Skip to main content
NCBI home page
Journal of Anatomy logoLink to Journal of Anatomy
. 2019 May 9;235(1):151–166. doi: 10.1111/joa.12987

The original boneheads: histologic analysis of the pachyostotic skull roof in Permian burnetiamorphs (Therapsida: Biarmosuchia)

Zoe T Kulik 1,✉, Christian A Sidor 1
PMCID: PMC6580075  PMID: 31070781

Abstract

Thickened, pachyostotic skulls are best known in pachycephalosaur dinosaurs, but evolved convergently in Permian burnetiamorphs as well as in some other stem‐mammal groups and Triassic archosauromorphs. Until now, only pachycephalosaur domes have been histologically sampled to reveal patterns of bone tissue microstructure and growth. Using computed tomography and osteohistology, we serially thin‐sectioned one of the smallest burnetiamorph skull caps ever recovered (estimated skull length = 10 cm), as well as an individual nearly twice as large, and here report the first cranial histological data from this clade. We recognize four highly vascularized histological zones visible in coronal thin‐sections, only one of which shares morphological similarities with the tripartite zonation previously reported in pachycephalosaur domes. Zone A forms the endocranial region of the skull cap and records disorganized primary osteons in a fibrolamellar complex. Zone B preserves a border of compact, avascular layers of parallel‐fibered bone surrounding an interior of partially remodeled vascular canals. Interestingly, the outline of Zone B resembles the shape of an incipient skull roof. Zone C forms the thickest portion of the skull cap and is composed of fast‐growing woven bone with minimal osteonal development. The superficial Zone D has a matrix of predominantly woven bone with narrower primary vascular canals than in deeper regions of the skull caps. Unlike in pachycephalosaurs, where primary vascular porosity is thought to decrease through ontogeny, both burnetiamorph skull caps preserve a thick Zone C of highly vascularized tissue. Additionally, the remnants of sutures are visible as radial struts that taper superficially, leaving no trace on the surface of the skull. Even in the smallest individual, the sutures are closed ectocranially, which is unusual, given that some large, presumably adult pachycephalosaur domes preserve open sutural gaps. Although pachycephalosaur and burnetiamorph skull domes are superficially similar, histological analysis reveals differences in their vascularity and construction that imply multiple evolutionary pathways to form an elaborate pachyostotic dome.

Keywords: Biarmosuchia, Burnetiamorpha, histology, pachyostosis, skull, sutures, Therapsida

Introduction

Cranial pachyostosis (thickening of the dermal skull roof bones) evolved several times among vertebrates, but is best documented in the extreme dome‐headed pachycephalosaur dinosaurs of the Late Cretaceous (Sternberg, 1945; Sues & Galton, 1987; Maryanska et al. 2004; see also Stocker et al. 2016). However, two early diverging clades allied with the mammalian stem lineage also developed unusually thickened skulls, the burnetiamorphs and tapinocephalids. Early burnetiamorphs coexisted with tapinocephalids but were much smaller in body size (Rubidge and Sidor 2001). The diverse but relatively short‐lived tapinocephalid clade is restricted to the Guadalupian (~ 260 Ma; middle Permian, Barghusen, 1975; Rubidge, 2005; Kammerer, 2009; Olroyd & Sidor, 2017), whereas burnetiamorphs persisted until the end of the Permian, with 11 genera currently recognized (Day et al. 2018). Importantly, burnetiamorphs developed a wide range of bony adornments on their skulls, including crests, bosses, and dome‐like thickenings, which can be tied to their relatively speciose fossil record (Padian & Horner, 2011; Sidor et al. 2017). The cranial elements that are typically thickened in pachycephalosaurs, burnetiamorphs, and tapinocephalids include the nasals, frontals, parietals, and squamosals (Boonstra, 1968; Smith et al. 2006; Horner & Goodwin, 2009).

Among vertebrate groups that evolved pachyostotic skulls, only pachycephalosaurs have been analyzed histologically. Pachycephalosaur domes show a stereotyped tripartite zonation and feature indicators of rapid bone growth accompanied by a remarkable degree of external remodeling (Goodwin & Horner, 2004; Evans et al. 2018). Unfortunately, no comparative data are available to show whether similar architecture and underlying developmental processes were present in the other groups with pachyostotic skulls.

Here we provide novel data on the anatomy and osteohistology of the burnetiamorph skull roof, using both histology and computed tomography (CT). We infer rapid bone deposition, which has interesting implications for endothermy in this basal group of therapsids, and describe a quadripartite zonation of the skull cap, with some similarity to what has been shown in juvenile to subadult pachycephalosaurs.

Materials

The fossils described here were recovered from the lower Madumabisa Mudstone Formation from the mid‐Zambezi Basin of southern Zambia. Based on the general stratigraphic framework developed for the area, Sidor et al. (2014) considered the vertebrate‐bearing horizons to be Guadalupian (middle Permian) in age. There are currently 11 burnetiamorph species recognized, based exclusively on cranial material with conspicuous cranial thickenings or horn‐like projections (Sidor et al. 2015). Several other specimens representing novel geographic records have been deemed too incomplete to be named formally, yet preserve distinctive anatomy that allows for their unambiguous identification as members of Burnetiamorpha (e.g. Sidor et al. 2010; Kammerer, 2016). Despite their incompleteness, Zambian specimens described here compare very closely in their anatomy to previously described burnetiamorphs and were collected alongside more complete specimens that are the subject of an upcoming publication. They are currently given National Heritage Conservation Commission (NHCC) LB project numbers but will ultimately be transferred to the Livingstone Museum as the permanent repository (Sidor & Nesbitt, 2018).

Methods

Histology

Two burnetiamorph skull roofs (NHCC LB373 and NHCC LB410) were histologically prepared following the hard tissue sampling techniques outlined by Lamm (2013). Specimens were embedded in Epothin Epoxy/Resin 2, sectioned to a thickness of approximately 2 mm on an Isomet 1000 and glued to glass slides using 2‐Ton epoxy. Slides were ground on a Metaserv 3000 lapidary plate until the specimen was 80 μm thick or until optical clarity was reached. The smaller of the two skulls (NHCC LB373) was serially sectioned in the coronal plane (Fig. 1). The other (NHCC LB410) was sectioned in the coronal plane in the rostral region as well as in the horizontal and sagittal planes more posteriorly to capture more completely the three‐dimensional architecture of the bone tissue and vascular canals.

Figure 1.

Figure 1

Open in a new tab

Cranial anatomy of a middle Permian burnetiamorph (NHCC LB373). Skull prior to sectioning shown in dorsal (A), ventral (B), right lateral (C), and posterior (D) views. The five arrows in (A) correspond with the coronal thin‐sections shown in E–I, which are arranged from anterior (E) to posterior (I). Enlargement (J) highlights the four distinctive histological zones seen in all thin‐sections. f, frontal; p, parietal; pf, postfrontal; po, postorbital; pp, preparietal.

Thin sections were imaged using a Nikon Eclipse LV100POL microscope under plain and cross‐polarized light with lambda compensation. Composite images were processed using Nikon NIS‐elements BR (version 4.3) imaging software. High‐resolution images have been uploaded to the online repository MorphoBank (Project Number 3383: http://morphobank.org/permalink?P3383).

Computed tomography

Prior to histological preparation, NHCC LB410 was CT‐scanned and digitally reconstructed to reveal the three‐dimensional organization of bone tissue types throughout the entire skull cap. An additional burnetiamorph skull roof (NHCC LB118) which is nearly identical in size and shape to NHCC LB373 was also CT‐scanned but not histologically sampled (Fig. 2). Measurements, CT images, and 3D rendering were obtained using avizo 9.2.

Figure 2.

Figure 2

Open in a new tab

Cranial anatomy of a middle Permian burnetiamorph (NHCC LB118). Illustrations of the skull in dorsal (A), ventral (B), and left lateral (C) views. CT reconstruction of the skull cap in dorsal view (F) with arrow indicating position of virtual thin‐sections (D, E). Lines of higher bone mineral density (white lines in D) are manually traced as red lines (E), which can be recognized as the remnants of sutures in the reconstruction (F) and drawings (A) and (C). f, frontal; fb, frontal boss; for, foramina; mr, midline ridge; nc, possible attachment area for nasal capsule; of, orbital fossa; pf, postfrontal; pfor, parietal foramen; po, postorbital; pob, postorbital bar; pp, preparietal; sob, supraorbital boss; tf, temporal fossa.

Specimens NHCC LB118 and NHCC LB410 were scanned at the University of Washington using a high‐resolution NSI Feinfocus scanner. No filters were used, and the voltage and current varied respectively from 225 kV and 250 μA for NHCC LB118, 225 kV and 80 μA for NHCC LB410. Reconstructions were compiled using eFx‐CT with a resulting voxel size of 60.6 μm for LB118 and 62.1 μm for LB410. A 3D isosurface model rendered in Avizo 9.2 was used to visualize the surface pattern of sutures of the two specimens (Figs 1 and 5).

