Scelidosaurus, Norman, 2020

Norman, David B, 2021, Scelidosaurus harrisonii (Dinosauria: Ornithischia) from the Early Jurassic of Dorset, England: biology and phylogenetic relationships, Zoological Journal of the Linnean Society 191 (1), pp. 1-86 : 31-34

Scelidosaurus

Contrary to Richard Owen’s interpretation that ‘In different parts of the matrix of blocks (Tabs. II and IV) are portions of long and slender bones, which are, most probably, abdominal ribs’ ( Owen, 1863: 6), there is no convincing evidence of gastralia in Scelidosaurus and there is no evidence of postcranial pneumatism ( Norman, 2020b). If Carrier & Farmer (2000a) are correct in their general thesis concerning archosaur aspiratory mechanisms, then the dinosauromorph ancestors of this lineage (represented by Herrerasaurus – see Fig. 27A ) are likely to have used a combination of costal and cuirassal aspiration. The additional presence of osteoderms (retention of an archosaur symplesiomorphy?), or even their secondary development in thyreophorans, may have influenced rib mobility and the extent to which the costal system could contribute to volume/pressure changes within the thoracic cavity (cf. crocodiles). Such constraints might have required thyreophorans (including Scelidosaurus ) to use additional (but so far unknown) mechanisms for ventilation.

Costal aspiration: Brocklehurst et al. (2017) examined rib mobility in juvenile alligators. One particular consideration was the arrangement of the para- and diapophyseal articular facets along the dorsal series and the degree to which the positioning of the facets affected rib motion. Archosaurs (and alligators are no exception; Brocklehurst et al., 2017: fig. 1; see also: Norman, 1980: figs 34, 37) display migration of these facets along the dorsal vertebral series: the parapophysis rises from the lateral surface of the centrum to the transverse process and then migrates toward the distal end of the latter before finally merging with the diapophysis. In Scelidosaurus , the parapophyses of the first three dorsals rise sequentially to occupy a position on the anteroventral edge at the base of the transverse process. However, once this position is attained, no further migration takes place until the last two dorsals in the series ( Norman, 2020b: figs 20, 21). The parapophyseal facets are all dimple-shaped and smoothly concave, indicating a normal synovial joint. The diapophyseal facets are more complex having a ventral half that is synovial and a dorsal half that is pockmarked by ligament pits ( Norman, 2020b: fig. 16, pits), suggesting that the upper half of the suture between the diapophysis and tuberculum formed a firmer, fibrous joint.

The ribs are notable for their span when placed in articulation ( Norman, 2020b: fig. 34): the back of the animal was broad, rather than narrow and deep. The distal ends of the thoracic ribs are bluntly truncated, indicating that their ends articulated with sternal rib cartilages (none of the latter are preserved). The presence of facets on the trailing edges of some of the longer anterior dorsal ribs, as well as what appears to be a partly mineralized plate ( Norman, 2020b: fig. 35A), shows that uncinate plates were present – but may have been localized to the middle of the shafts of the longer thoracic ribs – and similar indications of uncinate plates have been reported in other ankylosaurs (e.g. Brown, 1908; see also: Gilmore, 1930) and the basal stegosaur Huayangosaurus ( Zhou, 1984: fig. 19). The more specialized slender uncinate processes reported in some avialians have been linked to the mechanics of the respiratory system ( Codd et al., 2008). Recent work ( Codd et al., 2019) has also shown that in alligators the iliocostalis musculature is attached to the uncinate plates on the ribs and was capable of facilitating rib movement, augmenting thoracic compression (exhalation) under terrestrial conditions. Slender, uncinate processes are present on all the dorsal ribs of the extant lepidosaurian Sphenodon , but these have muscular connections to the gastralia and may be associated with cuirassal aspiration ( Codd et al., 2019).

In Scelidosaurus , the uncinate plates are mineralized sheets of tissue (probably cartilage) that are in their structure more similar to the flap or flangelike uncinate plates seen in crocodiles. These are here interpreted primarily as rib stabilizers or ‘spacers’, but there is the possibility that they also anchored iliocostalis musculature that facilitated exhalation, as described in alligators.

