Postmortem distortion between the left and right sides of the skull of UALVP 40. (A) Photograph of skull roof in dorsal view and with anterior to left, showing medial displacement of the left postorbital relative to the median parietal bar (white arrows). As a result, the frontal fontanelle (ff) is nearly closed and the bases of the paired postorbital horncores are closer to each other than they would have been in life (cf. same region in the undistorted skull depicted in Fig. 8). (B) Interpretive line drawing of median and left sides of skull shown in oblique anterodorsal view, and with anterior to bottom of figure, showing more pronounced distortion of left side relative to the right side. Arrows 1 and 2 show medioventral and anterior movement, respectively, of the left half of the skull relative to the right half. Note that the base of the left postorbital horncore (pohc) has been displaced anterior to and below the right postorbital bone (rpo). Abbreviations: aof, antorbital fenestra; ej, epijugal; en, external naris; ff, frontal fontanelle; itf, infratemporal fenestra; j, jugal; n, nasal; nhc, nasal horncore; o, orbit; p, parietal; pf, parietal fenestra; pm, premaxilla; pohc, postorbital horncore; q, quadrate; rpo, right postorbital; stf, supratemporal fenestra; sq, squamosal. Images are at different scales. 

Context in source publication

Context 1

... margin on the outer side that possibly correspond with alveoli. The dentary appears deep relative to the overall size of the skull, even without the missing ventral portion. The coronoid process is erect and pronounced, forming a right angle with the long axis of the dentary ramus. As is characteristic of ceratopsids, the top of the coronoid process overhangs the ramus (Fig. 5). This overhanging portion of the coronoid process is much closer to the alveolar margin than seen in other specimens (cf. Fig. 5 vs. Godfrey and Holmes 1995, fig. 7), which may be an artifact of the restoration work (note the plaster filling in Fig. 5). There is a plateau that slopes slightly anteriorly at the anterodorsal corner of the dentary, which could represent a sutural surface for the predentary, although the mandible would be very short relative to the rest of the skull if that was the case. Alternatively, much of this area may represent the edentulous gap between the predentary and the dental margin (see Godfrey and Holmes 1995, fig. 7). A distinct horizontal ridge extends anteriorly along the external side of the dentary at its mid- height, from the descending anterior border of the coronoid process, before gradually merging to the laterally flared surface of the bone (Fig. 5). There is at least one foramen opening near the ventral edge of and slightly posterior to the center of the dentary. The bone thickens and gradually slopes up anteriorly towards the probable articulation with the predentary. The bone is also thickened mediolaterally at the base of the coronoid process. Unfortu- nately, the medial side of the left mandible is so poorly preserved that no clear structures are evident. The edentulous predentary is typically chasmosaurine in exhibiting a nearly horizontal pair of cutting flanges (Figs. 2, 6; Lehman 1990, fig. 16.1). The element is nearly complete, missing only small portions of the anterior and posterior ends. The element is short overall and its straight lateral borders readily converging anteriorly, which leads to a lack of an anterior attenuation of the element, a juvenile character in Agujaceratops mariscalensis (Lehman, 1989) (Lehman 1990, fig. 16.8B). The length along the dorsal border of the predentary is only about 20% longer than the horizontal length of the external naris in UALVP 40 compared to 40% (Godfrey and Holmes 1995, fig. 1A) to 50% (Dodson and Currie 1990, figs. 29.3b, 29.3c) in some of the larger specimens of Chasmosaurus . The study reveals several important previously unreported cranial features of UALVP 40 and permits reinterpretation of certain earlier observations (e.g., Gilmore 1923; Tyson 1977; Lehman 1990). First, comparisons of cranial features between the right and the left side of the skull reveal several contrasting differences. For instance, the jugal–squamosal suture is clearly visible on the right side (Fig. 2), whereas it is not at all discernable on the left. Although Tyson (1977) showed almost all of the cranial sutures on her diagram of the left side of the skull (represented here as Fig. 3), in her description she stated (Tyson 1977, pp. 45, 48) that “except for the squamosal–epoccipital (= episquamosal) sutures,” all the sutures “bordering the squamosal of UA(LVP) 40 are closed (on the left side).” The contrasting presence of a clear and “open” jugal– squamosal suture on the right side of the specimen is intriguing, and it may be that the same diagenetic process that crushed the right squamosal near its contact with the postorbital had forced open the suture on this side. Whatever the cause may have been, the presence of this suture on one side of the skull demonstrates that at least some facial sutures had not coossified at the time of the animal’s death. Similarly, the presence of a possible nasal– epinasal suture on the right side of the nasal horncore may be diagenesis-induced given the lateral compression experienced at the base of this horncore on this side (Fig. 2). It is unlikely that this groove represents a vascular tract, as it is much coarser than any of the vascular tracts on the nasal. Otherwise, the only epinasal suture reported in the genus Chasmosaurus is in CMN 1254, the type specimen of E. canadensis and subsequently assigned to C. belli by Godfrey and Holmes (1995). Difference in the orientations of the greatest dimension across the left and right orbits also can be attributed to postmortem distortion. Hence this cranial character, as Godfrey and Holmes (1995) pointed out already, cannot constitute a diagnostic character when assigning any chasmosaurine taxon (cf. Lull 1933). The new observation is thus in agreement with Godfrey and Holmes’ (1995, p. 731) suggestion that “the nearly circular orbital outline...probably represents the natural configuration” in Chasmosaurus . Finally, differences in the inclination of the postorbital horncores between the right and the left side of the skull also can be explained by postmortem distortions (Fig. 4). New observations presented in this study of the dorsal side of the skull indicate that the skull has been compressed laterally in such a manner that the left half portion slid first medioventrally, breaking the connec- tion between the left postorbital and the median parietal bar, and then shifted slightly anteriorly. As a result, the left horncore is now situated not only closer to the midline than normal, but also more shallowly tilted anteriorly than the right (cf. Fig. 2 vs. 3). Because the contact between the right postorbital and median parietal bar is still intact, it is reasonable to assume that the right side of the skull shows a more natural condition, especially in terms of the configurations and contacts of bones in the circumor- bital region. Thus, the higher inclination of the right postorbital horncore at around 60° probably represents a more original condition than that of the left. This new interpretation subsequently raises a question about Lehman’s (1990) interpretation of UALVP 40 as a female, which relied on the lower inclination of the left postorbital horncore of about 45° from horizontal. Furthermore, and more importantly, the marked discrepancy in this particular cranial feature seen between the two sides of the skull cautions against uncritically correlating postorbital horncore orientations with taxonomic and (or) intraspecific differences, such as sexual dimorphism, in ceratopsids even in articulated specimens like UALVP 40. The longer the orbital horncore, the more susceptible it would be to having its inclination altered by diagenetic distortion. Along similar lines, Godfrey and Holmes (1995) pointed out a skull of Chasmosaurus characterized by a narrow snout and an expanded frill is especially susceptible to postmortem distortion. That too is a major factor that should be taken into account when inferring the significance of cranial variation (i.e., taxonomic, individual, and preservational) in Chasmosaurus . There are several cranial features of potential ontogenetic significance in UALVP 40. Largely on the basis of the closure of the cranial suture lines in the frontal and squamosal regions, Tyson (1977) considered the skull of UALVP 40 to be that of an older individual, even though it was slightly smaller than the skull in the E. canadensis holotype (CMN 1254) (the deepest part of the squamosal is 315 mm vs. 340 mm, respectively (Godfrey and Holmes 1995, table 1)). However, there are many features, especially on the right side of the skull and the predentary of UALVP 40, that indicate the specimen likely represents a young, probably subadult individual as suggested by Lehman (1990). First, the rostral bone is missing from the specimen along the sutural surface with the premaxilla. As Dodson and Currie (1990, p. 594) noted, the rostral in ceratopsians fuses to the premaxilla “only in very mature individuals” and “frequently fails to be pre- served in juveniles.” Thus, the absence of the rostral in UALVP 40, a relatively complete, articulated skull, is consistent with its immature status. Second, many clearly defined sutures are present on the right side of the skull, including the jugal–squamosal, maxilla–premaxilla, ventral premaxilla–nasal, and possible nasal–epinasal sutures. Open sutures have been reported in CMN 1254, a small specimen (see Godfrey and Holmes 1995, table 1 for measurements) considered to be an immature animal (e.g., Tyson 1977, p. 48; Lehman 1990; Godfrey and Holmes 1995), but are not present in larger, articulated skulls of Chasmosaurus that presumably belonged to mature individuals (Godfrey and Holmes 1995; but also see NHMUK R4948, a relatively large C . belli specimen with its maximum squamosal depth of 360 mm, that includes well-preserved, disarticulated cranial elements (Maidment and Barrett 2011)). Tyson (1977) argued that because the sutures in the frontal region were indistinct in UALVP 40, it is an older individual than CMN 1254, in which the sutures are more conspicuous. In his description of 10–15 individuals of A. mariscalensis , Lehman (1989, p. 141) reported that among those individuals, “sutures in the frontal region that are discernable in some large specimens are completely closed in some smaller specimens,” leading him to con- clude that presence or absence of the sutures in this region is unreliable for age determination. As such, and in contrast to Tyson’s (1977) suggestion, it is possible that UALVP 40 is as young as, if not younger than, CMN 1254 despite the former exhibiting a more completely fused frontal region. Third, the epiossifications are at best weakly attached to the skull. The right epijugal has separated from the right jugal, expos- ing an undamaged sutural surface. According to Lehman (1990, pp. 225–226), the “fully co-ossified epijugal and epoccipitals (= episquamosals and epiparietals)...appears to represent a fully adult animal” in Pentaceratops sternbergii regardless of the size of the specimen. The ...