Results

Morphological description of NHCC LB118 and LB373

Background

NHCC LB118 and LB373 represent two of the smallest burnetiamorphs yet recovered. Both specimens preserve only the skull cap (i.e. the interorbital and intertemporal part of the skull roof) and are strikingly similar in size and shape. LB118 is the more complete of the two and measures 54 mm in anteroposterior length along the midline and 48 mm in maximum width; this suggests a skull length of approximately 100 mm when complete (based on comparisons with Lemurosaurus and Lende; Sidor & Welman, 2003; Kruger et al. 2015). The left side of the skull roof, posterior to the parietal foramen, is broken in NHCC LB373, revealing the thick pachyostotic nature of the dermal elements (17 mm in maximum thickness; Fig. 1A–D). Mallon & Evans (2014) showed that the central dome of pachycephalosaur crania was the portion most commonly preserved, with less durable elements of the face and palate preferentially lost; based on recent collecting in Zambia and Tanzania, burnetiamorphs appear to show the same pattern (Sidor et al. 2015).

As in most burnetiamorph skulls, the external surface of both specimens lacks obvious sutures and has a spongy or porous surface texture (Rubidge & Sidor, 2002; Kammerer, 2016). However, as shown below, petrographic thin‐sections reveal the probable remnants of sutures (Fig. 2A). In CT‐based digital thin‐sections, remnant sutures appear as bright white lines representing areas of high bone mineral density; these lines can be manually traced in consecutive CT slices to reveal the cranial elements that make up the burnetiamorph skull roof (Fig. 2F). The resulting arrangement of cranial bones is very similar to what is seen in the closest relatives of burnetiamorphs – non‐pachyostotic biarmosuchians such as Hipposaurus and Herpetoskylax (Sidor & Rubidge, 2006).

Description

Dorsal view

The following description is primarily based on NHCC LB118, as it is the more complete of the two specimens. The dorsal view is dominated by a relatively large, circular, matrix‐filled parietal foramen located at the center of the broadly convex intertemporal region. A gentle midline ridge extends posteriorly from the broken anterior margin of the skull cap, broadening slightly in the interorbital region, to the intertemporal region, where it encompasses the parietal foramen. One notable difference between the specimens is the presence of a step between the skull roof and the anterior edge of the parietal foramen in NHCC LB373, whereas the anterior face of the parietal foramen is vertical in NHCC LB118. Anteriorly, the median frontal ridge is separated from the supraorbital bosses by shallow, parasagittally elongated fossae. Interestingly, an irregular series of small foramina open at the base of these fossae, suggesting a rich neurovascular supply. The parasagittal fossae decrease in depth posteriorly to a point in line with the anteroposterior midpoint of the orbit. However, a shallow depression still separates the supraorbital bosses from the thickening surrounding the parietal foramen. This is in contrast to some larger and presumably more adult burnetiamorph skulls, where the separate thickenings are merged into a single domed structure (Kammerer, 2016).

In burnetiamorphs, a median preparietal ossification was first recognized in Pachydectes elsi by Rubidge et al. (2006) and was later confirmed in an unnamed burnetiamorph by Kammerer (2016), based on sutures observed on the ventral surface of an isolated skull cap. Corroborating what was observed in these two previous examples, our thin‐section data suggest that the preparietal resembles those of non‐burnetiamorph biarmosuchians and most anomodont therapsids, where it forms the anterior margin of the parietal foramen and projects anteriorly between the paired frontals (Sidor & Rubidge, 2006; Olroyd et al., 2018).

Lateral view

In lateral view (Figs 1C and 2C), the specimens vary more in the degree of supraorbital thickening than might be expected given their remarkable overall similarity; maximum thickness is 11.3 mm in NHCC LB118 and 8.3 mm in NHCC LB373. Nonetheless, in both specimens the lateral face of the supraorbital boss is vertical and peaks in height just anterior to the orbit midpoint, tapering anteriorly and posteriorly. It is unclear whether a distinct boss was present at the dorsal end of the postorbital bar, as in some burnetiids (viz. Burnetia and Bullacephalus). Based on the inferred suture patterns, the supraorbital boss is formed primarily by the frontal, with a subequal contribution by the postfrontal posteriorly. The prefrontal appears to have disarticulated from the pachyostotic skull cap in both specimens.

Ventral view

Among burnetiamorphs, the ventral surface of the skull roof has been described only for National Museum of Tanzania (NMT) RB4, an unnamed burnetiid from the Permian of Tanzania, although this region is preserved in the natural molds of Proburnetia viatkensis as well (Sidor et al. 2010). The underside of NHCC LB118 is similar to NMT RB4 but is more triangular in overall outline and preserves better surface detail, including an array of small foramina shown in Fig. 2A. In ventral view, the orbits form large semicircular fossae that approach one another on the midline, but are separated by a midline trough, which likely housed the dorsal portion of the brain and olfactory tract (Benoit et al. 2017). The margins of the trough are gently curved inward, such that the channel expands in width both anteriorly and posteriorly. As noted by Sidor et al. (2010), we assume that the ridges bounding the trough articulated with the sphenethmoid ossification in life. Anteriorly, the ridges show a pronounced lateral curvature, thereby forming a broad area on the ventral midline of the skull cap. This area, which represents the contact area between the frontals and the nasals (medially) and prefrontals (laterally) has a central thickening bounded on either side by slight depressions (Fig. 2B). Romer & Price (1940) suggested that in Dimetrodon, these depressions, or pockets, housed the posteromedial corners of the nasal capsule.

On the better preserved right side, a transverse ridge separates the orbit from the dorsal reaches of the temporal fossa and ends laterally at the broken surface representing the origination of the postorbital bar. Compared with other burnetiamorphs, what is preserved of the temporal region makes it appear remarkably short anteroposteriorly. On both sides, at least one foramen is present at the dorsomedial apex of the temporal fossa. The posterior margin of the skull cap is gently M‐shaped in dorsal view, with a median projection bounded by lateral embayments. A small, poorly developed supratemporal ‘horn’ is preserved on the right side. Posteriorly, the skull roof and occiput meet at right angles, with just the dorsal portion of the occiput intact. The sutural configuration in the temporal region is difficult to reconstruct but, in addition to the parietal and postorbital, we suspect that a small section of squamosal is present at the posterodorsal reaches of the temporal fossa.

Flakes of bone obscure some of the matrix‐filled parietal foramen in ventral view. In contrast to NMT RB4, the parietal foramen in NHCC LB118 is almost perfectly circular both dorsally and ventrally, measuring approximately 7 mm in diameter. Posterior to the parietal foramen is a broad, U‐shaped fossa, the margins of which are formed by sharp ridges of bone that descend ventrally. A subdued midline ridge is present posterior to the parietal foramen on the ventral surface of the parietals.

Osteohistology of NHCC LB373

We employ the terminology of Francillon‐Vieillot et al. (1990) in the histological description of bone tissue types of NHCC LB373, the first burnetiamorph skull cap to have its histology sampled and documented. What follows is an overview of five coronal thin‐sections (Fig. 1E–I) that record an overall pattern of rapid primary bone growth with minimal secondary remodeling (Fig. 3A–H). Natural histovariation in the degree of vascular canal organization and orientation seen across the thin‐sections can be separated into four distinctive zones of bone tissue organization.

Figure 3.

Figure 3

Open in a new tab

Photomicrographs of burnetiamorph skull roof histology (NHCC LB373) showing abundant woven bone. Silhouettes correspond to the thin‐sections shown in Fig. 1E–I and indicate the location of each photomicrograph. (A) The endocranial region of the skull cap is made up of Zone A with disorganized and elongate (arrow) primary osteons in the fibrolamellar (fbl) complex. (B) Zone B sits superficial to Zone A and consists of a border of parallel‐fibered (pf) bone surrounding primary (po) and secondary (so) osteons. (C) Disorganized primary canals interrupt the compact layers of Zone B (yellow outline) in the thin‐section near the opening of the parietal foramen. (D) An unusually wide wedge‐shaped region of primary canals appears to separate the postorbital from the postfrontal bone in this posterior thin‐section. (E) The highly vascularized Zone C under cross‐polarized light. A scaffold of woven bone (wb) is infilled by lamellar bone (lb) in this radially oriented fibrolamellar bone. (F) Radial fibrolamellar bone continues ectocranially to form Zone D. (G) Primary vascular canals preserve a more scalloped texture in the anterior‐most thin‐section with posterior regions of the skull cap. (H) The interfrontal suture under cross‐polarized light made up of overlapping arching trabeculae that connect across the open sutural gap (arrow). Scale bars: 500 μm (A–C,H); 1 mm (D); 100 μm (E–G). Lambda compensator was used in (B,C,E,F,H). Plain‐polarized light was used in (A,D,G).