Ribs may also play a role in supporting the osteoderms that form conspicuous rows running the length of the torso. The most prominent osteoderms are found along the flanks of the animal and lie adjacent to the distal ends of the rib shafts. There is superficial similarity between the osteoderms covering the back of Scelidosaurus and those seen in crocodiles, but crocodilian osteoderms are more closely packed and articulate, creating a flexible dermal carapace, whereas those of this dinosaur do not articulate (apart from in the cervical region), being more widely spaced. The area between individual torso osteoderms comprises a semi-flexible dermis formed by a mosaic of much smaller osteoderms ( Norman, 2020c: fig. 36). Individual large osteoderms are lightly constructed, with a thin cortex and a cancellous medulla ( Norman, 2020c: fig. 43), so the weight of the dermal armour was unlikely to have been excessive. The extent to which osteoderms were anchored by ligaments to the ribcage and formed part of a tension system similar to that seen in extant crocodilians is unclear, but is considered to be unlikely, given their widely spaced and nonarticular arrangement in the thoracic region. The ossified tendon bundles that flanked the neural spines of the dorsal vertebrae would have tensioned the backbone in an analogous manner to the osteodermbased tension system of crocodilians. However, the larger osteoderms were likely to have been capped by keratinous sheaths ( Brown, 2017) and formed rows of defensive spikes. Firm anchorage in the dermis and to the underlying skeleton (ribs) might be expected – and this anchoring is perhaps reflected in the preservation of semi-natural arrays of these osteoderms in some articulated skeletons (e.g. BRSMG LEGL 0004 – Norman, 2020c: fig. 8).

Overall, the ribcage of Scelidosaurus has a broad span and there was a conspicuous uniformity in the articular relationships (and range of possible movements) of the thoracic ribcage. Rib motion is likely to have been of the bucket-handle type ( Brocklehurst et al., 2017), which would have permitted modest volumetric change in the thoracic cavity. This may have been augmented by the iliocostalis musculature that, in alligators, inserted on the uncinate plates ( Codd et al., 2019). The longer thoracic ribs were stabilized proximally by fibrous connections running across the diapophyseal–tubercular joints, and their shafts were ‘spaced’ distally by the presence of cartilaginous uncinate plates (that became mineralized in more mature individuals), so that the thoracic ribs moved as a parallelogram-like unit. The ribs supported and helped to anchor the larger osteoderms. However, there is no convincing evidence that the ribs and osteoderms formed a tensioning system that stiffened the backbone, as in crocodiles. Modest uniform flexure of the thoracic ribs provided a mechanism for costal aspiration.

Cuirassal aspiration: In the absence of gastralia, there is no osteological evidence for cuirassal aspiration

and this aspiratory mode is considered improbable by Carrier & Farmer (2000a).

Diaphragmatic aspiration: Aspiratory mechanisms akin to those seen in extant crocodilians cannot be entirely discounted because diaphragmatic muscles might have been anchored to the prepubic blades, although their lateral positioning counts against this possibility, provided that the crocodilian model of hepatic pistoning is an applicable comparator.

Pelvic aspiration: The observations of Carrier & Farmer (2000a) prompt brief consideration in relation to the respiratory capacity/potential in Scelidosaurus . Pelvic flexure (between the dorsal vertebrae and sacrum) can be excluded, given the presence of bundles of ossified tendons that run along the entire dorsal series and are anchored to the sacrum. These would inhibit flexure between the dorsal series and sacrum. The pubis and ischium are not fused to the ilium but articulate with the latter via thick connective tissue pads. The pubic shaft and ischium are equal in length, whereas the prepubic process forms a laterally directed blade. In basal ornithischians ( Lesothosaurus – Baron et al., 2017a; Heterodontosaurus – Santa Luca, 1980; Galton, 2014), the prepubic process forms a rectangular plate that projects anteriorly beneath the preacetabular process of the ilium to a greater extent than indicated in the reconstruction of Carrier & Farmer (2000a: fig. 11A). In Scelidosaurus , the prepubic process is short in juvenile individuals (NHMUK R6704: Norman, 2020b: fig. 73) but becomes a more substantial rectangular plate in larger (subadult) individuals ( Norman, 2020b: fig. 74).

The pubic shaft is a long rod, co-terminous with the distal end of the ischium and lies against the shaft of the ischium. The Carrier–Farmer model of rotation of the pubic shaft permits the prepubic blade, if adducted, to compress the broadly expanded posterior abdominal wall. If diverticula or air sacs were present (unknowable), respiration would have been augmented. A model involving a combination of costal (possibly diaphragmatic?) and pelvic aspiration (achieved by pubic mobility) is at least plausible for Scelidosaurus .

OPISTHOPUBIC PELVIC STRUCTURE: A REFLECTOR OF HERBIVORY OR RESPIRATORY BIOLOGY?

