the head bears a distinct, ‘U’-shaped ligament sulcus in proximal view (figure 1C, ls). More laterally, the head is distinctly separated from the greater trochanter by a second, more pronounced ‘U’-shaped depression (figure 1C, u). The lateral margin of the greater trochanter is laterally convex and abraded dorsally.

In posterior view, the ligament sulcus (figure 1B, ls) occupies most of the posterior surface of the femoral head and is broad and shallow. It extends from the dorsal margin of the head posteroventrally towards the shaft and merges into the medial margin of the shaft ventrally. This feature is widespread in early diverging cerapodans (e.g. Jeholosaurus, IVPP V15939 [48,49]; Hypsilophodon, NHMUK PV R 193 [38]; Thescelosaurus, NHMUK PV R 3675 [47]), dryosaurids (e.g. Valdosaurus, NHMUK PV R 184−185 [44]; Dysalotosaurus, NHMUK PV R 6861 [50]), early diverging ankylopollexians (e.g. Cumnoria, OUMNH J3303, [51]; Camptosaurus, Uteodon Callovosaurus, NHMUK PV R 1993 [17,41]) and early diverging neoceratopsians (e.g. Auroraceratops [52]; Leptoceratops, CMN 8889) but is absent in non-cerapodan ornithischians such as Hexinlusaurus and Lesothosaurus [30,34] and in hadrosauroids (e.g. Mantellisaurus, NHMUK PV R 5674 [43]). The ‘U’-shaped depression separating the head and greater trochanter in dorsal view continues onto the posterior surface of the shaft as a shallow sulcus (figure 1) that extends ventrally, bordered by ridges on either side (figure 1B, ur). This second sulcus is absent in early diverging cerapodans (e.g. Hypsilophodon, NHMUK PV R 193 [38]; Thescelosaurus, NHMUK PV R 3675), Cumnoria (OUMNH J3303 [17] and Callovosaurus (NHMUK PV R 1993), but is present to a degree in dryosaurids (e.g. Valdosaurus, NHMUK PV R 184−185 [44]; Dysalotosaurus, NHMUK PV R 6861) and perhaps Auroraceratops [52], E Morschhauser 2023, personal communication) and Leptoceratops (CMN 8889) although it is not as well defined as it is in USMBA 001 and it is not bordered by distinct ridges in any of these taxa. This sulcus, bordered by ridges, appears to be unique to the specimen and could be an autapomorphy, but due to the incomplete nature of the material and its uncertain systematic position, we refrain from naming the taxon here. The posterolateral surface of the greater trochanter is strongly convex, while its lateral surface is more gently convex.

4. Discussion

Ethics

This work did not require ethical approval from a human subject or animal welfare committee.

Data accessibility

This article has no additional data.

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors’ contributions

S.M.: conceptualization, funding acquisition, investigation, methodology, project administration, writing—original draft, writing—review and editing; D.O.: data curation, investigation, writing—review and editing; R.J.B.: conceptualization, investigation, writing—original draft, writing—review and editing; K.B.: data curation, investigation, writing—review and editing; A.O.: investigation, writing—review and editing; K.E.: investigation, writing—review and editing; A.E.K.: data curation, writing—review and editing; P.M.B.: investigation, writing—original draft, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

Fieldwork funding for SCRM and RJB was provided by the Science Investment Fund of the NHMUK, and our team’s ongoing work in the Middle Jurassic of the Middle Atlas Mountains is funded by a British Institute of Libyan and North African Studies Flagship Grant awarded to SCRM and DO.