Zones A and B

The deep endocranial tissue of the small burnetiamorph skull cap is histologically complex and includes two tissue types, which we recognize as Zone A and Zone B (Fig. 1J). Zone A is the deepest tissue type and is made up of disorganized primary osteons in the fibrolamellar complex (Zone A in Fig. 3A). Most of the primary canals are longitudinally oriented but some appear radial and extend towards the endocranial bone edge (arrow in Fig. 3A). Osteonal deposition has infilled each primary osteon, making this region of bone tissue the least porous across the entire skull cap. Zone A thickens posteriorly and forms the dorsal margin of the endocranial fossa where the brain was located.

Zone B, the second tissue type preserved in the endocranial region of the skull cap, is superficial to Zone A. It consists of a border of compact, avascular, parallel‐fibered to lamellar bone surrounding an interior of primary and secondary osteons (Fig. 3B). The layers of avascular tissue extend across the ventral region of the skull cap from each orbital margin and resemble the morphology of an embryonic skull cap (Fig. 1F,G).

Within the layers of parallel‐fibered and lamellar bone, cellular lacunae are well organized and lenticular in shape (see Supplemental image M657281 on MorphoBank). The bone tissue forming the interior of the embryonic skull outline consists of large secondary osteons and scarce primary osteons (Fig. 3B).

In the thin‐section that passes near the opening of the parietal foramen (Fig. 1H), the endocranial tissues do not resemble those of more anterior sections, as the compact layers of Zone B are absent from the midline region of the skull cap. Instead, the two layers of avascular parallel‐fibered bone are interrupted by an island of disorganized primary osteons that connect Zone A to Zone C (Fig. 3C). The absence of Zone B near the opening of the parietal foramen is not discernible in the digital thin‐sections of NHCC LB118, nor can its absence be confirmed in the more posterior thin‐sections of NHCC LB373 because the corresponding area is not preserved on that specimen.

The thin‐section just posterior to the parietal foramen (Fig. 1I) records a complicated internal structure, with a thicker portion of disorganized Zone A tissue types that extend laterally to the periphery of the skull roof. Here, Zones A and B contribute to wide, wedge‐shaped remnant suture patterns that extend dorsally to separate what appear to be five cranial elements on the right side. However, as the left side is not preserved, the identity of these elements is very difficult to confirm. From what is preserved, two median elements surround the posterior portion of the parietal foramen (Fig. 1I). In a very small burnetiamorph, Duhamel et al. (2018) suggested that the two median elements might represent multiple centers of ossification of the parietal bone, but this is difficult to confirm in our specimen without data from the left half of the skull cap. A deep sliver of what could be the frontal separates the parietal from the postfrontal (Fig. 1I). Laterally, the postfrontal is separated from the postorbital by a widely expanded region of reticular canals surrounded by woven and parallel‐fibered bone that does not appear to be a remnant suture since it is so wide and merges with deeper Zone A tissues (Fig. 3D). Unfortunately, not enough of the parietal region of the larger burnetiamorph specimen is preserved to clarify the complicated arrangement of elements observed in NHCC LB373.

Zone C

In the more superficial region of the skull cap, the bone tissue is drastically reorganized into a highly vascularized Zone C made up radial fibrolamellar bone. Zone C is the thickest of the four zones in all five thin‐sections and makes up the majority of the pachyostotic tissue (Fig. 1E–I). The radial fibrolamellar bone preserves minimal osteonal infilling. At high magnifications, thin layers of lamellar infill have an average thickness of only 43 μm. Cellular lacunae appear much more organized and lenticular within these thin layers of lamellar bone compared with the plump and polygonal lacunar shapes in the woven scaffold surrounding each vascular canal (Fig. 3E, Supplemental image M657284 on MorphoBank).

Zone D

Zone D is the most superficial region and is composed of radial fibrolamellar bone. The transition from Zone C to Zone D is not abrupt or easily discernible, as the radial canals transition from being largely open in Zone C to narrower and osteonally infilled in Zone D (Fig. 3F). As with Zone C, Zone D remains highly vascularized with a similar arrangement of woven and lamellar fibers making up the bone matrix. The average thickness of osteonal infill increases slightly from an average of 43 μm in Zone C to around 62 μm in Zone D. Additionally, canal lengths are shorter throughout Zone D and grade into reticular canal orientation (Fig. 3F).

Towards the snout, the texture of the ectocranial surface of the bone is more rugose than in more posterior thin‐sections (Fig. 3G). Although open vascular canals make up the ectocranial surface of all NHCC LB373 coronal thin‐sections (Fig. 3F), the anterior‐most section preserves slightly scalloped openings and larger foramina. Increased bone texture towards the rostral end of the skull cap could simply be a result of continued periosteal growth and resorption, or could suggest the presence of a thickened, epidermal covering in life (Hieronymus et al. 2009).

Remnant sutures

Throughout the pachyostotic tissue of NHCC LB373, the remnants of sutures are seen in thin‐section as thick, radial structures that extend from Zone B to the deeper portions of Zone D. These traceable struts are made up of overlapping arc‐shaped lamellae of woven and parallel‐fibered bone that occasionally form bone bridges connecting across the sutural gap (Fig. 3H). Most remnant sutures are obliterated towards the superficial portion of Zone D (Fig. 3F), whereas the deeper endocranial portion of the suture remains partially to completely open. CT data are of insufficient resolution to interpret the bone tissue microanatomy, with each remnant suture appearing as only a bright white structure. When viewed in histological thin‐sections, the complexity of the remnant sutures reveals bone tissues that change across the length of each suture and across anterior and posterior sections. Up to four remnant sutures are preserved with varying degrees of superficial obliteration (Fig. 1E–I). The configuration of these sutures allows for the recognition of the complement of cranial bones typically seen in Permian therapsids (Rubidge & Sidor, 2002; Rubidge & Kitching, 2003; Sidor & Welman, 2003; Sidor et al. 2004; Sidor & Smith, 2007; Kammerer, 2016).

Overall, the bone histology of the small burnetiamorph skull cap is predominantly made up of rapidly growing, highly vascularized, pachyostotic tissue. There is minimal remodeling in some of the deepest tissue of the skull cap, whereas the external surface consists of open vascular canals contributing to its spongy external surface.

Morphological description of NHCC LB410

Background

NHCC LB410 was selected for histological analysis because it represents a potential ‘adult version’ of the same taxon as NHCC LB118 and LB373. All three were collected from the same lower Madumabisa Mudstone Formation beds within 100 m of each other, and their skull caps preserve a similar set of thickenings and adornments. It should be noted, however, that our preliminary assessment of 13 burnetiamorph specimens from that area suggests that at least two distinct morphotypes are present. Nonetheless, the available evidence suggests that NHCC LB410 preserves external anatomy indicative of a later stage of burnetiamorph cranial ontogeny than the specimens previously described.

NHCC LB410 is a partial skull cap preserving the interorbital and intertemporal regions. As preserved, it measures 73.6 mm along the midline and 81.7 mm in maximum width. The left supraorbital boss measures 20.3 mm in maximum height (~ 80% thicker than in NHCC LB118) and the median frontal ridge is at least 33.7 mm in maximum height (~ 68% thicker than in NHCC LB118). Based on comparisons of the overall skull cap dimensions, we estimate NHCC LB410 to be just under twice as large as NHCC LB118.

Dorsal

The dorsal surface of this large skull cap is pitted and cracked but preserves both supraorbital bosses, a median frontal ridge, and a domed temporal region centered on a relatively small parietal foramen (Fig. 4A). The supraorbital boss is best preserved on the left side and has a vertical lateral face. The intertemporal region is relatively more domed than NHCC LB118, with less pronounced separation between the supraorbital and parietal thickenings. The spongy nature of the dorsal surface of the skull does not preserve obvious sutures. However, CT data reveal the same complement of bones as in NHCC LB118: paired frontals, postfrontals, and parietals, as well as a preparietal and part of the postorbital on the left side (Fig. 4C).

Figure 4.

Figure 4

Open in a new tab

Cranial anatomy of a middle Permian burnetiamorph (NHCC LB410). Skull prior to sectioning shown in dorsal view (A). (B) Digital reconstruction of the skull cap in dorsal view. (C) CT‐rendered sutures shown in red, highlight how the sutures change angle and shape in the pachyostotic vault. Virtual thin‐sections in the coronal plane (D, E) taken at the position of the arrow in (C). As in Fig. 2, areas of higher density were manually traced and then reconstructed to reveal suture patterns and organization of bones on the skull roof. f, frontal; pf, postfrontal; pfor, parietal foramen; po, postorbital; pp, preparietal

Ventral

The ventral surface of NHCC LB410 is strongly weathered and preserves surface detail only within the left orbit and a small section of the left temporal fossa. The median trough has been eroded away, as has much of the remaining morphology on the ventral face of the skull cap. A section through the parietal canal preserves a trapezoidal outline that tapers superficially.