Macaluso & Tschopp (2018) argued that it was necessary to demonstrate that respiration was more likely to be an ‘evolutionary driver’ of opisthopuby in dinosaurs than was herbivory. The basis for this proposition was a false premise: that opisthopuby in all dinosaurs had previously been causally linked to the adoption of an herbivorous diet. Furthermore, they claimed that this idea had been proposed by Weishampel & Norman (1989). However, Weishampel & Norman never made such a claim in that article. Rather opisthopuby, which characterizes Ornithischia , was proposed as a biomechanical adaptation that permitted small ornithischian herbivores to retain a bipedal posture and limb proportions indicative of cursoriality in the face of predation by coeval bipedal and cursorial theropods ( Norman & Weishampel, 1991). This latter proposition was never expanded by these authors to encompass all dinosaurian subclades. The evolution of an analogous form of opisthopuby among some avian-theropods, although interesting per se from an evolutionary perspective – particularly in light of the work of Baron et al., (2017b, c) – has always been considered (certainly by Norman and Weishampel) to be a functionally and physiologically unrelated matter.

An awareness of cranial adaptations that can be interpreted as indicators of a herbivorous or omnivorous diet among theropods (traditionally considered exclusively carnivorous) was highlighted in a general review by Barrett & Rayfield (2006). Their general thesis was further developed by Zanno et al., (2009) in which a wider range of cranial, as well as postcranial, morphologies, and their distribution among taxa, were mapped phylogenetically across a range of coelurosaurian theropods ( Zanno et al., 2009). They concluded that coelurosaurian theropods were not primitively ‘hypercarnivorous’ but were dietarily flexible, ranging from herbivory through omnivory to carnivory, and that strict carnivory was a secondary specialization found in one group of paravian theropods (dromaeosaurids). The recognition of dietary flexibility among coelurosaurians was posited as an evolutionary benefit because it allowed them to be dietary opportunists.

Returning to the issue of respiration vs herbivory as a driver of opisthopuby, Gatesy & Dial (1996a, b) demonstrated that the evolution of opisthopuby is coupled with a reconfiguration of body proportions and limb function among Avialae. This alteration presaged the ‘modularized’ bodies of extant birds. In short, the tail undergoes progressive reduction in its skeletal and muscle mass; as a consequence, the cantilevering effect of the tail is reduced and it simultaneously reduces its capacity to anchor the principal hindlimb retractor muscles (m. caudifemoralis longus). An anatomical marker reflecting the reduction of the femoral retractor musculature is the size and prominence of the femoral 4 th trochanter, which progressively reduces before disappearing completely in ‘paravian’ theropods (see Fig. 28 ). To maintain a bipedal pose, a number of subclades of shorter- or slender-tailed avian-theropods evolved degrees of pubic retroversion, ranging from intermediate (mesopuby) to full opisthopuby – followed by eventual separation of the pubes and ischia in the midline so that the abdomen can extend posteriorly beneath and behind the sacroiliac vault (see Figs 27 , 28 ). These changes reflect a rebalancing of the body to compensate for the loss of the cantilevering effect of the tail. There is also a consequential repurposing of the hindlimb. The femur becomes an exclusively anteriorly directed suspension member of the hindlimb, and the knee-joint adopts the role of a ‘neoacetabulum’. The anteriorly displaced knee-joint becomes the centre of balance and the locomotor stride of the hindlimb is achieved by swinging the elongated tibiotarsus and tarsometatarsus, pendulum-like, from the knee (as is the case in extant birds).

In this general context, gross expansion of the gut cavity and retroversion of the pubis (linked to a shortening of the tail), and the maintenance of a bipedal stance, can be correlated with herbivory in the highly modified opisthopubic condition described in therizinosaurs, such as Nothronychus ( Zanno et al., 2009) , but these are truly exceptional theropodans (see Fig. 27D ).

OBSERVATIONS

Macaluso & Tschopp (2018) undertook a study that was purposely restricted in scope: they limited the biological ‘drivers’ considered to just two. In terms of logic, the study is internally consistent, in that they consider whether dinosaurs exhibit carnivory, omnivory or herbivory, assign these traits to the taxa under consideration and then plot their assignments on a general phylogeny ( Macaluso & Tschopp, 2018: fig. 2). They indicate, on that phylogeny, the presence or absence of gastralia and whether pelvic anatomy was ‘propubic’, ‘mesopubic’ or ‘opisthopubic’ – employing the terminology adopted by Rasskin-Gutman & Buscalioni (2001). Dinosaur taxa (representative of selected dinosaurian subclades) are then scored according to the authors’ interpretation of pelvic morphology, diet and respiratory capacity. An analytical protocol was applied to their scores, which promotes the view that opisthopuby is more strongly correlated with respiratory mechanics than with herbivory in these dinosaurs.