Osteohistology of NHCC LB410

NHCC LB410 is approximately 45% thicker than NHCC LB343 or LB118, but the four zones of bone tissue types seen in these smaller burnetiamorph specimens can also be recognized here. However, less of Zone A is preserved due to the eroded ventral surface of this larger specimen. NHCC LB410 was sampled in the coronal, parasagittal, and horizontal planes to reveal more of the three‐dimensional nature of the pachyostotic bone tissue. The parasagittal thin‐section did not reveal new insights, as its tissue organization was very similar to the distribution of vascular canal orientations seen in the coronal thin‐section (see Supplemental image M657497 on MorphoBank).

The overall pattern of growth inferred for this larger burnetiamorph skull cap resembles that of the smaller specimen, with highly vascularized pachyostotic tissue consisting of woven and lamellar bone (Fig. 5A–H). Interestingly, as in many pachycephalosaur domes, extrinsic fibers are preserved along the ectocranial bone edge in this specimen, despite not being observed in the smaller burnetiamorph skull roof (Fig. 5A,B) (Goodwin & Horner, 2004; Horner & Goodwin, 2009; Evans et al. 2018).

Figure 5.

Figure 5

Open in a new tab

Photomicrographs of burnetiamorph cranial histology (NHCC LB410) showing evidence of rapidly growing bone with Sharpey's fibers near the dorsal surface of the skull roof. Silhouettes represent the cutting plane and location of each photomicrograph. (A) The left postorbital bone in frontal section shows a region of dense Sharpey's fibers (in brackets) near the external surface of Zone D. (B) Enlarged view of disorganized Sharpey's fibers (some highlighted with arrows) in frontal section. (C) Endocranial region of NHCC LB410 highlighting the arrangement of bone tissue types. (D) Enlargement of the reticular canals preserved throughout the deep region of Zone C, ectocranial edge of the bone is toward the upper right. (E) Enlargement of the radial vascular canals from Zone C made up of woven bone (wb) surrounded by lamellar infill (lb). (F) Remnants of the frontoparietal suture show extrinsic fibers (arrow) along the sutural gap in horizontal section. (G) Enlargement of (F) showing trabeculae that connect across the open sutural gap. (H) Vascular canals in horizontal section from Zone C near the parietal foramen (pfor, matrix‐filled semicircle at top right). Scale bars: 100 μm (A,B,E,G); 500 μm (D,F); 1000 μm (C,H). Lambda compensator was used in (F–H), all other images taken under plain‐polarized light.

Coronal section 1

The anterior region of NHCC LB410 was sectioned in the coronal plane. Here, the preparietal, frontal, and postfrontal bones are composed of primary bone tissue with scarce remodeling. Similar to NHCC LB373, compact, avascular bone tissue surrounding an interior of partially remodeled osteons (Zone B) is superficial to disorganized primary osteons (Zone A) in this larger specimen (Fig. 5C). Four remnant sutures extend from Zone B towards the superficial bone edge (Fig. 4E). Compared with the smaller specimen, NHCC LB410 preserves wider sutural areas that appear to extend farther towards the superficial edge of the bone, although the ectocranial surface is incomplete along most of the parietal and preparietal. Nonetheless, even when complete, the outer surface of the skull preserves no indication of sutures (Fig. 4A). Endocranially, the remnant sutures appear open and infilled with dark, opaque matrix (Fig. 5F,G, also see Supplemental images M657451, M657450 on MorphoBank). At high magnifications (10–20×), thin bony trabeculae connect across the sutural gap (Fig. 5G), which is similar to the condition observed in the smaller specimen where the sutures are open in the deep regions of the skull cap.

Zone C constitutes approximately 83% of the total thickness of this larger skull cap and is similarly made up of highly vascularized canals of woven and lamellar bone (Fig. 5E). Vascular canal orientations transition superficially from longitudinal and reticular in the deep region, where primary canals appear reworked, to open radial canals in the superficial region of Zone C (Fig. 5D,E). Average bone lamella thickness decreases across Zone C from an average thickness of 98.5 μm in the deep Zone C region to 42 μm in the superficial region. This pattern suggests that cranial growth occurred in the periosteal or superficial direction.

Within Zone C, lacunar morphology varies from elongate to ovoid in cross‐section and observable canaliculi are absent (Fig. 5E). In contrast with NHCC LB373, the lacunar morphologies of NHCC LB410 are not as plump or circular, but the density of lacunae remains high.

Primary radial canals continue to extend into Zone D, where canal orientation shifts to become more reticular (Fig. 5A,B). Vascular canal diameters are reduced due to osteonal infilling and are on average 35–60 μm thicker than in Zone C. The bone fiber organization in this region is predominantly woven with abundant, disorganized extrinsic, or Sharpey's fibers, seen in the horizontal plane (Fig. 5B). These fibers are concentrated along the outermost ectocranial bone surface, when preserved. The relatively thin margin of disorganized fibers could have served as attachment sites for muscle, skin or display structures (Hieronymus et al. 2009) but are not uniformly organized and instead resemble a dense plywood texture more similar to mineralizing fibers in fast‐growing bone tissue. It is also possible that these fibers are related to reworking or repair along the ectocranial surface. Without more of the bone surface preserved, the function of these fibers remains ambiguous.

Horizontal section 1

The left posterior region of NHCC LB410 was sectioned in the horizontal plane through the deeper reaches of Zone C. This portion of the skull roof is formed by the preparietal, frontal, postfrontal, postorbital, and parietal bones. The horizontal thin‐section records the mid‐dorsal region of the skull including the left half of the parietal foramen and therefore preserves none of the tissue types from Zone A or B (Fig. 5H). Instead, the majority of the tissue is highly vascularized woven and lamellar bone with abundant primary vascular canals oriented roughly perpendicular to the plane of section. More laterally on the skull cap, vascular canal orientation shifts to a more oblique angle remaining perpendicular to the curvature of the skull dome.

Throughout the entire horizontal thin‐section, numerous interdigitating sutures can be traced (see Supplemental image M657496 on MorphoBank). Unlike in the coronal section, which shows the sutures tapering dorsally, this top‐down view of the sutural gap records a relatively constant thickness (approximately 35–55 μm). Interestingly, the bone tissue that contributes to the sutures preserves extrinsic fibers that are not visible in the coronal plane. In some areas, the fibers are oriented perpendicular to the sutural margin (Fig. 5F). Densely fibrous sutural bone tissue has been reported in derived dicynodont therapsids (Jasinoski & Chinsamy‐Turan, 2012) and marginocephalians (Bailleul & Horner, 2016), and is common in modern mammals (e.g. humans; Cohen, 1993).

The horizontal thin‐section of NHCC LB410 provides a novel view that further confirms the location of the preparietal element in this burnetiamorph skull roof. Using the geometry and morphology of sutural margins in the frontal plane and reconstructed CT visualizations of the entire skull cap, the preparietal can be seen in direct contact with the anterior portion of the parietal foramen (Fig. 4C). Among biarmosuchians, the placement of the preparietal varies as well as the shape of the bone itself. Our data indicate that an arrowhead‐shaped preparietal forms the anterior margin of the parietal foramen. In NHCC LB373, the preparietal also contacts the parietal foramen, but preserves a more triangular outline in dorsal view.

Discussion

General observations of burnetiamorph skull histology

Both burnetiamorph skull caps show abundant, highly vascularized bone tissue, suggesting that both individuals were actively growing at their time of death. This forces us to reconsider the inferred ‘adult’ status of the larger specimen, as its histology is very similar to the smaller, presumably juvenile specimen. In both fossils, Zone B likely represents the earliest formed bone as it is the only tissue to show remodeling, a time‐dependent process that requires a foundation of primary bone. Additionally, the compact layers surrounding the remodeled tissue in Zone B form an outline that resembles that of an embryonic skull roof. Interestingly, Zone B can be recognized in CT images of a very small individual of the basal burnetiamorph Lemurosaurus (Benoit et al. 2017: fig. 6a). We suggest that after an initial region of compact bone formed, periosteal growth in the endocranial (Zone A) and ectocranial (Zones C and D) directions resulted in the expanded and pachyostotic morphology typical of the burnetiamorph skull cap.

In both burnetiamorph specimens, bone tissue is made up of the same distribution of woven and lamellar bone matrix, suggesting similar rates and patterns of growth. Additional similarities include the quadripartite zonation pattern of vascular canal orientations, with Zone C being the thickest in both specimens. Even though both skull roofs are thick, they are not very densely ossified: a typical thin‐section from NHCC LB373 is only 73% bone tissue and 77% bone tissue in NHCC LB410, the rest being open vascular space.

Despite the histological similarities in both burnetiamorph skull caps, the larger one preserves notable differences in the degree of reworking and osteonal development within the pachyostotic tissue. Unlike the smaller specimen, where Zone C is composed of radial vascular canals, the deeper Zone C tissue from NHCC LB410 has shorter, more disorganized radial canals with thicker osteonal development. This deep region was likely reworked while the skull was still growing, because the superficial‐most tissue consists of fast‐growing woven bone matrix and immature primary canals. The smaller specimen similarly preserves fast‐growing bone at its outermost surface but does not preserve reworking of the pachyostotic tissue, confirming its younger ontogenetic status.

Cranial histology in pachycephalosaurs

Background

Pachycephalosaur skull dome histology has been described by several authors (Goodwin & Horner, 2004; Horner & Goodwin, 2009; Jasinski & Sullivan, 2011; Schott et al. 2011; Goodwin & Evans, 2016; Evans et al. 2018) and provides a useful starting point for comparison to burnetiamorphs. In particular, Goodwin & Horner (2004) recognized three zones (1–3) with distinct histological characteristics in pachycephalosaur domes. In partially complete frontoparietal domes, the deepest Zone 1 consists of disorganized, deformed primary and secondary osteons that appear endochondral in origin. Zone 2 is made up of highly vascularized, spongy tissue that contributes to the thickest portion of pachyostotic bone in juvenile domes. Zone 3 records less vascularized bone tissue, especially in skeletally mature individuals where the thickness of Zone 3 has increased due to periosteal infilling of primary vascular spaces. These authors concluded that early in ontogeny, the domes of pachycephalosaurs were rapidly expanding until skeletal maturity was reached, at which point vascular porosity decreased, resulting in avascular and acellular regions in some of the largest domes (Goodwin & Horner, 2004; Horner & Goodwin, 2009; Evans et al. 2018). Horner & Goodwin (2009) hypothesized that pachycephalosaur frontoparietal domes formed through metaplastic processes because the lacunae they observed in coronal thin sections of adult Pachycephalosaurus wyomingensis lack observable canaliculi, indicating that they might represent fibrocytes rather than osteocytes. Metaplastic bone growth occurs without the presence of a periosteum (Haines & Mohuiddin, 1968); notably, we have not observed evidence of this growth pattern in burnetiamorph cranial tissue.

Horner & Goodwin (2009), Schott et al. (2011), and Schott & Evans (2012) described ontogenetic series of pachycephalosaur frontoparietal domes in Stegoceras validum and P. wyomingensis. They found major ontogenetic changes to include increased overall size, pronounced modifications in shape (e.g. remodeling of the squamosal nodes; Williamson et al. 2009), increased sutural fusion, decreased vascularity in Zones 2 and 3, and a surface texture that changed from exposed Sharpey's fibers forming a rugose surface relief to erosional pitting. Additionally, previous authors have suggested that as much as two‐thirds of the adult dome dimensions may have been attained during a phase of highly vascularized tissue growth prior to infilling of the porous vascular network (Goodwin & Horner, 2004; Lehman, 2010). The histological characteristics observed in these ontogenetic series have been used to interpret the ontogenetic stages of other pachycephalosaur domes (Evans et al. 2018) and offer a comparative dataset for burnetiamorph growth.

Pachycephalosaur comparison

Pachycephalosaurs and burnetiamorphs share thickened skull caps that are formed entirely of solid (albeit vascular) bone. This is in contrast to many extant vertebrates (e.g. proboscideans, bovids, cetaceans) that thicken their skull roof by expanding the spongiosa or by developing large sinuses between the endocranial cavity and periosteal surface of the skull. Overall, the pattern of bone deposition observed in the two burnetiamorph specimens suggests a rapidly growing and thickening skull cap. In pachycephalosaurs, vascular porosity decreases throughout ontogeny, as vascular canal spaces are centripetally infilled. Vascular porosity remains high in both our samples, a characteristic of rapid growth, with minimal centripetal infilling. Even in the larger specimen, despite it being almost twice as large as the small specimen, highly vascularized tissue makes up the majority of the skull cap. The retention of highly vascularized pachyostotic tissue in the larger specimen suggests either that both of these individuals had not reached somatic maturity or that burnetiamorphs employed a different growth strategy than pachycephalosaurs. Rapid growth commonly occurs in juveniles of both endothermic and ectothermic species (de Margerie et al. 2002; Tucker et al. 2007), but if NHCC LB410 represents a rapidly growing adult, then the proposition that early therapsids were endothermic gains support (Van Valen, 1960).

Zones of growth in burnetiamorphs and pachycephalosaurs

Macroscopically, the zonation pattern in burnetiamorphs resembles the pattern in pachycephalosaur frontoparietal domes first reported by Goodwin & Horner (2004). However, closer inspection indicates different tissue constructions in all zones except Zone C, which resembles Zone 2 in pachycephalosaurs. Compact, avascular bone akin to Zone B has not been reported from endocranial regions of pachycephalosaur domes nor are we able to distinguish this zone in dorso‐ventrally complete virtual thin‐sections (Jasinski & Sullivan, 2011; Schott et al. 2011; Williamson & Brusatte, 2016; Schott & Evans, 2016) . According to Evans et al. (2018) and Goodwin & Horner (2004), endochondral bone makes up the endocranial tissue in pachycephalosaur domes, and consists of ropey and disorganized primary and secondary osteons in longitudinal and reticular orientations in coronal cross‐section. Endocranial tissue (Zone A) from burnetiamorph skull caps is disorganized with elongated primary osteons and resembles Zone 1 tissues from the smallest pachycephalosaur domes (see Figure 9.A in Bailleul & Horner, 2016) but does not approach the degree of disorganization reported from sub‐adult to adult pachycephalosaur endocranial tissues (Goodwin & Horner, 2004; Evans et al. 2018). Zone D in burnetiamorphs differs from corresponding pachycephalosaur tissue types as it remains highly vascularized with anastomosing radial canals, unlike the dense, sparsely vascularized tissue common in large pachycephalosaur frontoparietal domes. Even in pachycephalosaur domes that are inferred to represent immature individuals, the ectocranial tissue is denser and less vascularized when compared with burnetiamorphs.

Suture development in thickened skulls

Compared with the pachycephalosaur model, burnetiamorph histology is striking in (1) the lack of dense, avascular tissue in the larger burnetiamorph and (2) the presence of completely closed sutures (at least towards the superficial surface of the dome). In relatively young pachycephalosaurs, the interfrontal and frontoparietal sutures remain open. At later ontogenetic stages, most sutures of the skull roof close, except for the frontoparietal suture, which remains open internally but is obliterated on the external surface of the skull (Williamson & Brusatte, 2016; Bailleul & Horner, 2016). The presence of open sutures in subadult pachycephalosaurs might not be surprising given the variation in sutural closure captured in extant archosaurs (Bailleul et al. 2016), but the fusion of sutures in the skull cap of one of the smallest burnetiamorphs ever found was unforeseen.

In humans, the premature fusion of cranial sutures (craniosynostosis) is pathological, with severe consequences for the infant (Moore & Persaud, 2003). By comparison, nearly all burnetiamorphs have been described as having obliterated sutures, at least externally in the region of skull cap (e.g. Sidor et al. 2004, 2010). However, the relative brain size of burnetiamorphs was much smaller than in humans and brain growth could have been accommodated by ventral expansion, so the fusion of cranial sutures likely did not have significant fitness consequences.

Cranial and craniofacial elements change shape and size through bone remodeling and cortical drift, where bone tissue is resorbed and redeposited towards one surface (Enlow, 1966; Martin et al. 2010). These processes obscure original patterns of growth but allow bones to maintain a proportional construction. Our burnetiamorph data strongly suggest that periosteal appositional growth resulted in the superficially obliterated suture patterns and thick, pachyostotic tissue. In juvenile pigs, sutural closure develops on the ectocranial side due to faster periosteal growth, whereas the endocranial side remains open (Sun et al. 2004). In fast‐growing, thick‐skulled burnetiamorphs, the same pattern likely took place. If growth at sutures contributed to the thickening of the skull cap, then sutures would have to be maintained towards the ectocranial surface (Opperman, 2000). Instead, suture patterns taper superficially, suggesting an initially wide sutural area that was later fused and obliterated during appositional expansion. Additionally, each cranial vault element is triangular or wedge‐shaped in coronal cross‐section, with Zones C and D predominantly made up of radially oriented canals and trabeculae highlighting the centrifugal and outward expansion of bone tissue.