Their approach conflates an objective analytical protocol with a set of subjective decisions concerning diet and respiratory capacity, and uses simplified twodimensional images of hip structure. They admit in the discussion section ‘that a change in the ventilatory system was [not] the only evolutionary force acting on the structure of the archosaurian pelvis. For instance, egg morphology, locomotion, nesting behaviour and reproductive organs could all have been equally influential’ ( Macaluso & Tschopp, 2018: 714). There was no mention of herbivory, but these other factors were not explored because they were considered by the authors to be ‘more difficult to recognize in the skeleton’ (p. 714).

On the basis of the published literature and available descriptions, it can be stated objectively that ornithischians and sauropods are the only dinosaur groups that show no evidence of gastralia. Representatives of all the theropod clades considered by Macaluso & Tschopp are known to possess gastralia [contra Macaluso & Tschopp, 2018: fig. 2 – note that node 4 in this figure implies that all ‘Pennaraptoran’ taxa (oviraptorosaurs, dromaeosaurs + Sinovenator ) possess gastralia and yet, paradoxically, were designated as non-cuirassal breathers]. Taken in total, the approach adopted in their article establishes a false premise and subsequently fails to account for the range and variety of anatomy, inferred biology and functional organization of these animals – all of which have a material bearing on our understanding. It has always been understood (certainly by Norman & Weishampel) that the unique evolution of the opisthopubic pelvis in ornithischians, and also seen to have arisen iteratively among some avian-theropod subclades, were independent, anatomically distinct and functionally unrelated events.

To provide an overview of the biological and functional issues associated with respiratory capacity and its linkage to the evolution of pelvic structure in dinosaurs (and their extant descendants birds), a set of summary comments is offered.

SUMMARY

Irrespective of basal dinosaur systematics ( Baron et al., 2017b, c; Langer et al., 2017), it can be agreed (following the work of Carrier & Farmer, 2000b) that stem-lineage taxa (dinosauromorph archosaurs) had mobile bicipital dorsal ribs and gastralia. This indicates that they were capable of using, to varying degrees, a combination of costal and cuirassal modes of aspiration. Gastralia, and by implication cuirassalstyle aspiration, are retained (symplesiomorphically) in Late Triassic dinosaurian taxa belonging to Sauropodomorpha and Theropoda ( Fig. 24 ), but this anatomical character and the inferred aspiratory mechanism is absent (synapomorphically) in the Early Jurassic clade that has a sister-taxon relationship with Theropoda : the Ornithischia ( Fig. 24 ) – although it does not matter from which basal dinosaurian clade the Ornithischia are derived in this regard. It is established that the iterative evolution of degrees of opisthopuby among subclades of Theropoda can be linked functionally to the reorganization of the bodies of these animals ( Gatesy & Dial, 1996a, b). It is only among Aves (flight-adapted birds) that gastralia are lost and the structural adaptations associated with the avian flow-through respiratory system can plausibly be inferred (see Fig. 27 ).

Figures

Figure 27. Theropod skeletal forms. A, Herrerasaurus (a dinosauriform, stem-dinosaur or a basal theropod – according to various analyses). B, Allosaurus a tetanuran theropod – see Fig. 28). C, Ornithomimus a coelurosaur. D, Nothronychus, a coelurosaur. E, Oviraptor, a maniraptoran. F, Deinonychus, a paravian. Herrerasaurus and all other theropods possess gastralia. Images kindly provided by Scott Hartman who retains the copyright of each. Scale bar in centimetres.

Figure 28. Simplified theropod phylogeny. All tetanuran theropods have gastralia; these are only lost in true birds (n.gast). The orientation of the pubis varies across these clades and is dependent upon multiple factors: dietary preference (herbivory, omnivory, carnivory); body proportions (notably the reduction in length and mass of the tail); and more specific locomotor adaptations and habits of individual theropods within each subclade. There is no simple and unambiguous correspondence between pelvic structure and posture, habit, locomotor style or putative diet and respiratory mechanics. Dinosaur images kindly provided by Scott Hartman, who retains the copyright of each. As illustrated here these animals are not strictly to the same scale. The raptor silhouette was made available through the following website: http://clipart-library.com/bird-silhouette.html

Figure24. Asimplified phylogenyofdinosaurs(afterBaron et al., 2017b). Note that gastralia are lost independently in sauropods and ornithischians but retained in all other clades [being also lost independently in the derived, powered flight-capable, Theropoda (= birds)].