Sutures and ontogenetic stage

Assessing the ontogenetic status of the burnetiamorphs sampled here is difficult. There is a common assumption that sutural fusion marks the cessation of growth, but craniofacial growth has been shown to continue after sutural fusion through periosteal apposition in both modern archosaurs and mammals (Cohen, 1993; Sun et al. 2004; Bailleul & Horner, 2016; Bailleul et al. 2016). In our sample, the presence of two mature individuals is not supported, as nearly the entirety of both skull caps is made up of immature, highly vascularized, primary tissue. If both skull caps were skeletally mature individuals, then we might expect to see more osteonal infilling and well‐developed primary osteons, similar to what is reported in mature pachycephalosaurs. Instead, both burnetiamorphs preserve highly vascularized primary tissue, with minimal periosteal infilling and abundant woven bone along the ectocranial bone edge.

Vascular pachyostosis

Among extant mammals, thickening of the skull roof is typically accomplished by the development of large spaces (diploë) between upper and lower compact layers of bone following an overall thickening of the skull bones in early ontogeny (Sun et al. 2004; Hall, 2005; Farke, 2010; Snively et al. 2011). One exception is in the extinct aquatic sloths, Thalassocnus littoralis and Thalassocnus carolomartini, where the cranial bones were recently shown to be remarkably more dense than normal skull elements (Amson et al. 2018). Adaptations to an aquatic environment are often related to pachyostosis, but this and related terms can encompass a variety of presentations across the skeleton (Houssaye et al. 2016). Indeed, pachyostosis is such a broadly applied term that its meaning has become unclear, especially histologically (Table 1). For example, it can describe completely different patterns of bone growth (i.e. the difference between increased bone compactness via periosteal infilling or expanded bone morphology due to thickened cortical bone growth). This forces authors to define and re‐define the term ‘pachyostosis’ based on how it relates to a specific change.

Table 1.

Previously used terms and definitions to describe thickened bone

Term Definition Citation
Pachyostosis Osseous specialization characterized by an increase in bone compactness or volume Houssaye (2009)
Pachyostosis sensu lato Non‐pathological bone hypertrophy Abel 1912 in Houssaye (2009)
Pachyostosis sensu stricto Hyperplasy of periosteal cortices that leads to a change in bone morphology by increasing volume Nopsca 1923 in Houssaye (2009)
Pachyostosis Unusually thickened bone Hall (2005)
Hyperostosis Deposition of abnormally high amounts of primary bone, typically pathological Francillon‐Vieillot et al. (1990)
Osteosclerosis Increased inner compactness either as a result of endochondral ossification, inhibition of remodeling or the filling of inner cavities with no effect on the external morphology of the bone Domning & de Buffrénil (1991)
Pachyosteosclerosis Thickening and greater density of bone resulting in heavy but fragile bones typical of manatees and dugongs Domning & de Buffrénil (1991)
Open in a new tab

There have been at least four major factors used to diagnose pachyostosis in the past. The two that are most commonly discussed are as follows: (1) whether a bone is enlarged, swollen or inflated compared with its normal form, as is common in sirenian ribs (de Buffrénil et al. 2010) or (2) whether a bone has increased density by removing vascular pore space, as mentioned above in aquatically adapted sloths (Amson et al. 2018). These two factors are not mutually exclusive, as increased bone volume can be accompanied by increased tissue density (termed pachyosteosclerosis in some extinct Sirenia), but more commonly results from cortical or periosteal hyperplasy (sometimes termed hyperostosis, but see Hall's, Abel's, and Nopsca's definitions, Table 1). A third aspect of variation is developmental, in terms of whether a bone is of endochondral or dermal origin, which can result in characteristic differences in bone thickening or greater density (Francillon‐Vieillot et al. 1990; Jin et al. 2016). Finally, the presence of enlarged sinuses within a bone, as seen in head‐butting ungulates like goats (Farke, 2008), is the fourth aspect of variation commonly mentioned. Considering these four axes of variation, we can define burnetiamorph pachyostosis as (1) enlarged, (2) not dense (i.e. vascularized), (3) dermal, and (4) of roughly uniform composition (i.e. not invaded by sinuses). What is unusual about the increased bone volume in burnetiamorph skulls is that much of the tissue remains highly vascularized, even after it is reworked as seen in the larger specimen.

Our reading of the literature suggests that the highly vascular, yet remarkably expanded, cranial bone seen in burnetiamorphs has unique histological characteristics shared with only the sub‐adult ontogenetic stages of pachycephalosaur domes. We suggest using ‘vascular pachyostosis’ to distinguish a hypertrophied cranial bone that is formed by a bony scaffold of primarily woven bone tissue supported by a large proportion of vascular canals. Additional histological work will be needed to confirm whether the anatomy observed in more adult burnetiamorphs, tapinocephalids, and Triopticus (a Triassic archosaur with a thickened skull) conforms to this definition.

Soft tissue inferences

The evolution of bizarre cranial structures in extinct species has long attracted interest among paleobiologists (e.g. Hopson, 1975). For example, the horns and frills of ceratopsian dinosaurs have been suggested to function in defense, mate choice, and species recognition (Sampson, 1999; Padian & Horner, 2011), but critical to any inference of function is an understanding of the soft‐tissue enveloping such bony features. By comparing soft tissues and their underlying osteohistological correlates in extant vertebrates, Hieronymus et al. (2009) were able to infer soft‐tissue structures in fossil centrosaurine ceratopsians. The same osteohistological correlates can be used to infer the presence of keratinous horns or other thickened display structures in burnetiamorphs.

Despite the elaborate bony morphology seen in some burnetiamorphs, we were unable to find evidence for thickened epidermal attachments. The smaller specimen (NHCC LB373) preserves open vascular canals at the superficial bone surface and is not pitted or reworked. At high magnifications, the rostral‐most thin‐section shows a slight resemblance to the rugose texture of cornified pad attachment surfaces in muskoxen (see Fig. 8D in Hieronymus et al. 2009) but the scale of the rugose pits are much larger in muskox. We interpret the slight pitting as vascular canals opening into the periosteum and skin. In the larger specimen (NHCC LB410), the degree of pitting and surface resorption indicative of horns, cornified pads or other elaborate display structures does not match the relatively smooth yet cracked (due to preservation) external surface. The presence of a thin and disorganized layer of Sharpey's fibers along the outermost surface of the skull cap suggests the possibility of a soft tissue display structure, but evidence for a pronounced display structure is lacking.

Conclusion

We report the first histological analysis for a burnetiamorph therapsid skull roof and show that it predominantly consists of highly vascularized, primary radial canals. The bone tissue construction of burnetiamorphs is unique when compared with other convergently evolved thickened skulls, as it maintains spongy, vascular bone across nearly a doubling of skull dimensions. Burnetiamorph bone tissue is similar in construction to that of immature, fast‐growing pachycephalosaur domes, and we suggest that the term vascular pachyostosis be applied to this anatomy. Rapid, vascular pachyostotic growth likely caused the cranial sutures to be obliterated in the ectocranial regions of burnetiamorphs. Further histological analyses should investigate whether vascular pachyostosis persists in larger and presumably more mature burnetiamorphs. In addition, sampling the thickened skulls of tapinocephalid dinocephalians would bolster the histological comparative sample of bizarrely adorned mammalian predecessors. Finally, while the presence of rapidly growing, highly vascularized tissue is consistent with hypotheses that early therapsids were endothermic, the evolution of thermoregulation in mammalian forebears remains difficult to infer with confidence (de Ricqlès, 1974; Hillenius, 1994; Kemp, 2006; Hopson, 2012; Rey et al. 2017).

Author contributions

The project was conceived by C. A. Sidor. Histological analysis and interpretation were completed by Z. T. Kulik. The manuscript was written and edited by both authors.

Supporting information

High‐resolution photomicrographs of NHCC LB373 and NHCC LB410 as well as full‐slide images are available on MorphoBank ( http://morphobank.org/permalink/?P3383).

Click here for additional data file. (1.1GB, zip)

Acknowledgements

We thank Adam Huttenlocker, Alida Bailleul, Tracy Popowics, Megan Whitney, Savannah Olroyd, and Sue Herring for thoughtful review and discussion. We also thank Lucas Legendre and an anonymous reviewer for improving the quality of this paper. Megan Whitney assisted in making thin‐sections at the University of Washington. We thank April Neander for her illustrations (Fig. 2A–C). We acknowledge the members of the field teams that worked in Zambia 2012 and 2014 for helping to collect the fossils described here. This study was supported by NSF EAR‐1337569 (to C.A.S.), EAR‐1337291 (to K. Angielczyk), NGS 8962‐11 (to C.A.S.), and a grant awarded to ZTK from the University of Washington Department of Biology.

References

  1. Amson E, Billet G, de Muizon C (2018) Evolutionary adaptation to aquatic lifestyle in extinct sloths can lead to systemic alteration of bone structure. Proc Biol Sci 285, 20180270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bailleul AM, Horner JR (2016) Comparative histology of some craniofacial sutures and skull‐base synchondroses in non‐avian dinosaurs and their extant phylogenetic bracket. J Anat 229, 252–285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bailleul AM, Scannella JB, Horner JR, et al. (2016) Fusion patterns in the skulls of modern archosaurs reveal that sutures are ambiguous maturity indicators for the Dinosauria. PLoS One 11, e0147687. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Barghusen HR (1975) A review of fighting adaptation in dinocephalians (Reptilia, Therapsida). Paleobiology 1, 295–311. [Google Scholar]
  5. Benoit J, Fernandez V, Manger PR, et al. (2017) Endocranial casts of pre‐mammalian therapsids reveal an unexpected neurological diversity at the deep evolutionary root of mammals. Brain Behav Evol 90, 311–333. [DOI] [PubMed] [Google Scholar]
  6. Boonstra LD (1968) The braincase, basicranial axis and median septum in the Dinocephalia. Ann S Afr Mus 50, 195–273. [Google Scholar]
  7. de Buffrénil V, Aurore C, D'Anastasio R, et al. (2010) Evolution of sirenian pachyosteosclerosis, a model‐case for the study of bone structure in aquatic tetrapods. J Mamm Evol 2, 101–120. [Google Scholar]
  8. Cohen MM (1993) Sutural biology and the correlates of craniosynostosis. Am J Med Genet 47, 581–616. [DOI] [PubMed] [Google Scholar]
  9. Day MO, Smith RMH, Benoit J, et al. (2018) A new species of burnetiid (Therapsida, Burnetiamorpha) from the early Wuchiapingian of South Africa and implications for the evolutionary ecology of the family Burnetiidae. Pap Palaeontol 4, 453–475. . [Google Scholar]
  10. Domning DP, de Buffrénil V (1991) Hydrostasis in the Sirenia – quantitative data and functional interpretations. Mar Mamm Sci 7, 331–368. [Google Scholar]
  11. Duhamel A, Benoit J, Rubidge JS (2018) Juvenile biarmosuchians from the Karoo Beaufort Group shed new light on basal therapsid ontogeny. African Paleobiology Symposium.
  12. Enlow DH (1966) A morphogenetic analysis of facial growth. Am J Orthod 52, 283–299. [DOI] [PubMed] [Google Scholar]
  13. Evans DC, Hayashi S, Chiba K, et al. (2018) Morphology and histology of new cranial specimens of Pachycephalosauridae (Dinosauria: Ornithischia) from the Nemegt Formation, Mongolia. Palaeogeogr Palaeoclimatol Palaeoecol 494, 121–134. [Google Scholar]
  14. Farke AA (2008) Frontal sinuses and head‐butting in goats: a finite element analysis. J Exp Biol 19, 3085–3094. [DOI] [PubMed] [Google Scholar]
  15. Farke AA (2010) Evolution and functional morphology of the frontal sinuses in Bovidae (Mammalia: Artiodactyla), and implications for the evolution of cranial pneumaticity. Zool J Linn Soc 159, 988‐1014. [Google Scholar]
  16. Francillon‐Vieillot H, de Buffrénil V, Castanet J, et al. (1990) Microstructure and mineralization of vertebrate skeletal tissues In: Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends, Vol. 1 (ed. Carter JG.), pp. 471–530. New York, NY: Van Nostrand Reinhold. [Google Scholar]
  17. Goodwin MB, Evans DC (2016) The early expression of squamosal horns and parietal ornamentation confirmed by new end‐stage juvenile Pachycephalosaurus fossils from the Upper Cretaceous Hell Creek Formation, Montana. J Vertebr Paleontol 36, e1078343. [Google Scholar]
  18. Goodwin MB, Horner JR (2004) Cranial histology of pachycephalosaurs (Ornithischia: Marginocephalia) reveals transitory structures inconsistent with head‐butting behavior. Paleobiology 30, 253–267. [Google Scholar]
  19. Haines RW, Mohuiddin A (1968) Metaplastic bone. J. Anat. 103, 527‐538. [PMC free article] [PubMed] [Google Scholar]
  20. Hall BK (2005) Bones and Cartilage: Developmental Evolutionary Skeletal Biology. San Diego, CA: Elsevier Academic Press. [Google Scholar]
  21. Hieronymus TL, Witmer LM, Tanke DH, et al. (2009) The facial integument of centrosaurine ceratopsids: morphological and histological correlates of novel skin structures. Anat Rec 292, 1370–1396. [DOI] [PubMed] [Google Scholar]
  22. Hillenius WJ (1994) Turbinates in therapsids: evidence for Late Permian origins of mammalian endothermy. Evolution 48, 207–229. [DOI] [PubMed] [Google Scholar]
  23. Hopson JA (1975) The evolution of cranial display structures in hadrosaurian dinosaurs. Paleobiology 1, 21–43. [Google Scholar]
  24. Hopson JA (2012) The role of foraging mode in the origin of therapsids: implications for the origin of mammalian endothermy In Fieldiana: Life and Earth Sciences. (eds Lombard RE, Anderson J, Ruta M, Sumida SS.), pp. 126–148. Chicago, IL, USA: Field Museum of Natural History. [Google Scholar]
  25. Horner JR, Goodwin MB (2009) Extreme cranial ontogeny in the Upper Cretaceous dinosaur Pachycephalosaurus . PLoS One 4, e7626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Houssaye A (2009) ‘Pachyostosis’ in aquatic amniote: a review. Integr Zool 4, 325–340. [DOI] [PubMed] [Google Scholar]
  27. Houssaye A, Sander PM, Klein N (2016) Adaptive patterns in aquatic amniote bone microanatomy – more complex than previously thought. Int Comp Biol 56, 1349–1369. [DOI] [PubMed] [Google Scholar]
  28. Jasinoski SC, Chinsamy‐Turan A (2012) Biological inferences of the cranial microstructure of the dicynodonts Oudenodon and Lystrosaurus In: Forerunners of Mammals: Radiation, Histology, Biology. (ed. Chinsamy‐Turan A.), pp. 149–176. Indianapolis, IN: Indiana University Press. [Google Scholar]
  29. Jasinski SE, Sullivan RM (2011) Re‐evaluation of pachycephalosaurids from the Fruitland‐Kirtland transition (Kirtlandian, Late Campanian), San Juan Basin, New Mexico, with a description of a new species of Stegoceras and a reassessment of Texacephale langstoni . New Mex Mus Nat Hist Sci Bull 53, 202–215. [Google Scholar]
  30. Jin S‐W, Sim K‐B, Kim S‐D (2016) Development and growth of the normal cranial vault: an embryologic review. J Korean Neurosurg Soc 59, 192–196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kammerer CF (2009) Cranial Disparity in the Non‐Mammalian Synapsida. pp. 1–466. Chicago, IL: University of Chicago. [Google Scholar]
  32. Kammerer CF (2016) Two unrecognized burnetiamorph specimens from historical Karoo collections. Palaeontol Afr 50, 64–75. [Google Scholar]
  33. Kemp TS (2006) The origin of mammalian endothermy: a paradigm for the evolution of complex biological structure. Zool J Linn Soc 147, 473–488. [Google Scholar]
  34. Kruger A, Rubidge BS, Abdala F, et al. (2015) New burnetiid therapsid from Malawi contributes to understanding of biarmosuchian phylogeny and biogeographic distribution. J Vertebr Paleontol 35, e1008698 10.1080/02724634.2015.1008698. [DOI] [Google Scholar]
  35. Lamm E‐T (2013) Preparation and sectioning of specimens In: Bone Histology of Fossil Tetrapods: Advancing Methods, Analysis, and Interpretation. (eds Padian K, Lamm E‐T.), pp. 55–160. Berkeley, CA: University of California Press. [Google Scholar]
  36. Lehman TM (2010) Pachycephalosauridae from the San Carlos and Aguja formations (Upper Cretaceous) of west Texas, and observations of the frontoparietal dome. J Vertebr Paleontol 30, 786–798. [Google Scholar]
  37. Mallon JC, Evans DC (2014) Taphonomy and habitat preference of North American pachycephalosaurids (Dinosauria, Ornithischia). Lethaia 47, 567–578. [Google Scholar]
  38. de Margerie E, Cubo J, Castanet J (2002) Bone typology and growth rate: testing and quantifying ‘Amprino's rule’ in the mallard (Anas platyrhynchos). CR Biol 325, 221–230. [DOI] [PubMed] [Google Scholar]
  39. Martin BR, Burr DB, Sharkey NA (2010) Skeletal biology In: Skeletal Tissue Mechanics. (ed. Smith R.), pp. 29–78. New York, NY: Springer. [Google Scholar]
  40. Maryanska T, Chapman RE, Weishampel DB (2004) Pachycephalosauria In: The Dinosauria. (eds Weishampel DB, Dodson P, Osmolska H.), pp. 464–477. Berkeley, CA: University of California Press. [Google Scholar]
  41. Moore KL, Persaud TVN (2003) The Developing Human: Clinically Oriented Embryology, 7th edn Philadelphia, PA: Saunders. [Google Scholar]
  42. Olroyd SL, Sidor CA (2017) A review of the Guadalupian (middle Permian) global tetrapod fossil record. Earth Sci Rev 171, 583–597. [Google Scholar]
  43. Olroyd SL, Sidor CA, Angielczyk KD (2018) New materials of the enigmatic dicynodont Abajudon kaayai (Therapsida: Anomodontia) from the lower Madumabisa Mudstone Formation, middle Permian of Zambia. J Vertebr Paleontol 37, e1403442 10.1080/02724634.2017.1403442. [DOI] [Google Scholar]
  44. Opperman L (2000) Cranial sutures as intramembranous bone growth sites. Dev Dyn 219, 472–485. [DOI] [PubMed] [Google Scholar]
  45. Padian K, Horner JR (2011) The evolution of ‘bizarre structures’ in dinosaurs: biomechanics, sexual selection, social selection or species recognition. J Zool 283, 3–17. [Google Scholar]
  46. Rey K, Amiot R, Fourel F, et al. (2017) Oxygen isotopes suggest elevated thermometabolism within multiple Permo‐Triassic therapsid clades. eLife 6, e28589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. de Ricqlès A (1974) Evolution of endothermy: histological evidence. Evol Theory 1, 51–80. [Google Scholar]
  48. Romer AS, Price LI (1940) Review of the Pelycosauria. Geol Soc Am Spec Pap 28, 1–538. [Google Scholar]
  49. Rubidge BS (2005) Re‐uniting lost continents – fossil reptiles from the ancient Karoo and their wanderlust. S Afr J Geol 108, 135–172. [Google Scholar]
  50. Rubidge BS, Kitching JW (2003) A new burnetiamorph (Therapsida: Biarmosuchia) from the lower Beaufort Group of South Africa. Palaeontology 46, 199–210. [Google Scholar]
  51. Rubidge BS, Sidor CA (2001) Evolutionary patterns among Permo‐Triassic therapsids. Ann Rev Ecol Syst 32, 449–480. [Google Scholar]
  52. Rubidge BS, Sidor CA (2002) On the cranial morphology of the basal therapsids Burnetia and Proburnetia (Therapsida: Burnetiidae). J Vertebr Paleontol 22, 257–267. [Google Scholar]
  53. Rubidge BS, Sidor CA, Modesto SP (2006) A new burnetiamorph (Therapsida: Biarmosuchia) from the middle Permian of South Africa. J Paleont 80, 740–749. [Google Scholar]
  54. Sampson SD (1999) Sex and destiny: the role of mating signals in speciation and macroevolution. Hist Biol 13, 173–197. [Google Scholar]
  55. Schott RK, Evans DC (2012) Squamosal ontogeny and variation in the pachycephalosaurian dinosaur Stegoceras validum Lambe, 1902, from the Dinosaur Park Formation, Alberta. J Vertebr Paleontol 32, 903–913. [Google Scholar]
  56. Schott RK, Evans DC, Goodwin MB, et al. (2011) Cranial ontogeny in Stegoceras validum (Dinosauria: Pachycephalosauria): a quantitative model of pachycephalosaur dome growth and variation. PLoS One 6, e21092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Schott RK, Evans DC (2016) Cranial variation and systematics of Foraminacephale brevis gen. nov. and the diversity of pachycephalosaurid dinosaurs (Ornithischia: Cerapoda) in the Belly River Group of Alberta. Canada. Zool J Linn Soc. 179, 865‐906. [Google Scholar]
  58. Sidor CA, Nesbitt SJ (2018) Introduction to vertebrate and climatic evolution in the Triassic rift basins of Tanzania and Zambia In: Vertebrate and Climatic Evolution in the Triassic Rift Basins of Tanzania and Zambia. (eds Sidor CA, Nesbitt SJ.), pp 1–7, Society of Vertebrate Paleontology Memoir 17. J Vertebr Paleontol 37 (6, supplement). [Google Scholar]
  59. Sidor CA, Rubidge BS (2006) Herpetoskylax hopsoni, a new biarmosuchian (Therapsida: Biarmosuchia) from the Beaufort Group of South Africa In: Amniote Paleobiology: Perspectives on the Evolution of Mammals, Birds, and Reptiles. (eds Carrano MT, Gaudin T, Blob R, Wible J.), pp. 76–113. Chicago, IL: University of Chicago Press. [Google Scholar]
  60. Sidor CA, Smith RMH (2007) A second burnetiamorph from the Permian Teekloof Formation of South Africa and its associated fauna. J Vertebr Paleontol 27, 420–430. [Google Scholar]
  61. Sidor CA, Welman J (2003) A second specimen of Lemurosaurus pricei (Therapsida: Burnetiamorpha). J Vertebr Paleontol 23, 631–642. [Google Scholar]
  62. Sidor CA, Hopson JA, Keyser AW (2004) A new burnetiamorph therapsid from the Teekloof Formation, Permian, of South Africa. J Vertebr Paleontol 24, 938–950. [Google Scholar]
  63. Sidor CA, Angielczyk KD, Weide DM, et al. (2010) Tetrapod fauna of the lowermost Usili Formation (Songea Group, Ruhuhu Basin) of southern Tanzania, with a new burnetiid record. J Vertebr Paleontol 30, 696–703. [Google Scholar]
  64. Sidor CA, Angielczyk KD, Smith RMH, et al. (2014) Tapinocephalids (Therapsida, Dinocephalia) from the Permian Madumabisa Mudstone Formation (Lower Karoo, Mid‐Zambezi Basin) of southern Zambia. J Vertebr Paleontol 34, 980‐986. [Google Scholar]
  65. Sidor CA, Knaub CR, Angielczyk KD, et al. (2015) Tanzania and Zambia yield an unprecedented fossil record of burnetiamorph therapsids. J Vertebr Paleontol 35 (electronic supplement), 213. [Google Scholar]
  66. Sidor CA, Angielczyk KD, Nesbitt SJ, et al. (2017) Burnetiamorphs did it first: cranial adornment and rates of speciation in a Permian lineage of therapsids. J Vertebr Paleontol 37 (electronic supplement), 194. [Google Scholar]
  67. Smith RMH, Rubidge BS, Sidor CA (2006) A new burnetiid (Therapsida: Biarmosuchia) from the upper Permian of South Africa and its biogeographic implications. J Vertebr Paleontol 26, 331–343. [Google Scholar]
  68. Snively E, Theodor JM, Carpenter K (2011) Common Functional Correlates of Head‐Strike Behavior in the Pachycephalosaur Stegoceras validum (Ornithischia, Dinosauria) and Combative Artiodactyls. PLoS ONE 6, e21422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Sternberg CM (1945) Pachycephalosauridae proposed for dome‐headed dinosaurs, Stegoceras lambei, n. sp., described. J Paleont 19, 534–538. [Google Scholar]
  70. Stocker MR, Nesbitt SJ, Criswell KE, et al. (2016) A dome‐headed stem archosaur exemplifies convergence among dinosaurs and their distant relatives. Curr Biol 26, 1–7. [DOI] [PubMed] [Google Scholar]
  71. Sues HD, Galton PM (1987) Anatomy and classification of the North American Pachycephalosauria (Dinosauria: Ornithischia). Palaeontographica (A) 198, 1–40. [Google Scholar]
  72. Sun Z, Lee E, Herring SW (2004) Cranial sutures and bones: growth and fusion in relation to masticatory strain. Anat Rec A Discov Mol Cell Evol Biol 276(A), 150–161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Sun Z, Lee E, Herring SW (2007) Cell proliferation and osteogenic differentiation of growing pig cranial sutures. J Anat 211, 280‐289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Tucker AD, Limpus CJ, McDonald KR, et al. (2007) Growth dynamics of freshwater crocodiles (Crocodylus johnstoni) in the Lynd River, Queensland. Aus J Zoo 6, 409–415. [Google Scholar]
  75. Van Valen L (1960) Therapsids as mammals. Evolution 14, 304–313. [Google Scholar]
  76. Williamson T, Brusatte S (2016) Pachycephalosaurs (Dinosauria: Ornithischia) from the Upper Cretaceous (upper Campanian) of New Mexico: a reassessment of Stegoceras novomexicanum . Cret Res 62, 29–43. [Google Scholar]
  77. Williamson TE, Carr TD, Williams SA, et al. (2009) Early ontogeny of pachycephalosaurine squamosals as revealed by juvenile specimens from the Hell Creek Formation, eastern Montana. J Vertebr Paleontol 29, 291–294. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

High‐resolution photomicrographs of NHCC LB373 and NHCC LB410 as well as full‐slide images are available on MorphoBank ( http://morphobank.org/permalink/?P3383).

Click here for additional data file. (1.1GB, zip)

Articles from Journal of Anatomy are provided here courtesy of Anatomical Society of Great Britain and Ireland

RESOURCES